Methods for Improved Efficacy in Segmentation and Tracking of Echocardiographic Images

Echo cardiography is one of the major imaging modalities for quantifying heart functionality, providing advantages such as real-time features, cost-reflectiveness, imaging efficiency, and safety compared to other image modalities. Many cardiac procedures for diagnosis and treatment of cardiac disorders require real-time segmentation of the endocardium walls and robustness to potential image imperfections (such as shadows, speckle noise and possible movements). Manually segmenting the cardiac boundary is a time-consuming process, and in emergency cases, the results are required to be available instantly. Therefore, providing an autonomous segmentation method is crucial in such situations. Also, tracking the cardiac boundary could provide valuable information about the heart condition. The primary goal of this work was to address issues with accuracy in segmentation and tracking of echocardiograms. Meanwhile, the robustness to noise and motion, as well as translation, involved in cardiac operations, were taken into account.