Locating cardiac walls with Adaboost and level sets

Locating the endocardium and epicardium in the echocardiograms is very hard with conventional segmentation methods in the lack of prior information. This paper presents a cardiac wall locating technique by introducing local and global priors through Adaboost and level sets. The global prior is incorporated into the system by regularly reinitializing the level set surface under the influence of the expert detected contours. The local priors with image and temporal information, extracted with 2D and 3D Haar filters, are learned through Adaboost. The system is validated on the echocardiograms and the results are found promising.