External Email - Use Caution
Hi Freesurfer Expert,
How can I know the power of my GLM test after I did the clusterwise correction? I got this question because I had a problem like this. I have 21 subjects for group A and B. Once I delete one subject from group A, one of the clusterwise correction results has disappeared. Can you help me on this? Thanks a lot!
Best wishes, Ting Li
I'm not sure what you mean by this. power with clusterwise corrections is quite tricky. You'd have to do a simulation under non-null conditions to see what the true positive rate is
On 07/10/2018 11:53 AM, Ting Li wrote:
External Email - Use CautionHi Freesurfer Expert,
How can I know the power of my GLM test after I did the clusterwise correction? I got this question because I had a problem like this. I have 21 subjects for group A and B. Once I delete one subject from group A, one of the clusterwise correction results has disappeared. Can you help me on this? Thanks a lot!
Best wishes, Ting Li
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
External Email - Use Caution
Hi Douglas,
Thanks for your response. I did the monte carlo simulation. Could you give me any instructions about how to perform a simulation under non-null conditions? Thanks a lot!
Best, Ting
On Jul 16, 2018, at 12:15 PM, Douglas N. Greve dgreve@mgh.harvard.edu wrote:
I'm not sure what you mean by this. power with clusterwise corrections is quite tricky. You'd have to do a simulation under non-null conditions to see what the true positive rate is
On 07/10/2018 11:53 AM, Ting Li wrote:
External Email - Use CautionHi Freesurfer Expert,
How can I know the power of my GLM test after I did the clusterwise correction? I got this question because I had a problem like this. I have 21 subjects for group A and B. Once I delete one subject from group A, one of the clusterwise correction results has disappeared. Can you help me on this? Thanks a lot!
Best wishes, Ting Li
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
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.
It is different than the MC simulations. When you run mri_glmfit, specify the --save-yhat option. This will create a file called yhat.mgh which will be used as the non-null basis of the simulation. You will have to add noise to this file and analyze through mri_glmfit and mri_glmfi-sim and see if you get your cluster. Do this many times (~1000) with different noise each time, then simply compute how often you see your cluster (this is the true positive rate). You can add the noise in matlab, something like
yhat = MRIread('yhat.mgh'); rstd = MRIread('rstd.mgh'); % standard deviation for n = 1:1000 y = yhat; y.vol = yhat.vol + randn(size(rstd.vol)).*rstd.vol; % add noise fname = sprintf('synth%04d.mgh',n); MRIwrite(y,fname); end
On 07/16/2018 01:43 PM, Ting Li wrote:
External Email - Use CautionHi Douglas,
Thanks for your response. I did the monte carlo simulation. Could you give me any instructions about how to perform a simulation under non-null conditions? Thanks a lot!
Best, Ting
On Jul 16, 2018, at 12:15 PM, Douglas N. Greve dgreve@mgh.harvard.edu wrote:
I'm not sure what you mean by this. power with clusterwise corrections is quite tricky. You'd have to do a simulation under non-null conditions to see what the true positive rate is
On 07/10/2018 11:53 AM, Ting Li wrote:
External Email - Use CautionHi Freesurfer Expert,
How can I know the power of my GLM test after I did the clusterwise correction? I got this question because I had a problem like this. I have 21 subjects for group A and B. Once I delete one subject from group A, one of the clusterwise correction results has disappeared. Can you help me on this? Thanks a lot!
Best wishes, Ting Li
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
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
External Email - Use Caution
Thanks for your detail information. I will try it now.
Ting
On Jul 16, 2018, at 1:26 PM, Douglas N. Greve dgreve@mgh.harvard.edu wrote:
It is different than the MC simulations. When you run mri_glmfit, specify the --save-yhat option. This will create a file called yhat.mgh which will be used as the non-null basis of the simulation. You will have to add noise to this file and analyze through mri_glmfit and mri_glmfi-sim and see if you get your cluster. Do this many times (~1000) with different noise each time, then simply compute how often you see your cluster (this is the true positive rate). You can add the noise in matlab, something like
yhat = MRIread('yhat.mgh'); rstd = MRIread('rstd.mgh'); % standard deviation for n = 1:1000 y = yhat; y.vol = yhat.vol + randn(size(rstd.vol)).*rstd.vol; % add noise fname = sprintf('synth%04d.mgh',n); MRIwrite(y,fname); end
On 07/16/2018 01:43 PM, Ting Li wrote:
External Email - Use CautionHi Douglas,
Thanks for your response. I did the monte carlo simulation. Could you give me any instructions about how to perform a simulation under non-null conditions? Thanks a lot!
Best, Ting
On Jul 16, 2018, at 12:15 PM, Douglas N. Greve dgreve@mgh.harvard.edu wrote:
I'm not sure what you mean by this. power with clusterwise corrections is quite tricky. You'd have to do a simulation under non-null conditions to see what the true positive rate is
On 07/10/2018 11:53 AM, Ting Li wrote:
External Email - Use CautionHi Freesurfer Expert,
How can I know the power of my GLM test after I did the clusterwise correction? I got this question because I had a problem like this. I have 21 subjects for group A and B. Once I delete one subject from group A, one of the clusterwise correction results has disappeared. Can you help me on this? Thanks a lot!
Best wishes, Ting Li
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
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
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Douglas N. Greve -
Ting Li