Stochastic active contour for cardiac MR image segmentation

We develop an energy based automatic image segmentation algorithm using a novel active contour scheme. The algorithm overcomes some unique challenges arising in cardiac MR images. Two features are particularly relevant. The first is that it uses region-based information captured by a stochastic model. As a result, our method is robust to assumed initial conditions and can be applied to a large range of images, particularly when the contrast is low. The second feature is the incorporation of prior knowledge on the shape of the organ to be segmented. For cardiac image segmentation, it is sufficient to assume that the shape resembles an ellipse.