An Information Theoretic Approach via IJM to Segmenting MR Images with MS Lesions

Abstract

Automated detection of brain pathologies from Magnetic Resonance (MR) images remains an outstanding problem. An information theoretic approach for automated segmentation of medical images called the Improved "Jump" Method (IJM) has been recently developed and validated. Here we extend this work by utilizing IJM to segment human brain MR images with multiple-sclerosis (MS) lesions in order to probe IJM's limitations and versatility.