MRI Segmentation via Active Contour Model Improved with Gaussian Mixture Model

Gaussian mixture model is used to approach the probability of the image's histogram for its veracity. By its feature of global optimization, the genetic algorithm is employed to calculate more accurate parameters of Gaussian mixture model. With these parameters a new sanction can be made to improve active contour models, to reduce the noise effect and prevent the curve over the weak edges. Therefore, the models fit better with MRI. Experiments on the segmentation of left ventricle magnetic resonance images show that this model performs better.