Automatic Cardiac Motion Quantification by Extracting Landmarks from an Echocardiography Sequence

Cardiovascular diseases are a major health concern world-wide. Analysis of the cardiac motion is a standard technique for studying and understanding the etiology of heart disease such as ischemic heart disease. Ultrasound imaging is a very important and competitive medical diagnostic tool, due to its low cost, short acquisition time, and non-invasive nature. Echocardiography is a widely used ultrasound imaging technique to examine myocardial function and cardiac motion in patients with known or suspected heart disease. Whereas echocardiographic imaging has reached a high level of technical sophistication, the myocardial function and cardiac motion analysis still relies mainly on the visual assessment and manual measurements by experienced cardiologists. Automating the analysis of echocardiographic images is therefore highly desirable but also challenging because of the low image quality and the high amount of speckle noise. In this paper, we present a fully automatic method that does not require segmentation, our approach is non invasive, because there is no injection needed. Besides, no markers are physically implanted on the surface of the ventricle wall. This approach has two parts, first, we extracted landmarks from echocardiography sequence, and this step is based on corner detection algorithms, especially Harris and SUSAN corner detectors. Second, landmarks using Gradient Vector Matcher (GVM) and Product Moment Coefficient Matcher (PMCM) were tracked.