A technique for Quantitative Coronary Analysis of Cine-Angiograms using segmentation and vessel path tracking

Abstract

Coronary Cine-angiogram (CCA) is one of the widely used preliminary invasive medical image modalities in Interventional Cardiology for the detection of luminal obstructions in Coronary Artery (CA) vasculature. It is apparent that the clinical judgments based on angiography are subjective and lead to overestimation and underestimation of the detected stenosis. Hence, such imprecise diagnosis can have an adverse impact on the patients' quality of life imperfectly. This study elaborates a novel method for Quantitative Coronary Analysis (QCA) for objectively assessing the severity of the detected stenosis in main CAs recorded in direct CCAs. The proposed QCA method is based on the segmented vessels' geometry and vessel path tracking and consists of five major implementation phases namely; preprocessing, segmentation, vessel isolation, vessel path tracking and diameter calculation. The result of the proposed method has successfully proven its ability to detect relevant features, clinical supportability for stenosis detection and capability to objectively assess the clinically approved stenosis regions. Moreover, this proposed method can be further extended to assess the stenosis severity based on the functional significance of the CAs.