Statistics and Its Interface

Volume 12 (2019)

Number 2

An integrative classification model for multiple sclerosis lesion detection in multimodal MRI

Pages: 193 – 202

DOI: https://dx.doi.org/10.4310/SII.2019.v12.n2.a1

Authors

Fengqing (Zoe) Zhang (Department of Psychology, Drexel University, Philadelphia, Pennsylvania, U.S.A.)

Ji-Ping Wang (Department of Statistics, Northwestern University, Evanston, Illinois, U.S.A.)

Wenxin Jiang (Department of Statistics, Northwestern University, Evanston, Illinois, U.S.A.)

Abstract

We study a classification problem of multiple sclerosis (MS) lesions in three dimensional brain magnetic resonance (MR) images. Segmentation of MS lesions is essential for MS diagnosis, assessment of disease progression and evaluation of treatment efficacy. Accurate identification of MS lesions in MR images is challenging due to variability in lesion location, size and shape in addition to anatomical variability between subjects. We propose a supervised classification algorithm for segmenting MS lesions, which integrates the intensity information from multiple MRI modalities, the texture information, and the spatial information in a Bayesian framework. A multinomial logistic regression is employed to learn the posterior probability distributions from the intensity information, combined from three MRI modalities. Texture features are selected by the Elastic Net model. The spatial information is then incorporated using a Markov random field prior. Finally, a maximum a posteriori segmentation is obtained by the graph cuts algorithm. We illustrate the effectiveness of our proposed model for lesion segmentation using both the synthetic BrainWeb data and the clinical neuroimaging data.

Keywords

supervised classification algorithm, multiple sclerosis, segmentation, multimodal MRI

2010 Mathematics Subject Classification

Primary 62H30. Secondary 62H35.

Full Text (PDF format)

Received 3 April 2018

Published 11 March 2019