Entropy based Fuzzy clustering algorithm for MR image segmentation

Derived a Fuzzy C-Means algorithm based on Shannon entropy for segmenting volumetric brain MR images. The model was evaluated on the Brainweb dataset and IBSR dataset before being utilized for the real-patient image volumes. The modified fuzzy clustering algorithm delivered relatively stable clustering results and was less prone to noisy MR images. The improvement in segmentation and tissue segmentation accuracy was found to be 4-7% in different scenarios over previous methods.