Assessment of Synchronism of Left Ventricular Wall Motion

Abstract

Experienced cardiologists reviewing the RAO 30/left ventricular angiogram of a patient will usually agree on normality or abnormality of the contraction/relaxation pattern, and on the location of the abnormality if present. Thus, angiographic sequences really contain information allowing for an objective discrimination between normal and abnormal wall-motion. A great variety of quantification models have already been proposed to this purpose [1]. This, and the fact that consensus is still not obtained indicates that designing the most adequate model is not easy [2, 3]. Stating that each model has advantages, or deciding arbitrarily that realignment of the LV silhouettes is not worthy to be further investigated, may not be the right way to solve the problem. On the contrary, progress should still be achievable by improving basic concepts. The results presented in this contribution may be a step in this direction. They do not demonstrate the superiority of our model over existing ones since no cross-comparisons were performed. Our goal was rather to investigate with a new approach the old debated, but nevertheless important point: should the silhouettes of the left ventricle (LV) be realigned in order to compensate for the translation and rotation of the LV relatively to the imaging system, or should a good model include these 2 motion components, and thus quantify the overall motion (or overall displacement) of the left ventricular wall?