A FAST AND ROBUST SEGMENTATION ALGORITHM FOR CEREBRAL T1-WEIGHTED MR IMAGES

摘要

In medical imaging, accurate segmentation of brain MR images is of interest for many brain manipulations. However, due to several factors such noise, imaging artefacts, intrinsic tissue variation and partial volume effects, tissue segmentation remains a challenging task. So, in this paper, a fast and accurate method for segmentation of brain MR images is presented. The method consists of five steps, First, noise removing by median filtering is done; second segmentation of brainon-brain tissue is performed by a Threshold Morphologic Brain Extraction method (TMBE). Then initial centroids estimation by gray level histogram analysis is executed. This step leads to the proposition of a Modified Fuzzy C-means (MFCM) Algorithm that is used for brain MRI tissues segmentation. Finally transformation of the fuzzy partition realised by MFCM to crisp one (defuzzyfication) is done using function of Decision by Focusing on the Neighbourhood (DFN). The TMBE efficiency is demonstrated by extensive segmentation experiments using simulated and real MR images and compared with three similar techniques through well-known performance measure. The accuracy of MFCM is evaluated on qualitatively and quantitatively.