Automatic segmentation of the epicardium and endocardium from MR image based on hough transform and geodesic active contour model

We propose an approach to segment the epicardium and endocardium in cardiac image automatically. Based on the knowledge on the circle-like shape of epicardium and endocardium, we first use Hough transform to detect circles for initial contours of the left ventricle. Based on the geodesic active contour model by integrating K-means clustering to provide regional information and the anatomical constraints, we then segment both the epicardium and endocardium automatically, from the initial contour detected by Hough transform. Experimental results demonstrate the effectiveness of our approach.