Left Ventricle MRI Segmentation Based on an Improved Fast Snake Model

An improved fast parametric active contour model based on greedy algorithm is proposed. The original model can find the optimization curve quickly but the initial contour must be set near the feature of the image. An area energy term is thus added to the target function for capturing the image feature in a wider region. By analyzing the greedy algorithm, a local area energy term is constructed and applied in the simplified target function for snake optimization. The experiment results demonstrate that this algorithm is characterized by its rapidity and capacity which can capture the image feature in a wider region and deal with the concave regions properly, and is an effective algorithm for segmenting the left ventricle MRI.