Machine learning with MRI and PET

4 views
Skip to first unread message

Remi Silva

unread,
Jul 8, 2024, 3:27:11 AM (9 days ago) Jul 8
to Clinica
Dear Clinica Support Team,
I am trying to use the machine learning classification pipeline to classify pMCI and sMCI patients. I have created a diagnosis tsv file but I am confused by the pipeline parameters.
Does this mean I cannot use both the T1 and PET together for classification? 
  • The fourth positional argument defines the studied modality (T1w or PET).
Also, given the prerequisite can this pipeline be used for surface-based preprocessing (obtained from the t1-freesurfer and pet-surface)

You need to have performed the t1-volume pipeline on your T1-weighted MR images and/or the pet-volume pipeline on your PET images.

Many thanks,
Renata

Alice Joubert

unread,
Jul 9, 2024, 2:37:49 PM (8 days ago) Jul 9
to Remi Silva, Clinica
Hello Renata,

To answer your questions :

- The ML pipeline from Clinica supports classification for T1 and PET separately so you do need to make a choice between the two when using it.
- Surface processing is not implemented in the pipeline so it cannot be used in this manner.

Best,
Alice



--
You received this message because you are subscribed to the Google Groups "Clinica" group.
To unsubscribe from this group and stop receiving emails from it, send an email to clinica-user...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/clinica-user/1f644caf-611a-40a2-9b23-ed6de4e4b18cn%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages