search

Mathilde-16 / STOP_MS_data

0 stars 0 forks source link

comparison of segmentation methods #12

Closed jcohenadad closed 5 months ago

jcohenadad commented 5 months ago

https://github.com/Mathilde-16/STOP_MS_data/blob/24e106b3ddfe728378e7a3854cacdec42fd345aa/Data_processing/ContrastAgnosticSeg_%26_Vertebral-labeling_BIDS-data/Batch_script.sh#L66

to cleanup the code I suggest you document the difference between the two algos, pick the best one, and remove the code related to the other method

Mathilde-16 commented 5 months ago

For me, I think the best one is the one which uses the sct_deepseg command (the contrast agnostic segmentation model), indeed the QC here shows a better segmentation with this model than with the sct_deepseg_sc command.

So you agree if I remove the code related to the sct_deepseg_sc ?

Mathilde-16 commented 5 months ago

Thanks for your approval, I've commited the cleaned code

jcohenadad commented 5 months ago

Thanks for your approval, I've commited the cleaned code

please always cross-ref to the commit SHA next to your comment "I've commited the cleaned code" to avoid confusion (ie: the code is constantly evolving)

Mathilde-16 commented 5 months ago

Yes of course sorry !

jcohenadad commented 5 months ago

please note that pointing to the diff of the commit is more useful because it enables to see what differs from the previous version (easier to quickly assess the changes)

jcohenadad commented 5 months ago

Issue can be closed

Mathilde-16 commented 5 months ago

please note that pointing to the diff of the commit is more useful because it enables to see what differs from the previous version (easier to quickly assess the changes)

Okay I understand, thank you !