External Email - Use Caution Thanks a lot Dr. Greve, that really helps.
On Thu, Jul 26, 2018 at 10:45 AM, Douglas Greve <dgr...@mgh.harvard.edu> wrote: > They give different results because they are supposed to. It may be that > the effect you are seeing without regressing out ICV is due to ICV, and > when you control for ICV the effect goes away. It could also be that the > ICV causes the variance to increase. This is why I asked you to check the > rstd (residual standard deviation). You said that did not change, so I > would expect that the ICV must be removing your desired effect as seen the > in gamma. Sorry, I don't know what else to tell you. > > > > On 7/26/18 1:24 PM, Martin Juneja wrote: > > External Email - Use Caution > Dear Dr. Greve, > > I am so sorry for annoying you with multiple emails. > I clearly got the difference between gamma.mgh files depending on whether > I use ICV or not as a covariate. Next, I checked the ratio of gamma/rstd > (Cohen's d) with and without ICV as covariate. Again, its totally different > with and without ICV. Here I am attaching screen shot for that: left one is > after I include ICV as covariate (minimum effect size) and the right one is > without ICV as covariate (maximum effect size). > > I really did not understand the idea behind comparing gamma.mgh and > rstd.mgh (and may be their ratio), especially, if it's always the case that > for surface area and volume, we need ICV as covariate. > > On Thu, Jul 26, 2018 at 10:13 AM, Douglas Greve <dgr...@mgh.harvard.edu> > wrote: > >> If you are using surface area or volume, then you need ICV >> >> >> On 7/25/18 1:51 PM, Martin Juneja wrote: >> >> External Email - Use Caution >> Dr. Greve, >> >> I am sorry if my questions were not clear in previous email. >> >> Basically, I do not know what to conclude from this gamma comparison i.e. >> with and without ICV as covariate. >> Clearly, adding ICV as covariate here, is reducing effect size all over >> the brain and without ICV effect size is higher at specific locations. >> >> So should I go ahead with or without ICV as covariate? >> >> Thanks. >> >> On Wed, Jul 25, 2018 at 7:33 AM, Douglas Greve <dgr...@mgh.harvard.edu> >> wrote: >> >>> >>> The gammas do look different, but it is hard to tell whether they are, >>> eg, changing sign. Not sure what you want me to comment on. >>> >>> >>> >>> On 7/24/18 2:17 PM, Martin Juneja wrote: >>> >>> External Email - Use Caution >>> Just to add some more info here: >>> The peak location of regions, X1 and X2, which I found without including >>> ICV as covariate are very close with the peak locations I found in >>> Gamma_Without_ICV >>> (~5.15), whereas Gamma_With_ICV is almost all over the brain (range >>> -0.6 to +0.6). >>> I am not sure if this additional info adds anything to interpret >>> gamma.mgh with and without ICV as covariate. >>> >>> On Tue, Jul 24, 2018 at 10:10 AM, Martin Juneja <mj70...@gmail.com> >>> wrote: >>> >>>> Hi Dr. Greve, >>>> >>>> So I checked both. The rstd.mgh files are very similar in both cases >>>> (with and without ICV as covariate), but gamma.mgh files are very different >>>> for both cases. Here I am attaching screen shot for both cases: >>>> Gamma_With_ICV as covariate and Gamma_Without_ICV as covariate. >>>> >>>> Could you please have a look at the attached screen shots and provide >>>> your thoughts/interpretation of this comparison? >>>> >>>> Thanks. >>>> >>>> On Tue, Jul 24, 2018 at 9:35 AM, Douglas N. Greve < >>>> dgr...@mgh.harvard.edu> wrote: >>>> >>>>> For noise compare the values in the rstd.mgh file, for effect size >>>>> look >>>>> in the gamma.mgh file >>>>> >>>>> On 07/24/2018 12:27 PM, Martin Juneja wrote: >>>>> > >>>>> > External Email - Use Caution >>>>> > >>>>> > Hi Dr. Greve, >>>>> > >>>>> > Thanks for your quick reply. Could you please give me more details >>>>> how >>>>> > can I check this whether its because of noise or its because of less >>>>> > CV difference? >>>>> > I am not sure what method/way is the best and commonly used to >>>>> confirm >>>>> > these factors. >>>>> > >>>>> > Thanks. >>>>> > >>>>> > On Tue, Jul 24, 2018 at 7:06 AM, Douglas Greve < >>>>> dgr...@mgh.harvard.edu >>>>> > <mailto:dgr...@mgh.harvard.edu>> wrote: >>>>> > >>>>> > your results could have vanished after ICV correction for one of >>>>> > two reasons: the CV difference became less or the values became >>>>> > noisier (or a combination). So check in your data which one of >>>>> > those things happened. >>>>> > >>>>> > >>>>> > On 7/23/18 8:30 PM, Martin Juneja wrote: >>>>> >> >>>>> >> External Email - Use Caution >>>>> >> >>>>> >> Hello experts, >>>>> >> >>>>> >> I am interested in identifying regions of interest by comparing >>>>> >> cortical volume (CV) between controls and patients. >>>>> >> >>>>> >> After including age and sex as my covariates, I identified >>>>> >> regions X1 and X2, which showed significantly lower CV for >>>>> >> patients (as compared to controls). >>>>> >> >>>>> >> But after I include ICV as another covariate, my results show >>>>> >> that for none of the areas there is any significant difference >>>>> in >>>>> >> CV, i.e. my results vanish. >>>>> >> >>>>> >> When I checked subjectwise ICV for each group, I found that >>>>> there >>>>> >> is almost significant difference (two-sampled t-test, p = 0.067) >>>>> >> in ICV between two groups, but interestingly mean group ICV for >>>>> >> patients group was larger compared than mean ICV for controls. >>>>> >> But as I said earlier, regions X1 and X2 had significantly lower >>>>> >> CV for patients (as compared to controls), when I didn't include >>>>> >> ICV as covariate. >>>>> >> >>>>> >> Could you please help me in interpreting these results? Is there >>>>> >> any advice regarding inclusion of ICV as covariate? Or my >>>>> results >>>>> >> are purely because of differences in ICV between groups, and >>>>> >> there is no real findings regarding the regions identified (X1 >>>>> >> and X2)? >>>>> >> >>>>> >> Thanks a lot ! >>>>> >> >>>>> >> >>>>> >> _______________________________________________ >>>>> >> Freesurfer mailing list >>>>> >> Freesurfer@nmr.mgh.harvard.edu >>>>> >> <mailto:Freesurfer@nmr.mgh.harvard.edu> >>>>> >> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>> >> <https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer> >>>>> > >>>>> > >>>>> > _______________________________________________ >>>>> > Freesurfer mailing list >>>>> > Freesurfer@nmr.mgh.harvard.edu <mailto:freesur...@nmr.mgh.har >>>>> vard.edu> >>>>> > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>> > <https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer> >>>>> > >>>>> > >>>>> > The information in this e-mail is intended only for the person to >>>>> > whom it is >>>>> > addressed. If you believe this e-mail was sent to you in error >>>>> and >>>>> > the e-mail >>>>> > contains patient information, please contact the Partners >>>>> > Compliance HelpLine at >>>>> > http://www.partners.org/complianceline >>>>> > <http://www.partners.org/complianceline> . If the e-mail was >>>>> sent >>>>> > to you in error >>>>> > but does not contain patient information, please contact the >>>>> > sender and properly >>>>> > dispose of the e-mail. >>>>> > >>>>> > >>>>> > >>>>> > >>>>> > _______________________________________________ >>>>> > Freesurfer mailing list >>>>> > Freesurfer@nmr.mgh.harvard.edu >>>>> > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>> >>>>> _______________________________________________ >>>>> Freesurfer mailing list >>>>> Freesurfer@nmr.mgh.harvard.edu >>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>> >>>> >>>> >>> >>> >>> _______________________________________________ >>> Freesurfer mailing >>> listfreesur...@nmr.mgh.harvard.eduhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>> >>> >>> >>> _______________________________________________ >>> Freesurfer mailing list >>> Freesurfer@nmr.mgh.harvard.edu >>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>> >>> >>> The information in this e-mail is intended only for the person to whom >>> it is >>> addressed. If you believe this e-mail was sent to you in error and the >>> e-mail >>> contains patient information, please contact the Partners Compliance >>> HelpLine at >>> http://www.partners.org/complianceline . If the e-mail was sent to you >>> in error >>> but does not contain patient information, please contact the sender and >>> properly >>> dispose of the e-mail. >>> >>> >> >> >> _______________________________________________ >> Freesurfer mailing >> listfreesur...@nmr.mgh.harvard.eduhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >> >> >> >> _______________________________________________ >> Freesurfer mailing list >> Freesurfer@nmr.mgh.harvard.edu >> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >> >> >> The information in this e-mail is intended only for the person to whom it >> is >> addressed. If you believe this e-mail was sent to you in error and the >> e-mail >> contains patient information, please contact the Partners Compliance >> HelpLine at >> http://www.partners.org/complianceline . If the e-mail was sent to you >> in error >> but does not contain patient information, please contact the sender and >> properly >> dispose of the e-mail. >> >> > > > _______________________________________________ > Freesurfer mailing > listfreesur...@nmr.mgh.harvard.eduhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer > > > > _______________________________________________ > Freesurfer mailing list > Freesurfer@nmr.mgh.harvard.edu > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer > > > The information in this e-mail is intended only for the person to whom it > is > addressed. If you believe this e-mail was sent to you in error and the > e-mail > contains patient information, please contact the Partners Compliance > HelpLine at > http://www.partners.org/complianceline . If the e-mail was sent to you in > error > but does not contain patient information, please contact the sender and > properly > dispose of the e-mail. > >
_______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.