A tower layered algorithm for medical image segmentation

This paper proposes an algorithm named Tower Layered FCM to improve the traditional FCM method. A Tower layered structure with a constraint is added into the objective function and pixel´s neighbor information is rational used. The Tower layered structure can speed up the algorithm and constraint the subordinate degree of pixels. So the subordinate degree of clustering center is made more reasonable. The experimental result shows that the new algorithm can save detail image information and it is more efficient than the traditional FCM.