Computerized segmentation of liver tumor using integrated fuzzy level set method

Abstract

CT abdominal image requires the automated diagnosis of part of the liver and lesions. It is challenging to segment the liver and the tumor due to the high strength resemblance between liver and other organs nearby. In this paper, an automatic method of segmenting liver from CT image using fuzzy level set algorithm is proposed. It can evolve immediately through spatial fuzzy clustering from preliminary segmentation. Reasonable initialization and effective specification of controlling parameters requiring significant manual intervention are subject to the efficiency of the level set segmentation. In the following ways, the algorithm is considerably improved. First during adaptive optimization, fuzzy clustering integrates spatial information, which removes intermediate morphological operations. Secondly, the level set segmentation controlling parameters are now extracted directly from the performance of fuzzy clustering. Thirdly, a approach is suggested to regularize the evolution of the level collection, which is distinct from other approaches, driven by fuzzy clustering. Finally, the fuzzy level set algorithm on CT liver was tested. Performance analysis of this algorithm was carried out in various modalities on medical images. The results supported its suitability for segmentation of liver tumor.