help
help > RE: Co-registration ROI with structural data
Jun 18, 2019 03:06 PM | Stephen L. - Coma Science Group, GIGA-Consciousness, Hospital & University of Liege
RE: Co-registration ROI with structural data
Originally posted by Mickael Tordjman:
Dear Mickael,
If I understand correctly, you have one lesion ROI per subject, done on the subject's T1. If that's the case, what you would need to do would be to first pre-process your whole dataset, and then apply the normalization transform that was calculated for the T1 onto the lesion ROI, this way you will get normalized lesion ROIs. Then, you can load these ROIs in CONN.
As I understand, you have an "anti" lesion ROI. I think it would work better if you compute the opposite mask, the mask covering only the lesion. In that case, after normalizing these ROIs, you can set them as regressors in CONN setup, and Denoising, so that any noisy signal coming from and to these lesional regions would be effectively regressed out.
Hope this can help,
Best regards,
Stephen
Hello,
I am a new user of Conn, and I have a question regarding ROI.
I want to process data with brain lesions. In order to do so, I wanted to add a ROI with a mask of the entire brain without the lesion.
These ROIs are not in the MNI space.
Should I co-register these ROIs (drawn on the original MPR T1) with the initial data, or after the pre-processing steps?
If I need to co-register the pre-processed data with my ROIs, where is the pre-processed MPR T1 saved?
Thanks a lot for your help!
Best,
Mickael
I am a new user of Conn, and I have a question regarding ROI.
I want to process data with brain lesions. In order to do so, I wanted to add a ROI with a mask of the entire brain without the lesion.
These ROIs are not in the MNI space.
Should I co-register these ROIs (drawn on the original MPR T1) with the initial data, or after the pre-processing steps?
If I need to co-register the pre-processed data with my ROIs, where is the pre-processed MPR T1 saved?
Thanks a lot for your help!
Best,
Mickael
Dear Mickael,
If I understand correctly, you have one lesion ROI per subject, done on the subject's T1. If that's the case, what you would need to do would be to first pre-process your whole dataset, and then apply the normalization transform that was calculated for the T1 onto the lesion ROI, this way you will get normalized lesion ROIs. Then, you can load these ROIs in CONN.
As I understand, you have an "anti" lesion ROI. I think it would work better if you compute the opposite mask, the mask covering only the lesion. In that case, after normalizing these ROIs, you can set them as regressors in CONN setup, and Denoising, so that any noisy signal coming from and to these lesional regions would be effectively regressed out.
Hope this can help,
Best regards,
Stephen
Threaded View
| Title | Author | Date |
|---|---|---|
| Mickael Tordjman | Jun 14, 2019 | |
| Stephen L. | Jun 18, 2019 | |
| Mickael Tordjman | Jun 18, 2019 | |
| Stephen L. | Jun 18, 2019 | |