S3DL Sparse Dictionary Learning based MR Image and Lesion Segmentation

S3DL (Subject Specific Sparse Dictionary Learning) is a software tool to generate whole brain segmentation as well as lesion segmentation from multi-contrast human brain MR images. It is a patch-based method, where similarities between patches from a subject and one or more atlases are exploited to create a segmentation of the subject.

There are two parts of the tool, one to create multi-class segmentation from T1-w MR images, another to segment MS lesions from T1-w and FLAIR images. They can be used independently. In addition to the software, some training atlases and test data are provided.

The toolbox is based on the following paper,
S. Roy, Q. He, E. Sweeney, A. Carass, D. S. Reich, J.L. Prince, and D. L. Pham, "Subject-Specific Sparse Dictionary Learning for Atlas-Based Brain MRI Segmentation", IEEE Journal of Biomedical and Health Informatics, 19(5):1598-1609, 2015.
Download Now Download Now
OR See All Files


Category:Volumetric Analysis
License:Attribution Non-Commercial
Programming Language:MATLAB
Supported Data Format:NIfTI-1


Recent Activity

Document Activity

Technical Publications documentation

MLMI paper posted by Snehashis Roy on Mar 7, 2017

Document Activity

Technical Publications documentation

JBHI Paper posted by Snehashis Roy on Mar 7, 2017

File Activity

s3dl: S3DL Brain Tissue Segmentation 1.0 release

S3DL_nolesion.zip posted by Snehashis Roy on Mar 7, 2017

File Activity

s3dl: S3DL Lesion Segmentation 1.0 release

S3DL_Lesion_Segmentation.zip posted by Snehashis Roy on Mar 6, 2017

Forum Activity

open-discussion forum

Welcome to Open-Discussion posted by Snehashis Roy on Mar 6, 2017

Forum Activity

help forum

Welcome to Help posted by Snehashis Roy on Mar 6, 2017