Multiple Sclerosis Lesions Detection from Noisy Magnetic Resonance Brain Images Tissue
Samah Yahia, Yassine Ben Salem, Abdelkrim Mohamed Naceur · 2018
Disease detection performance can be significantly influenced by Magnetic Resonance (MR) image quality. The noise is one of the major intensity modifiers in MR images. This paper is focused on Multiple Sclerosis (MS) Lesions detection from noisy MR brain images tissue. This is a chronic autoimmune inflammatory that appears in the central nervous system. This paper presents a new method for MR brain images analysis: the Decimal Descriptor Patterns (DDP). Experiments are performed over the 3D Brainweb database. The classification is done based on the classifier multiclass Support Vector Machines (SVM). The Grey Level Co-occurrence Matrix (GLCM) and the Local Binary Patterns (LBP) two methods of texture analysis, always considered as references in image analysis are used for comparison. Experimental results show that the DDP approach is very effective to detect MS Lesions; and demonstrate its stability in front of the noise with respect to two different Tl- and T2-weiahted MR images.