Segmentation of Medical Imagery with Wavelet Based Active Contour Model

A novel method for Medical image segmentation is proposed in this paper. Segmentation is one of the important key tools in medical image analysis. The main application of segmentation is in delineating an organ reliably, quickly, and effectively. In this paper, we have proposed efficient region based segmentation with wavelet transform based Active Contour (WAC) model. The proposed algorithm is segmentation of the brain and bone tissue sarcoma (BTS) present in 2D medical images. The 2D medical images large amounts of in homogeneities are present in the foreground and background. WAC model can easily distinguish the image regions in the interior, exterior, background, edges of tissues by enhancing the wavelet coefficients. The proposed WAC model utilizes the energy minimization function for solving energy functional inside and outside the contours to ensure stability. After that, it eliminates the costly re-initialization and complexity from Level Set Equation. The proposed model is stable, accurate, and immune from boundary anti-leakage and easy to implement. We get promising results obtained on real world medical images over the conventional methods.