integrating generate renders to pipeline

nimh-dsst / dsst-defacing-pipeline

Defacing algorithm to improve and evaluate accuracy for large datasets.

4 stars 2 forks source link

integrating generate renders to pipeline #32

Closed Arshitha closed 1 year ago

Arshitha commented 1 year ago

This PR includes the following changes:

visualqc t1_mri prep code removed
integrated generate renders function into the defacing scripts
some light code refactoring to consistently use pathlib library

I've tested the code on two sample datasets from start to end, and its working as expected.

Things to do before it can be merged:

README edits

Arshitha commented 1 year ago

README has now been updated.

ericearl commented 1 year ago

I spoke too soon! I got a lot of:

Has the render been created? False

I was testing over an SSH connection on curium from my VPN'ed PC. I don't know whether or not that should have worked... I will try direct on curium one more time and then I'm done. But this begs the question, "how can a user ONLY generate renders if they fail the first time and save all that defacing again?" we may want to file an issue to add --only-generate-renders to skip the defacing over again.

Arshitha commented 1 year ago

@ericearl I'll try this out on curium too but just to confirm, did you have fsleyes on your env?

ericearl commented 1 year ago

Oy. I did forget to activate conda again. Testing over right now...

ericearl commented 1 year ago

It was the conda environment, got me again.

Arshitha commented 1 year ago

@ericearl Great, thank you! I'll merge this to main and do some more testing with the main branch.