International Journal of Artificial Intelligence and Neural Networks
Author(s) : CHE-LUN HUNG, CHIA-CHEN LIN, JIEH-SHAN YEH, YAW-LING LIN, YU-CHEN HU, YUAN-HUAI WU
In the computer aided medical image process, image segmentation is always required as a preprocess stage. Fuzzy c-means (FCM) clustering algorithm has been commonly used in many medical image segmentations, particularly in the analysis of magnetic resonance (MR) brain image. However, all of these FCM methods are computation consuming that is difficult to be used in real time application. In the paper, we proposed a Parallel FCM algorithm based on graphic process units (GPUs) to accelerate computation speed of timeconsuming FCM applications. The experimental results show that the proposed algorithm can reduce the computational cost dramatically.