Automated detection of serial endocardial borders from 2D-Magnetic Resonance Images of the left ventricle

The purpose of this study was to evaluate a new semi-automated method to identify left ventricular borders from cardiac Magnetic Resonance Images. The method is based on the use of cubic B-spline to modelize the shape of the cardiac cavity. It presents a good suppleness for the tracing and the interactive deformation of the LV borders, allowing a true clinical implementation of cardiac function quantitation.