Magnetic resonance imaging (MRI) is an advanced medical imaging technique providing rich information about the human soft tissue anatomy. The goal of magnetic resonance (MR) image segmentation is to accurately identify the principal tissue structures in these image volumes.

A new unsupervised MR image segmentation method based on self-organizing feature map (SOFM) network is presented. The algorithm includes spatial constraints by using a Markov Random Field (MRF) model.

Implemented in matlab.


How to edit:

1) SOM_MRF is a matlab file contains the implementation of SOFM and MRF model.

2)ImageC.m will process the SOM_MRF file and displays the different segmentations of the image.