Abstract
Non-invasive techniques for measuring the dynamic behavior of the cardiac ventricle can be invaluable tool for diagnosing heart disease. In this paper, we present methods for quantitative analysis of cardiac function using magnetic resonance images.
A 256×256 magnetic resonance transaxial image of the cardiac ventricle was obtained from a healthy subject. After gray level thresholding and region segmentation, the boundary of the ventricular chamber was extracted. Then, the boundaries of the ventricular chamber were displayed three-dimensionally using the Z-buffer algorithm. Thus, ventricular volume and ejection fraction were calculated. In our example, left ventricular ejection fraction was 61%. These results agree well with the corresponding data obtained by the echo-cardiography.
