Efficient Segmentation of Liver Image using Spatial FCM with Level Sets enhanced by Mumford-Shah Approach

Liver segmentations and tumor detection plays a vital role in medical environments. The performance of the segmentation using level sets is subjected to appropriate initialization and optimal configuration of controlling parameters. Basically, standard FCM does not fully utilize the spatial information in the clustering of the image. In this paper, initial segmentation is done by spatial fuzzy clustering and the controlling parameters of level set evolution are also estimated from the results of fuzzy clustering. Moreover the fuzzy level set algorithm is enhanced with locally regularized evolution. Such improvements facilitate level set manipulation and lead to more robust segmentation. An innovation in this paper is we are using Mumford shah approach to increase the robustness. The advantages of the new method is as follows, it automatically detects the interior contours, increase the robustness with respect to noise, detects blurred contours and allows for automatical change of topology which is not possible in other techniques. It is expected that we can minimize the Mumford-Shah functional and the results produced by this approach confirm its effectiveness for both single and multiple feature data with spatial information.