http://www.lsoft.com/images/listserv-community.gifhttp://www.lsoft.com/images/listserv_small.gifSPM ListSPM List Archiveshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?RSS&L=SPM&v=ATOM1.0
This is an Atom formatted XML site feed. It is intended to be viewed in a Newsreader.
Alternatively you can view the HTML archives at https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A0=SPM.
LISTSERV Web Interface 16.52024-03-28T09:28:29ZRik Henson2024-03-28T09:28:13+00:002024-03-28T09:28:13+00:00Re: How to determine the optimal stimulus sequence for a slow-event fMRI experiment?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;6b8da5c6.2403Actually, you do not need additional jitter to optimise fMRI efficiency with such long and varied SOAs – your HRF-convolved regressors will already be reasonably decorrelated. Jitter becomes more important when you have a short minimal SOA, and needs to be over a much larger range than 1-2 seconds to have any effect on correlation between regressors (eg 2-12 seconds)*. And even then, such jitter is only needed if you want to estimate the response versus inter-event baseline; if you only care about differences between two types of events, you do not need to jitter, but you do need to randomise the order of those event-types (in which case, the shortest SOA possible is optimal, assuming no nonlinear saturation of the HRF). So if you don’t care about contrasts versus baseline (and are happy to assume a fixed canonical HRF), then I would ask why you are using such long SOAs (unless you have a limited number of stimuli, since more TRs is nearly always more efficient)?<br><br>With 6 conditions (event-types) and such long SOAs (assuming they need to be long for psychological reasons), some orders of event-types can more optimal than others, to avoid too much signal (for your contrasts of interest) being lost in the low-frequency fMRI noise (eg if you highpass filter). That optimal order depends on which contrasts (effects) in your 2x3 design are most important – you want such contrasts to have the highest frequency (ie vary fastest).<br><br>This webpage is probably easiest to start with for a conceptual overview: https://imaging.mrc-cbu.cam.ac.uk/imaging/DesignEfficiency, then if you want to simulate efficiencies of different designs, try this function: https://www.mrc-cbu.cam.ac.uk/wp-content/uploads/www/sites/3/2013/09/fMRI_GLM_efficiency.m, which accompanies this more technical paper: https://www.mrc-cbu.cam.ac.uk/wp-content/uploads/www/sites/3/2015/03/Henson_EN_15_Efficiency.pdf.<br><br>Hope this helps<br>Rik<br><br>* If you have a long TR>2s, and your SOAs are multiples of your TR, you could jitter by 0-TR/2 secs to ensure a higher effective sampling rate, but that is a different reason – see above webpage.Negin Javaheri2024-03-28T10:18:23+01:002024-03-28T10:18:23+01:00Help Needed: fMRI Analysis on Event Design and Scanner Drift Effectshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;81187c23.2403Dear SPM Experts,<br><br>I am reaching out to seek your guidance on a matter pertaining to my<br>fMRI experiment.<br><br>I want to analyse data with the following structure:<br><br>* Event related design, within subject, 40 subjects<br>* two blocks with 15 min. duration, no break in between<br>* 21 events A in block 1 (pseudrandomly distributed within the block)<br>* 21 events B in block 2 (pseudrandomly distributed within the block)<br>* duration of all events between 0.2-4 seconds (dependent on<br>participant response time)<br><br>I want to contrast condition A and B in the context of a general linear<br>model.<br><br>My questions are the following:<br><br>1. I am aware that scanner drift rate can make problems in this<br>setup. What effect would I observe in my results? Can I estimate how<br>strong the effects of scanner drift rate on my results are?<br>2. Due to the somewhat block-like nature: How problematic would a<br>high-pass filter as part of my analysis be? I currently have a<br>filter with a cutoff of 4320 seconds.<br>3. How much trust would you put in any results of this analysis?<br>4. Is there a different analysis approach that would lead to more<br>trustworthy results.<br><br>Thank you in advance and warm regards,<br><br>Negin Javaheri<br><br>--<br>Negin Javaheri<br>PhD Candidate<br>Department of Neuropsychology and Behavioral Neurobiology<br>University of Bremen | Cognium Building<br>Hochschulring 18 | D-28359 Bremen<br>Phone: +49-421-218-68750<br>E-Mail:javaheri@uni-bremen.de<br>Web:www.neuropsychologie.uni-bremen.deFrank VAN OVERWALLE2024-03-28T08:58:25+00:002024-03-28T08:58:25+00:00Re: 2nd level covariate t contrast vs. biased ROI t testhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;370cac2c.2403Hi Hannah,<br><br>It seems that the SPM contrasts of the clusters were not very strong, significant only at p<.05. Hence, selecting specific data/voxel from each cluster might have weakened that contrast, e.g., by a larger variation around the mean, an effect that might be smoothed by using a larger amount of voxels (Note, I am not a statistical or SPM expert, so take this interpretation with some salt).<br><br>In our lab, we never single out a single voxel from the clusters, but usually take a larger ROI using Marsbar (e.g., a sphere with radius 5, 10, 15 depending on the volume of the clusters), and that typically confirms the SPM contrast.<br><br>Hope this helps, Frank<br><br>Van: SPM (Statistical Parametric Mapping) <SPM@JISCMAIL.AC.UK> Namens Doyle, Hannah<br>Verzonden: dinsdag 26 maart 2024 21:39<br>Aan: SPM@JISCMAIL.AC.UK<br>Onderwerp: [SPM] 2nd level covariate t contrast vs. biased ROI t test<br><br>Hi all,<br><br>I am running into some confusion in my analyses that I'm hoping someone can shed light on. I have two groups of participants. I specify the groups as a 2nd level covariate, and then specify t contrasts as follows: group1>group2 = [1 -1] and group2>group1 = [-1 1]. I see FWE cluster corrected activity at p<0.05 in both contrasts.<br><br>However, when I take a biased ROI from one of those contrasts, pull beta values from that ROI separately for each group, and then run a t test on those values, it does not come out significant. To verify I was pulling values correctly, I took a peak voxel of activity from that ROI and recorded the exact beta value for each participant in each group, and the t test still was not significant. I expected based on contrast maps that one group would be significantly greater than the other in a biased ROI analysis. I tried this for different peak voxels and the closest result was marginally greater for one group over the other. This also occurs if I pull T values rather than beta values.<br><br>I was under the impression when those t contrasts are calculated by SPM that a t test is performed, and therefore I should have similar results from both types of analyses, but maybe I am misunderstanding how t contrasts for 2nd level covariates are handled by SPM? Does anyone have insight into this?<br><br>Thank you in advance!<br><br>Best,<br>HannahFrank VAN OVERWALLE2024-03-28T08:49:03+00:002024-03-28T08:49:03+00:00Re: How to determine the optimal stimulus sequence for a slow-event fMRI experiment?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;594f1ea9.2403Luna,<br><br>I do not understand entirely your question.<br><br>In principle, you build your experiment as you normally would do for a behavioral psychological experiment, with the addition of jittering (a random addition between say 1-2 seconds) in the inter trial/stimulus intervals for better estimation of the BOLD function.<br><br>In SPM, you determine for each condition a regressor with as onset times of the stimulus onsets in each condition, and the duration of the stimulus (which you can interpret as a single peak, hence duration = 0; or for the whole length of the stimulus, hence duration = 2 s)<br><br>I hope this answers your question,<br><br>Frank<br><br>Van: SPM (Statistical Parametric Mapping) <SPM@JISCMAIL.AC.UK> Namens Luna Sato<br>Verzonden: woensdag 27 maart 2024 20:38<br>Aan: SPM@JISCMAIL.AC.UK<br>Onderwerp: [SPM] How to determine the optimal stimulus sequence for a slow-event fMRI experiment?<br><br>Hi all,<br><br>In my experiment, there are six conditions belong to two independent variables (2x3).<br><br>We are interested into the interaction effect of 2x3 ANOVA on activation (beta values) of six conditions.<br><br>A slow event design will be used: each event lasting two seconds, with intervals between events ranging from 8 to 14 seconds. Each session comprises 32 events.<br><br>Are there any tools or MATLAB codes available to help determine the best stimulus sequence for this slow event design?<br><br>Thank you!<br>Best,<br>LunaDavid Kennedy2024-03-27T17:44:12-04:002024-03-27T17:44:12-04:00SOBP Hackathon, May 7-8, 2024https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;4f87b6bd.2403Hi all, sorry for the multiple postings, but...<br>Going to Society of Biological Psychiatry (SOBP) annual meeting this May?<br>Want more time to think? Want more time to code? Want more time to<br>collaborate? The SOBP BrainHack is the place to be! May 7th&8th, 2024.<br><br>Details at https://www.repronim.org/SOBPHack-052024...<br><https://www.repronim.org/SOBPHack-052024/;><br>Registration at https://www.eventbrite.com/e/sobp-brain-...<br><https://www.eventbrite.com/e/sobp-brain-hack-tickets-857174581247?aff=oddtdtcreator.><br><br>Space is limited, register soon!Luna Sato2024-03-27T20:38:27+01:002024-03-27T20:38:27+01:00How to determine the optimal stimulus sequence for a slow-event fMRI experiment?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;eccb4543.2403Hi all,<br><br>In my experiment, there are six conditions belong to two independent variables (2x3).<br><br>We are interested into the interaction effect of 2x3 ANOVA on activation (beta values) of six conditions.<br><br>A slow event design will be used: each event lasting two seconds, with intervals between events ranging from 8 to 14 seconds. Each session comprises 32 events.<br><br>Are there any tools or MATLAB codes available to help determine the best stimulus sequence for this slow event design?<br><br>Thank you!<br>Best,<br>LunaZeidman, Peter2024-03-27T15:26:49+00:002024-03-27T15:26:49+00:00Re: Subjects with less than 10% explained variance in DCM_PEBhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;fb5b9ea7.2403Hi Debora<br>It depends on the models in your comparison. In general, Bayesian model comparison will favour the simplest model, if there’s no useful information in the data. So including subjects without experimental effects could increase the evidence for models with few experimental effects (e.g. in the limit, if you include a null model that has no modulations).<br><br>Best<br>Peter<br><br>From: Veronica Debora TORO <veronicadebora.toro@studenti.unipr.it><br>Sent: 22 March 2024 10:48<br>To: Zeidman, Peter <peter.zeidman@ucl.ac.uk>; SPM@JISCMAIL.AC.UK<br>Subject: Re: [SPM] Subjects with less than 10% explained variance in DCM_PEB<br><br>Dear Peter,<br>Thank you for your quick reply!<br>And in your experience, does the exclusion of these subjects impact the model comparison result? For example: I assumed two alternative models to the full One, can the posterior probability of compared models increase and/or decrease as a result of excluding these subjects?<br><br>Best<br>Debora<br><br>Inviato da Outlook per Android<https://aka.ms/AAb9ysg>Elaine Bearer2024-03-26T16:56:32-07:002024-03-26T16:56:32-07:00Neurological to radiological after SPM8 normalize functionhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;932af6e5.2403All,<br>We are using mouse brain MR images.<br>The normalize function in SPM8 converts our neurological nifti images into<br>radiological orientation.<br><br>Can we remove that from the processing algorithm? Or change the headers so<br>that the images are consistently in the neurological format?<br><br>We cannot find the relevant command that is doing this conversion during<br>nonlinear alignment.<br><br>Thank you<br><br>Elaine L. Bearer, MD-PhD, FAAAS, FCAP<br>The Harvey Family Professor, Dept of Pathology<br>Professor, Department of Music (secondary)<br>University of New Mexico Health Sciences Center<br>Albuquerque, NM 87131<br>https://hsc.unm.edu/medicine/departments/pathology/research/labs/bearer.html<br><http://pathology.unm.edu/faculty/faculty/ebearer.html><br>https://en.wikipedia.org/wiki/Elaine_Bearer<br><br>Also, Visitor<br>California Institute of TechnologyDoyle, Hannah2024-03-26T16:38:37-04:002024-03-26T16:38:37-04:002nd level covariate t contrast vs. biased ROI t testhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;57fc53ac.2403Hi all,<br><br>I am running into some confusion in my analyses that I'm hoping someone can<br>shed light on. I have two groups of participants. I specify the groups as a<br>2nd level covariate, and then specify t contrasts as follows: group1>group2<br>= [1 -1] and group2>group1 = [-1 1]. I see FWE cluster corrected activity<br>at p<0.05 in both contrasts.<br><br>However, when I take a biased ROI from one of those contrasts, pull beta<br>values from that ROI separately for each group, and then run a t test on<br>those values, it does not come out significant. To verify I was pulling<br>values correctly, I took a peak voxel of activity from that ROI and<br>recorded the exact beta value for each participant in each group, and the t<br>test still was not significant. I expected based on contrast maps that one<br>group would be significantly greater than the other in a biased ROI<br>analysis. I tried this for different peak voxels and the closest result was<br>marginally greater for one group over the other. This also occurs if I pull<br>T values rather than beta values.<br><br>I was under the impression when those t contrasts are calculated by SPM<br>that a t test is performed, and therefore I should have similar results<br>from both types of analyses, but maybe I am misunderstanding how t<br>contrasts for 2nd level covariates are handled by SPM? Does anyone have<br>insight into this?<br><br>Thank you in advance!<br><br>Best,<br>HannahLevi2024-03-26T15:32:54+00:002024-03-26T15:32:54+00:00Group level analysis and contrast weightshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;8d584f5f.2403Dear SPM experts,<br><br>I am encountering several difficulties in understanding how to set up my<br>analyses with SPM, as I am unable to find appropriate information for what<br>I'd like to do in the manual or other documents. I am working with scalp x<br>time EEG data.<br><br>I have four conditions (A, B, C, D) and four regressors, one for each<br>condition (A1, B1, C1, D1), which are trial-by-trial learning estimates.<br><br>WHAT I WANT TO DO: I want to investigate if the EEG signal of a given<br>condition covaries with its own regressor (e.g., if A covaries with A1, B<br>with B1, and so on). It would also be useful to then make some contrasts<br>(e.g., A+A1 vs. B+B1).<br><br>WHAT I HAVE DONE:<br>So far, I have performed single subject one-way ANOVA, populating the<br>"Cells" field with four cells, one per condition that I've populated with<br>my scalp x time maps, and loading the file with the vector of learning<br>estimates (following the order of the single trial maps loaded in Cells)<br>under "Multiple covariates", selecting "Interaction with Factor 1". This<br>gives me 8 Beta maps per subject. Have I proceeded correctly so far?<br><br>GROUP ANALYSIS: I am not clear on how I should proceed, meaning that I<br>don't know what data should I use (all 8 beta maps? And how should I<br>populate the fields then?) nor how to specify any contrasts, in the sense<br>that I am not sure how to choose the weights when setting a contrast<br>(likewise for visualizing the results on single subjects).<br><br>SUMMARY OF REQUESTS:<br><br>Is the single subject analysis correct?<br>How do I set up the group analysis?<br>How do I assign weights to contrasts (e.g., if I wanted to visualize,<br>precisely, if A covaries with A1)?<br><br>I realize I am asking a lot, but your help would be truly valuable to me.<br>With very best wishes,<br><br>LeviOostenveld, R. (Robert)2024-03-26T12:32:05+00:002024-03-26T12:32:05+00:00FieldTrip MEG/EEG toolkit at Donders: registration openhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;4f9a1244.2403Dear all,<br><br>From May 27-31 2024 we will again host the FieldTrip MEG/EEG toolkit course at the Donders in Nijmegen. The toolkit course is aimed at PhD and postdoctoral researchers that already have some experience with EEG and/or MEG data acquisition (eithert a pilot or a full study) and that have a good understanding of their own experimental design. Furthermore, for successful participation in the advanced topics you need to know the basics of MATLAB and already have some experience with MEG/EEG preprocessing and analysis.<br><br>We will demonstrate EEG, MEG and OPM data acquisition and teach you advanced data analysis methods and cover preprocessing, frequency analysis, source reconstruction, analysis methods such as RSA and TRFs, and various statistical methods. Furthermore, we will give attention to good practices for reproducible research and open science. The toolkit will consist of a number of interactive online lectures, followed by Q&A sessions. Besides lectures we will have interactive hands-on sessions in which you will be tutored through the complete analysis of a MEG/EEG data set using the FieldTrip toolbox. There will be plenty of opportunity to interact and also ask questions about your research and data. On the final day you will have the opportunity to work on your own dataset under supervision of skilled tutors.<br><br>There are 38 places available for this toolkit. From past experience we expect the course to be oversubscribed, hence we will start with pre-registration. The selection of participants will be based on the background experience, the research interest and the expectations on what you will learn. We prefer the group to be reasonably homogeneous in level and expertise of the participants, as this improves the overall interaction. The deadline for pre-registration is April 15 2024.<br><br>More information, including registration details and the tentative program can be found at http://www.fieldtriptoolbox.org/workshop/toolkit2024/.<br><br>best regards<br>Robert Oostenveld and Jan-Mathijs Schoffelen Zach H2024-03-26T05:00:16+01:002024-03-26T05:00:16+01:00Code for gPPI analysishttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;49393eda.2403Hi experts,<br><br>gPPI is a commonly used analysis method. However, the SPM manual does not provide the code for gPPI. Therefore, I'm wondering if we can utilize existing batch modules related to PPI within SPM to conduct gPPI analysis.<br><br>Suppose an experiment involves four conditions ('cond1', 'cond2', 'cond3', 'cond4'), and the psychological effect of interest is: cond1 + cond2 - cond3 + cond4. In this case, the gPPI analysis code for this experiment would be as follows:<br><br>conditionstr={'cond1','cond2','cond3','cond4'};<br>% generate PPI regressers<br>for con=1:length(conditionstr)<br> weight_matrix=[];<br> weight_matrix(:,1)=1:length(conditionstr);<br> weight_matrix(:,2)=1;<br> weight_matrix(con,3)=1;<br> matlabbatch{1}.spm.stats.ppi.spmmat = {[ GLMpath '\SPM.mat']};<br> matlabbatch{1}.spm.stats.ppi.type.ppi.voi = {[ GLMpath '\VOI_seed_1.mat']};%voi of seed region<br> matlabbatch{1}.spm.stats.ppi.type.ppi.u = weight_matrix;<br> matlabbatch{1}.spm.stats.ppi.name =[ 'seed_con' num2str(con)];<br> matlabbatch{1}.spm.stats.ppi.disp = 0; <br> spm_jobman('run', matlabbatch);clear matlabbatch<br>end<br><br>% GLM spec<br>matlabbatch{1}.spm.stats.fmri_spec.dir = dirpath;<br>matlabbatch{1}.spm.stats.fmri_spec.timing.units = 'secs';<br>matlabbatch{1}.spm.stats.fmri_spec.timing.RT = TR;<br>matlabbatch{1}.spm.stats.fmri_spec.timing.fmri_t = nslice;<br>matlabbatch{1}.spm.stats.fmri_spec.timing.fmri_t0 =referenceslice;<br>matlabbatch{1}.spm.stats.fmri_spec.sess.scans = filenames;<br>matlabbatch{1}.spm.stats.fmri_spec.sess.cond = struct('name', {}, 'onset', {}, 'duration', {}, 'tmod', {}, 'pmod', {}, 'orth', {});<br>matlabbatch{1}.spm.stats.fmri_spec.sess.multi = {''};<br><br>% interaction terms<br>for con=1:length(conditionstr)<br>load([ GLMpath '\PPI_seed_con' num2str(con) '.mat']);<br>matlabbatch{1}.spm.stats.fmri_spec.sess.regress(con).name = ['PPI_Interaction' conditionstr{con}];<br>matlabbatch{1}.spm.stats.fmri_spec.sess.regress(con).val = PPI.ppi;<br>end<br><br>% psychological terms<br> for con=1:length(conditionstr)<br>load([ GLMpath '\PPI_seed_con' num2str(con) '.mat']);<br>matlabbatch{1}.spm.stats.fmri_spec.sess.regress(end+1).name = ['Psy' conditionstr{con}];<br>matlabbatch{1}.spm.stats.fmri_spec.sess.regress(end).val = PPI.P;<br> end<br><br> % physiological<br>matlabbatch{1}.spm.stats.fmri_spec.sess.regress(end+1).name =['Phys'];<br>matlabbatch{1}.spm.stats.fmri_spec.sess.regress(end).val = PPI.Y;<br>matlabbatch{1}.spm.stats.fmri_spec.sess.multi_reg = headmotionfile;<br>matlabbatch{1}.spm.stats.fmri_spec.sess.hpf = 128;<br>matlabbatch{1}.spm.stats.fmri_spec.fact = struct('name', {}, 'levels', {});<br>matlabbatch{1}.spm.stats.fmri_spec.bases.hrf.derivs = [0 0];<br>matlabbatch{1}.spm.stats.fmri_spec.volt = 1;<br>matlabbatch{1}.spm.stats.fmri_spec.global = 'None';<br>matlabbatch{1}.spm.stats.fmri_spec.mthresh = 0.8;<br>matlabbatch{1}.spm.stats.fmri_spec.mask = {''};<br>matlabbatch{1}.spm.stats.fmri_spec.cvi = 'AR(1)';<br>spm('defaults', 'FMRI');<br>spm_jobman('run', matlabbatch);clear matlabbatch<br><br>% estimation<br>matlabbatch{1}.spm.stats.fmri_est.spmmat = {[dirpath '\SPM.mat']};<br>matlabbatch{1}.spm.stats.fmri_est.method.Classical = 1;<br>spm('defaults', 'FMRI');<br>spm_jobman('run', matlabbatch);clear matlabbatch<br><br>%contrast<br>matlabbatch{1}.spm.stats.con.spmmat = {[dirpath '\SPM.mat']};<br>matlabbatch{1}.spm.stats.con.consess{1}.tcon.name = 'ppi_interaction_difference';<br>matlabbatch{1}.spm.stats.con.consess{1}.tcon.weights = [1 1 -1 -1];<br>matlabbatch{1}.spm.stats.con.consess{1}.tcon.sessrep = 'none';<br>matlabbatch{1}.spm.stats.con.delete = 1;<br>spm('defaults', 'FMRI');<br>spm_jobman('run', matlabbatch);clear matlabbatch<br><br>The spmT_0001.nii file generated by this contrast represents the gPPI result for the participant, while the con_0001.nii file can be used for group analysis in GLM.<br><br>Please correct my mistakes.<br>Best,<br><br>ZachJames Lee2024-03-25T12:20:45-06:002024-03-25T12:20:45-06:00Re: Extracting beta values of sig clusters/regionshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;c9b4cae2.2403Lena,<br><br>The field of fmri statistical analysis is a bit controversial, so you may<br>get replies that disagree with mine, but here goes!<br><br>I think the reviewer is concerned that you are selectively reporting the<br>most active part of the cluster by limiting the analysis to the data<br>surrounding the peak voxel. That introduces bias into the results. Lots of<br>early fmri papers did this, and people are more careful about it now.<br><br>Strictly speaking, decisions about analysis should be made before the data<br>is acquired, i.e. the statistical threshold that will be used, and the<br>regions of interest that will be analyzed.<br><br>Two methods that avoid bias are:<br><br>1) choosing the spherical region (or regions) for analysis before you look<br>at the data, based on previous publications.<br><br>2) run two identical tasks on each patient, and use one of the datasets, at<br>a chosen statistical threshold, to generate masks that are then used to<br>limit the analysis of the second set of data.<br><br>The data-generated mask can also be combined with a previously-chosen<br>anatomic mask so that you analyze the most active part of the chosen<br>anatomic mask. This way the experiment itself limits the region of<br>interest, so there is no operator bias.<br><br>Hope this helps!<br><br>Jim<br><br>On Sat, Mar 23, 2024 at 9:21 AM Lena Lim <<br>000063c14c8ef627-dmarc-request@jiscmail.ac.uk> wrote:<br><br>> Dear SPM Experts,<br>><br>> Is it okay to extract beta values of sig clusters/regions using MarsBaR,<br>> defined using spherical masks with a radius of 8mm around the peak<br>> coordinates? I have seen many earlier papers extracting beta values using<br>> spherical masks around peak coordinates too, but a reviewer commented<br>> that a sphere centered at the peak coordinate is not representative of the<br>> whole cluster, and that I should use the whole SPM clusters in their<br>> original form instead... could you please kindly advise if the latter<br>> approach is the better one please?<br>><br>><br>><br>> Many thanks,<br>><br>><br>> Lena<br>><br>><br>>ELENA GROSSO2024-03-25T12:06:43+01:002024-03-25T12:06:43+01:00Re: How to invert a Normalizationhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;c920ec6e.2403Dear Carles,<br><br>Thanks for your quick answer!<br>We tried with the code below but in this way we were not able to save<br>anything, there was no output. We tried also with normalise.write but in<br>this way the output atlas wasn't in the right space, do you have any<br>suggestions?<br><br>Thank you very much,<br>Best regards<br>Elena<br><br>*Segmentation: *<br>matlabbatch{1}.spm.spatial.preproc.channel.vols = {'T13D_nu2nodif.nii,1'};<br>matlabbatch{1}.spm.spatial.preproc.channel.biasreg = 0.001;<br>matlabbatch{1}.spm.spatial.preproc.channel.biasfwhm = 60;<br>matlabbatch{1}.spm.spatial.preproc.channel.write = [1 1];<br>matlabbatch{1}.spm.spatial.preproc.tissue(1).tpm = {<br>'/home/bcc/matlab/spm12/tpm/TPM.nii,1'};<br>matlabbatch{1}.spm.spatial.preproc.tissue(1).ngaus = 1;<br>matlabbatch{1}.spm.spatial.preproc.tissue(1).native = [1 0];<br>matlabbatch{1}.spm.spatial.preproc.tissue(1).warped = [0 0];<br>matlabbatch{1}.spm.spatial.preproc.tissue(2).tpm = {<br>'/home/bcc/matlab/spm12/tpm/TPM.nii,2'};<br>matlabbatch{1}.spm.spatial.preproc.tissue(2).ngaus = 1;<br>matlabbatch{1}.spm.spatial.preproc.tissue(2).native = [1 0];<br>matlabbatch{1}.spm.spatial.preproc.tissue(2).warped = [0 0];<br>matlabbatch{1}.spm.spatial.preproc.tissue(3).tpm = {<br>'/home/bcc/matlab/spm12/tpm/TPM.nii,3'};<br>matlabbatch{1}.spm.spatial.preproc.tissue(3).ngaus = 2;<br>matlabbatch{1}.spm.spatial.preproc.tissue(3).native = [1 0];<br>matlabbatch{1}.spm.spatial.preproc.tissue(3).warped = [0 0];<br>matlabbatch{1}.spm.spatial.preproc.tissue(4).tpm = {<br>'/home/bcc/matlab/spm12/tpm/TPM.nii,4'};<br>matlabbatch{1}.spm.spatial.preproc.tissue(4).ngaus = 3;<br>matlabbatch{1}.spm.spatial.preproc.tissue(4).native = [1 0];<br>matlabbatch{1}.spm.spatial.preproc.tissue(4).warped = [0 0];<br>matlabbatch{1}.spm.spatial.preproc.tissue(5).tpm = {<br>'/home/bcc/matlab/spm12/tpm/TPM.nii,5'};<br>matlabbatch{1}.spm.spatial.preproc.tissue(5).ngaus = 4;<br>matlabbatch{1}.spm.spatial.preproc.tissue(5).native = [1 0];<br>matlabbatch{1}.spm.spatial.preproc.tissue(5).warped = [0 0];<br>matlabbatch{1}.spm.spatial.preproc.tissue(6).tpm = {<br>'/home/bcc/matlab/spm12/tpm/TPM.nii,6'};<br>matlabbatch{1}.spm.spatial.preproc.tissue(6).ngaus = 2;<br>matlabbatch{1}.spm.spatial.preproc.tissue(6).native = [0 0];<br>matlabbatch{1}.spm.spatial.preproc.tissue(6).warped = [0 0];<br>matlabbatch{1}.spm.spatial.preproc.warp.mrf = 1;<br>matlabbatch{1}.spm.spatial.preproc.warp.cleanup = 1;<br>matlabbatch{1}.spm.spatial.preproc.warp.reg = [0 0.001 0.5 0.05 0.2];<br>matlabbatch{1}.spm.spatial.preproc.warp.affreg = 'mni';<br>matlabbatch{1}.spm.spatial.preproc.warp.fwhm = 0;<br>matlabbatch{1}.spm.spatial.preproc.warp.samp = 3;<br>matlabbatch{1}.spm.spatial.preproc.warp.write = [1 1];<br>matlabbatch{1}.spm.spatial.preproc.warp.vox = NaN;<br>matlabbatch{1}.spm.spatial.preproc.warp.bb = [NaN NaN NaN<br>NaN NaN NaN];<br>spm('defaults', 'FMRI');<br>spm_jobman('run',matlabbatch);<br><br>*Deformation tool: *<br><br>matlabbatch{1}.spm.util.defs.comp{1}.def = {'iy_T13D_nu2nodif.nii,1'};<br>matlabbatch{1}.spm.util.defs.out{1}.pull.fnames = {'Atlas.nii,1'};<br>matlabbatch{1}.spm.util.defs.out{1}.pull.savedir.saveusr = pwd;<br>matlabbatch{1}.spm.util.defs.out{1}.pull.interp = 0;<br>matlabbatch{1}.spm.util.defs.out{1}.pull.mask = 1;<br>matlabbatch{1}.spm.util.defs.out{1}.pull.fwhm = [0 0 0];<br>spm('defaults', 'FMRI');<br>spm_jobman('run',matlabbatch);<br><br>Il giorno ven 22 mar 2024 alle ore 17:25 carles falcon <<br>carlesfalcon@gmail.com> ha scritto:<br><br>> Hi Elena<br>> you can use deformation tools:<br>> matlabbatch{1}.spm.util.defs.comp{1}.def = {LDEFORM_FILE};<br>> matlabbatch{1}.spm.util.defs.out{1}.pull.fnames = {FULLPATH_ATLASFILENAME};<br>> matlabbatch{1}.spm.util.defs.out{1}.pull.savedir.saveusr = {OUTFOLDER};<br>> matlabbatch{1}.spm.util.defs.out{1}.pull.interp = 0;<br>> matlabbatch{1}.spm.util.defs.out{1}.pull.mask = 1;<br>> matlabbatch{1}.spm.util.defs.out{1}.pull.fwhm = [0 0 0];<br>><br>> defining LDEFORM_FILE as the file containing subject's deformation from<br>> MNI to native. You can get this transformation by several ways, the most<br>> usual is to ask to store iy file in segmentation or using deformation tools<br>> to invert dartel's flow fields<br>><br>> Hope this helps<br>> Carles<br>><br>> Missatge de ELENA GROSSO <elena.grosso01@universitadipavia.it> del dia<br>> dv., 22 de març 2024 a les 17:09:<br>><br>>> Dear all,<br>>><br>>> I'm experiencing trouble inverting the normalization to MNI space I've<br>>> computed with the code above. I need it to register an atlas into the<br>>> subject space.<br>>> Thanks to anybody who can help!<br>>><br>>> Elena<br>>> matlabbatch{1}.spm.spatial.normalise.estwrite.subj.vol = {T1dir};<br>>> matlabbatch{1}.spm.spatial.normalise.estwrite.subj.resample = {Gdir<br>>> Gdir1<br>>> Gdir2<br>>> Gdir3<br>>> Gdir4<br>>> Gdir5};<br>>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.biasreg = 0.0001;<br>>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.biasfwhm = 60;<br>>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.tpm = {<br>>> '/home/bcc/matlab/spm12/tpm/TPM.nii'};<br>>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.affreg = 'mni';<br>>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.reg = [0 0.001 0.5<br>>> 0.05 0.2];<br>>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.fwhm = 0;<br>>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.samp = 3;<br>>> matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.bb = [-91 -126 -72<br>>> 90 91 109];<br>>> matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.vox = [1 1 1];<br>>> matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.interp = 4;<br>>> matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.prefix = 'MNI_1mm'<br>>> ;<br>>> spm('defaults', 'FMRI');<br>>> spm_jobman('run',matlabbatch);<br>>><br>>><br>>>Michael Zyphur2024-03-25T21:00:00+11:002024-03-25T21:00:00+11:00Run your Seminars, Workshops, and Conferences with Instats - Live Q&Ahttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;fefc3717.2403Hi everyone<br><br>Instats is offering a free introduction to the Instats platform<br><https://instats.org/seminar/partner-with-instats-to-freely-create-yo2> for<br>potential partners, running March 28 / 29. Our platform is designed to<br>support research communities and promote academic freedom by providing a<br>free platform for research training and community building with our global<br>network of partners. Instats is methodologically pluralistic and open to<br>partnerships across disciplines, and our partners include individual<br>researchers with a PhD or equivalent, as well as professional<br>associations/societies and universities.<br><br>It's always free to become an Instats partner, which allows creating and<br>offering free or fee-based livestreaming seminars, workshops, or<br>conferences through instats.org. Our partners have free access to our<br>platform's many back-end features, including AI-enabled content creation<br>and automated video/audio editing that allows turning livestreamed<br>recordings into on-demand streaming content. By building a library of<br>on-demand streaming content, you can support your knowledge community and<br>create a lasting body of research training content that can be useful for<br>years to come (while retaining ownership of your intellectual property).<br><br>Register today<br><https://instats.org/seminar/partner-with-instats-to-freely-create-yo2> for<br>this free introduction to the Instats platform, and for more information<br>about our partnerships please feel free to get in touch with me directly.<br><br>Best wishes<br><br>Michael Zyphur<br>Director<br>Institute for Statistical and Data Science<br>*instats.org* <http://instats.org><br><http://instats.org>BENEDETTA CECCONI2024-03-25T09:56:33+00:002024-03-25T09:56:33+00:00Re: SPM EEG - Displaying contrast and T maps in SPM ?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;17061c1d.2403Dear Vladimir,<br><br>right, sorry I meant thresholded t-maps.. That toolbox I shared last time, unfortunately, plots only peaks but not all the thresholded significant t-values.. Any chance you could share the codes you used in your Science 2011 paper (https://pubmed.ncbi.nlm.nih.gov/21566197/) to plot thresholded t-maps at selected time-points?<br><br>Thanks again for your help and time,<br><br>BenedettaDavid Pascucci2024-03-25T09:27:31+01:002024-03-25T09:27:31+01:00PhD and Postdoc Positions in Psychology/Neuroscience/Electroencephalographyhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;ea7dd149.2403A project, funded by the Swiss National Science Foundation as part of a<br>Starting Grant (Principal Investigator: David Pascucci<br><https://scholar.google.com/citations?user=kvHIlLwAAAAJ&hl=en>, *CHUV/Unil/The<br>Sense Innovation Center*), is currently seeking:<br><br>- *1 PhD candidate* (4 years, *starting in 2024*)<br><br>- *2 Postdoctoral* researchers (2 years each, *starting late 2024<br>– early-mid 2025 preferred*)<br><br>*Project aim*: To understand the relationship between individual-specific<br>temporal structures of neural activity—particularly focusing on *alpha-band<br>neuronal oscillations*—and variability in *perceptual and cognitive<br>functions*.<br>Specific segments of the project will explore temporal structures of neural<br>activity linked to inter-individual variability in *contextual effects in<br>perception*.<br><br>*Project methods*: The project involves the *collection of large-scale<br>datasets* comprising resting-state and event-related electroencephalography<br>(*EEG*) data, recorded alongside behavioral performance in a wide<br>range of *visual<br>tasks*. The analysis will combine human psychophysics, EEG and EEG inverse<br>modeling with advanced multivariate statistical and predictive modeling<br>techniques.<br><br>*PhD position<br><https://recrutement.chuv.ch/vacancy/phd-position-in-psychology-neuroscience-eeg-radiology-department-297545.html><br>(starting as of 2024):*<br><br>*Main tasks*:<br><br>1. Recruitment of study participants, collection, and analysis of<br>EEG+behavioral data.<br><br>2. Dissemination of research through conference presentations and<br>peer-reviewed publications.<br><br>*Requirements and ideal skills*:<br><br>- Master’s degree in Cognitive Psychology / Neuroscience /<br>Bioengineering / Physiology.<br><br>- Strong motivation and commitment to data collection and analysis.<br><br>- Previous experience with psychophysics and EEG/signal processing<br>(a plus, not mandatory).<br><br>*Postdoc positions<br><https://recrutement.chuv.ch/vacancy/postdoctoral-positions-in-psychology-neuroscience-eeg-radiology-department-297525.html><br>(starting late 2024 – early-mid 2025 preferred):*<br><br>*Main tasks*:<br><br>1. Analysis and modeling of EEG+behavioral data.<br><br>2. Leading sub-projects (from experimental design to data analysis).<br><br>3. Development and validation of codes and toolboxes for public release<br>(Matlab, Python, R).<br><br>4. Dissemination of research through conference presentations and<br>peer-reviewed publications.<br><br>*Requirements and ideal skills*:<br><br>- Proficiency in neuroimaging techniques (EEG, inverse solution)<br>and signal processing (ERP, spectral analysis), as well as psychophysics.<br><br>- Good/strong statistical, computational, and programming skills.<br><br>The project involves collaboration with a broad network of partners both<br>within and outside Switzerland.<br><br>*Positions will remain open until filled, more information and the<br>application form can be found here (click on the links): PhD<br><https://recrutement.chuv.ch/vacancy/phd-position-in-psychology-neuroscience-eeg-radiology-department-297545.html><br>– PostDoc<br><https://recrutement.chuv.ch/vacancy/postdoctoral-positions-in-psychology-neuroscience-eeg-radiology-department-297525.html>.*<br><br>For further details and formal/informal enquiries, please contact:<br><br>psc.dav@gmail.com<br><br>david.pascucci@chuv.chJorge Almeida2024-03-24T11:10:32+00:002024-03-24T11:10:32+00:00Deadline approaching: 3 Post-Doctoral positions at the Faculty of Psychology, University of Coimbra, Portugal, to work with Alfonso Caramazza and Jorge Almeidahttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;f5e7e743.2403DEADLINE APPROACHING<br><br>The Faculty of Psychology and Educational Sciences of the University of<br>Coimbra Portugal (FPCE-UC) invites applications from rising and<br>enthusiastic researchers in Cognitive Science and Cognitive Neuroscience<br>for 3 Post-Doctoral positions to work with Alfonso Caramazza and Jorge<br>Almeida.<br><br>These positions are part of a transformative ERA Chair grant CogBooster<br>from the European Union to FPCE-UC led by Alfonso Caramazza. The goal of<br>CogBooster is to implement a strong and international line of research in<br>Basic Cognitive Science and Cognitive Neuroscience to contribute to the<br>ongoing renewal of the Psychological Sciences in Portugal over the next<br>decade.<br><br>Positions:<br><br>As part of this expansion and renewal, we are searching for Post-doctoral<br>applicants with expertise related to the following areas:<br><br>· 1 Post-doctoral position in any of the following areas: lexical<br>processing, visual object recognition, reading, or action recognition.<br><br>· 1 Post-doctoral position on visual object recognition, in particular on<br>how object knowledge is organized and represented neurally and cognitively,<br>using fMRI (potentially with ultra-high field MRI), and potentially<br>ECoG/sEEG.<br><br>· 1 Post-doctoral position on object dimensionality and dimensional<br>mapping, using population receptive field analysis/connective field<br>modeling to work on topics related to the following preprint:<br>https://www.biorxiv.org/content/10.1101/2023.11.29.568856v1<br><br>The selected applicants will work directly with Alfonso Caramazza and Jorge<br>Almeida. They will be based in Coimbra, but will have the opportunity to<br>spend some time at Harvard University, at Alfonso Caramazza’s laboratory.<br><br>Qualifications:<br><br>Applicants should be recent graduates – the date in their PhD diploma<br>should not be before October 2021 (i.e., no more than 3 years from the<br>signing of the contract). They should also have their diploma recognized in<br>Portugal at the time of the signing of the contract (diploma recognition<br>could be requested here). The start date is negotiable (but should be no<br>later than fall 2024).<br><br>The successful applicant should have a strong record of research for their<br>career level, and meet, or show promise of meeting, the following<br>qualifications:<br><br>· Research potential through publications with impact in the field;<br><br>· Proficiency in English, both written and spoken (speaking Portuguese is<br>not necessary);<br><br>· Good written and spoken communication skills;<br><br>Offer:<br><br>The positions do not involve formal teaching. They do involve, however, lab<br>mentoring.<br><br>The salary is competitive for Portuguese standards – 1800 euro per month<br>net value. According to Numbeo, 1800 euro in Coimbra correspond to a local<br>purchasing power of about 3230 euro in Paris, 3520 pounds in London (UK),<br>2685 euro in Brussels, 2385 euro in Rome, 2890 euro in Munich, about 4750<br>USD in Los Angeles, Boston or Washington, 4830 CAD in Toronto, or 5940 AUD<br>euro in Sydney.<br><br>The position will be for a maximum of 3 years, renewable every year.<br><br>About FPCE-UC and Coimbra:<br><br>The University of Coimbra is a 700-year-old University and is a UNESCO<br>world Heritage site. Coimbra is one of the liveliest university cities in<br>the world, and it is a beautiful city with easy access to beaches and<br>mountains. The Faculty of Psychology and Educational Sciences has been<br>consistently ranked as the best, or one of the best Psychology Departments<br>in Portugal. In the last decade it has become the leading department in<br>Portugal on Psychological research, holding, for instance the only ERC<br>grants in Psychology in Portugal. FPCE-UC has a major laboratory for<br>Cognitive Science and Cognitive Neuroscience research – the Proaction Lab.<br>We have access locally to two 3T MRI scanners, and access to one 7T MRI<br>remotely, to tDCS, to a 256 channel EEG, and to a fully set behavioral lab.<br><br>Application Instructions:<br><br>All candidates should submit the following documents in English: (1) a<br>curriculum vitae; (2) a motivation letter describing their interest in the<br>position, and their track record; (3) at least two letters of reference<br>submitted before the application deadline.<br><br>Full consideration will be given to applications received by March 31, 2023.<br><br>If you are interested in applying for one of the positions or know of any<br>suitable applicants in your network of colleagues and former students who<br>may be suitable for any of the positions, please direct them to the<br>advertisement.<br><br>Equal Employment Opportunity statement:<br><br>The University of Coimbra is an equal opportunity/affirmative action<br>employer and has a Gender Equality Plan in place. We are committed to<br>fostering a diverse and inclusive academic global community at the<br>University. We particularly encourage applications from women, and from<br>other under-represented groups in the University of Coimbra’s workforce and<br>in the cognitive and brain sciences.<br><br>Further information:<br><br>If you want to know more about the positions you can informally contact<br>Alfonso Caramazza (caram@wjh.harvard.edu) and Jorge Almeida (<br>jorgecbalmeida@gmail.com). Please mention the profile(s) you would fit,<br>from the three proposed.Lena Lim2024-03-23T15:10:50+00:002024-03-23T15:10:50+00:00Extracting beta values of sig clusters/regionshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;2e1c2ad6.2403Dear SPM Experts,<br><br>Is it okay to extract beta values of sig clusters/regions using MarsBaR, defined using spherical masks with a radius of 8mm around the peak coordinates? I have seen many earlier papers extracting beta values using spherical masks around peak coordinates too, but a reviewer commented that a sphere centered at the peak coordinate is not representative of the whole cluster, and that I should use the whole SPM clusters in their original form instead... could you please kindly advise if the latter approach is the better one please?<br><br>Many thanks,<br><br>Lena carles falcon2024-03-22T17:25:16+01:002024-03-22T17:25:16+01:00Re: How to invert a Normalizationhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;cd71fe89.2403Hi Elena<br>you can use deformation tools:<br>matlabbatch{1}.spm.util.defs.comp{1}.def = {LDEFORM_FILE};<br>matlabbatch{1}.spm.util.defs.out{1}.pull.fnames = {FULLPATH_ATLASFILENAME};<br>matlabbatch{1}.spm.util.defs.out{1}.pull.savedir.saveusr = {OUTFOLDER};<br>matlabbatch{1}.spm.util.defs.out{1}.pull.interp = 0;<br>matlabbatch{1}.spm.util.defs.out{1}.pull.mask = 1;<br>matlabbatch{1}.spm.util.defs.out{1}.pull.fwhm = [0 0 0];<br><br>defining LDEFORM_FILE as the file containing subject's deformation from MNI<br>to native. You can get this transformation by several ways, the most usual<br>is to ask to store iy file in segmentation or using deformation tools to<br>invert dartel's flow fields<br><br>Hope this helps<br>Carles<br><br>Missatge de ELENA GROSSO <elena.grosso01@universitadipavia.it> del dia dv.,<br>22 de març 2024 a les 17:09:<br><br>> Dear all,<br>><br>> I'm experiencing trouble inverting the normalization to MNI space I've<br>> computed with the code above. I need it to register an atlas into the<br>> subject space.<br>> Thanks to anybody who can help!<br>><br>> Elena<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.subj.vol = {T1dir};<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.subj.resample = {Gdir<br>> Gdir1<br>> Gdir2<br>> Gdir3<br>> Gdir4<br>> Gdir5};<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.biasreg = 0.0001;<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.biasfwhm = 60;<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.tpm = {<br>> '/home/bcc/matlab/spm12/tpm/TPM.nii'};<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.affreg = 'mni';<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.reg = [0 0.001 0.5<br>> 0.05 0.2];<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.fwhm = 0;<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.samp = 3;<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.bb = [-91 -126 -72<br>> 90 91 109];<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.vox = [1 1 1];<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.interp = 4;<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.prefix = 'MNI_1mm';<br>> spm('defaults', 'FMRI');<br>> spm_jobman('run',matlabbatch);<br>><br>><br>>ELENA GROSSO2024-03-22T17:08:59+01:002024-03-22T17:08:59+01:00How to invert a Normalizationhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;409c4b75.2403Dear all,<br><br>I'm experiencing trouble inverting the normalization to MNI space I've<br>computed with the code above. I need it to register an atlas into the<br>subject space.<br>Thanks to anybody who can help!<br><br>Elena<br>matlabbatch{1}.spm.spatial.normalise.estwrite.subj.vol = {T1dir};<br>matlabbatch{1}.spm.spatial.normalise.estwrite.subj.resample = {Gdir<br>Gdir1<br>Gdir2<br>Gdir3<br>Gdir4<br>Gdir5};<br>matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.biasreg = 0.0001;<br>matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.biasfwhm = 60;<br>matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.tpm = {<br>'/home/bcc/matlab/spm12/tpm/TPM.nii'};<br>matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.affreg = 'mni';<br>matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.reg = [0 0.001 0.5<br>0.05 0.2];<br>matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.fwhm = 0;<br>matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.samp = 3;<br>matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.bb = [-91 -126 -72<br>90 91 109];<br>matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.vox = [1 1 1];<br>matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.interp = 4;<br>matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.prefix = 'MNI_1mm';<br>spm('defaults', 'FMRI');<br>spm_jobman('run',matlabbatch);Tamrin Holloway2024-03-22T11:13:29+00:002024-03-22T11:13:29+00:00PhD Vacancies: Cusack Lab, Trinity College Dublinhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;1f7a6354.2403Two PhD Studentships in Deep Neural Networks for Medical Imaging<br><br>Brain development in infants and its disruption by preterm birth or perinatal injury, can be measured with functional MRI (fMRI). Unfortunately, infants move in the scanner and half the images are discarded, precluding clinical application. In recent years, deep neural networks (DNNs) have led to breakthroughs in artificial intelligence and are finding growing application in biomedical imaging. DNNs have considerable potential to correct head motion in fMRI, as they can learn complex mappings, and exploit knowledge of brain structure. The PhD Candidates will develop DNNs to motion correct fMRI data.<br><br>Candidates must have expertise in at least one of the two following areas and must be willing to develop skills in the other:<br>design and optimisation of deep neural networks<br>neuroimaging<br>Candidates must have expertise in programming in python or another language.<br><br>The studentships comprise EU fees, stipend and conference travel.<br><br>See further details: www.cusacklab.org/vacancies<br><br>Deadline 12 noon on April 22, 2024.Tamrin Holloway2024-03-22T11:12:16+00:002024-03-22T11:12:16+00:002 Postdoc Vacancies: Cusack Lab, Trinity College Dublinhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;c2bb4a5e.2403Two Senior Postdoctoral Researchers in Deep Neural Networks for Medical Imaging<br><br>Brain development in infants and its disruption by preterm birth or perinatal injury, can be measured with functional MRI (fMRI). Unfortunately, infants move in the scanner and half the images are discarded, precluding clinical application. In recent years, deep neural networks (DNNs) have led to breakthroughs in artificial intelligence and are finding growing application in biomedical imaging. DNNs have considerable potential to correct head motion in fMRI, as they can learn complex mappings, and exploit knowledge of brain structure. The postdoctoral researchers will develop DNNs to motion correct fMRI data.<br><br>Candidates must have expertise in at least one of the two following areas and must be willing to develop skills in the other:<br>design and optimisation of deep neural networks<br>neuroimaging with fMRI<br>Candidates must have a strong level of expertise in programming in python or another language.<br><br>Full time, €50,540-€54,965 per annum. Benefits include a pension contribution and PRSI social insurance.<br>Theses posts will be for two years but there may arise the opportunity to extend them.<br><br>See further details: www.cusacklab.org/vacancies<br><br>Deadline 12 noon on April 22, 2024.Tamrin Holloway2024-03-22T11:07:26+00:002024-03-22T11:07:26+00:00Postdoc Vacancy: Cusack Lab, Trinity College Dublinhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;f0e540ec.2403Experienced Postdoctoral Fellow in Infant Developmental Cognitive Neuroscience<br><br>The Foundations of Cognition project is studying the emergence of cognition in infants, and how this is disrupted by brain injury. It has acquired functional MRI in a substantial cohort of 2-and 9-month-old infants (N=134), awake watching pictures and movies. The postholder will work with Professor Cusack and the team to use this rich dataset to characterise infant’s emerging cognition: critically evaluating the literature, developing a hypothesis, implementing the analysis, interpreting the results, and writing the manuscripts(s).<br><br>Full time, €50,540-€54,965 per annum, post ends June 2025.<br>Benefits include a pension contribution and PRSI social insurance.<br><br>See further details: www.cusacklab.org/vacancies<br><br>Deadline 12 noon on April 22, 2024.Veronica Debora TORO2024-03-22T10:48:15+00:002024-03-22T10:48:15+00:00Re: Subjects with less than 10% explained variance in DCM_PEBhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;a15e5bd2.2403Dear Peter,<br>Thank you for your quick reply!<br>And in your experience, does the exclusion of these subjects impact the model comparison result? For example: I assumed two alternative models to the full One, can the posterior probability of compared models increase and/or decrease as a result of excluding these subjects?<br><br>Best<br>Debora<br><br>Inviato da Outlook per Android<https://aka.ms/AAb9ysg>Zeidman, Peter2024-03-22T10:39:03+00:002024-03-22T10:39:03+00:00Re: Subjects with less than 10% explained variance in DCM_PEBhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;39bbf77e.2403Dear Debora<br>I'm assuming the subjects with low explained variance had little or no activation in their SPM analyses (particularly in the region(s) that receive driving input).<br><br>You can either include the subjects with little or no activation - acknowledging that a representative sample of the population will include some people for whom experimental effects couldn't be detected - or you can exclude them, focussing your results on just those who do show activation. Both approaches have been used and can reasonably be justified.<br><br>All the best<br>Peter<br><br>-----Original Message-----<br>From: SPM (Statistical Parametric Mapping) <SPM@JISCMAIL.AC.UK> On Behalf Of Debora Toro<br>Sent: 22 March 2024 09:28<br>To: SPM@JISCMAIL.AC.UK<br>Subject: [SPM] Subjects with less than 10% explained variance in DCM_PEB<br><br>⚠ Caution: External sender<br><br>Dear SPM and DCM experts,<br>I'm running a DCM analysis. While checking diagnostics I noticed the presence of some subjects with poor explained variance (less than 10%). After checking their single-subject First level SPM Maps, am I entitled to esclude these subjects from the 2nd level DCM analysis? Does the esclusion of these subjects affect the results obtained from the model comparison?<br><br>Best regards<br>DeboraDebora Toro2024-03-22T09:27:33+00:002024-03-22T09:27:33+00:00Subjects with less than 10% explained variance in DCM_PEBhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;ef0b73dd.2403Dear SPM and DCM experts,<br>I'm running a DCM analysis. While checking diagnostics I noticed the presence of some subjects with poor explained variance (less than 10%). After checking their single-subject First level SPM Maps, am I entitled to esclude these subjects from the 2nd level DCM analysis? Does the esclusion of these subjects affect the results obtained from the model comparison?<br><br>Best regards<br>DeboraVolkmar Glauche2024-03-22T08:55:44+00:002024-03-22T08:55:44+00:00Re: MRI volume "origin": how it is defined during scanning ?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;325e656b.2403Dear Vadim,<br><br>the best thing to do at the scanner side would be to check for consistent positioning of subjects on the scanner table and within the head coil and for consistent positioning of the subject within the magnet. Unless there is a manual override, subject placement within the scanner should be quite automated. As for subject placement in the head coil, a good rule of thumb is to have the lateral center markings of the head coil in between eyes and eyebrows of the subject.<br><br>Volkmar<br><br>Am Donnerstag, dem 21.03.2024 um 23:59 +0200 schrieb Vadim Axelrod:<br>Dear Volkmar,<br><br>Thank you for your detailed answer. My problem is that unless I set the origins in the center of the brain (somewhere around the AC), the segmentation & normalization is completely messed up. I attach EPIs of two subjects which have very different origins and I have no idea why this happened (two subjects were scanned one after another on the same day). In the left subject segmentation & normalization worked just fine. For the right subject, only after I reoriented the volumes by positioning the origins at the AC (similar to the left subject), the segmentation & normalization started to work. Obviously, the origins were the same throughout all volumes of the same subject, including the T1. So, I wondered whether something can be done at the scanner side...<br><br>Thank you,<br>Vadim<br><br>On Tue, Mar 19, 2024 at 2:48 PM Volkmar Glauche <volkmar.glauche@uniklinik-freiburg.de<mailto:volkmar.glauche@uniklinik-freiburg.de>> wrote:<br>Dear Vadim,<br><br>in measured data, the origin is defined by the hardware and image reconstruction parameters of the measurement device. MR scanners usually set the origin to the center of the magnet. If you position the head of your subject in a standard head coil and position this coil correctly in the scanner, images should actually be quite close to what MNI/Talairach define as origin.<br>If it is not absolutely necessary, I would not mess around with this. Unless your data are really far off the origin, SPM does a pretty good job moving your data to the correct location during spatial normalisation or segmentation.<br><br>Best regards.Vadim Axelrod2024-03-21T23:59:39+02:002024-03-21T23:59:39+02:00Re: MRI volume "origin": how it is defined during scanning ?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;6a504943.2403Dear Volkmar,<br><br>Thank you for your detailed answer. My problem is that unless I set the<br>origins in the center of the brain (somewhere around the AC), the<br>segmentation & normalization is completely messed up. I attach EPIs of two<br>subjects which have very different origins and I have no idea why this<br>happened (two subjects were scanned one after another on the same day). In<br>the left subject segmentation & normalization worked just fine. For the<br>right subject, only after I reoriented the volumes by<br>positioning the origins at the AC (similar to the left subject), the<br>segmentation & normalization started to work. Obviously, the origins were<br>the same throughout all volumes of the same subject, including the T1. So,<br>I wondered whether something can be done at the scanner side...<br><br>Thank you,<br>Vadim<br><br>On Tue, Mar 19, 2024 at 2:48 PM Volkmar Glauche <<br>volkmar.glauche@uniklinik-freiburg.de> wrote:<br><br>> Dear Vadim,<br>><br>> in measured data, the origin is defined by the hardware and image<br>> reconstruction parameters of the measurement device. MR scanners usually<br>> set the origin to the center of the magnet. If you position the head of<br>> your subject in a standard head coil and position this coil correctly in<br>> the scanner, images should actually be quite close to what MNI/Talairach<br>> define as origin.<br>> If it is not absolutely necessary, I would not mess around with this.<br>> Unless your data are really far off the origin, SPM does a pretty good job<br>> moving your data to the correct location during spatial normalisation or<br>> segmentation.<br>><br>> Best regards.<br>>Alfonso Nieto-Castanon2024-03-20T23:03:33+00:002024-03-20T23:03:33+00:00last call CONN functional connectivity workshop registrationhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;a545590f.2403Just a quick announcement to let you know that the CONN 2024 functional connectivity online workshop will be starting in a couple of weeks (April 5th) and there are still a few spots available to register.<br><br>CONN (https://www.conn-toolbox.org) is a popular SPM toolbox for functional connectivity analyses. It has been used in over 5000 published studies to date, it is ranked in the top 1% most frequently viewed and downloaded neuroimaging tools in NITRC, and it enjoys a large and active user community, with over 1,900,000 page-views and 14,000 user-support posts. CONN implements preprocessing, quality assurance, denoising, first-level and second-level functional connectivity analyses, for both resting-state as well as task-based functional data.<br><br>The CONN toolbox workshop offers intensive hands-on and highly interactive courses covering all aspects of functional connectivity analyses in CONN, and both beginners and advanced users are welcome. Workshop topics include preprocessing fMRI data, denoising and quality control, seed-based and ROI-to-ROI connectivity analyses, gPPI for task designs, ICA, fc-MVPA, graph theoretical and dynamic connectivity measures, group-level GLM analyses, and network- and cluster-level inferences, among others. Classes take place over Zoom, and Zoom meetings of each class are recorded and made available to students to view at any time during the course. There are also practice homework assignments, and faculty is available during office-hour segments for additional student questions or support.<br><br>This workshop will take place ONLINE over ten consecutive Fridays, with 3-½ hour classes once a week from 11am to 2:30pm EDT (UTC−04:00), starting on Friday April 5 and ending on Friday June 7 2024 (see https://courses.conn-toolbox.org/course-schedule for course schedule details).<br><br>For additional information and registration see https://courses.conn-toolbox.org<br><br>Best regards<br>Alfonso Nieto-CastanonZhichao Xia2024-03-20T20:54:45+00:002024-03-20T20:54:45+00:00Postdoc position(s) at UConn brainLENS Lab (https://www.brainlens.org)https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;4ff89d08.2403Apologies for cross-posting!<br><br>The Hoeft Lab (http://brainLENS.org; PI: Fumiko Hoeft MD PhD, Campus Dean & CAO; Professor of the Department of Psychological Sciences) at the University of Connecticut (UConn) is looking for exceptional postdoc(s) in the field of cognitive neuroscience/neurolinguistics with advanced neuroimaging, computational, programming, and writing skills.<br><br>The postdocs are expected to analyze functional and structural MRI (and EEG) data and publish primarily from three NIH-funded projects: (1) the INTERGENERATIONAL NEUROIMAGING project that examines transmission of neurocognitive features of language/reading and executive function across generations (e.g., Ho et al. TiNS 2016); (2) the MULTILINGUAL project that investigates reading development and the underlying neural mechanisms in early multilingual children who were followed up since pre-literate grades 3/5 (e.g., Kepinska et al. Sci Rep 2023); and (3) the NEURAL NOISE HYPOTHESIS project that tests the hypothesis using multimodal MRI and EEG (e.g., Hancock et al. TiCS 2017). There are also opportunities to write grants and publish using other existing (e.g., TMS+MRI+MRS) and publicly available datasets in the lab on the neuroscience of language and literacy.<br><br>The candidate must have strong research experience in (1) reading / dyslexia or related fields of cognitive neuroscience and (2) MRI- (and/or EEG) based neuroimaging. Strong management, collaboration, communication, and writing skills are required. A strong publication record and expertise in programming, open-science approaches, and network / machine learning approaches are a plus.<br><br>The positions can begin immediately. Please email brainlens@uconn.edu with a cover letter describing qualifications and a current CV. Please add “[Postdoc job] First & Last Name” to the email's subject. Qualified candidates will be asked to have three letters of reference forwarded.<br><br>Thank you!<br><br>Best,<br>ZhichaoTaru Flagan2024-03-20T19:52:37+00:002024-03-20T19:52:37+00:00UCSF postdoctoral fellowship position in clinical neuroimaginghttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;36db24b9.2403The mission of the Dementia Imaging Genetics lab at the UCSF Memory and Aging Center is to understand the underlying biology of genetic dementias to improve the diagnosis and monitoring of preclinical and early-stage disease. Using neuroimaging techniques, we map the neural circuits targeted in genetic neurodegenerative diseases such as frontotemporal dementia (FTD) and Alzheimer’s disease (AD).<br><br>The postdoctoral fellow will contribute to ongoing experiments and pursue innovative clinical research projects. Studies include neuroimaging research in patients with genetic neurodegenerative diseases and their asymptomatic family members. The fellow will work with multimodal structural and functional neuroimaging methods, such as task-free MRI and diffusion tensor imaging, and will integrate neuroimaging techniques with behavioral and biofluid marker data to study the clinical, anatomical, and neuropathological correlates of disease. These projects are funded by grants from the National Institutes of Health, the Tau Consortium, and the Bluefield Project.<br><br>This position provides an exciting opportunity to become a key player in developing novel research projects with a dynamic team of lab members whose backgrounds and interests lie in neurodegenerative disease, brain development and neuroscience. Our multidisciplinary team includes neurologists, neuropsychologists, and neuroscientists. Research recruitment is supported by an extensive network of collaborative research projects at the UCSF Memory and Aging Center, with large datasets of archived MRI data available for analysis.<br><br>We are seeking creative, motivated, fun people to help improve the lives of our patients and families.<br><br>Qualifications: A PhD in neuroscience, psychology, or related field and expertise in neuroimaging and computer programming (Matlab, R, Python, etc.) are required. Competitive candidates will have a strong interest in neurodegenerative diseases. Previous experience working with clinical neuroimaging is preferred.<br><br>About UCSF: The UCSF Memory and Aging Center is located at the Mission Bay campus in San Francisco, CA. The Memory and Aging Center has over 300 employees and is the largest center in the United States focused on the study of neurodegenerative diseases. Our center includes a multidisciplinary team of neurologists, neuroscientists, neuropsychologists, postdoctoral fellows, nurses and research coordinators who form a rich and highly collaborative learning environment for trainees. The current National Institutes of Health rankings show that UCSF Neurology is ranked #1 nationally among departments of neurology at US medical schools.<br><br>UCSF is an Equal Opportunity/Affirmative Action Employer. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, age, protected veteran or disabled status, or genetic information.<br><br>Application: To apply, please send a cover letter describing your research interests and qualifications, a curriculum vitae, and at least three references to suzee.lee@ucsf.edu. Applications will be reviewed on an ongoing basis. Salary will be competitive and commensurate with experience.Model-based Neuroscience Summer School2024-03-20T11:28:03+01:002024-03-20T11:28:03+01:00Model-based Neuroscience Summer Schoolhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;6f4fb90f.2403Dear colleagues,<br><br>The tenth Model-Based Neuroscience and Cognition Summer School will be held<br>from July 29 to August 2, 2024, at the University of Amsterdam. The summer<br>school follows the annual meetings of the Society for Mathematical<br>Psychology and the Cognitive Science Society, both held in the Netherlands.<br><br>The Summer School will provide participants with hands-on experience in<br>both cognitive modeling using evidence-accumulation models and cognitive<br>neuroscience methods. This year, the summer school includes a three day<br>introduction to evidence-accumulation modeling using Bayesian inference. In<br>the last two days, participants can choose between a neuroscience and an<br>advanced cognition track. In the neuroscience track, participants will<br>learn how to create joint models of behavior and EEG/fMRI data. In the advanced<br>cognition track, participants will learn how to use the<br>evidence-accumulation framework to formalize performance in complex<br>experimental paradigms, how to construct joint models of multiple tasks using<br>techniques such as hierarchical factor models, and how to combine evidence-<br>accumulation and reinforcement-learning models.<br><br>The program includes a series of lectures on model-based cognitive and<br>neuroscience<br>from experts, including Andrew Heathcote, Dora Matzke, Michael Nunez,<br>Steven Miletić, and Birte Forstmann.<br><br>The summer school is aimed at PhD-students and early postdocs who wish to<br>acquire the basics of Bayesian evidence-accumulation modeling, and combine<br>that knowledge with 1) experimental and theoretical neuroscience research;<br>or 2) advanced modeling techniques aimed at capturing performance in<br>complex and dynamic paradigms and the integration of evidence-accumulation<br>with reinforcement-learning principles.<br><br>*Application and registration*<br><br>The registration fee for the Summer School is € 550. Participants should<br>make their own housing arrangements. Space is limited, therefore we ask<br>participants to provide a statement of interest. We will select<br>participants based on the relevance of the Summer School to their research.<br>Participants should be familiar with general programming concepts and state<br>in their statement of interest which programming languages and software<br>packages they typically use. Familiarity with R is particularly<br>recommended. The application deadline is May 15, 2024.<br><br>More information about the Summer School including lecturers, preliminary<br>program, and application can be found at https://modelbasedneurosci.com/.<br><br>We look forward to welcoming you to Amsterdam!<br><br>*Organization*<br><br>Birte Forstmann (University of Amsterdam)<br><br>Dora Matzke (University of Amsterdam)<br><br>Andrew Heathcote (University of Amsterdam)<br><br>Michelle Donzallaz (University of Amsterdam)<br><br>Niek Stevenson (University of Amsterdam)Brad Wyble2024-03-19T22:21:45-04:002024-03-19T22:21:45-04:00Hello, We're thrilled to announce our lineup of immersive courses at Neuromatch and Climatematch Academies! Whether you're a student eager to dive into cutting-edge topics or a seasoned researcher looking to contribute as a Teaching Assistant, we have opportunities tailored for you. 3-week Courses (July 8-26, 2024): Computational Neuroscience: Explore the intricacies of the brain's computational processes and join an engaging community of learners. Deep Learning: Delve into the world of artificial intelligence to uncover the principles and applications of deep learning. 2-week Courses (July 15-26, 2024): Computational Tools for Climate Science: Uncover the tools and techniques driving climate science research in this dynamic two-week course. NeuroAI (Inaugural Year!): Be part of history as we launch our first-ever NeuroAI course, designed to explore the intersection of neuroscience and artificial intelligence. Key Dates: Student & TA Applications Close: Sunday, March 24th, midnigh...https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;db224b62.2403Hello,<br>We're thrilled to announce our lineup of immersive courses at Neuromatch<br>and Climatematch Academies! Whether you're a student eager to dive into<br>cutting-edge topics or a seasoned researcher looking to contribute as a<br>Teaching Assistant, we have opportunities tailored for you.<br>*3-week Courses (July 8-26, 2024):*<br><br>- *Computational Neuroscience:* Explore the intricacies of the brain's<br>computational processes and join an engaging community of learners.<br>- *Deep Learning:* Delve into the world of artificial intelligence to<br>uncover the principles and applications of deep learning.<br><br>*2-week Courses (July 15-26, 2024):*<br><br>- *Computational Tools for Climate Science:* Uncover the tools and<br>techniques driving climate science research in this dynamic two-week course.<br>- *NeuroAI (Inaugural Year!):* Be part of history as we launch our<br>first-ever NeuroAI course, designed to explore the intersection of<br>neuroscience and artificial intelligence.<br><br>*Key Dates:*<br><br>- *Student & TA Applications Close:* Sunday, March 24th, midnight in the<br>last time zone on Earth<br><br>*Application Link:* Apply on Neuromatch Academy<br><https://neuromatch.io/courses/><br>*Additional Information:*<br><br>- All courses will be conducted exclusively online, providing<br>flexibility and accessibility.<br>- Detailed information about course fees can be found here<br><https://neuromatchacademy.github.io/widgets/cola.html>.<br>- Teaching Assistants will lead small groups of students, fostering<br>collaborative learning.<br>- TA positions offer financial compensation; detailed rates can be found<br>here <https://neuromatchacademy.github.io/widgets/ta_cola.html>.<br><br>Don't miss out on this incredible opportunity to expand your knowledge,<br>connect with experts in the field, and contribute to the success of our<br>courses. Apply today and join us on a journey of discovery!<br>Should you have any questions or need further information, feel free to<br>reach out to us.<br>We look forward to welcoming you to the Neuromatch Academy family!<br>Best regards,Fani Golemi2024-03-19T20:10:10+13:002024-03-19T20:10:10+13:00Re: DCM after EEG source reconstruction informed by fMRI priorshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;6056bd87.2403Dear experts,<br><br>I have to revise at some point for a project. Do I have to pay ? How much<br>does the course cost? Is there a discount if you have done it before?<br><br>Are there any free data I can use for a prototype?<br><br>I am aPhD student at the University of Otago and cyrrently writing a<br>proposal for one little study.<br><br>Many thanks<br>Fani<br><br>On Wed, 13 Mar 2024, 3:55 am Júlia Soares, <julii.f.soares@gmail.com> wrote:<br><br>> Dear SPM experts,<br>><br>> I performed EEG source reconstruction informed by fMRI priors. I intend to<br>> use this signal to do DCM analysis however, after reading the section<br>> "Dynamic Causal Modelling for M/EEG" in SPM manual I have a few questions I<br>> hope you can clarify:<br>><br>> 1) Is it only possible to do DCM in MNI space since the prior source<br>> locations should be given in MNI coordinates or is it possible to conduct<br>> DCM analysis in native space for each specific subject ?<br>><br>> 2) In source reconstruction I inverted a continuous signal, i.e., I did<br>> not separate the signal into epochs (trials). However I have a task which<br>> has 3 conditions in which I intend to study connectivity in each of them.<br>> Is there a way to separate my signal after source reconstruction so I can<br>> include them in the DCM model?<br>><br>> Thank you in advance.<br>> Regards,<br>> Júlia Soares<br>>Volkmar Glauche2024-03-19T12:47:54+00:002024-03-19T12:47:54+00:00Re: MRI volume "origin": how it is defined during scanning ?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;d8fe3819.2403Dear Vadim,<br><br>in measured data, the origin is defined by the hardware and image reconstruction parameters of the measurement device. MR scanners usually set the origin to the center of the magnet. If you position the head of your subject in a standard head coil and position this coil correctly in the scanner, images should actually be quite close to what MNI/Talairach define as origin.<br>If it is not absolutely necessary, I would not mess around with this. Unless your data are really far off the origin, SPM does a pretty good job moving your data to the correct location during spatial normalisation or segmentation.<br><br>Best regards.Hammerer, Dorothea2024-03-19T12:13:30+00:002024-03-19T12:13:30+00:00Locus Coeruleus meeting 2024 in Innsbruckhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;dc3e4f17.2403Dear SPMlers,<br><br>This email is to inform you about the upcoming 3rd Locus Coeruleus meeting which will take place at the Grillhof in Innsbruck from 9th to the 11th of September 2024.<br><br>As in the last meetings, we aim to bring together researchers with a background in clinical research, cognitive neuroscience, MRI Physics, neuroscience in animal models and other relevant disciplines with an interest in neuromodulatory systems and in particular the noradrenergic locus coeruleus in health and disease.<br><br>For more information on the meeting and to register please see the meeting website<https://www.uibk.ac.at/de/congress/lc-meeting-2024/>. (https://www.uibk.ac.at/de/congress/lc-meeting-2024/)<br><br>Registration will be open until the 15th of April 2024. Please do not hesitate to email us (using LCMeeting2024 ‘at’ uibk.ac.at) should you have any questions!<br><br>Please feel also free to forward information about the meeting to other people you think might be interested.<br><br>We hope to be able to welcome you at our 3rd LC Meeting!<br><br>Best wishes,<br><br>Dorothea Haemmerer<br><br>Professor for Developmental Psychology<br>Department of Psychology, University of Innsbruck<br><br>Principal Investigator,<br>Institute of Cognitive Neurology and Dementia Research,<br>Medical Faculty, Otto von Guericke University Magdeburg<br><br>Associate research fellow,<br>Institute of Cognitive Neuroscience, University College London<br><br>lab homepage<https://www.hammerer-lab.com> @d_haemmerer google scholar<https://scholar.google.de/citations?hl=de&user=XQT5B3QAAAAJ&view_op=list_works&sortby=pubdate> Research gate<https://www.researchgate.net/profile/Dorothea_Haemmerer> Zeidman, Peter2024-03-19T11:00:08+00:002024-03-19T11:00:08+00:00Re: What is the DCM.U.idx parameter for?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;44ff74c1.2403Hi Luna<br>DCM.U.idx is used by the “specify group” batch GUI when replicating a template DCM over subjects. It allows the GUI to know which experimental conditions to import from the subjects’ SPMs. The first column in DCM.U.idx is the condition number (within the Matlab structure SPM.Sess.U in your SPM.mat), and the second column is the regressor number (1 for the main regressor, 2 for the first parametric regressor, 3 for the second parametric regressor etc).<br><br>It will only have an impact on your results if you are using the “specify group” batch to generate your subject-specific DCMs from a template. Is that the case here?<br><br>Best<br>Peter<br><br>From: lunamitsukisato@keemail.me <lunamitsukisato@keemail.me><br>Sent: 17 March 2024 08:50<br>To: Zeidman, Peter <peter.zeidman@ucl.ac.uk><br>Cc: SPM@JISCMAIL.AC.UK<br>Subject: RE: [SPM] What is the DCM.U.idx parameter for?<br><br>Hi Peter,<br><br>Thank you!<br><br>Initially, I thought I understood the logic behind the numbers in the DCM.U.idx generated by the DCM GUI. The matrix looked like this:<br>DCM.U.idx=[1 1<br>2 1<br>3 1<br>4 1]<br><br>I interpreted it as follows: the first column ranging from 1 to 4 corresponds to the four experimental conditions, and the second column consists entirely of ones.<br><br>Based on this understanding, I defined all DCMs in my code using the same idx matrix. Namely, I applied this matrix to DCM.U.idx in all subjects' DCM files.<br><br>However, I later discovered that the DCM GUI generates the DCM.U.idx for another DCM (including parametric modulation regressors) in this format:<br><br>DCM.U.idx=[1 1<br>1 2<br>1 3<br>1 4]<br><br>This discrepancy has left me uncertain whether my approach of defining the same matrix for all subjects' DCM.U.idx was incorrect.<br><br>Would defining the DCM.U.idx matrix incorrectly in the code have a significant impact on the results?<br><br>Cheers, Luna<br><br>12 Mar 2024, 18:07 by peter.zeidman@ucl.ac.uk<mailto:peter.zeidman@ucl.ac.uk>:<br>Hi Luna<br>This is nothing to worry about. If I remember correctly, I added that field to help with specifying DCMs using the batch editor. Its omission won’t alter your results.<br><br>Best<br>Peter<br><br>From: SPM (Statistical Parametric Mapping) <SPM@JISCMAIL.AC.UK<mailto:SPM@JISCMAIL.AC.UK>> On Behalf Of Luna Sato<br>Sent: 06 March 2024 03:23<br>To: SPM@JISCMAIL.AC.UK<mailto:SPM@JISCMAIL.AC.UK><br>Subject: [SPM] What is the DCM.U.idx parameter for?<br><br>⚠ Caution: External sender<br><br>Hi experts,<br><br>When checking DCM results, I found certain subjects' DCM fields include the parameter DCM.U.idx, while others don't. I suspect this variation might be due to different SPM versions used.<br><br>I'm wondering about the significance of DCM.U.idx. Can I combine subjects with and without this parameter in group analysis? Or should I consider redoing some DCM analyses?<br><br>Best regards,<br>LunaDiederick C Niehorster2024-03-19T10:13:07+00:002024-03-19T10:13:07+00:00Three day course in eye-tracking methodology - A practical introduction to eye trackinghttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;cdc3f9a0.2403Dear colleagues,<br><br>For the eighth time, the Lund University Humanities Lab, Lund, Sweden, offers a three-day intensive course in eye-tracking methodology: A practical introduction to eye tracking. An important part of the course comprises hands-on exercises, where participants will work individually with state-of-the-art eye tracker hardware and software.<br>The course takes place May 29 - 31, 2024 at the Humanities Lab.<br><br>For more information about the course, visit the course webpage:<br>https://www.humlab.lu.se/education/commissioned-education/<br><br>Please feel free to share the link with people who you think would be interested.<br><br>Best wishes,<br>Marcus Nyström and Diederick Niehorster<br>Lund University Humanities Lab Zeidman, Peter2024-03-19T09:56:38+00:002024-03-19T09:56:38+00:00Re: Inquiry regarding interpretation of self-connection in one-state fMRI DCMhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;e03368da.2403Dear Zach<br>I hope you don’t mind me CC’ing the SPM mailing list so others can benefit.<br><br>You asked how to interpret the self-connections in SPM. In brief, writing out the DCM for fMRI neural model (ordinary differential equation) for a self-connection:<br><br>dz/dt = (-0.5 * exp[A + Bu(t)] ) * z + CU<br><br>Where u(t)>0 at times when the modulatory input is switched on and u(t)=0 at other times. Details of the terms of the equation are in the DCM tutorial paper (https://doi.org/10.1016/j.neuroimage.2019.06.031) . Equivalently this can be written as:<br><br>dz/dt = (-0.5 * exp(A) * exp(Bu(t))) * z + Cu<br><br>So you can see the default connection strength is -0.5Hz, which is scaled by the exponential of the A matrix, and by the exponential of the B matrix (at times when modulation is switched on).<br><br>To answer your direct question, you are correct, and you can confirm this by plugging in numbers to the formula. E.g. if the modulatory input is on, so u(t)=1, and if A=0 and B=0, then the self-connection is -0.5 * exp(0) * exp(0) = -0.5Hz (the default value for a self-connection). If you increase the B parameter from zero to two, then it’s -0.5 * exp(0) * exp(2) = -3.7Hz, so much more inhibitory.<br><br>I hope that helps<br>Peter<br><br>Subject: Inquiry regarding interpretation of self-connection in one-state fMRI DCM<br><br>⚠ Caution: External sender<br><br>Dear Professor Zeidman,<br><br>I apologize for reaching out once again, but I've been grappling with the interpretation of self-connection in one-state DCM for quite some time.<br><br>There is the equation:<br>[cid:1e567sppqqqe]<br>So, large 'a' parameter will cause faster neural decay than a small 'a' parameter.<br><br>For one-state fMRI DCM, the relationship between the values on the diagonal of matrix A and 'a' is:<br>[cid:jsj7gdns7cf]<br>So large values on the diagonal of matrix A (DCM.Ep.A) will cause slower decay than small values on the diagonal of matrix A.<br><br>However, the values on the diagonal of matrix of B matrices are NOT log scaling parameters.<br>So larger values on the diagonal of matrix of B matrices (DCM.Ep.B) indicate the corresponding experimental effects faster decay than small values on the diagonal of matrix of B.<br><br>Am I correct?<br><br>Thank you!<br><br>Best,<br>ZachVladimir Litvak2024-03-18T17:53:12+00:002024-03-18T17:53:12+00:00Re: DCM after EEG source reconstruction informed by fMRI priorshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;51ffff8a.2403On Mon, 18 Mar 2024 at 14:56, Vladimir Litvak <litvak.vladimir@gmail.com><br>wrote:<br><br>> Dear Julia,<br>><br>> On Mon, Mar 18, 2024 at 2:45 PM Júlia Soares <julii.f.soares@gmail.com><br>> wrote:<br>><br>>><br>>> What do you mean with "native images coregistered with MNI" and " write<br>>> out the results of your source analysis in the native space on a template<br>>> MNI mesh"?<br>>><br>><br>> I mean coregistering your images to the MNI template (Coregister/Estimate)<br>> so that their coordinate system doesn't have a big offset or some rotation<br>> with respect to MNI coordinates. Then in my opinion locations specified in<br>> this MNI-aligned native space will be similar enough to MNI coordinates for<br>> the purposes of EEG analysis. Also by default your results image for source<br>> reconstruction is generated in MNI template space no matter how your priors<br>> are specified.<br>><br>><br>><br>>> Does this mean to transform my structural images to MNI using the<br>>> "normalise write" function from SPM,<br>>><br>> to build the head models and then do source reconstruction in this space?<br>>><br>><br>> No, that's not what I mean.<br>><br>><br>>><br>>> About the epoching step, just to make sure: do you suggest separating my<br>>> continuous signal into the periods of time of my conditions and then<br>>> separating each condition into epochs of 1-2s like a sub epoching step ?<br>>><br>>><br>> I'm not sure how your conditions are recorded. If they are in separate<br>> files then you could just epoch each one into 1 sec epochs and then merge<br>> the resulting files. Otherwise you could convert each epoch separately (by<br>> specifying a time window) and then epoch it and merge. The most<br>> straightforward way to do this kind of custom epoching is write your own<br>> function for specifying the trl and conditionlabels variables that the<br>> epoching function takes as the input. But if you want to only use the GUI,<br>> you could do as suggested above.<br>><br>> Best,<br>><br>> Vladimir<br>><br>><br>><br>><br>><br>><br>>><br>>> Em sex., 15 de mar. de 2024 às 16:28, Vladimir Litvak <<br>>> litvak.vladimir@gmail.com> escreveu:<br>>><br>>>> Dear Julia,<br>>>><br>>>> On Fri, Mar 15, 2024 at 4:20 PM Júlia Soares <julii.f.soares@gmail.com><br>>>> wrote:<br>>>><br>>>>> 1) Regarding the source locations in MNI I didn't quite understand why<br>>>>> this doesn't matter. The DCM model requires an EEG signal after source<br>>>>> reconstruction, right? So the space will be the source space instead of the<br>>>>> sensor space (the actual electrodes), right? If so, how come the resolution<br>>>>> of the coordinates doesn't make a difference? Isn't it possible that<br>>>>> sources are several mm misaligned with corresponding locations in MNI ?<br>>>>><br>>>><br>>>> The kind of differences in source locations that make a difference in<br>>>> EEG are on the order of cm so if your native images are coregistered to MNI<br>>>> and the head sizes are not unusually large or small I wouldn't expect the<br>>>> mm differences to matter. But you can always write out the results of your<br>>>> source analysis in the native space on a template MNI mesh and then you<br>>>> won't have that problem at all.<br>>>><br>>>><br>>>><br>>>><br>>>>> 2) About data epoching: I have a continuous signal acquired during<br>>>>> performance of a task constituted by 4 conditions: 8 periods of "baseline"<br>>>>> (22 seconds), 5 periods of "condition A" (18 seconds), 4 periods of<br>>>>> "condition B" (18 seconds) and 3 periods of "condition C" (18 seconds). I<br>>>>> was thinking about separating my continuous signal into epochs of equal<br>>>>> length to the periods of each condition. So, for example for "condition A"<br>>>>> I would have 5 epochs of 18 seconds each corresponding to "condition A "<br>>>>> which would then be averaged into one single epoch. Does this make sense?<br>>>>><br>>>>><br>>>> The implementation assumes short epochs 1-2 sec at most so I'd suggest<br>>>> you epoch your conditions into epochs of that length and then the<br>>>> differences in duration won't matter.<br>>>><br>>>> Best,<br>>>><br>>>> Vladimir<br>>>><br>>>><br>>>>> Regards,<br>>>>> Júlia Soares<br>>>>><br>>>>><br>>>>><br>>>>> Em ter., 12 de mar. de 2024 às 15:24, Vladimir Litvak <<br>>>>> litvak.vladimir@gmail.com> escreveu:<br>>>>><br>>>>>> Dear Julia,<br>>>>>><br>>>>>> On Tue, Mar 12, 2024 at 2:55 PM Júlia Soares <julii.f.soares@gmail.com><br>>>>>> wrote:<br>>>>>><br>>>>>>> 1) Is it only possible to do DCM in MNI space since the prior source<br>>>>>>> locations should be given in MNI coordinates or is it possible to conduct<br>>>>>>> DCM analysis in native space for each specific subject ?<br>>>>>>><br>>>>>>><br>>>>>> I think this distinction is too fine to matter for DCM if you are<br>>>>>> doing it at the sensor level. So I'd just define source locations in MNI<br>>>>>> space and not worry too much about it.<br>>>>>><br>>>>>><br>>>>>><br>>>>>>> 2) In source reconstruction I inverted a continuous signal, i.e., I<br>>>>>>> did not separate the signal into epochs (trials). However I have a task<br>>>>>>> which has 3 conditions in which I intend to study connectivity in each of<br>>>>>>> them. Is there a way to separate my signal after source reconstruction so I<br>>>>>>> can include them in the DCM model?<br>>>>>>><br>>>>>><br>>>>>> Both source analysis and DCM were not intended to work on long<br>>>>>> continuous data segments. I'd suggest you epoch your data into arbitrary<br>>>>>> 1-2 sec epochs. There is a way to do it in the epoching tool. Then I would<br>>>>>> do both steps on these epoched data.<br>>>>>><br>>>>>> Best,<br>>>>>><br>>>>>> Vladimir<br>>>>>><br>>>>>><br>>>>>><br>>>>>>><br>>>>>>> Thank you in advance.<br>>>>>>> Regards,<br>>>>>>> Júlia Soares<br>>>>>>><br>>>>>>Sam Javidi2024-03-18T15:22:06+00:002024-03-18T15:22:06+00:00Research Assistant Position in the Neuroscience of Memoryhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;7de24a69.2403Research Assistant Position in the Neuroscience of Memory<br><br>A research assistant position is available in Dr. Noa Herz’s laboratory in the Department of Neurology at Thomas Jefferson University. Research in the lab focuses on the neural substrates underlying episodic memory. We use direct brain recording and stimulation collected from neurosurgical patients who have implanted electrodes for seizure mapping. Our research focuses on characterizing memory deficits in psychopathological (depression, anxiety, post-traumatic stress disorder) and neurological (epilepsy) disorders and on developing direct stimulation approaches to address them. <br><br>We are closely collaborating and holding routine meetings with Prof. Michael Kahana's research group at the University of Pennsylvania. Duties will include assisting with all aspects of data collection, experiment preparation, data postprocessing and report generation. Data analyses and manuscript writing are offered to interested individuals.<br><br>Review of applications will start immediately and will continue until the position is filled. <br><br> <br><br>Requirements:<br><br>- BA/BS in cognitive science, neuroscience, biology, psychology, computer science, engineering, or other related scientific fields.<br>- Strong computing skills (knowledge of python/R/Matlab is a plus)<br>- An ability to solve technical problems independently<br>- Strong organization skills and high attention to detail<br>- High motivation and work commitment<br>- Ability to work well with patients in a hospital environment<br>- At least one, but preferably a 2-year commitment<br><br>The Department of Neurology, located in the city center of Philadelphia, is among the ten best neuroscience departments in the country. The work includes collaboration with top neurologists, neurosurgeons, and neuropsychologists and is, therefore, ideal for students thinking about an MD. The Herz lab is currently under development, allowing the selected applicant to shape future work in the lab and assist in forming new research collaborations.<br><br>For inquiries, please email: noa.herz@jefferson.edu<br><br>To apply, please submit a resume (including a description of computer skills, relevant coursework, grades, previous research, and contact information for at least two references) and a cover letter describing academic and research interests on:<br><br>https://recruit.jefferson.edu/psc/hcmp/EMPLOYEE/HRMS/c/HRS_HRAM_FL.HRS_CG_SEARCH_FL.GBL?Page=HRS_APP_JBPST_FL&Action=U&FOCUS=Applicant&SiteId=1&JobOpeningId=9298395&PostingSeq=1Sabrina Golde2024-03-18T15:21:17+00:002024-03-18T15:21:17+00:00Problem with A matrix's PEB-model for DCMhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;44214972.2403Hi everyone,<br><br>I am reaching out due to a curious problem with DCM to seek any insights you might offer. Our analysis involves a comprehensive first-level full model that includes 10 regions of interest (ROIs). To streamline the model's complexity, we strategically reduced the connections from an the initial pool of 10x10 = 100 down to 57. (This reduction was based on the posterior probability (>0.95) obtained by inverting the A matrix in a preparatory step for the actual analysis.)<br><br>Consequently, in the single-subject full models, we set these 57 connections as active (value of 1), while the remaining connections were inactivated / pruned (value of 0) in our A and B matrices.<br><br>Following Zeidman et al. 2019, we then ran BMA+PEB on the inverted single subject model for both, the A and the B matrix. The results from the B matrix's PEB model appeared sensible and only included those 57 connections. However, the A matrix presented an unexpected outcome: 21 connections, which were previously pruned and set to 0 in the original single-subject DCMs, were part of the A matrix PEB model. I'm confused why PEB would have different parameters for A and B matrices because it is done on the same single-subject models. If I understand correctly, it should only calculate the Bayesian averages of those parameters.<br><br>I double checked the specifications etc. and I am not sure whether something has gone wrong. I would greatly appreciate any input, advice, or suggestions. Thank you very much in advance for your time.<br><br>Best,<br>Sabrina Sam Javidi2024-03-18T15:19:03+00:002024-03-18T15:19:03+00:00Postdoctoral Fellowship in Cognitive Electrophysiology of Human Memoryhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;cb3bc02c.2403Postdoctoral Fellowship in Cognitive Electrophysiology of Human Memory<br><br>We are looking for a postdoctoral scientist to join our emerging team and carry out a research program in electrophysiology. The successful applicant will collaborate with the research group of Dr. Noa Herz and Dr. Michael Sperling at the Comprehensive Epilepsy Center of Thomas Jefferson University.<br><br>The purpose of the post is to conduct cutting-edge research on the neural basis of episodic memory in humans using direct brain recordings and stimulation. We are seeking to develop a memory mapping procedure to reduce the risk of memory loss following neurosurgery, as well as a direct stimulation intervention to treat memory-related disorders such as post-traumatic stress disorder. We use data collected from epilepsy patients undergoing seizure monitoring at the hospital as well as patients implanted with a responsive neurostimulation system (RNS).<br><br>The successful candidate will undertake project management, analyze data and will be expected to submit publications to top journals, assist in applying for external research funding, and promote collaborations with research groups outside of Jefferson.<br><br>This post offers an excellent opportunity for those interested in state-of-the-art translational research, bridging cognitive and clinical neuroscience. The selected researcher will become an early member of an inclusive and interdisciplinary team of neurologists, neurosurgeons, and neuroscientists working to study and develop treatment interventions for memory problems in psychopathological and neurological disorders. We are closely collaborating and holding routine meetings with Prof. Michael Kahana's research group at the University of Pennsylvania.<br><br>The ideal candidate will have expertise in analyzing electrophysiological data (either in humans or animals) and strong computing skills, including coding experience. Knowledge of machine learning methods will be an advantage. <br><br>We are located at the Center City Campus of Thomas Jefferson University (The Vickie & Jack Farber Institute for Neuroscience). Applications will be reviewed on a rolling basis and salary will be based on the NIH postdoctoral scale.<br><br>Applicant must have a Ph.D. in neuroscience/biology/psychology/computer science or a related field. Applicants close to completion of their PhDs will also be considered where experience is directly relevant.<br><br>Queries relating to the position should be directed to Dr. Noa Herz: noa.herz@jefferson.edu.<br><br> <br><br>Apply here:<br>https://recruit.jefferson.edu/psc/hcmp/EMPLOYEE/HRMS/c/HRS_HRAM_FL.HRS_CG_SEARCH_FL.GBL?Page=HRS_APP_SCHJOB_FL&Action=U<br><br> <br><br>Key Responsibilities:<br><br>- Analyses of behavioral and intracranial data<br>- Take initiative in the planning and conduct of research<br>- Acquire and interpret research data and results<br>- Ensure the validity and reliability of collected data<br>- Prepare publications for submission in top refereed journals<br>- Assist in the preparation of research grant proposals<br>- Assist with designing and building experimental tasks, as well as in data collection<br><br>Required Experience:<br><br>- Proven track record of electrophysiology research.<br>- Experience in data analyses and coding (e.g., using Python/Matlab/R/C++).<br>- Experience in delivering research project results, as exemplified by a track record of peer-reviewed publications in a relevant area.<br><br>Skills and Abilities<br><br>- Capable of working collaboratively with neurologists and neurosurgeons in a hospital environment.<br>- Capable of working independently, exercising a high degree of initiative, and demonstrating a proactive approach to work.<br>- Strong quantitative background (multivariate methods such as machine learning are a plus)<br>- Ability to conduct a detailed review of recent literature.<br>- Demonstrated ability to conduct independent research.<br>- Creative approach to problem-solving.<br>- Excellent written communication skills in scientific English and the ability to write clearly and succinctly at a level suitable for international conferences and peer-reviewed journal publications.<br>- Self-motivation and ability to exercise initiative and judgment in carrying out research tasks.<br><br> <br><br>Interested applicants should upload a cover letter including a statement of research interests (describing how past experience and future plans fit the advertised position), a CV, and the details of at least two referees.<br><br> <br><br>We are committed to equality of opportunity, eliminating discrimination, and creating an inclusive working environment for all. We therefore encourage candidates to apply irrespective of age, disability, marriage or civil partnership status, pregnancy or maternity, race, religion and belief, gender reassignment, sex, or sexual orientation. Vladimir Litvak2024-03-18T14:56:21+00:002024-03-18T14:56:21+00:00Re: DCM after EEG source reconstruction informed by fMRI priorshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;8ff560ae.2403Dear Julia,<br><br>On Mon, Mar 18, 2024 at 2:45 PM Júlia Soares <julii.f.soares@gmail.com><br>wrote:<br><br>><br>> What do you mean with "native images coregistered with MNI" and " write<br>> out the results of your source analysis in the native space on a template<br>> MNI mesh"?<br>><br><br>I mean coregistering your images to the MNI template (Coregister/Estimate)<br>so that their coordinate system doesn't have a big offset or some rotation<br>with respect to MNI coordinates. Then in my opinion locations specified in<br>this MNI-aligned native space will be similar enough to MNI coordinates for<br>the purposes of EEG analysis. Also by default your results image for source<br>reconstruction is generated in MNI template space no matter how your priors<br>are specified.<br><br>> Does this mean to transform my structural images to MNI using the<br>> "normalise write" function from SPM,<br>><br>to build the head models and then do source reconstruction in this space?<br>><br><br>No, that's not what I mean.<br><br>><br>> About the epoching step, just to make sure: do you suggest separating my<br>> continuous signal into the periods of time of my conditions and then<br>> separating each condition into epochs of 1-2s like a sub epoching step ?<br>><br>><br>I'm not sure how your conditions are recorded. If they are in separate<br>files then you could just epoch each one into 1 sec epochs and then merge<br>the resulting files. Otherwise you could convert each epoch separately (by<br>specifying a time window) and then epoch it and merge. The most<br>straightforward way to do this kind of custom epoching is write your own<br>function for specifying the trl and conditionlabels variables that the<br>epoching function takes as the input. But if you want to only use the GUI,<br>you could do as suggested above.<br><br>Best,<br><br>Vladimir<br><br>><br>> Em sex., 15 de mar. de 2024 às 16:28, Vladimir Litvak <<br>> litvak.vladimir@gmail.com> escreveu:<br>><br>>> Dear Julia,<br>>><br>>> On Fri, Mar 15, 2024 at 4:20 PM Júlia Soares <julii.f.soares@gmail.com><br>>> wrote:<br>>><br>>>> 1) Regarding the source locations in MNI I didn't quite understand why<br>>>> this doesn't matter. The DCM model requires an EEG signal after source<br>>>> reconstruction, right? So the space will be the source space instead of the<br>>>> sensor space (the actual electrodes), right? If so, how come the resolution<br>>>> of the coordinates doesn't make a difference? Isn't it possible that<br>>>> sources are several mm misaligned with corresponding locations in MNI ?<br>>>><br>>><br>>> The kind of differences in source locations that make a difference in EEG<br>>> are on the order of cm so if your native images are coregistered to MNI and<br>>> the head sizes are not unusually large or small I wouldn't expect the mm<br>>> differences to matter. But you can always write out the results of your<br>>> source analysis in the native space on a template MNI mesh and then you<br>>> won't have that problem at all.<br>>><br>>><br>>><br>>><br>>>> 2) About data epoching: I have a continuous signal acquired during<br>>>> performance of a task constituted by 4 conditions: 8 periods of "baseline"<br>>>> (22 seconds), 5 periods of "condition A" (18 seconds), 4 periods of<br>>>> "condition B" (18 seconds) and 3 periods of "condition C" (18 seconds). I<br>>>> was thinking about separating my continuous signal into epochs of equal<br>>>> length to the periods of each condition. So, for example for "condition A"<br>>>> I would have 5 epochs of 18 seconds each corresponding to "condition A "<br>>>> which would then be averaged into one single epoch. Does this make sense?<br>>>><br>>>><br>>> The implementation assumes short epochs 1-2 sec at most so I'd suggest<br>>> you epoch your conditions into epochs of that length and then the<br>>> differences in duration won't matter.<br>>><br>>> Best,<br>>><br>>> Vladimir<br>>><br>>><br>>>> Regards,<br>>>> Júlia Soares<br>>>><br>>>><br>>>><br>>>> Em ter., 12 de mar. de 2024 às 15:24, Vladimir Litvak <<br>>>> litvak.vladimir@gmail.com> escreveu:<br>>>><br>>>>> Dear Julia,<br>>>>><br>>>>> On Tue, Mar 12, 2024 at 2:55 PM Júlia Soares <julii.f.soares@gmail.com><br>>>>> wrote:<br>>>>><br>>>>>> 1) Is it only possible to do DCM in MNI space since the prior source<br>>>>>> locations should be given in MNI coordinates or is it possible to conduct<br>>>>>> DCM analysis in native space for each specific subject ?<br>>>>>><br>>>>>><br>>>>> I think this distinction is too fine to matter for DCM if you are doing<br>>>>> it at the sensor level. So I'd just define source locations in MNI space<br>>>>> and not worry too much about it.<br>>>>><br>>>>><br>>>>><br>>>>>> 2) In source reconstruction I inverted a continuous signal, i.e., I<br>>>>>> did not separate the signal into epochs (trials). However I have a task<br>>>>>> which has 3 conditions in which I intend to study connectivity in each of<br>>>>>> them. Is there a way to separate my signal after source reconstruction so I<br>>>>>> can include them in the DCM model?<br>>>>>><br>>>>><br>>>>> Both source analysis and DCM were not intended to work on long<br>>>>> continuous data segments. I'd suggest you epoch your data into arbitrary<br>>>>> 1-2 sec epochs. There is a way to do it in the epoching tool. Then I would<br>>>>> do both steps on these epoched data.<br>>>>><br>>>>> Best,<br>>>>><br>>>>> Vladimir<br>>>>><br>>>>><br>>>>><br>>>>>><br>>>>>> Thank you in advance.<br>>>>>> Regards,<br>>>>>> Júlia Soares<br>>>>>><br>>>>>Júlia Soares2024-03-18T14:44:45+00:002024-03-18T14:44:45+00:00Re: DCM after EEG source reconstruction informed by fMRI priorshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;4e198a66.2403Dear Vladimir,<br><br>What do you mean with "native images coregistered with MNI" and " write out<br>the results of your source analysis in the native space on a template MNI<br>mesh"? Does this mean to transform my structural images to MNI using the<br>"normalise write" function from SPM, to build the head models and then do<br>source reconstruction in this space?<br><br>About the epoching step, just to make sure: do you suggest separating my<br>continuous signal into the periods of time of my conditions and then<br>separating each condition into epochs of 1-2s like a sub epoching step ?<br><br>I apologize for my insistence and appreciate your help,<br>Júlia Soares<br><br>Em sex., 15 de mar. de 2024 às 16:28, Vladimir Litvak <<br>litvak.vladimir@gmail.com> escreveu:<br><br>> Dear Julia,<br>><br>> On Fri, Mar 15, 2024 at 4:20 PM Júlia Soares <julii.f.soares@gmail.com><br>> wrote:<br>><br>>> 1) Regarding the source locations in MNI I didn't quite understand why<br>>> this doesn't matter. The DCM model requires an EEG signal after source<br>>> reconstruction, right? So the space will be the source space instead of the<br>>> sensor space (the actual electrodes), right? If so, how come the resolution<br>>> of the coordinates doesn't make a difference? Isn't it possible that<br>>> sources are several mm misaligned with corresponding locations in MNI ?<br>>><br>><br>> The kind of differences in source locations that make a difference in EEG<br>> are on the order of cm so if your native images are coregistered to MNI and<br>> the head sizes are not unusually large or small I wouldn't expect the mm<br>> differences to matter. But you can always write out the results of your<br>> source analysis in the native space on a template MNI mesh and then you<br>> won't have that problem at all.<br>><br>><br>><br>><br>>> 2) About data epoching: I have a continuous signal acquired during<br>>> performance of a task constituted by 4 conditions: 8 periods of "baseline"<br>>> (22 seconds), 5 periods of "condition A" (18 seconds), 4 periods of<br>>> "condition B" (18 seconds) and 3 periods of "condition C" (18 seconds). I<br>>> was thinking about separating my continuous signal into epochs of equal<br>>> length to the periods of each condition. So, for example for "condition A"<br>>> I would have 5 epochs of 18 seconds each corresponding to "condition A "<br>>> which would then be averaged into one single epoch. Does this make sense?<br>>><br>>><br>> The implementation assumes short epochs 1-2 sec at most so I'd suggest you<br>> epoch your conditions into epochs of that length and then the differences<br>> in duration won't matter.<br>><br>> Best,<br>><br>> Vladimir<br>><br>><br>>> Regards,<br>>> Júlia Soares<br>>><br>>><br>>><br>>> Em ter., 12 de mar. de 2024 às 15:24, Vladimir Litvak <<br>>> litvak.vladimir@gmail.com> escreveu:<br>>><br>>>> Dear Julia,<br>>>><br>>>> On Tue, Mar 12, 2024 at 2:55 PM Júlia Soares <julii.f.soares@gmail.com><br>>>> wrote:<br>>>><br>>>>> 1) Is it only possible to do DCM in MNI space since the prior source<br>>>>> locations should be given in MNI coordinates or is it possible to conduct<br>>>>> DCM analysis in native space for each specific subject ?<br>>>>><br>>>>><br>>>> I think this distinction is too fine to matter for DCM if you are doing<br>>>> it at the sensor level. So I'd just define source locations in MNI space<br>>>> and not worry too much about it.<br>>>><br>>>><br>>>><br>>>>> 2) In source reconstruction I inverted a continuous signal, i.e., I did<br>>>>> not separate the signal into epochs (trials). However I have a task which<br>>>>> has 3 conditions in which I intend to study connectivity in each of them.<br>>>>> Is there a way to separate my signal after source reconstruction so I can<br>>>>> include them in the DCM model?<br>>>>><br>>>><br>>>> Both source analysis and DCM were not intended to work on long<br>>>> continuous data segments. I'd suggest you epoch your data into arbitrary<br>>>> 1-2 sec epochs. There is a way to do it in the epoching tool. Then I would<br>>>> do both steps on these epoched data.<br>>>><br>>>> Best,<br>>>><br>>>> Vladimir<br>>>><br>>>><br>>>><br>>>>><br>>>>> Thank you in advance.<br>>>>> Regards,<br>>>>> Júlia Soares<br>>>>><br>>>>Fani Golemi2024-03-18T11:52:22+13:002024-03-18T11:52:22+13:00Re: DCM after EEG source reconstruction informed by fMRI priorshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;ba42100a.2403Are you all having fun?<br><br>On Wed, 13 Mar 2024, 4:25 am Vladimir Litvak, <litvak.vladimir@gmail.com><br>wrote:<br><br>> Dear Julia,<br>><br>> On Tue, Mar 12, 2024 at 2:55 PM Júlia Soares <julii.f.soares@gmail.com><br>> wrote:<br>><br>>> 1) Is it only possible to do DCM in MNI space since the prior source<br>>> locations should be given in MNI coordinates or is it possible to conduct<br>>> DCM analysis in native space for each specific subject ?<br>>><br>>><br>> I think this distinction is too fine to matter for DCM if you are doing it<br>> at the sensor level. So I'd just define source locations in MNI space and<br>> not worry too much about it.<br>><br>><br>><br>>> 2) In source reconstruction I inverted a continuous signal, i.e., I did<br>>> not separate the signal into epochs (trials). However I have a task which<br>>> has 3 conditions in which I intend to study connectivity in each of them.<br>>> Is there a way to separate my signal after source reconstruction so I can<br>>> include them in the DCM model?<br>>><br>><br>> Both source analysis and DCM were not intended to work on long continuous<br>> data segments. I'd suggest you epoch your data into arbitrary 1-2 sec<br>> epochs. There is a way to do it in the epoching tool. Then I would do both<br>> steps on these epoched data.<br>><br>> Best,<br>><br>> Vladimir<br>><br>><br>><br>>><br>>> Thank you in advance.<br>>> Regards,<br>>> Júlia Soares<br>>><br>>William Penny2024-03-18T10:31:34+00:002024-03-18T10:31:34+00:00Job Ad: Lecturer in Psychology @ UEA, Norwich, UKhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;4da81854.2403*Lecturer in Psychology (2 posts) REF: ATR1682*<br><br>*Salary on appointment will be £46,974 per annum, dependent on skills and<br>experience, with an annual increment up to £54,395 per annum. *<br><br>An exciting opportunity has arisen for two new Lecturers to join the School<br>of Psychology. The posts will augment our current excellence in research<br>and take an active role in the delivery of research-led teaching.<br><br>In REF 2021, 96% of our research in Psychology, Psychiatry and Neuroscience<br>was recognised as world-leading or internationally excellent, with a 100%<br>4* world leading research environment. The School has outstanding<br>laboratory equipment and facilities including a Siemens Prisma 3T MRI<br>scanner in the UEA Wellcome-Wolfson Brain Imaging Centre (UWWBIC:<br>https://uwwbic.uea.ac.uk/).<br><br>You will join a network of researchers within the School of Psychology,<br>with clinical colleagues in the School of Medicine and others across the<br>University and the wider Norwich Research Park, consistent with the<br>multidisciplinary ethos at UEA. You will have the opportunity to develop<br>your research profile which should compliment the existing research<br>strengths, producing high quality proposals to secure external research<br>funding, disseminate finding through academic publications and external<br>impact.<br><br>Teaching is a key part of this role and as such you will be expected to<br>plan, teach and assess at undergraduate and postgraduate levels and<br>supervise PhD students.<br><br>We are seeking candidates specialising in any area of Psychology who will<br>expand and enhance our existing research strengths. Those working in Social<br>Psychology, Cognitive and/or Social Development, Cognitive Neuroscience,<br>Social Neuroscience, or Developmental Neuroscience would complement the<br>School’s approach to psychological science and would be especially welcome.<br><br>You must have a PhD (or equivalent) in a relevant subject area or<br>equivalent experience, with experience of undergraduate teaching and<br>student assessment. A strong publication record is also essential, and the<br>ability to supervise PhD students would be advantageous.<br><br>These full-time posts are available from 1 August 2024 on an indefinite<br>basis.<br><br>UEA offers a variety of flexible working options and although these roles<br>are advertised on a full-time basis, we encourage applications from<br>individuals who would prefer a flexible working pattern including part<br>time, or job share arrangements. Details of any preferred flexible hours<br>should be stated in the personal statement and will be discussed further at<br>interview.<br><br>Benefits include:<br><br>- *44 days annual leave* inclusive of Bank Holidays and University<br>Customary days (pro rata for part-time).<br>- *Family and Work-life balance policies *including hybrid working and<br>considerable maternity, paternity, shared parental leave and adoption<br>leave.<br>- *Generous pension scheme* with life cover for dependants, plus<br>incapacity cover.<br>- *Health and Wellbeing: *discounted access to Sportspark facilities,<br>relaxation rooms, 320 acres of rolling parkland, wellbeing walks, Wellbeing<br>Ambassador network, medical centre, Occupational Health and a 24/7 Employee<br>Assistance Programme.<br>- *Campus Facilities:* Sportspark, library, nursery, supermarket, post<br>office, bars and catering outlets.<br>- Exclusive shopping *discounts* to help cut the cost of household<br>bills, childcare salary sacrifice scheme, Cycle to Work scheme and public<br>transport discounts.<br>- *Personal Development: *unlimited access to LinkedIn Learning courses,<br>specialist advice and training from our Organisational Development and<br>Professional Learning Team.<br><br>*Closing date: 12 April 2024*<br><br>*We strongly encourage applicants from Black, Asian or other minority<br>ethnic backgrounds and welcome applications from all protected groups as<br>defined by the Equality Act 2010. Appointment will be made on merit.*<br><br>*The University holds an Athena Swan Silver Institutional Award in<br>recognition of our advancement towards gender equality.*<br>Further Information<br><br>For further information, including the Job Description and Person<br>Specification, please see the attached Candidate Brochure.<br><br>For an informal discussion about the post please contact the Head of<br>School, Professor Neil Cooper via neil.cooper@uea.ac.uk.Jamy Andy2024-03-18T10:29:32+00:002024-03-18T10:29:32+00:00parametric modulationshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;8fc3bed5.2403Dear all,<br>I was wondering if there is a state of the art procedure for subjects with no parametric modulation.<br>One subject has same ratings for one contrast (therefore no modulation).<br>Should the contrast be excluded, or can we add some slight modulation..<br>Or are there some polynomial expansion that can deal with such cases?<br>Thanks<br>J.A.Vladimir Litvak2024-03-18T09:30:58+00:002024-03-18T09:30:58+00:00Re: SPM EEG - Displaying contrast and T maps in SPM ?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;f2da4e85.2403Dear Benedetta,<br><br>I think the issue there was not that it was a t-map but that it was<br>thresholded so there was a sharp transition that the plotting tool didn't<br>handle well. With a continuous t-map there should be no problem. Thanks for<br>posting your tools.<br><br>Best,<br><br>Vladimir<br><br>On Mon, Mar 18, 2024 at 9:22 AM BENEDETTA CECCONI <bcecconi@wisc.edu> wrote:<br><br>> Dear Vladmir,<br>><br>> thank you for your answer.<br>><br>> I managed to plot multiple topographies (for con images) modifying the SPM<br>> function "spm12/toolbox/MEEGtools/spm_eeg_img2maps.m". In case it might be<br>> useful to others, I attach the modified function here, and an example<br>> script for plotting 16 topographies at specified time points, customizing<br>> colormap, style and scale.<br>><br>> Regarding the t-maps, I saw that in other thread you had recommended not<br>> to use the function spm_eeg_img2maps.m because of biased outputs:<br>> https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=spm;b735448.1202.. Is this<br>> still the case? If yes, I found this very nice and easy to use toolbox to<br>> plot t-maps. Again, I attach the links in case they may be useful to<br>> others: (<br>> https://link.springer.com/article/10.1007/s12021-019-09447-6#Sec24) and<br>> scripts:https://github.com/JeremyATaylor/Porthole<br>> <https://github.com/JeremyATaylor/Porthole><br>> GitHub - JeremyATaylor/Porthole<br>> <https://github.com/JeremyATaylor/Porthole><br>> Contribute to JeremyATaylor/Porthole development by creating an account on<br>> GitHub.<br>> github.com<br>><br>> <https://link.springer.com/article/10.1007/s12021-019-09447-6#Sec24><br>> Porthole and Stormcloud: Tools for Visualisation of Spatiotemporal M/EEG<br>> Statistics - Neuroinformatics<br>> <https://link.springer.com/article/10.1007/s12021-019-09447-6#Sec24><br>> Electro- and magneto-encephalography are functional neuroimaging<br>> modalities characterised by their ability to quantify dynamic<br>> spatiotemporal activity within the brain. However, the visualisation<br>> techniques used to illustrate these effects are currently limited to<br>> single- or multi-channel time series plots, topographic scalp maps and<br>> orthographic cross-sections of the spatiotemporal data structure. Whilst<br>> these methods each have their own strength and weaknesses, they are only<br>> able to show a subset of the data and are suboptimal at articulating one or<br>> both of the space-time components.Here, we propose Porthole and Stormcloud,<br>> a set of data visualisation tools which can automatically generate context<br>> appropriate graphics for both print and screen with the following graphical<br>> capabilities: Animated two-dimensional scalp maps with dynamic timeline<br>> annotation and optional user interaction; Three-dimensional construction of<br>> discrete clusters within sparse spatiotemporal volumes, rendered with<br>> ‘cloud-like’ appe<br>> link.springer.com<br>><br>> Best,<br>> Benedetta<br>><br>> ------------------------------<br>> *From:* Vladimir Litvak <litvak.vladimir@gmail.com><br>> *Sent:* Friday, March 15, 2024 5:19 AM<br>> *To:* BENEDETTA CECCONI <bcecconi@wisc.edu><br>> *Cc:* SPM@jiscmail.ac.uk <SPM@jiscmail.ac.uk><br>> *Subject:* Re: [SPM] SPM EEG - Displaying contrast and T maps in SPM ?<br>><br>> Dear Benedetta,<br>><br>><br>><br>> On Thu, Mar 14, 2024 at 3:42 PM BENEDETTA CECCONI <<br>> 00006e6f00386e18-dmarc-request@jiscmail.ac.uk> wrote:<br>><br>><br>><br>> 1. MEEG Tools > `Plot Scalp maps from M/EEG image` and input my<br>> contrast image* but I can't adjust the scale *(nor select<br>> multiple time points, but this is less of a problem)<br>><br>><br>> The function has an option S.clim that you can use to adjust the scale.<br>> You would need to run it from a script to specify that.<br>><br>><br>><br>> 1. for T maps, I opened the SPM.mat in Results and saved the<br>> thresholded image. I tried to open it with MEEG Tools > `Plot Scalp maps<br>> from M/EEG image` again but it didn't work. I then tried to collapse the<br>> thresholded t image in time using `Images > Collapse time` (input: original<br>> spmT-map) and then tried to input the output image in MEEG Tools > `Plot<br>> Scalp maps from M/EEG image` (selecting time 1 1) but it doesn't work...<br>><br>><br>><br>><br>> I'm not sure what you mean by 'didn't work'. If you send your error<br>> message I might be able to say more.<br>><br>> Best,<br>><br>> Vladimir<br>><br>><br>>Paparella Ilenia2024-03-18T09:29:59+00:002024-03-18T09:29:59+00:00CMM_NMMhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;916b65a9.2403Dear SPM experts,<br><br>We used the default CMM_NMDA model in SPM12 to model the first 50ms of a TMS-evoked response registered with EEG. With PEB we then tested which (if any) parameter of the model was related to an external variable.<br>I have issues interpreting the effect we see on the GABAa parameter as I do not fully understand what information it is conveying in the model:<br><br>1.<br>Is GABAa mediating all the inhibitory connections allowed in the model (thus both the inhibitory feedback loops and the projections from the inhibitory interneuron to the other subpopulations)?<br>2.<br>Is it mediating just the inhibitory connections from the inhibitory interneuron to the other subpopulations?<br>3.<br>Or is it conveying another type of information?<br><br>Thank you a lot in advance for your time and help.<br><br>Best,<br>Ilenia BENEDETTA CECCONI2024-03-18T09:22:38+00:002024-03-18T09:22:38+00:00Re: SPM EEG - Displaying contrast and T maps in SPM ?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;63f1c83f.2403Dear Vladmir,<br><br>thank you for your answer.<br><br>I managed to plot multiple topographies (for con images) modifying the SPM function "spm12/toolbox/MEEGtools/spm_eeg_img2maps.m". In case it might be useful to others, I attach the modified function here, and an example script for plotting 16 topographies at specified time points, customizing colormap, style and scale.<br><br>Regarding the t-maps, I saw that in other thread you had recommended not to use the function spm_eeg_img2maps.m because of biased outputs: https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=spm;b735448.1202.. Is this still the case? If yes, I found this very nice and easy to use toolbox to plot t-maps. Again, I attach the links in case they may be useful to others: (https://link.springer.com/article/10.1007/s12021-019-09447-6#Sec24) and scripts:https://github.com/JeremyATaylor/Porthole<br>[https://opengraph.githubassets.com/d44765b8123e0c83a8ae181f9f3735dda58dc72f409d278ad696fe609a3fa306/JeremyATaylor/Porthole]<https://github.com/JeremyATaylor/Porthole><br>GitHub - JeremyATaylor/Porthole<https://github.com/JeremyATaylor/Porthole><br>Contribute to JeremyATaylor/Porthole development by creating an account on GitHub.<br>github.com<br><br>[https://static-content.springer.com/image/art%3A10.1007%2Fs12021-019-09447-6/MediaObjects/12021_2019_9447_Fig1_HTML.png]<https://link.springer.com/article/10.1007/s12021-019-09447-6#Sec24><br>Porthole and Stormcloud: Tools for Visualisation of Spatiotemporal M/EEG Statistics - Neuroinformatics<https://link.springer.com/article/10.1007/s12021-019-09447-6#Sec24><br>Electro- and magneto-encephalography are functional neuroimaging modalities characterised by their ability to quantify dynamic spatiotemporal activity within the brain. However, the visualisation techniques used to illustrate these effects are currently limited to single- or multi-channel time series plots, topographic scalp maps and orthographic cross-sections of the spatiotemporal data structure. Whilst these methods each have their own strength and weaknesses, they are only able to show a subset of the data and are suboptimal at articulating one or both of the space-time components.Here, we propose Porthole and Stormcloud, a set of data visualisation tools which can automatically generate context appropriate graphics for both print and screen with the following graphical capabilities: Animated two-dimensional scalp maps with dynamic timeline annotation and optional user interaction; Three-dimensional construction of discrete clusters within sparse spatiotemporal volumes, rendered with ‘cloud-like’ appe<br>link.springer.com<br><br>Best,<br>BenedettaMichael Zyphur2024-03-18T19:05:00+11:002024-03-18T19:05:00+11:00Using ChatGPT and Copilot for Efficient Data Analysis in R - livestream seminarhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;11ec8e46.2403Hi everyone<br><br>Instats is incredibly pleased to be offering a 2-day workshop on Using<br>ChatGPT and Copilot for Efficient Data Analysis in R<br><https://instats.org/seminar/using-chatgpt-and-github-copilot-for-eff2>,<br>running March 20 - 21, by professor Peter Gruber (who holds dual PhDs in<br>physics and economics). This workshop provides a 21st-century introduction<br>to Statistical Analysis with R, focusing on the efficient use of AI<br>assistants including ChatGPT and Github Copilot to automate R coding with<br>plain language requests. Because R is free, this revolution will help<br>democratize access to basic and advanced analysis tools without having to<br>suffer the steep learning curve of coding in R. Participants will learn<br>step by step how to install AI tools and how to harness their power for<br>efficient data analysis in R, making them many times more efficient. They<br>will be able to create R code in the blink of an eye and with unprecedented<br>ease of use, while learning some of the underlying principles of the R<br>language as this relates to competently assessing and using AI-generated<br>code.<br><br>Register now<br><https://instats.org/seminar/using-chatgpt-and-github-copilot-for-eff2> and<br>don't miss out on this unique opportunity to learn how to easily and<br>rapidly code all of your analyses in R, and please feel free to tell your<br>friends and colleagues!<br><br>Best wishes<br><br>Michael Zyphur<br>Director<br>Institute for Statistical and Data Science<br>*instats.org* <http://instats.org><br><http://instats.org>Vadim Axelrod2024-03-17T17:32:56+02:002024-03-17T17:32:56+02:00MRI volume "origin": how it is defined during scanning ?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;24825219.2403Dear all,<br><br>Clicking on the "Origin" button in the display window brings the crosshair<br>to a coordinate which is a so-called origin. What is this "origin"<br>coordinate from the point of view of the scanning operator? What's the name<br>of this original coordinate? Maybe someone knows how to set its location<br>in 7T SIEMENS MAGNETOM Terra. I am asking because for different subjects I<br>get different origin locations. While the volumes can obviously be<br>reoriented in SPM, a more straightforward way would be to fix in the<br>location at the scanner side.<br><br>Thank you for the help.<br>VadimLuna Sato2024-03-17T09:50:28+01:002024-03-17T09:50:28+01:00Re: What is the DCM.U.idx parameter for?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;2394d8bd.2403Hi Peter,<br><br>Thank you!<br><br>Initially, I thought I understood the logic behind the numbers in the DCM.U.idx generated by the DCM GUI. The matrix looked like this:<br><br>DCM.U.idx=[1 1 2 1<br> 3 1<br> 4 1]<br><br>I interpreted it as follows: the first column ranging from 1 to 4 corresponds to the four experimental conditions, and the second column consists entirely of ones.<br><br>Based on this understanding, I defined all DCMs in my code using the same idx matrix. Namely, I applied this matrix to DCM.U.idx in all subjects' DCM files.<br><br>However, I later discovered that the DCM GUI generates the DCM.U.idx for another DCM (including parametric modulation regressors) in this format:<br><br>DCM.U.idx=[1 1<br> 1 2<br> 1 3<br> 1 4]<br><br>This discrepancy has left me uncertain whether my approach of defining the same matrix for all subjects' DCM.U.idx was incorrect.<br><br>Would defining the DCM.U.idx matrix incorrectly in the code have a significant impact on the results?<br><br>Cheers,Luna<br><br>12 Mar 2024, 18:07 by peter.zeidman@ucl.ac.uk:<br><br>><br>> Hi Luna<br>><br>><br>> This is nothing to worry about. If I remember correctly, I added that field to help with specifying DCMs using the batch editor. Its omission won’t alter your results.<br>><br>><br>> <br>><br>><br>> Best<br>><br>><br>> Peter<br>><br>><br>> <br>><br>><br>> From:> SPM (Statistical Parametric Mapping) <SPM@JISCMAIL.AC.UK> > On Behalf Of > Luna Sato<br>> > Sent:> 06 March 2024 03:23<br>> > To:> SPM@JISCMAIL.AC.UK<br>> > Subject:> [SPM] What is the DCM.U.idx parameter for?<br>><br>> <br>><br>><br>> ⚠> Caution: External sender<br>><br>><br>> <br>><br>><br>> Hi experts,<br>><br>><br>> <br>><br>><br>> When checking DCM results, I found certain subjects' DCM fields include the parameter DCM.U.idx, while others don't. I suspect this variation might be due to different SPM versions used.<br>><br>><br>> <br>><br>><br>> I'm wondering about the significance of DCM.U.idx. Can I combine subjects with and without this parameter in group analysis? Or should I consider redoing some DCM analyses?<br>><br>><br>> <br>><br>><br>> Best regards,<br>><br>><br>> Luna<br>><br>><br>> <br>><br>>Jia-Hou, Poh2024-03-15T17:42:11-04:002024-03-15T17:42:11-04:00Clinical Research Coordinator Position at Duke University - Adcock Labhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;ec1935e5.2403The *Motivated Memory Laboratory* <https://www.adcocklab.org/> in the<br>Center for Cognitive Neuroscience and Department of Psychiatry at Duke<br>University is seeking a full-time *Clinical Research Coordinator* with<br>strong interest in understanding how motivation shapes human experience to<br>work on an NIH-funded project (*PI: R. Alison Adcock*). This project aims<br>to uncover the neural mechanisms of motivation and its regulation using<br>fMRI neurofeedback and advanced statistical modeling.<br><br>Ongoing projects include: i) using real-time fMRI neurofeedback for<br>endogenous regulation of dopaminergic midbrain activity, ii) understanding<br>how midbrain neuromodulation contributes to downstream behavioral and<br>cognitive outcomes (e.g. memory and decision-making), iii) identifying how<br>different motivational states influence learning and memory formation.<br><br>Review of applications will start immediately and will continue until the<br>position is filled.<br><br>*Required qualifications*<br>Education: Completion of an Associate's degree<br>Experience: Work requires a minimum of two years of relevant research<br>experience. A Bachelor's degree may substitute for 2 years required<br>experience.<br><br>For inquiries please email: alison.adcock@duke.edu.<br><br>Please see full postings and apply using the following link:<br>Clinical Research Coordinator - Psychiatry - Beh Med Div - Adcock Team<br><https://careers.duke.edu/job-invite/242300/>Jia-Hou, Poh2024-03-15T17:42:03-04:002024-03-15T17:42:03-04:00Postdoctoral Position at Duke University - Adcock Labhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;72f9dec.2403*Postdoctoral Researcher in the Cognitive Neuroscience of Motivation and<br>Memory at Duke University*<br>The *Motivated Memory Laboratory* <https://www.adcocklab.org/> in the<br>Center for Cognitive Neuroscience and Department of Psychiatry at Duke<br>University is seeking a full-time Cognitive Neuroscience *Postdoctoral<br>Researcher* with strong interest in understanding how motivation shapes<br>human experience to join the team and lead an NIH-funded project (*PI: R.<br>Alison Adcock*). This project aims to uncover the neural mechanisms of<br>motivation and its regulation using fMRI neurofeedback and advanced<br>statistical modeling.<br><br>Ongoing projects include: i) using real-time fMRI neurofeedback for<br>endogenous regulation of dopaminergic midbrain activity, ii) understanding<br>how midbrain neuromodulation contributes to downstream behavioral and<br>cognitive outcomes (e.g. memory and decision-making), iii) identifying how<br>different motivational states influence learning and memory formation.<br><br>Review of applications will start immediately and will continue until the<br>position is filled. Salary is commensurate with NIH guidelines and<br>applicant experience.<br><br>*Required qualifications*<br>Candidates must have a PhD in Psychology, Cognitive Science, Neurobiology,<br>Neuroscience, or a related field by the position start date.<br><br>*Required skills and experience*<br><br>- Experience collecting and analyzing behavioral and fMRI data<br>(preferably in humans)<br>- Ability to work with and provide guidance to junior lab members and<br>trainees<br>- Ability to work effectively with mentor(s) to conduct and implement<br>research projects<br>- Demonstrated ability to conduct independent research - including<br>formulating hypotheses, designing experiments, collecting data, analyzing<br>data, and communication of results<br>- Demonstrated ability to independently write manuscripts and grant<br>application with guidance from mentor(s)<br>- Experience with programming and troubleshooting experimental tools -<br>preferably in Matlab and/or Python (e.g. for stimulus presentation, data<br>cleaning, computational modeling)<br><br>*Strongly preferred but not required qualifications*<br><br>- Experience working with real-time fMRI and/or neuro/biofeedback in<br>other modalities (e.g. EEG, Pupillometry, HRV)<br>- Experience using advanced quantitative method for analysis of<br>behavioral and imaging data (e.g. computational modeling, machine learning,<br>mixed-models)<br>- Strong background in neuroanatomy and neurobiology<br>- Interest in translating basic science towards clinical applications<br><br>For inquiries please email: alison.adcock@duke.edu.<br><br>Please see full postings and apply using the following link:<br>Postdoctoral Associate - Psychiatry -<br><http://careers.duke.edu/job/Durham-Postdoctoral-Associate-Psychiatry-Behavioral-Adcock-Team-NC-27710/1135094700/?from=email&refid=22020353400&utm_source=J2WEmail&source=2&eid=148200-202424230624-28752723000&locale=en_US>Behavioral<br>- Adcock Team<br><http://careers.duke.edu/job/Durham-Postdoctoral-Associate-Psychiatry-Behavioral-Adcock-Team-NC-27710/1135094700/?from=email&refid=22020353400&utm_source=J2WEmail&source=2&eid=148200-202424230624-28752723000&locale=en_US>Vladimir Litvak2024-03-15T16:28:08+00:002024-03-15T16:28:08+00:00Re: DCM after EEG source reconstruction informed by fMRI priorshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;1922268e.2403Dear Julia,<br><br>On Fri, Mar 15, 2024 at 4:20 PM Júlia Soares <julii.f.soares@gmail.com><br>wrote:<br><br>> 1) Regarding the source locations in MNI I didn't quite understand why<br>> this doesn't matter. The DCM model requires an EEG signal after source<br>> reconstruction, right? So the space will be the source space instead of the<br>> sensor space (the actual electrodes), right? If so, how come the resolution<br>> of the coordinates doesn't make a difference? Isn't it possible that<br>> sources are several mm misaligned with corresponding locations in MNI ?<br>><br><br>The kind of differences in source locations that make a difference in EEG<br>are on the order of cm so if your native images are coregistered to MNI and<br>the head sizes are not unusually large or small I wouldn't expect the mm<br>differences to matter. But you can always write out the results of your<br>source analysis in the native space on a template MNI mesh and then you<br>won't have that problem at all.<br><br>> 2) About data epoching: I have a continuous signal acquired during<br>> performance of a task constituted by 4 conditions: 8 periods of "baseline"<br>> (22 seconds), 5 periods of "condition A" (18 seconds), 4 periods of<br>> "condition B" (18 seconds) and 3 periods of "condition C" (18 seconds). I<br>> was thinking about separating my continuous signal into epochs of equal<br>> length to the periods of each condition. So, for example for "condition A"<br>> I would have 5 epochs of 18 seconds each corresponding to "condition A "<br>> which would then be averaged into one single epoch. Does this make sense?<br>><br>><br>The implementation assumes short epochs 1-2 sec at most so I'd suggest you<br>epoch your conditions into epochs of that length and then the differences<br>in duration won't matter.<br><br>Best,<br><br>Vladimir<br><br>> Regards,<br>> Júlia Soares<br>><br>><br>><br>> Em ter., 12 de mar. de 2024 às 15:24, Vladimir Litvak <<br>> litvak.vladimir@gmail.com> escreveu:<br>><br>>> Dear Julia,<br>>><br>>> On Tue, Mar 12, 2024 at 2:55 PM Júlia Soares <julii.f.soares@gmail.com><br>>> wrote:<br>>><br>>>> 1) Is it only possible to do DCM in MNI space since the prior source<br>>>> locations should be given in MNI coordinates or is it possible to conduct<br>>>> DCM analysis in native space for each specific subject ?<br>>>><br>>>><br>>> I think this distinction is too fine to matter for DCM if you are doing<br>>> it at the sensor level. So I'd just define source locations in MNI space<br>>> and not worry too much about it.<br>>><br>>><br>>><br>>>> 2) In source reconstruction I inverted a continuous signal, i.e., I did<br>>>> not separate the signal into epochs (trials). However I have a task which<br>>>> has 3 conditions in which I intend to study connectivity in each of them.<br>>>> Is there a way to separate my signal after source reconstruction so I can<br>>>> include them in the DCM model?<br>>>><br>>><br>>> Both source analysis and DCM were not intended to work on long continuous<br>>> data segments. I'd suggest you epoch your data into arbitrary 1-2 sec<br>>> epochs. There is a way to do it in the epoching tool. Then I would do both<br>>> steps on these epoched data.<br>>><br>>> Best,<br>>><br>>> Vladimir<br>>><br>>><br>>><br>>>><br>>>> Thank you in advance.<br>>>> Regards,<br>>>> Júlia Soares<br>>>><br>>>Júlia Soares2024-03-15T16:19:58+00:002024-03-15T16:19:58+00:00Re: DCM after EEG source reconstruction informed by fMRI priorshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;9dcdb947.2403Dear Vladimir,<br><br>First of all, thank you for your prompt response and valuable suggestions.<br><br>1) Regarding the source locations in MNI I didn't quite understand why this<br>doesn't matter. The DCM model requires an EEG signal after source<br>reconstruction, right? So the space will be the source space instead of the<br>sensor space (the actual electrodes), right? If so, how come the resolution<br>of the coordinates doesn't make a difference? Isn't it possible that<br>sources are several mm misaligned with corresponding locations in MNI ?<br><br>2) About data epoching: I have a continuous signal acquired during<br>performance of a task constituted by 4 conditions: 8 periods of "baseline"<br>(22 seconds), 5 periods of "condition A" (18 seconds), 4 periods of<br>"condition B" (18 seconds) and 3 periods of "condition C" (18 seconds). I<br>was thinking about separating my continuous signal into epochs of equal<br>length to the periods of each condition. So, for example for "condition A"<br>I would have 5 epochs of 18 seconds each corresponding to "condition A "<br>which would then be averaged into one single epoch. Does this make sense?<br><br>Regards,<br>Júlia Soares<br><br>Em ter., 12 de mar. de 2024 às 15:24, Vladimir Litvak <<br>litvak.vladimir@gmail.com> escreveu:<br><br>> Dear Julia,<br>><br>> On Tue, Mar 12, 2024 at 2:55 PM Júlia Soares <julii.f.soares@gmail.com><br>> wrote:<br>><br>>> 1) Is it only possible to do DCM in MNI space since the prior source<br>>> locations should be given in MNI coordinates or is it possible to conduct<br>>> DCM analysis in native space for each specific subject ?<br>>><br>>><br>> I think this distinction is too fine to matter for DCM if you are doing it<br>> at the sensor level. So I'd just define source locations in MNI space and<br>> not worry too much about it.<br>><br>><br>><br>>> 2) In source reconstruction I inverted a continuous signal, i.e., I did<br>>> not separate the signal into epochs (trials). However I have a task which<br>>> has 3 conditions in which I intend to study connectivity in each of them.<br>>> Is there a way to separate my signal after source reconstruction so I can<br>>> include them in the DCM model?<br>>><br>><br>> Both source analysis and DCM were not intended to work on long continuous<br>> data segments. I'd suggest you epoch your data into arbitrary 1-2 sec<br>> epochs. There is a way to do it in the epoching tool. Then I would do both<br>> steps on these epoched data.<br>><br>> Best,<br>><br>> Vladimir<br>><br>><br>><br>>><br>>> Thank you in advance.<br>>> Regards,<br>>> Júlia Soares<br>>><br>>Vladimir Litvak2024-03-15T10:19:04+00:002024-03-15T10:19:04+00:00Re: SPM EEG - Displaying contrast and T maps in SPM ?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;68a4e896.2403Dear Benedetta,<br><br>On Thu, Mar 14, 2024 at 3:42 PM BENEDETTA CECCONI <<br>00006e6f00386e18-dmarc-request@jiscmail.ac.uk> wrote:<br><br>><br>><br>> 1. MEEG Tools > `Plot Scalp maps from M/EEG image` and input my<br>> contrast image* but I can't adjust the scale *(nor select<br>> multiple time points, but this is less of a problem)<br>><br>><br>The function has an option S.clim that you can use to adjust the scale. You<br>would need to run it from a script to specify that.<br><br>><br>> 1. for T maps, I opened the SPM.mat in Results and saved the<br>> thresholded image. I tried to open it with MEEG Tools > `Plot Scalp maps<br>> from M/EEG image` again but it didn't work. I then tried to collapse the<br>> thresholded t image in time using `Images > Collapse time` (input: original<br>> spmT-map) and then tried to input the output image in MEEG Tools > `Plot<br>> Scalp maps from M/EEG image` (selecting time 1 1) but it doesn't work...<br>><br>><br>><br><br>I'm not sure what you mean by 'didn't work'. If you send your error message<br>I might be able to say more.<br><br>Best,<br><br>Vladimir<br><br>>Zeidman, Peter2024-03-14T16:26:45+00:002024-03-14T16:26:45+00:00Open Science Room at OHBM 2024https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;3bafa25b.2403Full message available at: <a href="https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;3bafa25b.2403">Open Science Room at OHBM 2024</a>BENEDETTA CECCONI2024-03-14T15:42:47+00:002024-03-14T15:42:47+00:00SPM EEG - Displaying contrast and T maps in SPM ?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;94e35393.2403Hello everyone,<br><br>I would like to display the results of my group analysis by showing contrast maps and thresholded t maps, something like this:<br><br>[cid:04ca0dfa-84fa-450b-bc8a-bfcce5029ce0]<br><br>I tried<br><br>1.<br>MEEG Tools > `Plot Scalp maps from M/EEG image` and input my contrast image but I can't adjust the scale (nor select multiple time points, but this is less of a problem)<br><br>2.<br>for T maps, I opened the SPM.mat in Results and saved the thresholded image. I tried to open it with MEEG Tools > `Plot Scalp maps from M/EEG image` again but it didn't work. I then tried to collapse the thresholded t image in time using `Images > Collapse time` (input: original spmT-map) and then tried to input the output image in MEEG Tools > `Plot Scalp maps from M/EEG image` (selecting time 1 1) but it doesn't work...<br><br>Can someone please help me?<br><br>Any suggestions are greatly appreciated.<br><br>Thanks a lot in advance,<br><br>Benedetta Criscuolo, Antonio (PSYCHOLOGY)2024-03-14T15:17:40+00:002024-03-14T15:17:40+00:00Body-Brain Waves: Neuroscience events in southern Italyhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;cd3aad4c.2403Dear colleagues,<br><br>With this email, I'd like to bring to your attention Waves’24<http://waves-conference.com>.<br><br>Are you interested in brain, respiratory, cardiovascular, oculomotor and/or gastrointestinal signals?<br><br>After the success of the first Body-Brain Waves<https://waves-conference.com/373-2/waves-23/> conference, this year we are planning a 4-day Summer School, followed by a 2-day Symposium.<br><br>Waves’24<https://waves-conference.com/> takes place in Salerno<https://www.google.com/search?sca_esv=23e333b9b80ac6d8&rlz=1C1PRFI_enNL888NL888&sxsrf=ACQVn09G6TvG74jIEW4fwftrawAyqAes4w:1707127739095&q=salerno&uds=AMwkrPuANOudBFD0X4ERrATo2Mm2JH8T11cwFXV0RzxWGAKrtKKJ25Jk4baHHXzQCHa8RWWUYlbN3cTieUIda9QMyvGI_t9_nU8cp2Tnf0Etmj0DGaj3gZU&udm=2&sa=X&ved=2ahUKEwiuzOSz-pOEAxVCxQIHHSZ0CAMQtKgLegQICRAB&biw=1920&bih=929&dpr=1> <https://www.google.com/search?sca_esv=23e333b9b80ac6d8&rlz=1C1PRFI_enNL888NL888&sxsrf=ACQVn09G6TvG74jIEW4fwftrawAyqAes4w:1707127739095&q=salerno&uds=AMwkrPuANOudBFD0X4ERrATo2Mm2JH8T11cwFXV0RzxWGAKrtKKJ25Jk4baHHXzQCHa8RWWUYlbN3cTieUIda9QMyvGI_t9_nU8cp2Tnf0Etmj0DGaj3gZU&udm=2&sa=X&ved=2ahUKEwiuzOSz-pOEAxVCxQIHHSZ0CAMQtKgLegQICRAB&biw=1920&bih=929&dpr=1> (my hometown in southern Italy) from the 23rd to 28th September 2024.<br><br>1. Summer School: 23-26 September 2024;<br>2. Symposium: 27-28 September 2024.<br><br>At the core of both events, the following questions:<br><br>1. How to best measure Body-Brain physiological signals?<br>2. How to best investigate Body-Brain interactions and their influence on cognition?<br><br>Preliminary program<br><br>* During the Summer School, young and emerging early career researchers will offer several courses and workshops on how to best acquire and analyze bodily physiological signals next to brain activity and/or behavior.<br><br>Participants will learn best practices in acquiring respiratory, cardiovascular, oculomotor and gastrointestinal signals, and how to holistically examine dynamic Body-Brain interactions and their modulatory role in cognition. Real hands-on data!<br><br>* The Summer School will be followed by a 2-day Symposium featuring talks and workshops on Body-Brain Waves and mobile body-brain imaging applications.<br><br>At the end of each day, social activities such as aperitivo on the seaside, cruise to Amalfi<https://www.google.com/search?q=amalfi&rlz=1C1CHBF_itNL979NL979&sxsrf=ALiCzsbUyyoZfyx72u2PsKC13Cm8CEd0cA:1663767583484&source=lnms&tbm=isch&sa=X&ved=2ahUKEwjrn7Wegab6AhXPzqQKHZJ6CMkQ_AUoAnoECAIQBA&biw=1666&bih=943&dpr=1.8> and visit of the Greek temples in Paestum<https://www.google.com/search?q=paestum&tbm=isch&ved=2ahUKEwj9ks-ngab6AhUt57sIHX6IC1oQ2-cCegQIABAA&oq=paestum&gs_lcp=CgNpbWcQAzIECAAQQzIFCAAQgAQyBAgAEEMyBAgAEEMyBAgAEEMyBQgAEIAEMgUIABCABDIFCAAQgAQyBAgAEEMyBQgAEIAEOgQIIxAnUPYEWJAMYKUNaABwAHgAgAFqiAH7BJIBAzUuM5gBAKABAaoBC2d3cy13aXotaW1nwAEB&sclient=img&ei=MhQrY_2xL63O7_UP_pCu0AU&bih=943&biw=1666&rlz=1C1CHBF_itNL979NL979>.<br><br>If this initiative is of interest to you and colleagues, please spread the info and apply following these<https://waves-conference.com/#applications> instructions.<br>Please, note that there are separate applications for the Summer School and Symposium.<br><br>Don't hesitate to get in touch for any doubts and questions.<br><br>Looking forward to welcoming you at Waves'24.<br><br>Thank you in advance for attention and for helping us spreading the email and info around.<br><br>Have a nice day,<br><br>Antonio Criscuolo<br><br>Faculty of Psychology and Neuroscience<br><br>Dept NP&PP, Maastricht University<br><br>Office: UNS40 2.749<br><br>Web: band-lab.com<http://www.band-lab.com/><br><br>[cid:ef3fc0eb-175f-4582-9209-c675a16df60b]<http://waves-conference.com> Chen Yu2024-03-13T23:38:04+00:002024-03-13T23:38:04+00:00Exciting Postdoctoral Position Available at UCSF's Tee Lab - Neuroimaging Researchhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;c0ea8175.2403We are thrilled to announce an opening for a Postdoctoral Position at the Tee Lab, Department of Neurology, University of California, San Francisco (UCSF). This is a unique opportunity for a passionate researcher to contribute to groundbreaking work in the field of neuroimaging and neurodegenerative diseases.<br><br>Job Summary: The selected candidate will participate in projects that focus on establishing MRI protocols with cutting-edge methodologies and applying neuroimaging techniques, such as tractography, graph theory analysis, and machine learning algorithms. Our focus spans a range of neurodegenerative diseases, including Alzheimer’s disease, primary progressive aphasia, and frontotemporal dementia, particularly in the context of diverse populations and bilingualism.<br><br>Laboratory Mission: The Tee Lab is dedicated to promoting equal representation in cognitive and dementia research, enhancing our understanding of brain aging and neurodegenerative diseases. We collaborate internationally, aiming to understand bilingualism and dementia syndromes and promote language diversity in cognitive research.<br><br>Required Qualifications:<br><br>* Ph.D. in neurology, radiology, neuropsychology, cognitive neuroscience, biomedical engineering, or related fields.<br>* Experience with MRI/PET data analysis.<br>* Proficiency in neuroimaging tools and programming/scripting languages.<br><br>Preferred Qualifications:<br><br>* Experience in connectomics, tractography, functional MRI, bilingualism, and cross-linguistic studies.<br>* Proficiency in Mandarin and/or Cantonese is a plus.<br><br>Application Process: Interested candidates should send a cover letter, CV, and contact information for three references to BoonLead.Tee@ucsf.edu<mailto:BoonLead.Tee@ucsf.edu> and Stephanie.Kwan3@ucsf.edu<mailto:Stephanie.Kwan3@ucsf.edu>.<br><br>For more information about the Tee Lab and our projects, please visit our website: Tee Lab - UCSF<https://teelab.ucsf.edu/><br><br>To view the detailed job posting, please visit: Tee Lab Open Positions - UCSF<https://teelab.ucsf.edu/open-positions><br><br>We look forward to welcoming a new member to our dynamic team! Cryns, Noah2024-03-13T15:51:56+00:002024-03-13T15:51:56+00:00UCSF Postdoctoral Fellowship Positionhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;20fa36d4.2403Dr. David Perry (https://perrylab.ucsf.edu/) is now inviting applications for a NIH-funded postdoctoral fellowship position in his lab at the University of California, San Francisco (UCSF) Memory and Aging Center. The goal of our lab’s research is to elucidate brain-behavior relationships in neurodegenerative disease in order to improve diagnostic certainty and identify therapeutic targets.<br><br>The UCSF Memory and Aging Center (memory.ucsf.edu<https://memory.ucsf.edu/>) is part of the Department of Neurology and Weill Institute for Neurosciences. It has an extensive research infrastructure, with over 250 full-time research faculty and staff. The postdoctoral fellow will have the opportunity to participate in our innovative, interdisciplinary research environment. We are looking for candidates who have a background in neuroimaging, strong statistical training, and programming experience. The start date is flexible; review of applications is ongoing. Applicants should send a brief cover letter describing interests and relevant prior experience, CV, and contact information for three references to (david.perry@ucsf.edu).<br><br>The postdoctoral fellow will work on our lab's study investigating abnormalities in reward processing in neurodegenerative diseases and mood disorders. Reward processing involves a determination of what an individual will work for or pursue, such as food, money, or social approval. Patients with neurodegenerative and mood disorders have profound changes in their reward valuation. We propose that a greater understanding of reward-seeking behavior in these illnesses and their underlying neural mechanisms will improve diagnostic accuracy and lead to therapeutic targets for behavioral symptoms that currently have no adequate treatment. Our studies of reward processing use behavioral paradigms with tools such as psychophysiology, as well as structural and functional neuroimaging.<br><br>Noah Cryns | Assistant Clinical Research Coordinator<br><br>Memory and Aging Center<br><br>University of California – San Francisco<br><br>Phone: (415) 514-7580<br><br>https://perrylab.ucsf.edu/<br><br>https://decisionlab.ucsf.edu/<br><br>[cid:3f0db6cf-1e1f-4894-8797-f6d565b85484] Topper, Mackenzie2024-03-12T16:13:37-04:002024-03-12T16:13:37-04:00Postdoctoral Position on PET and structural MRI/DTI at Brown Universityhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;3c533532.2403Position: Postdoctoral Position in Neuroimaging Research on Brain<br>Aging and Alzheimer’s disease<br><br>Area: Brain Aging and Alzheimer’s disease<br><br>Department: Department of Psychiatry and Human Behavior<br><br>Institution: Brown University<br><br>We are seeking a highly motivated postdoctoral researcher to lead an<br>NIH-funded project in the Laboratory for Cognitive and Translational<br>Neuroscience (Director: Dr. Hwamee Oh) at Brown University. The<br>project is to use amyloid and tau PET and structural MRI/DTI to study<br>the impact of Alzheimer’s disease pathologies on structural and<br>functional networks among cognitively normal older adults and patients<br>with cognitive impairment.<br><br>Applicants with a Ph.D. degree in cognitive neuroscience,<br>computational neuroscience, biomedical engineering, or a relevant field<br>are encouraged to apply. Prior experience with human neuroimaging in<br>PET and/or structural MRI, and familiarity with programming in relevant<br>languages (e.g., Python, MATLAB) are required. Applicants with<br>experience with amyloid and tau PET or diffusion weighted imaging are<br>especially encouraged to apply. The successful candidate is expected<br>to demonstrate communication skills, motivation and interest in the<br>area of neuroscience of cognitive and brain aging and Alzheimer’s<br>disease, and the ability to independently develop research questions<br>and work in collaboration with other team members.<br><br>A preferred start date is Summer 2024, although it can be flexible. The<br>position is primarily affiliated with the Department of Psychiatry and<br>Human Behavior at Brown University.<br><br>Interested applicants should email a brief description of research<br>background and career goals, and CV with a list of 3 references to Dr.<br>Hwamee Oh at hwamee_oh@brown.edu.<br><br>Contact Website:<br>https://sites.brown.edu/oh-ctnlab/available-positions/<br><br>--<br>Mackenzie Topper<br>Research Assistant & Lab Manager<br>Oh Lab<br>Brown UniversityVladimir Litvak2024-03-12T15:24:47+00:002024-03-12T15:24:47+00:00Re: DCM after EEG source reconstruction informed by fMRI priorshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;b473c382.2403Dear Julia,<br><br>On Tue, Mar 12, 2024 at 2:55 PM Júlia Soares <julii.f.soares@gmail.com><br>wrote:<br><br>> 1) Is it only possible to do DCM in MNI space since the prior source<br>> locations should be given in MNI coordinates or is it possible to conduct<br>> DCM analysis in native space for each specific subject ?<br>><br>><br>I think this distinction is too fine to matter for DCM if you are doing it<br>at the sensor level. So I'd just define source locations in MNI space and<br>not worry too much about it.<br><br>> 2) In source reconstruction I inverted a continuous signal, i.e., I did<br>> not separate the signal into epochs (trials). However I have a task which<br>> has 3 conditions in which I intend to study connectivity in each of them.<br>> Is there a way to separate my signal after source reconstruction so I can<br>> include them in the DCM model?<br>><br><br>Both source analysis and DCM were not intended to work on long continuous<br>data segments. I'd suggest you epoch your data into arbitrary 1-2 sec<br>epochs. There is a way to do it in the epoching tool. Then I would do both<br>steps on these epoched data.<br><br>Best,<br><br>Vladimir<br><br>><br>> Thank you in advance.<br>> Regards,<br>> Júlia Soares<br>>Júlia Soares2024-03-12T14:54:31+00:002024-03-12T14:54:31+00:00DCM after EEG source reconstruction informed by fMRI priorshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;8c87c916.2403Dear SPM experts,<br><br>I performed EEG source reconstruction informed by fMRI priors. I intend to<br>use this signal to do DCM analysis however, after reading the section<br>"Dynamic Causal Modelling for M/EEG" in SPM manual I have a few questions I<br>hope you can clarify:<br><br>1) Is it only possible to do DCM in MNI space since the prior source<br>locations should be given in MNI coordinates or is it possible to conduct<br>DCM analysis in native space for each specific subject ?<br><br>2) In source reconstruction I inverted a continuous signal, i.e., I did not<br>separate the signal into epochs (trials). However I have a task which has 3<br>conditions in which I intend to study connectivity in each of them. Is<br>there a way to separate my signal after source reconstruction so I can<br>include them in the DCM model?<br><br>Thank you in advance.<br>Regards,<br>Júlia SoaresZeidman, Peter2024-03-12T08:07:21+00:002024-03-12T08:07:21+00:00Re: What is the DCM.U.idx parameter for?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;c3fd8a86.2403Hi Luna<br>This is nothing to worry about. If I remember correctly, I added that field to help with specifying DCMs using the batch editor. Its omission won’t alter your results.<br><br>Best<br>Peter<br><br>From: SPM (Statistical Parametric Mapping) <SPM@JISCMAIL.AC.UK> On Behalf Of Luna Sato<br>Sent: 06 March 2024 03:23<br>To: SPM@JISCMAIL.AC.UK<br>Subject: [SPM] What is the DCM.U.idx parameter for?<br><br>⚠ Caution: External sender<br><br>Hi experts,<br><br>When checking DCM results, I found certain subjects' DCM fields include the parameter DCM.U.idx, while others don't. I suspect this variation might be due to different SPM versions used.<br><br>I'm wondering about the significance of DCM.U.idx. Can I combine subjects with and without this parameter in group analysis? Or should I consider redoing some DCM analyses?<br><br>Best regards,<br>LunaZeidman, Peter2024-03-12T08:03:19+00:002024-03-12T08:03:19+00:00Re: Searching over PEB models in DCMhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;41c7e002.2403Dear Lukas<br>Yes you’ve done this correctly. The A-matrix represents the average connectivity over conditions if you mean-centre your design matrix (in DCM for fMRI, DCM.options.centre=true), otherwise it represents the baseline or intercept. The B-matrix add to this for each experimental condition. So a value of zero in the A-matrix, and a non-zero value in the B-matrix, means that the connectivity was zero on average (or at baseline), but there was a difference between experimental conditions.<br><br>All the best<br>Peter<br><br>From: SPM (Statistical Parametric Mapping) <SPM@JISCMAIL.AC.UK> On Behalf Of Lorentz, Lukas Kay<br>Sent: 06 March 2024 10:06<br>To: SPM@JISCMAIL.AC.UK<br>Subject: [SPM] Searching over PEB models in DCM<br><br>⚠ Caution: External sender<br><br>Dear experts,<br><br>I am currently employing DCM on a task-based dataset and ran into a problem when interpreting the results. As a framework, we are following the two Zeidman et al. papers from 2019 on Parametric Empirical Bayes (https://doi.org/10.1016/j.neuroimage.2019.06.031 and https://doi.org/10.1016/j.neuroimage.2019.06.032).<br><br>For inference regarding our modulatory inputs (B matrix), we conducted an automatic "search over reduced PEB models" as described in section 4.7 in the 2nd Zeidman paper. This yielded significant results for several connections that were very much in line with our hypotheses.<br>However, when we then conducted another automatic search on our A matrix to derive average connectivity parameters (described in section 4.8 in 2nd Zeidman paper), we found that two connections were pruned away, even though our first analysis suggests that these connections would be significantly modulated by specific conditions.<br><br>Can anyone explain to me how to interpret this? Is searching over reduced A matrix models the correct way to estimate average effective connectivity through Bayesian Model Averaging?<br><br>Best regards,<br>LukasThomas HINAULT2024-03-11T19:44:32+00:002024-03-11T19:44:32+00:00Postdoc on multimodal integration (TEP-IRM-EEG) and computational modelling : • INSERM U1077 – Brain imaging center, Cyceron, Caen, Francehttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;e94749f0.2403Duration: Two years (renewable)<br><br>The INSERM-UNICAEN-EPHE U1077 research unit based in Caen (Normandy, France, https://nimh.unicaen.fr) invites applications for a position as a Postdoc in the field of brain aging and time processing.<br><br>The recruited research will join the TIMES research program “Time processing changes with aging”, led by Dr. Thomas Hinault (Ph.D). TIMES aims to understand the cognitive and neural mechanisms of temporal cognition and their evolution with aging.<br><br>The Unit 1077 and Cyceron Neuroimaging Platform offer an exciting and friendly multi-disciplinary research environment, with ample opportunities for training and collaboration, and excellent technical facilities. Cyceron is a structure devoted to multimodal imaging (pre- clinical and clinical) and provides a stimulating work environment as it groups several research units and several research instruments, such as a cyclotron for molecular marking, 2 PET-CT, 2 MRI (including a brand-new GE 3T), and a molecular and cellular imaging department. Caen is a friendly environment with an excellent work-life balance. We are located 12 km away from the Normandy coast and beaches. Caen is a young and vibrant city with many venues for music and culture.<br><br>We are looking for a postdoc with a strong expertise in neuroimaging (EEG, PET, MRI data, at least two of these methods), multiscale brain modelling (knowledge about The Virtual Brain is recommended), and excellent programming skills.<br><br>Review of applications will continue until the position is filled.<br><br>Salary: approximately 2900 euros (gross salary i.e. salary before taxes) per month<br><br>Application (motivation letter + CV + at least one recommendation letter) should be sent to Dr. Thomas Hinault (thomas.hinault@inserm.fr<mailto:thomas.hinault@inserm.fr>). Ashburner, John2024-03-11T13:58:46+00:002024-03-11T13:58:46+00:00Re: Normalization quality assessmenthttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;9aec1fc4.2403Unfortunately, there is no automated way that I can think of for doing this. Approaches for validating spatial normalisation methods normally involve manually defining regions on images and then assessing how closely these regions can be made to overlap using the estimated warps.<br><br>Perhaps you could identify an appropriate method from the literature. The Klein et al paper might be a good place to start, but remember that the technology has progressed in the last 15 years:<br>Klein, A., Andersson, J., Ardekani, B.A., Ashburner, J., Avants, B., Chiang, M.C., Christensen, G.E., Collins, D.L., Gee, J., Hellier, P. and Song, J.H., 2009. Evaluation of 14 nonlinear deformation algorithms applied to human brain MRI registration. Neuroimage, 46(3), pp.786-802.<br><br>This paper includes some more recent comparisons among methods:<br>Brudfors, Mikael, Yaël Balbastre, Guillaume Flandin, Parashkev Nachev, and John Ashburner. "Flexible Bayesian modelling for nonlinear image registration." In Medical Image Computing and Computer Assisted Intervention–MICCAI 2020: 23rd International Conference, Lima, Peru, October 4–8, 2020, Proceedings, Part III 23, pp. 253-263. Springer International Publishing, 2020.<br><br>Some people use the mean squares difference between aligned images to assess performance. By itself, this is not a good way to do things.<br>Rohlfing, T., 2011. Image similarity and tissue overlaps as surrogates for image registration accuracy: widely used but unreliable. IEEE transactions on medical imaging, 31(2), pp.153-163.<br><br>Best regards,<br>-JohnLisa Jeschke2024-03-11T13:15:28+00:002024-03-11T13:15:28+00:00Group contrast distorted after mni resamplinghttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;a33f4dc5.2403Dear experts,<br><br>I currently have an issue with my 2nd level analyses of fmri data of a brain slab.<br><br>I have run my single-subject/1st level analysis all in native space, which worked fine. We then used ANTs to morph all the contrast files into mni space, and those resampled contrast files were then used to calculate our group contrast.<br><br>Transforming the native contrast files into mni-space worked just fine and the native-space as well as mni-space contrasts all look correct. We did not observe any resampling issues with ANTs there. However, when we the use SPM to calculate the 2nd level/group analyses with the mni-contrasts, the result looks very noisy and not identifiable.<br><br>For demonstration purposes, I have attached an image of a dummy-like group contrast calculated of three contrast files of the same subject in native space (A) and then the group contrast with the same files that were transformed into mni space (B). As you can see, the voxels look much smaller and frizzy.<br><br>It would be great if anyone could give me some advice on how to deal with this issue - or ideally check two example files to identify what changed in those aside from morphing the blobs into mni space.<br><br>Thank you very much for your help and advice in advance!<br><br>Best,<br><br>Lisa<br><br>Lisa Jeschke<br>PhD student<br>lisa.jeschke@tu-dresden.de<br>+49 351 463-43897<br><br>Cognitive and Clinical Neuroscience / Professur für Kognitive und Klinische Neurowissenschaft<br>TUD Dresden University of Technology<br>https://tu-dresden.de/mn/psychologie/ifap/kknw Lei Zhang2024-03-11T12:29:52+00:002024-03-11T12:29:52+00:00[reminder] Birmingham-Leiden Summer School in Computational Social Cognition 2024https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;679ee9f9.2403(Apologies for cross-posting)<br><br>FINAL REMINDER:<br><br>We are delighted to welcome applications for the first edition of the<br>Birmingham-Leiden<br>summer school in Computational Social Cognition (CSC).<br><https://www.compsoccog.com/> The summer school is hosted at the University<br>of Birmingham (UK), in collaboration with Leiden University (NL), and will<br>take place from 15th-17th July 2024. Apply by *14th March 2024* (see<br>information below)!<br><br>Attending the Birmingham-Leiden CSC Summer School will equip a diverse<br>cohort of early career researchers (trainees through to junior faculty<br>members) with the ability to understand, program and interpret the output<br>of a range of computational models of social cognition. Attendees will<br>receive different types of training aimed at understanding modelling as<br>well as the theoretical and practical inferences that can be drawn from<br>computational models. For more information about this year’s training<br>program, including criteria and application instructions, please visit our<br>website <https://www.compsoccog.com/>. The deadline is March 14, 2024. We<br>hope to see you this summer in Birmingham!<br><br>Keynote speakers:<br><br>- Cecilia Heyes, University of Oxford (UK)<br><br>- Christian Ruff, University of Zurich (CH)<br><br>- Wolfram Schultz, University of Cambridge (UK)<br><br>Instructors (alphabetical):<br><br>- Matt Apps, University of Birmingham (UK)<br><br>- Jo Cutler, University of Birmingham (UK)<br><br>- Anna Van Duijvenvoorde, Leiden University (NL)<br><br>- Romy Froemer, University of Birmingham (UK)<br><br>- Arkady Konovalov, University of Birmingham (UK)<br><br>- Patricia Lockwood, University of Birmingham (UK)<br><br>- Ili Ma, Leiden University (NL)<br><br>- Lei Zhang, University of Birmingham (UK)<br><br>Anna & Lei, on behalf of the CSC 2024 organization team<br><br>---<br>Dr. Lei Zhang<br>Associate Professor<br>Centre for Human Brain Health, University of Birmingham<br>w: lei-zhang.net<br>t: @lei_zhang_lz <https://twitter.com/lei_zhang_lz>Christian Büchel2024-03-11T09:36:32+01:002024-03-11T09:36:32+01:00AW: [EXT] [SPM] Finite Impulse Response and event durations in SPMhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;547a5c14.2403Dear Batiah, you have to be careful here. FIR is an onset-related model, where I guess only duration=0 makes sense. The length of the modelled response and the granularity is determined by "order" and "length", where the granularity/resolution is length ./ order. Note that the resolution cannot be infinitely high, because you then might end up with empty regressors due to sampling. Usually people define the granularity as a TR i.e. order=10, then length = 10*TR gives you 10 bins after stimulus onset with a resolution of one TR per bin. PS. The units depend on whether you specify everything in seconds or scans (...fmri_spec.timing.units). I hope this helps, Christian -- Prof. Dr. Christian Büchel Institut für Systemische Neurowissenschaften Haus W34, Universitätsklinikum Hamburg-Eppendorf Martinistr. 52, D-20246 Hamburg, Germany Tel.: +49-40-7410-54726 Fax.: +49-40-7410-59955 buechel@uke.de http://www.uke.uni-hamburg.de/institute/systemische-neurowissenschaften/ > -----Ursprüngliche Nachricht----- > Von: SPM (Statistical Parametric Mapping) [mailto:SPM@JISCMAIL.AC.UK] Im > Auftrag von Batiah Keissar > Gesendet: Montag, 11. März 2024 09:12 > An: SPM@JISCMAIL.AC.UK > Betreff: [EXT] [SPM] Finite Impulse Response and event durations in SPM > > Hello SPM experts, > > I am using Finite Impulse Response in an SPM fMRI analysis, and I wanted to > ask if it is possible to do this with a window containing events with durations of > 0 seconds? What would be the implications for my analysis in comparison to > longer durations? > Also, If anyone has recommendations for further learning materials on this > topic I would greatly appreciate it. > > Thank you all kindly, > > Batiah --Batiah Keissar2024-03-11T08:12:00+00:002024-03-11T08:12:00+00:00Finite Impulse Response and event durations in SPMhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;8bd69116.2403Hello SPM experts,<br><br>I am using Finite Impulse Response in an SPM fMRI analysis, and I wanted to ask if it is possible to do this with a window containing events with durations of 0 seconds? What would be the implications for my analysis in comparison to longer durations?<br>Also, If anyone has recommendations for further learning materials on this topic I would greatly appreciate it.<br><br>Thank you all kindly,<br><br>BatiahChristian Gaser2024-03-10T19:40:00+00:002024-03-10T19:40:00+00:00Re: ROI analysis: VBM/DBM errorhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;8e672951.2403Dear Pavlina,<br><br>the ROI tool checks the filenames of the original (VBM) analysis and should contain one of the following patterns<br>mwp<br>m0wp<br>wp<br><br>These are the patterns usually found in the filenames after segmentation. In your case, I assume that you have renamed or moved the files.<br><br>Best,<br><br>Christian<br><br>On Thu, 7 Mar 2024 14:18:24 +0100, Pavlina Lieskovsky <pavlina.lieskovsky@GMAIL.COM> wrote:<br><br>>Dear all,<br>><br>>I am currently facing an issue while attempting ROI analysis in CAT12.<br>>Specifically, I keep encountering the following error message: "ROI<br>>analysis is only supported for VBM of GM/WM/CSF. No ROI values for DBM will<br>>be estimated." I followed the VBM longitudinal data CAT12 manual during<br>>preprocessing.<br>><br>>I have attached my script and screenshots of my batch. I would greatly<br>>appreciate it if someone could review these and offer insights into why<br>>this error is occurring and if there is some reason this data could be DBM.<br>><br>>Thank you very much for your attention and assistance.<br>><br>>Warm regards,<br>><br>>Pavlina<br>>%-----------------------------------------------------------------------<br>>%%<br>>matlabbatch{1}.spm.tools.cat.long.datalong.timepoints = {<br>>{<br>>..............baseline files<br>>}<br>>{<br>>..............follow up files'<br>>}<br>>}';<br>>%%<br>>matlabbatch{1}.spm.tools.cat.long.longmodel = 2;<br>>matlabbatch{1}.spm.tools.cat.long.enablepriors = 1;<br>>matlabbatch{1}.spm.tools.cat.long.prepavg = 2;<br>>matlabbatch{1}.spm.tools.cat.long.bstr = 0;<br>>matlabbatch{1}.spm.tools.cat.long.avgLASWMHC = 0;<br>>matlabbatch{1}.spm.tools.cat.long.nproc = 4;<br>>matlabbatch{1}.spm.tools.cat.long.opts.tpm = {<br>>'/Users/Downloads/spm12/tpm/TPM.nii'};<br>>matlabbatch{1}.spm.tools.cat.long.opts.affreg = 'mni';<br>>matlabbatch{1}.spm.tools.cat.long.opts.biasacc = 0.5;<br>>matlabbatch{1}.spm.tools.cat.long.extopts.restypes.optimal = [1 0.3];<br>>matlabbatch{1}.spm.tools.cat.long.extopts.setCOM = 1;<br>>matlabbatch{1}.spm.tools.cat.long.extopts.APP<br>><http://spm.tools.cat.long.extopts.app/> = 1070;<br>>matlabbatch{1}.spm.tools.cat.long.extopts.affmod = 0;<br>>matlabbatch{1}.spm.tools.cat.long.extopts.spm_kamap = 0;<br>>matlabbatch{1}.spm.tools.cat.long.extopts.LASstr = 0.5;<br>>matlabbatch{1}.spm.tools.cat.long.extopts.LASmyostr = 0;<br>>matlabbatch{1}.spm.tools.cat.long.extopts.gcutstr = 2;<br>>matlabbatch{1}.spm.tools.cat.long.extopts.WMHC = 2;<br>>matlabbatch{1}.spm.tools.cat.long.extopts.registration.shooting.shootingtpm<br>>= {<br>>'/Users/Downloads/spm12/toolbox/cat12/templates_MNI152NLin2009cAsym/Template_0_GS.nii'<br>>};<br>>matlabbatch{1}.spm.tools.cat.long.extopts.registration.shooting.regstr =<br>>0.5;<br>>matlabbatch{1}.spm.tools.cat.long.extopts.vox = 1.5;<br>>matlabbatch{1}.spm.tools.cat.long.extopts.bb = 12;<br>>matlabbatch{1}.spm.tools.cat.long.extopts.SRP = 22;<br>>matlabbatch{1}.spm.tools.cat.long.extopts.ignoreErrors = 1;<br>>matlabbatch{1}.spm.tools.cat.long.output.BIDS.BIDSno = 1;<br>>matlabbatch{1}.spm.tools.cat.long.output.surface = 1;<br>>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.neuromorphometrics = 1;<br>>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.lpba40 = 1;<br>>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.cobra = 1;<br>>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.hammers = 0;<br>>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.thalamus = 1;<br>>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.thalamic_nuclei = 1;<br>>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.suit = 1;<br>>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.ibsr = 0;<br>>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.ownatlas = {''};<br>>matlabbatch{1}.spm.tools.cat.long.longTPM = 1;<br>>matlabbatch{1}.spm.tools.cat.long.modulate = 1;<br>>matlabbatch{1}.spm.tools.cat.long.dartel = 0;<br>>matlabbatch{1}.spm.tools.cat.long.printlong = 2;<br>>matlabbatch{1}.spm.tools.cat.long.delete_temp = 1;<br>><br>><br>><br>>[image: image.png]<br>><br>>[image: image.png]<br>>Luna Sato2024-03-10T02:28:47+01:002024-03-10T02:28:47+01:00How to Remove Motor Response Artifacts in fMRI Experimentshttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;ada07fe6.2403Hi SPM experts,<br><br>In my fMRI experiment, participants were instructed to perform motor responses during blocks. These responses, however, were not of interest as the task solely aimed at improving concentration levels. So no timestamps were recorded for these motor responses.<br><br>However, in our group analysis, we observed significant effects of these motor responses: all experimental conditions exhibited negative activations in regions of interest. This could potentially be attributed to motor responses between blocks, leading to a higher baseline in the GLM.<br><br>Hence, I’m wondering if there are any noise-reduction techniques available to eliminate such noise from the data?<br><br>Best,Stephan Klaus DEHEN2024-03-09T21:50:28+00:002024-03-09T21:50:28+00:00Normalization quality assessmenthttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;5acb78e3.2403Hi all,<br><br>For my master's thesis I need to normalize several PDw-MRIs from different subjects to a MNI template. For this I would like to compare the quality between first creating a DARTEL template or normalizing the MRIs directly to the MNI template.<br>Is there an automatic way to assess and quantify the quality of the normalization? Or is there a standardized procedure?<br><br>Thank you very much for your support.<br><br>Stephan Gaël Cordero Otero2024-03-08T16:30:08+01:002024-03-08T16:30:08+01:00Re: [EXT] [SPM] Normalization shifting image upwards (axially)https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;c51d1ead.2403Dear Christian,<br>Thank you very much for taking the time to answer my question.<br>After tinkering a bit more I found that running 'normalization (estimate<br>and write)' as opposed to just the write option took care of the problem.<br>Best,<br>Gaël<br><br>On Thu, 7 Mar 2024 at 13:46, Christian Büchel <buechel@uke.de> wrote:<br><br>> Dear Gael,<br>><br>> if the template is "higher" than your images this would be the expected<br>> behavior of spatial normalization. Open a template in the same "checkreg"<br>> to see whether this is the case. In general I would point you to the<br>> excellent PDF in the SPM distribution under /man which explains all these<br>> concepts.<br>><br>> I hope this helps,<br>><br>> Christian<br>> --<br>> Prof. Dr. Christian Büchel<br>> Institut für Systemische Neurowissenschaften Haus W34,<br>> Universitätsklinikum Hamburg-Eppendorf Martinistr. 52, D-20246 Hamburg,<br>> Germany<br>> Tel.: +49-40-7410-54726<br>> Fax.: +49-40-7410-59955<br>> buechel@uke.de<br>> http://www.uke.uni-hamburg.de/institute/systemische-neurowissenschaften/<br>><br>><br>><br>><br>> > -----Ursprüngliche Nachricht-----<br>> > Von: SPM (Statistical Parametric Mapping) [mailto:SPM@JISCMAIL.AC.UK] Im<br>> > Auftrag von Ga ël Cordero Otero<br>> > Gesendet: Donnerstag, 7. März 2024 13:38<br>> > An: SPM@JISCMAIL.AC.UK<br>> > Betreff: [EXT] [SPM] Normalization shifting image upwards (axially)<br>> ><br>> > Dear experts,<br>> ><br>> > During preprocessing, normalization seems to be moving my images upwards<br>> > (axially speaking). To better illustrate what I mean I've attached an<br>> image of a<br>> > realigned & unwarped image (left) and the same image after normalization<br>> > (right). We acquired the volumes with a slice tilt since there is<br>> evidence that<br>> > suggests that this increases the SNR of temporal lobes, that's why there<br>> isn't<br>> > whole brian coverage. I'm using SPM12 on matlab 2021a, if that is of any<br>> help.<br>> ><br>> > Has anyone run into this issue previously? If so, how can it be solved?<br>> > Thank you very much for your time,<br>> > Gaël<br>><br>> --<br>><br>> _____________________________________________________________________<br>><br>> Universitätsklinikum Hamburg-Eppendorf; Körperschaft des öffentlichen<br>> Rechts; Gerichtsstand: Hamburg | www.uke.de<br>> Vorstandsmitglieder: Prof. Dr. Christian Gerloff (Vorsitzender), Joachim<br>> Prölß, Prof. Dr. Blanche Schwappach-Pignataro, Matthias Waldmann (komm.)<br>> _____________________________________________________________________<br>><br>> SAVE PAPER - THINK BEFORE PRINTING<br>><br><br>--<br>Gaël Cordero Otero<br>Department of Basic Sciences<br>Faculty of Medicine and Health Sciences<br>UIC Barcelona<br>Telf. 93 504 20 00 (ext. 5240)Michael Zyphur2024-03-08T22:30:00+11:002024-03-08T22:30:00+11:00Last chance: Machine Learning and AI for Research in R - livestream seminarhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;f8ff4790.2403Hi everyone<br><br>Instats is pleased to present an upcoming seminar introducing Machine<br>Learning and AI for Research in R<br><https://instats.org/seminar/machine-learning-and-ai-for-researchers-4>,<br>running March 12 - 13. This seminar is being led by professor Giovanni<br>Cerulli who has extensive experience teaching this material, and will<br>follow along with the core topics in his new book *Fundamentals of<br>Supervised Machine Learning: With Applications in Python, R, and Stata*<br><https://www.amazon.com/Fundamentals-Supervised-Machine-Learning-Applications/dp/3031413369>*.<br>*The seminar provides a comprehensive introduction to Machine Learning and<br>Artificial Intelligence methods for the social, economic, and health<br>sciences using R. After introducing the subject, the seminar will cover the<br>following methods: (i) model selection and regularization (Lasso, Ridge,<br>Elastic-net); (ii) discriminant analysis and nearest-neighbor<br>classification; and (iii) artificial neural networks. The course will offer<br>various instructional examples using real datasets in R and Python. An<br>Instats certificate of completion is provided at the end of the seminar, and<br>2 ECTS equivalent points are offered.<br><br>Register today<br><https://instats.org/seminar/machine-learning-and-ai-for-researchers-4> to<br>secure your spot, and please feel free to tell your colleagues and friends.<br><br>Best wishes and we hope to see you there!<br><br>Michael Zyphur<br>Director<br>Institute for Statistical and Data Science<br><http://goog_711907693>*instats.org* <http://instats.org>Kyoungeun Lee2024-03-07T15:57:59-06:002024-03-07T15:57:59-06:001st level global normalization - none vs. scalinghttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;4b057432.2403Dear SPM users,<br><br>I have a question regarding the 1st level global normalization option (none<br>vs. scaling) for fMRI analysis.<br>[image: image.png]<br><br>I just have a very broad understanding that it does the mean centering each<br>volume in the time series, but wondering:<br>1) What's the implication of doing it (none vs. scaling)?<br>2) When is it preferable to use scaling, and when is it not recommended?<br>3) Is there a relationship between this choice and the heterogeneity of the<br>sample?<br><br>In my case, I am working with a lifespan sample ranging from ages 18 to 79.<br>I am wondering if it is optimal to use scaling or stick with the default<br>option (none).<br><br>I experimented this with my dataset, and it impacted my results (2nd level)<br>a lot (though the most significant regions seem consistent).<br><br>[image: image.png]<br><br>Any insight would be greatly appreciated.<br>Thank you!<br><br>Best,<br>KyoungeunPavlina Lieskovsky2024-03-07T14:18:24+01:002024-03-07T14:18:24+01:00ROI analysis: VBM/DBM errorhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;c1e0eb8f.2403Dear all,<br><br>I am currently facing an issue while attempting ROI analysis in CAT12.<br>Specifically, I keep encountering the following error message: "ROI<br>analysis is only supported for VBM of GM/WM/CSF. No ROI values for DBM will<br>be estimated." I followed the VBM longitudinal data CAT12 manual during<br>preprocessing.<br><br>I have attached my script and screenshots of my batch. I would greatly<br>appreciate it if someone could review these and offer insights into why<br>this error is occurring and if there is some reason this data could be DBM.<br><br>Thank you very much for your attention and assistance.<br><br>Warm regards,<br><br>Pavlina<br>%-----------------------------------------------------------------------<br>%%<br>matlabbatch{1}.spm.tools.cat.long.datalong.timepoints = {<br>{<br>..............baseline files<br>}<br>{<br>..............follow up files'<br>}<br>}';<br>%%<br>matlabbatch{1}.spm.tools.cat.long.longmodel = 2;<br>matlabbatch{1}.spm.tools.cat.long.enablepriors = 1;<br>matlabbatch{1}.spm.tools.cat.long.prepavg = 2;<br>matlabbatch{1}.spm.tools.cat.long.bstr = 0;<br>matlabbatch{1}.spm.tools.cat.long.avgLASWMHC = 0;<br>matlabbatch{1}.spm.tools.cat.long.nproc = 4;<br>matlabbatch{1}.spm.tools.cat.long.opts.tpm = {<br>'/Users/Downloads/spm12/tpm/TPM.nii'};<br>matlabbatch{1}.spm.tools.cat.long.opts.affreg = 'mni';<br>matlabbatch{1}.spm.tools.cat.long.opts.biasacc = 0.5;<br>matlabbatch{1}.spm.tools.cat.long.extopts.restypes.optimal = [1 0.3];<br>matlabbatch{1}.spm.tools.cat.long.extopts.setCOM = 1;<br>matlabbatch{1}.spm.tools.cat.long.extopts.APP<br><http://spm.tools.cat.long.extopts.app/> = 1070;<br>matlabbatch{1}.spm.tools.cat.long.extopts.affmod = 0;<br>matlabbatch{1}.spm.tools.cat.long.extopts.spm_kamap = 0;<br>matlabbatch{1}.spm.tools.cat.long.extopts.LASstr = 0.5;<br>matlabbatch{1}.spm.tools.cat.long.extopts.LASmyostr = 0;<br>matlabbatch{1}.spm.tools.cat.long.extopts.gcutstr = 2;<br>matlabbatch{1}.spm.tools.cat.long.extopts.WMHC = 2;<br>matlabbatch{1}.spm.tools.cat.long.extopts.registration.shooting.shootingtpm<br>= {<br>'/Users/Downloads/spm12/toolbox/cat12/templates_MNI152NLin2009cAsym/Template_0_GS.nii'<br>};<br>matlabbatch{1}.spm.tools.cat.long.extopts.registration.shooting.regstr =<br>0.5;<br>matlabbatch{1}.spm.tools.cat.long.extopts.vox = 1.5;<br>matlabbatch{1}.spm.tools.cat.long.extopts.bb = 12;<br>matlabbatch{1}.spm.tools.cat.long.extopts.SRP = 22;<br>matlabbatch{1}.spm.tools.cat.long.extopts.ignoreErrors = 1;<br>matlabbatch{1}.spm.tools.cat.long.output.BIDS.BIDSno = 1;<br>matlabbatch{1}.spm.tools.cat.long.output.surface = 1;<br>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.neuromorphometrics = 1;<br>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.lpba40 = 1;<br>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.cobra = 1;<br>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.hammers = 0;<br>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.thalamus = 1;<br>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.thalamic_nuclei = 1;<br>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.suit = 1;<br>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.ibsr = 0;<br>matlabbatch{1}.spm.tools.cat.long.ROImenu.atlases.ownatlas = {''};<br>matlabbatch{1}.spm.tools.cat.long.longTPM = 1;<br>matlabbatch{1}.spm.tools.cat.long.modulate = 1;<br>matlabbatch{1}.spm.tools.cat.long.dartel = 0;<br>matlabbatch{1}.spm.tools.cat.long.printlong = 2;<br>matlabbatch{1}.spm.tools.cat.long.delete_temp = 1;<br><br>[image: image.png]<br><br>[image: image.png]Christian Büchel2024-03-07T13:46:29+01:002024-03-07T13:46:29+01:00AW: [EXT] [SPM] Normalization shifting image upwards (axially)https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;bcc669ad.2403Dear Gael, if the template is "higher" than your images this would be the expected behavior of spatial normalization. Open a template in the same "checkreg" to see whether this is the case. In general I would point you to the excellent PDF in the SPM distribution under /man which explains all these concepts. I hope this helps, Christian -- Prof. Dr. Christian Büchel Institut für Systemische Neurowissenschaften Haus W34, Universitätsklinikum Hamburg-Eppendorf Martinistr. 52, D-20246 Hamburg, Germany Tel.: +49-40-7410-54726 Fax.: +49-40-7410-59955 buechel@uke.de http://www.uke.uni-hamburg.de/institute/systemische-neurowissenschaften/ > -----Ursprüngliche Nachricht----- > Von: SPM (Statistical Parametric Mapping) [mailto:SPM@JISCMAIL.AC.UK] Im > Auftrag von Ga ël Cordero Otero > Gesendet: Donnerstag, 7. März 2024 13:38 > An: SPM@JISCMAIL.AC.UK > Betreff: [EXT] [SPM] Normalization shifting image upwards (axially) > > Dear experts, > > During preprocessing, normalization seems to be moving my images upwards > (axially speaking). To better illustrate what I mean I've attached an image of a > realigned & unwarped image (left) and the same image after normalization > (right). We acquired the volumes with a slice tilt since there is evidence that > suggests that this increases the SNR of temporal lobes, that's why there isn't > whole brian coverage. I'm using SPM12 on matlab 2021a, if that is of any help. > > Has anyone run into this issue previously? If so, how can it be solved? > Thank you very much for your time, > Gaël --Gaël Cordero Otero2024-03-07T12:37:46+00:002024-03-07T12:37:46+00:00Normalization shifting image upwards (axially)https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;af481129.2403Dear experts,<br><br>During preprocessing, normalization seems to be moving my images upwards (axially speaking). To better illustrate what I mean I've attached an image of a realigned & unwarped image (left) and the same image after normalization (right). We acquired the volumes with a slice tilt since there is evidence that suggests that this increases the SNR of temporal lobes, that's why there isn't whole brian coverage. I'm using SPM12 on matlab 2021a, if that is of any help.<br><br>Has anyone run into this issue previously? If so, how can it be solved?<br>Thank you very much for your time,<br>GaëlELENA GROSSO2024-03-07T10:41:08+01:002024-03-07T10:41:08+01:00Re: SPM registration to MNI152 1 mmhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;c0ad7fe8.2403Thank you very much John!<br>It works!!!<br><br>BW<br>Elena<br><br>Il giorno gio 7 mar 2024 alle ore 09:09 Ashburner, John <<br>j.ashburner@ucl.ac.uk> ha scritto:<br><br>> You can get the voxel sizes and bounding box for an axial image by:<br>><br>> P = spm_select(1,'nifti');<br>> [bb,vx]=spm_get_bbox(P)<br>><br>> Note that for historical reasons, I think it still rounds to the origin to<br>> the closest voxel, so the dimensions may not quite be exactly the same. I<br>> think it should work for one of the MNI average images (in NIfTI format)<br>> though.<br>><br>> Also note that the vx and bb formulation only works for exactly axial<br>> images (imho it is better to specify this using image dimensions and a<br>> voxel-to-world matrix). Again, this should be fine for the MNI data.<br>><br>> Best regards,<br>> -John<br>><br>><br>> ------------------------------<br>> *From:* ELENA GROSSO <elena.grosso01@universitadipavia.it><br>> *Sent:* 06 March 2024 16:23<br>> *To:* Ashburner, John <j.ashburner@ucl.ac.uk><br>> *Cc:* SPM@JISCMAIL.AC.UK <SPM@jiscmail.ac.uk><br>> *Subject:* Re: [SPM] SPM registration to MNI152 1 mm<br>><br>> Thank you John for your fast reply.<br>><br>> Changing the voxel size I don't obtain the same dimensions as in MNI152<br>> (as you can see from the screenshots attached).<br>> In fact, my normalized spm image has dimensions 181x217x181, while MNI152<br>> has dimensions 182x218x182. How could I obtain the same dimensions?<br>> I also copy the batch if can be useful:<br>><br>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.biasreg<br>> = 0.0001;<br>><br>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.biasfwhm = 60;<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.tpm =<br>> {'spm12/tpm/TPM.nii'};<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.affreg<br>> = 'mni';<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.reg =<br>> [0 0.001 0.5 0.05 0.2];<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.fwhm =<br>> 0;<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.samp =<br>> 3;<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.bb =<br>> [-90 -126 -7 90 90 108];<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.vox =<br>> [1 1 1];<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.interp<br>> = 4;<br>> matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.prefix<br>> = 'bb';<br>><br>> Thanks!!<br>> Elena<br>><br>><br>> Il giorno mer 6 mar 2024 alle ore 16:50 Ashburner, John <<br>> j.ashburner@ucl.ac.uk> ha scritto:<br>><br>> If I understand your question, you want to be able to specify a bounding<br>> box for generating images spatially normalised to 1 mm isotropic resolution.<br>><br>> The bounding box is in units of mm, and specifies coordinates within MNI<br>> space that define the corners of your normalised images. I think you just<br>> need to change the voxel sizes for the normalised images to [1 1 1] instead<br>> of their current default values.<br>><br>> Best regards,<br>> -John<br>><br>> ------------------------------<br>> *From:* SPM (Statistical Parametric Mapping) <SPM@JISCMAIL.AC.UK> on<br>> behalf of ELENA GROSSO <elena.grosso01@UNIVERSITADIPAVIA.IT><br>> *Sent:* 06 March 2024 15:02<br>> *To:* SPM@JISCMAIL.AC.UK <SPM@JISCMAIL.AC.UK><br>> *Subject:* [SPM] SPM registration to MNI152 1 mm<br>><br>><br>> ⚠ Caution: External sender<br>><br>> Hi all,<br>><br>> Have you ever registered maps with 1x1x1 mm3 resolution in MNI space with<br>> SPM?<br>> I can' t find anywhere the bounding box to do it!<br>><br>> Thanks,<br>> Elena<br>><br>><>2024-03-07T08:59:08+00:002024-03-07T08:59:08+00:00PhD position availablehttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;d1ce5181.2403PhD in psychology/ neuroscience/multi-modal neuroimaging<br><br>Start date: April 2024 or later<br>Duration: 3 years, optional extension by another year<br><br>We offer a PhD position in a project funded by the Swiss National Science Foundation (Schweizerischer Nationalfonds, SNF): „The extended metabolic phenotype of preclinical Huntington’s disease: Whole body PET studies of glucose metabolism“.<br><br>The group of Prof. Michael Orth, MD, PhD (University Hospital for Old Age Psychiatry and Psychotherapy) explores the relationship between brain structure, brain function, glucose metabolism and behaviour in Huntington’s disease, a hereditary neurodegenerative disease. The goal is to better understand what happens just before people carrying the HD mutation develop clinical signs of manifest HD. This can help to better predict the age at onset and improve the timing of disease modifying interventions. The current project is a collaboration with the Department of Nuclear Medicine at Inselspital Bern (head Prof. Axel Rominger), Prof. Jessica Peter at University Hospital for Old Age Psychiatry and Psychotherapy and Prof. Christian Wolf, Department of Psychiatry, Heidelberg University Hospital, Germany. We will examine carriers of the HD mutation who have no clinical signs of HD. They will undergo a whole-body glucose PET, and structural and resting state functional 3T MRI. The main question is whether HD mutation carriers differ from healthy volunteers in dynamic glucose uptake in the brain and/or peripheral tissues like skeletal muscle, and, if so, if there is a relationship between glucose metabolism and structural, or functional, changes in the brain. The project employs state-of-the-art PET and MRI methods, and multi-modal biostatistical methods for data analysis.<br><br>Tasks<br><br>*<br>Recruitment of study participants (healthy volunteers: HD participants are recruited via the HD clinic at the Swiss HD centre by the PI)<br>*<br>Generation of PET and MRI data; data analysis<br>*<br>Publication of results at conferences and in peer-reviewed journals<br><br>You have<br><br>*<br>A master degree in neurosciences, psychology, or a related field<br>*<br>Keen interest in research<br>*<br>Proficiency in German and English<br>*<br>Ability to work independently and self-driven<br>*<br>Knowledge in empirical methods and biostatistics<br>*<br>Previous experience in imaging research (PET or MRI) will be helpful<br><br>We offer<br><br>*<br>Close PhD supervision by the PI and the multidisciplinary team<br>*<br>Opportunities for training in neuroimaging and the analysis of complex multi-modal data<br>*<br>Opportunities for international networking in HD research<br>*<br>Salary according to the guidelines of the Swiss National Science Foundation (Schweizerischer Nationalfonds,SNF)<br><br>Contact and application<br><br>Please apply in writing to michael.orth@unibe.ch<mailto:michael.orth@unibe.ch> (deadline 15 March 2024) and include CV, cover letter, and references (as pdf no larger than 5MB). Inquiries can also be sent to the above email address.<br><br><br>Antworten<br><br>Weiterleiten<br>Teilnehmerbereich geschlossen<br><br>Universitäre Psychiatrische Dienste Bern (UPD) AG<br>Universitätsklinik für Alterspsychiatrie und Psychotherapie<br><br>Prof. Dr. Jessica Peter<br>Leiterin Forschung<br><br>Bolligenstrasse 111, 3000 Bern 60<br>Tel.: +41(0)31 932 89 03<br>Mail: jessica.peter@unibe.ch<mailto:jessica.peter@upd.unibe.ch><br>Webseite: http://www.upd.unibe.ch/research/research_groups/group_peter<http://www.upd.unibe.ch/research/research_groups/group_peter/index_eng.html>Ashburner, John2024-03-07T08:08:58+00:002024-03-07T08:08:58+00:00Re: SPM registration to MNI152 1 mmhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;560acbe3.2403You can get the voxel sizes and bounding box for an axial image by:<br><br>P = spm_select(1,'nifti');<br>[bb,vx]=spm_get_bbox(P)<br><br>Note that for historical reasons, I think it still rounds to the origin to the closest voxel, so the dimensions may not quite be exactly the same. I think it should work for one of the MNI average images (in NIfTI format) though.<br><br>Also note that the vx and bb formulation only works for exactly axial images (imho it is better to specify this using image dimensions and a voxel-to-world matrix). Again, this should be fine for the MNI data.<br><br>Best regards,<br>-JohnFalko Mecklenbrauck2024-03-07T08:40:33+01:002024-03-07T08:40:33+01:00Re: How to identify subjects with unacceptable motion?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;d3520415.2403Hey Olivia,<br><br>there are multiple ways to check for movement.<br>The easiest way would be to visually inspect the movement plots that SPM<br>generates based on the text file. (If there was no visual output and SPM<br>also did not save any postscript files, please see the MATLAB Code at<br>the bottom).<br>There you would look for spikes, so movement between two images that<br>exceeds half a voxel size (at least that is the recommondation of<br>Poldrack et al. in their book on MRI data analysis) or you can look at<br>the total movment of the participant across the experiment or the runs.<br>We usually consider movement as too much, if they moved more that one<br>voxel across the entire experiment, which is a rather conservative<br>measure, since SPM's movement correction is fairly good correcting<br>slower movement drifts.<br>Of course you can also just read the text-file with a program of your<br>choice and calculate the maximal movement and spikes.<br><br>Beyond this inspection there is also the concept of Framewise<br>Displacement that was introduced in Power et al. (2012, Neuroimage,<br>https://doi.org/10.1016%2Fj.neuroimage.2011.10.018) which commbines the<br>6 columns into one measure to identify frames with too much movement.<br><br>But please consider, that throwing away an entire subject just because<br>there is a spike is usually not necessary, you can still exclude<br>individual images that might be "contaminated".<br><br>Good luck and best wishes,<br>Falko<br><br>%% MATLAB CODE FOR REPRODUCING PLOTS<br><br> rp_move = readmatrix(textFileName);<br><br> pics = 1:size(rp_move,1);<br><br> figure;<br> h(1) = subplot(2,1,1); % upper plot<br> plot(pics, rp_move(:,1), 'Color', [0,0,1], 'DisplayName', 'x');<br>hold on;<br> plot(pics, rp_move(:,2), 'Color', [0,1,0], 'DisplayName', 'y');<br> plot(pics, rp_move(:,3), 'Color', [1,0,0], 'DisplayName', 'z');<br>hold off;<br><br> xlabel('Scans');<br> ylabel('Translation in mm');<br> legend(gca,'show');<br><br> h(2) = subplot(2,1,2); % lower plot<br> plot(pics, rp_move(:,4)*(180/pi), 'Color', [0,0,1], 'DisplayName',<br>'pitch'); hold on;<br> plot(pics, rp_move(:,5)*(180/pi), 'Color', [0,1,0], 'DisplayName',<br>'roll');<br> plot(pics, rp_move(:,6)*(180/pi), 'Color', [1,0,0], 'DisplayName',<br>'yaw'); hold off;<br><br> xlabel('Scans');<br> ylabel('Rotation in deg');<br> legend('show');<br><br> linkaxes(h,'x'); % link the axes in x direction (just for convenience)<br><br> saveas(gcf, 'movement_plot.png')<br><br>%%%%%%<br><br>Am 07.03.2024 um 04:05 schrieb Olivia Yang:<br>> Hi all,<br>><br>> After realignment processing in SPM, we obtain a file named rp_*.txt<br>> containing a matrix with six columns.<br>><br>> How can we determine if a subject’s head motion is unacceptable using<br>> this matrix?<br>><br>> Thank you for your help.<br>><br>> I wish u good.<br>> OliviaOlivia Yang2024-03-07T03:05:44+00:002024-03-07T03:05:44+00:00How to identify subjects with unacceptable motion?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;7e7f38b5.2403Hi all, After realignment processing in SPM, we obtain a file named rp_*.txt containing a matrix with six columns. How can we determine if a subject’s head motion is unacceptable using this matrix? Thank you for your help. I wish u good. OliviaELENA GROSSO2024-03-06T17:23:39+01:002024-03-06T17:23:39+01:00Re: SPM registration to MNI152 1 mmhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;4c49634f.2403Thank you John for your fast reply.<br><br>Changing the voxel size I don't obtain the same dimensions as in MNI152 (as<br>you can see from the screenshots attached).<br>In fact, my normalized spm image has dimensions 181x217x181, while MNI152<br>has dimensions 182x218x182. How could I obtain the same dimensions?<br>I also copy the batch if can be useful:<br><br>matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.biasreg<br>= 0.0001;<br>matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.biasfwhm<br>= 60;<br>matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.tpm =<br>{'spm12/tpm/TPM.nii'};<br>matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.affreg =<br>'mni';<br>matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.reg = [0<br>0.001 0.5 0.05 0.2];<br>matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.fwhm = 0;<br>matlabbatch{1}.spm.spatial.normalise.estwrite.eoptions.samp = 3;<br>matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.bb =<br>[-90 -126 -7 90 90 108];<br>matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.vox = [1<br>1 1];<br>matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.interp =<br>4;<br>matlabbatch{1}.spm.spatial.normalise.estwrite.woptions.prefix =<br>'bb';<br><br>Thanks!!<br>Elena<br><br>Il giorno mer 6 mar 2024 alle ore 16:50 Ashburner, John <<br>j.ashburner@ucl.ac.uk> ha scritto:<br><br>> If I understand your question, you want to be able to specify a bounding<br>> box for generating images spatially normalised to 1 mm isotropic resolution.<br>><br>> The bounding box is in units of mm, and specifies coordinates within MNI<br>> space that define the corners of your normalised images. I think you just<br>> need to change the voxel sizes for the normalised images to [1 1 1] instead<br>> of their current default values.<br>><br>> Best regards,<br>> -John<br>><br>> ------------------------------<br>> *From:* SPM (Statistical Parametric Mapping) <SPM@JISCMAIL.AC.UK> on<br>> behalf of ELENA GROSSO <elena.grosso01@UNIVERSITADIPAVIA.IT><br>> *Sent:* 06 March 2024 15:02<br>> *To:* SPM@JISCMAIL.AC.UK <SPM@JISCMAIL.AC.UK><br>> *Subject:* [SPM] SPM registration to MNI152 1 mm<br>><br>><br>> ⚠ Caution: External sender<br>><br>> Hi all,<br>><br>> Have you ever registered maps with 1x1x1 mm3 resolution in MNI space with<br>> SPM?<br>> I can' t find anywhere the bounding box to do it!<br>><br>> Thanks,<br>> Elena<br>>Ashburner, John2024-03-06T15:50:17+00:002024-03-06T15:50:17+00:00Re: SPM registration to MNI152 1 mmhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;bf750aca.2403If I understand your question, you want to be able to specify a bounding box for generating images spatially normalised to 1 mm isotropic resolution.<br><br>The bounding box is in units of mm, and specifies coordinates within MNI space that define the corners of your normalised images. I think you just need to change the voxel sizes for the normalised images to [1 1 1] instead of their current default values.<br><br>Best regards,<br>-JohnELENA GROSSO2024-03-06T16:02:59+01:002024-03-06T16:02:59+01:00SPM registration to MNI152 1 mmhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;bc9a60c0.2403Hi all,<br><br>Have you ever registered maps with 1x1x1 mm3 resolution in MNI space with<br>SPM?<br>I can' t find anywhere the bounding box to do it!<br><br>Thanks,<br>ElenaTorben Lund2024-03-06T12:50:57+01:002024-03-06T12:50:57+01:00Re: A fundamental question about spm's high pass filteringhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;fcb65dfc.2403Dear Mayank<br><br>I assume that y1 in your simulation is a BOLD signal of interest, and y2 is some form nuisance e.g. drift you want to filter away. If this is the case your simulation would apply to a situation where the SNR is around 1/50 (+white noise) which luckily is not quite the situation we typically are dealing with in fMRI. I have extended your simulation to SNR around 1 and SNR around 1/10. As you can see from the figure (run the attached code if it is removed due to its size) the filter performs pretty consistently for the SNR range from 0.1 to 10, but you do get some ringing if your SNR is around 0.02. In that scenario I am afraid you will have other problems :-) The ringing can be reduced if the length of the signal is extended with a factor of 100 - but that would be equally irrelevant in practice.<br><br>I hope this helps<br><br>Best<br>Torben<br><br>%%<br>L = 1024; %<length of signal<br>filter_100s = ... %< filter with hpf_cutoff = 100s<br>spm_filter( struct('RT', 1, 'HParam', 100, 'row', 1:L) );<br><br>y1 = sin(2*pi*[1:L]/50 )'; %< sinusoid with period = 50s, shouldn't be filtered<br>y2 = sin(2*pi*[1:L]/350)'; %< sinusoid with period = 350s, should be filtered<br><br>y_filter_test1 = spm_filter(filter_100s, y1+50*y2); %< 'a' = 50<br>y_filter_test2 = spm_filter(filter_100s, y1+.1*y2); %< 'a'=0.1<br><br>y_filter_test3 = spm_filter(filter_100s, y1+10*y2); %< 'a' = 10<br>y_filter_test4 = spm_filter(filter_100s, y1+1*y2); %< 'a'=1<br><br>figure;<br>subplot(2,1,1); plot([y1+50*y2 y1+10*y2 y1+1*y2 y1+.1*y2]),<br>l=legend({'SNR=0.02' 'SNR=0.1' 'SNR=1' 'SNR=10' }),xlim([1 L])<br>l.Location='NorthEastOutside';<br>title('Unfiltered signals')<br>xlabel('Time [s]')<br>subplot(2,1,2);<br>plot([y_filter_test1 y_filter_test3 y_filter_test4 y_filter_test2])<br>title('Highpass filtered T=100s')<br>xlabel('Time [s]')<br>ylim([-5 5]),xlim([1 L])<br>l=legend({'SNR=0.02' 'SNR=0.1' 'SNR=1' 'SNR=10' }),xlim([1 L])<br>l.Location='NorthEastOutside';<br><br>> Den 25. feb. 2024 kl. 21.49 skrev Mayank Jog <mayankjog@GMAIL.COM>:<br>><br>> Hello experts!<br>> I was trying to understand an oddity I observed with high-pass filtering in spm.<br>><br>> Basically, I constructed a signal = y1+ a*y2;<br>> y1 = sinusoid whose freq > hpf_cutoff, ie. it shouldn't be filtered out<br>> y2 = sinusoid whose freq < hpf_cutoff, ie. it should be filtered out.<br>><br>> The issue I'm having is that the filter gives different results based on "a" above (MATLAB code @ end of this email). Thinking from a brick wall** -type filtering POV, this shouldn't happen... the result of filtering "signal" above should be y1 everytime, independent of "a".<br>><br>> 1. Reading the documentation, I realized that SPM implements high-pass filtering using DCT.... why do we prefer filtering fMRI data with a DCT filter, since as the above case shows, a brick wall filter seems to be more accurate?<br>> 2. Thinking of y2 as "noise", it's almost as if the output is dependent on the scale of noise (captured by the scaling factor "a" above). Is this the right way to think about it/ Am I missing something here?<br>><br>> Thank you!<br>> Mayank<br>><br>><br>> **By brick wall, I mean doing an fft, and nulling all frequencies above hpf_cutoff, followed by an inverse fft.<br>><br>> MATLAB Code: ===================<br>> L = 1024; %<length of signal<br>> filter_100s = ... %< filter with hpf_cutoff = 100s<br>> spm_filter( struct('RT', 1, 'HParam', 100, 'row', 1:L) );<br>><br>><br>> y1 = sin(2*pi*[1:L]/50 )'; %< sinusoid with period = 50s, shouldn't be filtered<br>> y2 = sin(2*pi*[ylim1:L]/350)'; %< sinusoid with period = 350s, should be filtered<br>><br>><br>> y_filter_test1 = spm_filter(filter_100s, y1+50*y2); %< 'a' = 50<br>> y_filter_test2 = spm_filter(filter_100s, y1+.1*y2); %< 'a'=0.1<br>><br>> figure; subplot(3,1,1); plot([y1 y2]);<br>> subplot(3,1,2); plot(y_filter_test1);<br>> subplot(3,1,3); plot(y_filter_test2);<br>> %============================ Lorentz, Lukas Kay2024-03-06T10:05:48+00:002024-03-06T10:05:48+00:00Searching over PEB models in DCMhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;162e1215.2403Dear experts,<br><br>I am currently employing DCM on a task-based dataset and ran into a problem when interpreting the results. As a framework, we are following the two Zeidman et al. papers from 2019 on Parametric Empirical Bayes (https://doi.org/10.1016/j.neuroimage.2019.06.031 and https://doi.org/10.1016/j.neuroimage.2019.06.032).<br><br>For inference regarding our modulatory inputs (B matrix), we conducted an automatic "search over reduced PEB models" as described in section 4.7 in the 2nd Zeidman paper. This yielded significant results for several connections that were very much in line with our hypotheses.<br>However, when we then conducted another automatic search on our A matrix to derive average connectivity parameters (described in section 4.8 in 2nd Zeidman paper), we found that two connections were pruned away, even though our first analysis suggests that these connections would be significantly modulated by specific conditions.<br><br>Can anyone explain to me how to interpret this? Is searching over reduced A matrix models the correct way to estimate average effective connectivity through Bayesian Model Averaging?<br><br>Best regards,<br>Lukas Luna Sato2024-03-06T04:23:09+01:002024-03-06T04:23:09+01:00What is the DCM.U.idx parameter for?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;44923acd.2403Hi experts,<br><br>When checking DCM results, I found certain subjects' DCM fields include the parameter DCM.U.idx, while others don't. I suspect this variation might be due to different SPM versions used.<br><br>I'm wondering about the significance of DCM.U.idx. Can I combine subjects with and without this parameter in group analysis? Or should I consider redoing some DCM analyses?<br><br>Best regards,<br>LunaDennis Thompson2024-03-05T10:07:52-08:002024-03-05T10:07:52-08:00Re: A fundamental question about spm's high pass filteringhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;b08bd69d.2403https://en.wikipedia.org/wiki/Ringing_artifacts#Introduction<br><br>Brickwall filtering in the frequency domain can introduce ringing in<br>the time domain.<br><br>On Mon, Mar 4, 2024 at 5:47 PM Mayank Jog <mayankjog@gmail.com> wrote:<br>><br>> Dear experts,<br>> Just following up on my query regarding the spm implementation of high pass filtering, in case anyone had insights,<br>> Thank you,<br>> Mayank<br>><br>> On Sun, Feb 25, 2024 at 12:49 PM Mayank Jog <mayankjog@gmail.com> wrote:<br>>><br>>> Hello experts!<br>>> I was trying to understand an oddity I observed with high-pass filtering in spm.<br>>><br>>> Basically, I constructed a signal = y1+ a*y2;<br>>> y1 = sinusoid whose freq > hpf_cutoff, ie. it shouldn't be filtered out<br>>> y2 = sinusoid whose freq < hpf_cutoff, ie. it should be filtered out.<br>>><br>>> The issue I'm having is that the filter gives different results based on "a" above (MATLAB code @ end of this email). Thinking from a brick wall** -type filtering POV, this shouldn't happen... the result of filtering "signal" above should be y1 everytime, independent of "a".<br>>><br>>> 1. Reading the documentation, I realized that SPM implements high-pass filtering using DCT.... why do we prefer filtering fMRI data with a DCT filter, since as the above case shows, a brick wall filter seems to be more accurate?<br>>> 2. Thinking of y2 as "noise", it's almost as if the output is dependent on the scale of noise (captured by the scaling factor "a" above). Is this the right way to think about it/ Am I missing something here?<br>>><br>>> Thank you!<br>>> Mayank<br>>><br>>><br>>> **By brick wall, I mean doing an fft, and nulling all frequencies above hpf_cutoff, followed by an inverse fft.<br>>><br>>> MATLAB Code: ===================<br>>> L = 1024; %<length of signal<br>>> filter_100s = ... %< filter with hpf_cutoff = 100s<br>>> spm_filter( struct('RT', 1, 'HParam', 100, 'row', 1:L) );<br>>><br>>><br>>> y1 = sin(2*pi*[1:L]/50 )'; %< sinusoid with period = 50s, shouldn't be filtered<br>>> y2 = sin(2*pi*[ylim1:L]/350)'; %< sinusoid with period = 350s, should be filtered<br>>><br>>><br>>> y_filter_test1 = spm_filter(filter_100s, y1+50*y2); %< 'a' = 50<br>>> y_filter_test2 = spm_filter(filter_100s, y1+.1*y2); %< 'a'=0.1<br>>><br>>> figure; subplot(3,1,1); plot([y1 y2]);<br>>> subplot(3,1,2); plot(y_filter_test1);<br>>> subplot(3,1,3); plot(y_filter_test2);<br>>> %============================Volkmar Glauche2024-03-05T14:10:11+00:002024-03-05T14:10:11+00:00Re: How can I convert the SPM12 to an exe file?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;936013a3.2403Dear Sanaz,<br><br>the SPM batch system and the various extension mechanisms (toolboxes, spm_orthviews, fieldtrip...) require more knowledge than just compiling spm.m as a target function. SPM comes with a MATLAB .m-file spm_make_standalone.m that will take care of the necessary steps to compile SPM from within MATLAB. You should give it a try on your SPM installation. If you want to customize your standalone SPM (by e.g. adding extra toolboxes etc) you may need to make some changes to the code.<br><br>Hope this helps<br>VolkmarRomy Lorenz2024-03-05T15:07:42+01:002024-03-05T15:07:42+01:002 PhD positions in layer-fMRI of high-level cognition at the Max Planck Institute for Biological Cybernetics in Tübingen, Germanyhttps://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;556316f2.2403The newly established Cognitive Neuroscience & Neurotechnology group led by Dr. Romy Lorenz is looking for two enthusiastic PhD students (m/f/d) to join our growing team at the Max Planck Institute for Biological Cybernetics in Tübingen, Germany.<br><br>Our lab focuses on advancing our understanding of the frontoparietal brain network mechanisms that underpin high-level cognition and adaptive behaviour. For this, we pursue an interdisciplinary research programme that allows studying this brain system at multiple levels of granularity. Our methodology involves subject-specific brain-computer interface technology, fMRI at 3T and ultrahigh (i.e., 7T and 9.4T) magnetic field strengths (for resolving cortical layers), EEG, non-invasive brain stimulation as well as machine learning. You can find out more about our work at: https://www.kyb.tuebingen.mpg.de/711763/cognitive-neuroscience-neurotechnology<br><br>We are seeking two ambitious PhD students who will work on the exciting field of ultrahigh resolution fMRI that allows to investigate the human cortex at the scale of layers and columns.<br><br>The ideal candidates should have a master’s degree in cognitive (neuro)science, psychology, computer science, biomedical or electrical engineering, physics, or related disciplines. A strong background in fMRI data analysis (e.g., FSL, Freesurfer, ANTS) and very good programming skills in Bash on Linux, Matlab and/or Python are required. Prior experience in MRI data acquisition and experience with ultrahigh resolution fMRI (e.g., at 7T) is desirable but not necessary. Equally, experience with machine learning-methods, code sharing platforms (e.g. GitHub) and high-performance computing clusters are highly desirable.<br><br>The Max Planck Institute for Biological Cybernetics offers a world-leading research environment with access to the latest cutting-edge MRI hardware (including a Siemens 9.4T and Prisma 3T for humans as well as a 14.2T small animal system) and other excellent research facilities (EEG, eye-tracking, fMRI-TMS). The PhD student will receive generous support for professional travel and research needs (~2500€/year). Additionally, the student will have the opportunity to become part of the Graduate Training Centre of Neuroscience that provides training courses, summer schools and conferences to further educate doctoral students. Further, the Institute is part of the TübingenNeuroCampus (with more than 100 active groups), offering a vibrant community of international researchers and enriching environment of collaboration.<br><br>The position is available from May 2024 on and remains open until filled. The salary is paid in accordance with the collective agreement for the public sector (65% TVL-E13, amounting to ~2000€ net per month).<br><br>For more details about the two advertised PhD positions and how to apply, please see: https://www.kyb.tuebingen.mpg.de/729399/join-the-lab<br><br>Dr. Romy Lorenz<br><br>Max Planck Research Group Leader<br><br>Research Group Cognitive Neuroscience & Neurotechnology<br><br>Max Planck Institute for Biological Cybernetics<br><br>Tübingen, Germany<br><br>romy.lorenz@tuebingen.mpg.de<br><br>www.kyb.tuebingen.mpg.de/711763/cognitive-neuroscience-neurotechnologySanaz Hariri2024-03-05T08:29:06+00:002024-03-05T08:29:06+00:00Re: How can I convert the SPM12 to an exe file?https://www.jiscmail.ac.uk/cgi-bin/wa-jisc.exe?A2=SPM;b615e45.2403Dear experts,<br><br>Based on page 643 of EANM guideline for brain PET imaging (Guedj, Eur JNucl Med Mol Imaging, 2022, 49:632-651), “since version 8, SPM is available asa stand-alone tool”. However, I have not found the executable file of SPM 12but the m files. Please kindly let me know if the exe file is available and themethod to obtain it.<br><br>Best regards,<br><br>Sanaz Hariri<br><br>On Saturday, March 2, 2024 at 03:40:28 PM GMT+3:30, Sanaz Hariri <shanraiz@yahoo.com> wrote:<br><br>Dear experts,<br><br>I want to change the SPM12 to a standalone executable file. I used thedeploytool of MATLAB and added all files and directories of SPM folder to itwhile the main file was spm_Menu.m. However, running the built exe file resultedin an error message: “Can’t obtain SPM Revision information. Error in = >spm_Menu.m at line 25”.<br><br>In addition, I checked the built exe file when spm.m was selected as themain file. Again, an error message was appeared as “Can’t obtain SPM Revisioninformation. Error in = > spm.m at line 299”. In this line the spm_Welcome functionis called which had been added to the list of accompanying files, before.<br><br>Any help will be appreciated.<br><br>Best regards,<br><br>Sanaz