MRI brain tissue clustering segmentation method

Abstract

The invention discloses an MRI (Magnetic Resonance Imaging) brain tissue clustering segmentation method. The method comprises the following steps of: setting tissues except spinal fluid, grey matter and white matter in an MRI brain image as an image background and removing the image background; calculating the intensity and position similarity between pixels in the MRI brain image to form a bilateral similarity matrix Bsm; segmenting pixels of the MRI brain image into a set C = {Ck, k = 1, 2,... P} by using an FCM algorithm; calculating a weight coefficient set Vk = {gamma 1, gamma 2,... gamma N} of pixels in the Ck; sorting the weight coefficients in the Vk to obtain a first R weight coefficient set VkR = {gamma'1, gamma'2,... gamma'R}, wherein the set V = {VkR, k = 1, 2,... P} is a main multi-clustering center set of C; distributing labels in a K neighbor mode according to the bilateral similarity matrix Bsm and distribution of the set V, determining a point with the farthest distance in the K neighbor range of elements in the V as a secondary clustering center, forming a secondary clustering center set A1, enabling a plurality of times of secondary clustering to form a set Az, wherein the set A = {A1, A2,... Az} is a set of all the secondary clustering centers of the C. Compared with the prior art, the scheme of the invention avoids falling into local optimum, and the MRI brain image segmentation result distribution is more balanced.

Description

technical field [0001] The invention relates to the technical field of MRI brain image segmentation, in particular to an MRI brain tissue clustering and segmentation method. Background technique [0002] Magnetic Resonance Imaging (MRI) has the advantages of high soft tissue resolution, no ionizing radiation damage, free selection of sections, and multi-sequence imaging. It has been widely used in brain and soft tissue imaging. The human brain is mainly composed of gray matter, white matter and cerebrospinal fluid. In order to make the analysis, diagnosis and treatment of brain diseases more convenient and accurate for doctors, it is of great significance for clinical diagnosis and treatment to divide MRI brain images into cerebrospinal fluid, gray matter and white matter. [0003] In the prior art, the segmentation of MRI brain images into cerebrospinal fluid, gray matter, and white matter faces three types of technical problems: Partial Volume (PV) effect, noise, and offse...

AI Technical Summary

Problems solved by technology

However, the technical scheme based on FCM and its extended algorithm is easy to fall into local optimum, which makes the segmentation effect of MRI brain images poor

Images