Quantitative regional curvature analysis: An application of shape determination for the assessment of segmental left ventricular function in man

Abstract

All traditional techniques of regional ventricular function analysis depend upon one or more assumptions about coordinate, reference, or indexing systems, idealized ventricular geometry, and the uniformity of ventricular contraction. Therefore, a method of shape analysis was developed that allows the quantitation of regional curvature and is independent of the assumptions outlined. This was implemented on a commercial image processing unit and applied to silhouettes of 30-degree right anterior oblique left ventriculograms. Three groups with abnormal wall motion (anterior abnormality, N = 23; inferior abnormality, N = 23; anterior and inferior abnormalities, N = 22) were analyzed and compared to a group with normal regional function (n = 22). Relatively few significant quantitative curvature differences were noted at end diastole among the groups. These few abnormalities described a slight increase in curvature or globularity of the anterior and inferior walls. More marked and extensive aberrations were detected at end systole. The group with anterior wall motion disturbances showed four distinct areas of curvature abnormality. Excessive curvature was present on either side of the apex (anterior and inferoapical regions) and apical curvature was less than normal. The fourth region was in the inferior zone, which showed curvature values that were less than normal, suggesting increased inward motion contraiateral to the anterior abnormality. The group with inferior wall motion abnormalities also showed excessive end-systolic curvature on either side of the apex (diaphragmatic and anteroapical zones) and deficient curvature at the apex. A combination of these regional morphologic abnormalities was noted in the group with both anterior and inferior dysfunction. Regional shape change (end-diastolic minus end-systolic curvature) largely reflected the same abnormalities as the analysis of end-systolic curvature alone. In conclusion, shape abnormalities secondary to regional dysfunction can be quantitated on a segmental basis. This form of analysis provides a conceptually different approach to the quantitation of regional ventricular function that is divorced from the assumptions mandated by traditional techniques.