Image-based morphometric studies of human coronary artery bifurcations with/without coronary artery disease

It is of great clinical significance to study the relationship between coronary bifurcation's morphometrical feature change and coronary artery disease (CAD) lesion. The purpose of this study is to determine the morphological changes in patients with CAD lesion when compared with non-CAD subjects and to find indicators that may be used for cardiovascular disease diagnosis. Computed tomography angiography images from Southern Chinese populations were used to reconstruct three-dimensional coronary arterial trees. Murray's law was introduced to assess the level of deviation of the realistic vascular networks from their optimal state. The results showed CAD Left had the highest deviation values of $\mathrm{ARR}$ ($0.2552\pm 0.0071)$ and $\mathrm{DERR}$ ($0.5059\pm 0.0098$), while non-CAD Right had the lowest values ($\mathrm{ARR}:$ $0.1892\pm 0.0066$ and$\mathrm{DERR}:$ $0.3733\pm 0.0092,$ respectively). Moreover, the slope values of the ratio between $D_m^3$ and $D_s^3+D_l^3$ for non-CAD Left, CAD Left, non-CAD Right, and CAD Right were 0.7428, 0.7004, 0.7628, and 0.7577, respectively. Theoretically, the slope value should equal to 1 when the bifurcation structure is in its optimal state. Therefore, these results indicated that coronary bifurcations with CAD lesion deviated from the optimal structure further than those without CAD lesion and coronary bifurcations in right were closer to the optimal structure than those in left. More importantly, the present study found that $\mathrm{DERR}$ and $\mathrm{AER}$ depended only on the diseased state, but not age, suggesting that $\mathrm{DERR}$ and $\mathrm{AER}$ were potentially used as two novel indicators for early CAD diagnosis.