search

Australian-Imaging-Service / pipelines

Scripts to generate analysis pipelines that can be run in XNAT's container service
Other
1 stars 3 forks source link

[STORY] Summary of QC methods #222

Open tclose opened 1 year ago

tclose commented 1 year ago

Epic: #35 Feature: #66 Feature Release: Required knowledge: mid-level

Description

As a member of pipeline team, I want to prepare a summary of the most recent and popular dMRI QC methods, so that I can choose the proper dMRI QC method and find out the gaps to address.

Acceptance Criteria

Mahdieh-dst commented 1 year ago

dMRI QC.pptx

Please find the "dMRI QC.pptx" file attached.

Some points:

(i) All the introduced pipelines do some preprocessing and generate single subject or study-wise QC report.

(ii) The generated report should be manually reviewed or the threshold of outliers should be set by the user.

(iii) I found "PreQUAL" pipeline (slides 2-4) the most interesting one as it is very recent (2021), does a comprehensive pre-processing on dMRI images, and generates a comprehensive QC report including graphs and images.

(IV) "dMRIQCpy" (slide 6) is a very recent (2022) python-based dMRI QC toolbox for dMRI QC. I did not find much documentation on that.

(V) The deep learning based QC methods (last slide) have been trained to do slice artifact detection (e.g. ghosting, motion, chemical shift , etc.). However, (I guess) the user should manually go through the detected slices and evaluate the amount of artifact (accept or reject).