A computational approach for the simultaneous generation of competing and optimized arterial trees based on...

geometric modeling of arterial trees is fundamental for computational studies of the cardiovascular system and may contribute to applications such as hemodynamic analysis and vascular modeling. However, constructing realistic synthetic arterial networks beyond the resolution of current medical imaging remains challenging due to the limited availability of anatomical data.

we propose the computational algorithm Competing Optimized Arterial Trees (COAT), designed to generate three-dimensional models of arterial networks composed of multiple competing trees under physiological flow and pressure conditions. In COAT, intravascular volume is minimized through local Constrained Constructive Optimization (CCO)-based growth mechanisms, while multiple trees evolve simultaneously within a shared perfusion domain. Unlike the classical CCO method, which generates a single arterial tree per execution, COAT enables the coordinated generation of multiple interacting trees through a flow-driven competition mechanism combined with adaptive domain subdivision based on weighted Voronoi diagrams.

Computational simulations indicate that the proposed framework generates vascular forests that reproduce qualitative morphometric trends reported in the literature, show a strong association between prescribed flow fractions and occupied perfusion territories, and exhibit emergent scaling trends that are compatible with classical allometric arguments in vascular biology.

The proposed framework provides an efficient and flexible approach for generating physiologically plausible multi-tree vascular networks which enables the study of flow competition and territorial organization and has potential applications in multiscale modeling.