Clinical validation of an automated boundary tracking algorithm using cardiac MR images

Manual tracing of ventricular borders in MRI images is a tedious process. The goal of this research was to develop an algorithm for tracking the borders of the left ventricle (LV) in a cine-MRI gradient-echo temporal data set. The algorithm was validated on healthy volunteers and patients with dilated cardiomyopathy (DCM) before and after surgical remodeling of the LV. A full tomographic set (~11 slices/case) of short axis images were obtained through systole. Initial endocardial and epicardial contours for the end-diastolic (ED) and end-systolic (ES) phases were manually traced on the computer by an expert. The initial tracing of the ED borders were used as starting point. These borders were then temporally tracked through each phase of the temporal data set, until the ES phase was reached (~7 phases/slice). Peak gradients along equally spaced chords calculated perpendicular to centerline determined midway between the endocardial and epicardial borders were used for border detection. This approach was tested by comparing the LV epicardial and endocardial volumes calculated at ES to those based on the tracings by an expert. The LV volumes based on the borders obtained from the automated algorithm correlated well with those calculated from the expert manual tracings