PROJECT TITLE :

An Efficient and Robust Method for Automatically Identifying the Left Ventricular Boundary in Cine Magnetic Resonance Images

ABSTRACT:

Economical and strong identification of the left ventricular borders remains a challenging problem in cardiology. During this paper, we have a tendency to proposed an automatic methodology to segment the left ventricles and then establish their borders robustly. The proposed method is known as as “ABDC” as a result of it utilizes the strengths of 4 techniques: Automatic threshold choice; Boundary extraction, Deformation flow tracking, and Convex form modeling. We have a tendency to compared the proposed methodology with the PDE optical flow technique on 1660 pictures which are obtained from 10 complete short-axis cine MRI datasets (five traditional subjects and 5 patients). As it turned out, the proposed methodology is a lot of efficient and sturdy than the benchmark in segmenting LV borders.