Greg, my suggestion was an aggressive removal of the physiological noise components without needing to rerun ICA+FIX.
It is also worth noting that the extant ³motion estimates² may not perform as expected in multi-band fMRI data. Matt. On 6/1/16, 10:08 AM, "Burgess, Greg" <[email protected]> wrote: >Hello Raquel, > >As soon as you wade into this work, you¹ll realize that there are still >tons of unresolved questions. For instance, I would suggest doing >physiological regression in the context of FIX a bit differently. >Re-running the FIX-ICA pipeline on the MPP data, including the full space >of the physiological regressors in the multiple regression (similar to >how 24 motion parameters are regressed) will remove all BOLD signal >correlated with the physiological regressors. In contrast, the prior >suggestion will effectively preserve the portion of the physiological >regressors that are contained in the ICA signal components. > >If you think that your signal components may be contaminated by noise, >you should consider an ³aggressive² regression of your noise regressors >(24-motion, noise ICA components, physiological regressors, etc.) from >the BOLD timeseries. If you think that your noise regressors might be >contaminated by signal that you don¹t want to lose, you might consider a >³soft² regression of your noise regressors in order to preserve the full >space of your signal components. > >Either way, I¹d also suggest running correlations between the amount of >motion in the scan and your estimates of functional connectivity, similar >to Power et al 2014 (Figure 12) or Pruim et al. 2015 (Figure 3). This >will make you aware of how the estimated motion relates to your estimates >of functional connectivity before and after your chosen denoising >strategy. > >‹Greg > >Power, J. D., Mitra, A., Laumann, T. O., Snyder, A. Z., Schlaggar, B. L., >& Petersen, S. E. (2014). Methods to detect, characterize, and remove >motion artifact in resting state fMRI. NeuroImage, 84, 320341. >http://doi.org/10.1016/j.neuroimage.2013.08.048 >Pruim, R. H. R., Mennes, M., Buitelaar, J. K., & Beckmann, C. F. (2015). >Evaluation of ICA-AROMA and alternative strategies for motion artifact >removal in resting state fMRI. NeuroImage, 112, 278287. >http://doi.org/10.1016/j.neuroimage.2015.02.063 >____________________________________________________________________ >Greg Burgess, Ph.D. >Staff Scientist, Human Connectome Project >Washington University School of Medicine >Department of Neuroscience >Phone: 314-362-7864 >Email: [email protected] > >> On May 31, 2016, at 5:03 PM, Glasser, Matthew <[email protected]> >>wrote: >> >> WARNING: This e-mail has been altered by MIMEDefang. Following this >> paragraph are indications of the actual changes made. For more >> information about your site's MIMEDefang policy, contact >> MIMEDefang Administrator's Full Name <postmaster@localhost>. For more >>information about MIMEDefang, see: >> >> http://www.roaringpenguin.com/mimedefang/enduser.php3 >> >> An attachment named PreprocessPhysio.sh was removed from this document >>as it >> constituted a security hazard. If you require this document, please >>contact >> the sender and arrange an alternate means of receiving it. >> >> >> _______________________________________________ >> HCP-Users mailing list >> [email protected] >> http://lists.humanconnectome.org/mailman/listinfo/hcp-users >> For rfMRI gradients I would not use MGTR as it moves some of the >>gradients in strange ways and truly global noise will not affect >>gradients (so it appears this occurs because MGTR is removing spatially >>specific signal from the data). It is the case that there is some >>global, likely physiological, noise in the HCP data after ICA+FIX >>(spatial ICA is blind to global signals). >> >> One solution for connectivity matrices (after parcellation) is to use >>partial correlation. >> >> Another would be to remove some of the physiological noise with the >>provided physiological regressors (though physiological regressors are >>likely not going to be a perfect model of the effects of physiology on >>BOLD signals). I have attached some scripts that allow you to quickly >>preprocess the released physiological data (make sure it is downloaded >>in 2016 as there were problems with previous versions). After >>preprocessing, you would need to run the hp 2000 detrending filter, >>regress out the 24 motion parameters, and the ICA+FIX noise ICA >>timecourses on the resultant physiological EVs before regressing them >>out of your ICA+FIX cleaned fMRI data (i.e. do the same steps as were >>done to clean the data to the regressors). >> >> This is an area of active work though. >> >> Peace, >> >> Matt. >> >> From: <[email protected]> on behalf of Raquel Viejo >>Sobera <[email protected]> >> Date: Tuesday, May 31, 2016 at 3:50 PM >> To: "[email protected]" <[email protected]> >> Subject: [HCP-Users] MGTR >> >> Hello HCP users, >> >> I'm planning on doing some funcional connectivity analysis to perform a >>parcellation >> with a method similar to the one presented in Cohen 2008 (1). >> >> >> >> Although my intention is to use the ICA-FIX denoised data, I'm confused >>about the convenience >> of making some extra processing resgressing the mean gray timecourse. >>I know that HCP team usually doesn't recommend to do that, but reading >>the Smith 2013 (2) paper and the results presented in one poster from >>the last OHBM by them, it seems that I should. >> >> >> >> Smith says: "Preliminary analyses of measures of motion-related >>artefacts indicate that the >> ICA-FIX process greatly reduces but does not, in some datasets, >>totally eliminate motion artefacts that are frame- specific and >>non-spatially-specific." and also "... some analyses, such as >>gradient-based investigation of functional boundaries (Cohen et al., >> 2008),may benefit from MGT regression, as it may help highlight >>transitions in connectivity patterns across the cortex" >> >> >> >> The OHBM poster is this: >> >> >> https://ww4.aievolution.com/hbm1501/index.cfm?do=abs.viewAbs&abs=3394 >> , and the results show that "FIX and MGTR appear to improve >>relationships of individual differences with resting state functional >>connectivity " >> >> >> Taking this into account, do you still recommend not to perform MGTR? >> >> >> Otherwise, can you offer some guidelines to perform this regression? >> >> >> Thank you very much in advance! >> >> >> (1) Cohen, A. L., Fair, D. A., Dosenbach, N. U. F., Miezin, F. M., >>Dierker, D., Van Essen, >> D. C., Š Petersen, S. E. (2008). Defining functional areas in >>individual human brains using resting functional connectivity MRI. >>NeuroImage, 41(1), 4557. >> >> http://doi.org/10.1016/j.neuroimage.2008.01.066 >> >> >> (2) Smith, S. M., Beckmann, C. F., Andersson, J., Auerbach, E. J., >>Bijsterbosch, J., Douaud, >> G., Š Glasser, M. F. (2013). Resting-state fMRI in the Human >>Connectome Project. NeuroImage, 80, 144168. >> >> http://doi.org/10.1016/j.neuroimage.2013.05.039 >> >> >> >> >> >> Raquel Viejo Sobera, PhD >> Cognitive NeuroLab >> Hospital de la Santa Creu i Sant Pau >> www.cognitiveneurolab.org >> ____________________________________ >> >> Universitat Oberta de Catalunya >> Faculty of Health Sicences >> Avinguda Tibidabo 39-43, 08035 Barcelona >> 93 450 (5370) | [email protected] >> >> >> >> Este mensaje se dirige exclusivamente a quien se destina y puede >>contener información privilegiada o confidencial y datos de carácter >>personal, cuya difusión está regulada por la Ley orgánica de protección >>de datos y la Ley de servicios de la sociedad de la información. Si >>usted no es la persona destinataria indicada (o la responsable de la >>entrega a quien se destina), no debe copiar este mensaje ni entregarlo a >>terceros bajo ningún concepto. Si ha recibido este mensaje por error o >>lo ha conseguido por otros medios, le rogamos que nos lo comunique >>inmediatamente por esta misma vía y proceda a su eliminación >>irreversible. >> >> >> Antes de imprimir este mensaje electrónico piense en el medio ambiente. >> _______________________________________________ >> HCP-Users mailing list >> [email protected] >> http://lists.humanconnectome.org/mailman/listinfo/hcp-users >> >> >> The materials in this message are private and may contain Protected >>Healthcare Information or other information of a sensitive nature. If >>you are not the intended recipient, be advised that any unauthorized >>use, disclosure, copying or the taking of any action in reliance on the >>contents of this information is strictly prohibited. If you have >>received this email in error, please immediately notify the sender via >>telephone or return mail. >> _______________________________________________ >> HCP-Users mailing list >> [email protected] >> http://lists.humanconnectome.org/mailman/listinfo/hcp-users >> <custom_slice_timing.txt><PreprocessPhyiso.txt> > ________________________________ The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail. _______________________________________________ HCP-Users mailing list [email protected] http://lists.humanconnectome.org/mailman/listinfo/hcp-users
