AUTOMATIC MEDICAL IMAGE SEGMENTATION BY INTEGRATING KFCM CLUSTERING AND LEVEL SET BASED FTC MODEL

Recently, researchers integrate available methods in order to automate medical image segmentation, resolve their drawbacks and enjoy their benefits. Geometric deformable models based on level set have shown promising results in this area. Because, they can capture objects topologies and automatically adapt to topological changes. Among all extensions of them, Fast Two Cycle (FTC) model is the fastest one while retaining significant accuracy. But contour initialization affects its efficiency extremely. This paper proposes an integrative segmentation approach including two successive stages as follows. Firstly, the KFCM clusters input image. Then ROI's fuzzy membership matrix is injected to next stage as an initial contour. Ultimately, FTC model segments the image by curve evolution based on level set. This approach has valuable benefits including automation, noise invariant and high efficiency in terms of accuracy and speed. Simulation results show promising outputs in segmentation of different modalities of medical images.