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 16.52024-03-18T17:53:26ZVladimir 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