fig1

Figure 1. From medical DTs to digital vessel. (A) Architecture of an integrated DT framework. This model combines data-driven and mechanistic approaches for multiscale physiological simulation. A Graph Neural Network (GNN) predicts organ-level hemodynamics, while a Generative Adversarial Network (GAN) simulates molecular-level transcriptomic crosstalk. These are integrated with Ordinary Differential Equations (ODEs) to capture system dynamics, enabling a holistic patient model^[7]; (B) A conceptual workflow for generating a virtual patient database from clinical data using a one-dimensional hemodynamic model^[42]; (C) A detailed 55-branch arterial network representing the geometric foundation of such a model^[43]; (D) The core biophysical rule, which calculates vessel stiffness from diameter and age^[43]. This figure is reproduced from Refs^[7,42,43] under a Creative Commons Attribution 4.0 International (CC BY 4.0) license. DT: Digital twin.