Multiple Sclerosis Lesion Segmentation using Magnetic Resonance
Imaging (MRI)
Description
Trains and makes predictions from the OASIS method, described in
detail in the paper "OASIS is Automated Statistical Inference for Segmentation,
with applications to multiple sclerosis lesion segmentation in MRI"
.
OASIS is a method for multiple sclerosis (MS)
lesion segmentation on structural magnetic resonance image (MRI) studies. OASIS
creates probability maps of lesion presence using the FLAIR, T2, T1, and PD
structural MRI volumes. This packages allows for training of the OASIS model
and prediction of OASIS probability maps from a trained model with user supplied
studies that have a gold standard lesion segmentation masks. The package will
also create OASIS probability maps for MRI studies using the OASIS model from
the OASIS paper if no gold standard lesion segmentation masks are available.