This automated pipeline can be used for accurate Corpus Callosum (CC) segmentation across multiple MR modalities (T1, T2 and FLAIR) and extract a variety of features to describe the shape of the CC. We also include an automatic quality control function to detect poor segmentations using Machine Learning.

Clone the github directory using:

git clone https://github.com/ShrutiGadewar/smacc.git

Navigate to the "smacc" folder and then create a virtual environment using the requirements.txt file:

conda create -n smacc python==3.11 -y
conda activate smacc
pip install -r requirements.txt
pip install .

All the MR images should be registred to MNI 1mm template(182 X 218 X 182) with 6dof. You can use the template provided in the "model" folder on github. You can use the FSL's flirt command for linear registration:

flirt -in ${inpdir}/${subj}.nii.gz \
	-ref ${MNI_1mm_template} \
  	-out ${outdir}/${subj} \
 	-dof 6 \
  	-cost mutualinfo \
  	-omat ${outdir}/matrices/${subj}_MNI_6p.xfm

smacc -f ./subject_list.txt -o ./smacc_output -m t1

-f : Text file with a list of absolute paths to the niftis to be processed and names to save the outputs for each subject. Check example text file "subject_list.txt" provided.
-o : Absolute path of output folder
-m : Modality of the images to be processed (t1/t2/flair)
-q : Optional flag to perform Automated QC on the segmentations
The final output is a csv which will contain all the extracted shape metrics and a column "QC label" indicating whether the segmentations were accurate(0)/fail(1) if the QC flag is provided.

Gadewar SP, Nourollahimoghadam E, Bhatt RR, Ramesh A, Javid S, Gari IB, Zhu AH, Thomopoulos S, Thompson PM, Jahanshad N. A Comprehensive Corpus Callosum Segmentation Tool for Detecting Callosal Abnormalities and Genetic Associations from Multi Contrast MRIs. Annu Int Conf IEEE Eng Med Biol Soc. 2023 Jul;2023:1-4. doi: 10.1109/EMBC40787.2023.10340442. PMID: 38083493.