On 2/22/19 11:43 AM, john Anderson wrote:

External Email - Use Caution

Dear Dr Greve,

Thank you so much for the response! Indeed, I used MRTM2 for HB pathological regions and the results were very confusing.... Based on your response bellow, I understand that for pathological regions I can feed "bp.nii.gz" files from "MRTM1" not "MRTM2" to a surface based analyses. Is this correct? You can actually use either MRTM1 or MRTM2 for further analysis, but I think the way I have it documented, you would use MRTM2.

Also for the flag "--km-ref", I have interest in doing whole brain normalization (i.e. all ROIs in wmparc.mgz) in this case do I need to feed all the numbers relevant to all ROIs in gtmseg.ctab. Is there any way to let --km-ref know that I want to do whole brain normalization without feeding a very long list of numbers? No, I don't think so. Sorry, I never considered this case. Are you sure that is what you want to do?

Thank you so much for help John

On 2/22/19 7:58 AM, john Anderson wrote:

External Email - Use Caution

Hi Dr Greve,

I would like to use petsurfet to do kinetic modeling (KM), the pipeline is straightforward and easy to use. Thank you so much! I would appreciate any clarifications relevant to my questions bellow:

1) I understand that the flag "--km-ref" define the reference region for normalizing PET signal. in the linear below the reference region is cerebellum cortex. If we need to normalize to occipital just we change the numbers based on labels in aparc+aseg atlas?

Yes.

--km-ref 8 47 --km-hb 11 12 13 50 51 52 --no-rescale

2) --km-hb define the high binding regions. Are these regions represent the regions that are pathologically involved where we expect it to have higher PET signal OR it represent the regions where PET tracer has the maximum binding in general regardless of pathology? I mean every tracer have high level in specific regions in the brain is this what the flag is referring too?

I have not dealt with pathological cases with MRTM2. I would avoid using them for the HB region as their kinetics may be different than healthy tissue.

Thanks for any clarification,

John

_______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

On 2/22/19 1:47 PM, john Anderson wrote:

External Email - Use Caution

Hi Dr Greve,

Yes, in my lab we normalize DPA713 tracer to whole brain mean using the command fslmaths something like fslmaths pet.nii -mask mask -inm 1 pet_normalized_whole_brain.nii.gz

For kinetic modeling.. kindly I would like to be sure that I got it right. I am doing kinetic modeling for two groups of subjects patients and controls. I am following PET surfer tutorial. In the case of patients and healthy controls comparison I can use MRTM1 and MRTM2 correct? I don't know about DPA713. Generally, there will be a lot of studies done to verify the kinetic models that can be used. MRTM2 makes more assumptions than MRTM1, so it is not applicable to everything.

Also for the flag "--km-hb" are the IDs that I need to feed to this flag represent regions of high binding affinity in patients or healthy controls or it doesn't matter just regions known to have high binding. For example some tracers show regions of high binding affinity on healthy scans as well as patients scans. Again, you should look through the literature and see what people have done to verify MRTM2 for your tracer.

On 2/22/19 11:43 AM, john Anderson wrote:

External Email - Use Caution

Dear Dr Greve,

Thank you so much for the response! Indeed, I used MRTM2 for HB pathological regions and the results were very confusing.... Based on your response bellow, I understand that for pathological regions I can feed "bp.nii.gz" files from "MRTM1" not "MRTM2" to a surface based analyses. Is this correct?

You can actually use either MRTM1 or MRTM2 for further analysis, but I think the way I have it documented, you would use MRTM2.

Also for the flag "--km-ref", I have interest in doing whole brain normalization (i.e. all ROIs in wmparc.mgz) in this case do I need to feed all the numbers relevant to all ROIs in gtmseg.ctab. Is there any way to let --km-ref know that I want to do whole brain normalization without feeding a very long list of numbers?

No, I don't think so. Sorry, I never considered this case. Are you sure that is what you want to do?

Thank you so much for help

John

On 2/22/19 7:58 AM, john Anderson wrote:

External Email - Use Caution

Hi Dr Greve,

I would like to use petsurfet to do kinetic modeling (KM), the pipeline is straightforward and easy to use. Thank you so much! I would appreciate any clarifications relevant to my questions bellow:

1) I understand that the flag "--km-ref" define the reference region for normalizing PET signal. in the linear below the reference region is cerebellum cortex. If we need to normalize to occipital just we change the numbers based on labels in aparc+aseg atlas?

Yes.

--km-ref 8 47 --km-hb 11 12 13 50 51 52 --no-rescale

2) --km-hb define the high binding regions. Are these regions represent the regions that are pathologically involved where we expect it to have higher PET signal OR it represent the regions where PET tracer has the maximum binding in general regardless of pathology? I mean every tracer have high level in specific regions in the brain is this what the flag is referring too?

I have not dealt with pathological cases with MRTM2. I would avoid using them for the HB region as their kinetics may be different than healthy tissue.

Thanks for any clarification,

John

_______________________________________________

Freesurfer mailing list

Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu

https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

_______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edumailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer