A statistical 3-D segmentation algorithm for classifying brain tissues in multiple sclerosis

Deterministic annealing (DA) algorithm has been successfully used to segment the simulated magnetic resonance normal brain images [1]. This paper presents the results of applying it to simulated and actual clinical multiple sclerosis (MS) magnetic resonance (MR) brain data with the objective of developing a computer-aided diagnostic (CAD) tool for early detection and follow-up for MS lesions. Multiple sclerosis lesions on T1 simulated brain images [2] can be obtained by segmenting the image data using deterministic annealing algorithm and then performing further arithmetic manipulations on these segmented images. Lesions in clinical T2 multiple sclerosis MR images are isolated entities in the segmented images of white matter, gray matter and cerebrospinal fluid. The achieved results demonstrate the ability of deterministic annealing algorithm to isolate MS lesions from clinical MR data, thus providing a potential CAD tool for the clinicians.