Automatic white matter lesion detection and segmentation on Magnetic Resonance Imaging: A review of past and current state-of-the-art

摘要

White matter lesion (WML) is an abnormal tissue occurring in white matter. It indicated the damage of the myelin sheath that used to surround the axon of a neurone. This resulting neurological and vascular disorder occur in the patient, also commonly developed in the healthy brain of elderly. Magnetic Resonance Imaging is a non-invasive medical equipment preferred choice by the clinician to diagnose and observed the injury of brain tissue. However, WML quantitative assessment and analyse on the large volume of MR imaging is a challenge. In this paper, we provide an intensive review of the past and recent WML delineation and detection methods. This review included visual scoring assessment, a common preprocessing step for WML segmentation, false positive elimination, and the latest automatic WML segmentation approaches will be presented.