AFID QC Feature #24
AFID QC Feature #24
Conversation
…toAFIDs placements
Dhananjhay
force-pushed
the
afidqc-feature
branch
from
79bfba3 to
4d544b5
Compare
|
Amazing work here @ataha24! I was successfully able to run this feature and it looks great! I've resolved the merge conflicts and updated the dag, but would need @mackenziesnyder to work her magic once again to create the final version of the dag. |
This worked well for me!! One thing I noticed (pictured above) is that everything in the qc is labelled Landmark 1-32. In the lab meeting I think I remember a comment about naming them by the landmark name like AC, PC - I just wanted to make sure this naming was intentional for this PR in case the most recent changes aren't pushed yet.
|
autoafids/workflow/scripts/utils.py
Outdated
|
|
||
| html_content += f""" | ||
| <div class="container"> | ||
| <h2>Landmark: {label}</h2> |
There was a problem hiding this comment.
ahh great catch @mackenziesnyder -- we probably just need to index the dictionary to print the name here.
(i.e., {afids_labels[label]}
There was a problem hiding this comment.
I updated code to now call the short form AFID. I think they probably would mean the same to a new user (i.e., they may not be familiar with the namings). But, I guess short form may still be a bit more informative than just a number?
There was a problem hiding this comment.
Yes just tested it out, I agree that makes sense to me
|
@Dhananjhay & @mackenziesnyder looks like this one is good 2 go! |
|
Great work guys! Merging it now. |
🚀 PR: AFID QC Feature
🧠 Overview
This PR enables a QC feature which takes as input the AFIDs predicted via AutoAFIDs and generates an interactive .html for the user. The user can toggle different MRI slices and visualize the placements at different anatomical slices with a red crosshair showing the location of the AFID of interest. Furthrmore, the user can pan between the template defined placement and the subject placement to enable for a more informed QC process.
🔍 How It Works
New functions which slice MRI at various levels around the coordinates of interest.
📂 Outputs of stereotaxy.smk
.html file: Contains the predicted coordinates and on native space raw MRI image.
🔧 Changes to the Codebase
utils.py
Added new helper functions for *.html file generation, and various helper function.
fidqc.smk
New rule to take the input image of the workflow and corresponding .fcsv file predictions.
📁 Resource Additions
None.
📚 Documentation Updates
Instructions on how to use the fidqc feature in config file.
None.