Fractional Active Contour Model for Edge Detector on Medical Image Segmentation

In computer terms, segmentation is a process to partition or to divide an image based on the number of objects within the image. The process of segmentation can be easy depending on the quality of the image such as the level of noise, the image contrast and etc. Segmentation on medical images has its own importance such as to extract an importance object like tumor or others, to assist the physician in making decision for surgery purposes. Currently many methods have been developed but to get accuracy in segmenting multi modalities of medical images are still remain unsuccessful. Among all methods, Active Contour Model shows good potential in medical image segmentation. But, accuracy in detecting edges along the object boundary is still remain unsuccessful. The used of fractional calculus that act as the first order integer to extract the missing pixels along the object boundary is seen to have the potential in solving the problem. This paper proposed a method called, Fractional Active Contour (FAC) model. The proposed method tends to highlight the role of fractional as an operator to detect and preserve the missing edges as well as giving the bending capability to the contour of the model. Experiments on several medical images from MRI, CT Scan and X-ray images demonstrates that the proposed FAC with the usage of the powerful fractional calculus as the edge detector model realizes an accurate boundary segmentation although under the constraint of missing edges within the environment of intensity inhomogeneity.