Cardiovascular disease remains the leading cause of death worldwide, and clinical decisions increasingly rely on extracting functional and hemodynamic information from medical images. Image-derived assessment, such as fractional flow reserve from CT (FFR-CT), myocardial perfusion analysis, and patient-specific blood-flow simulation, is moving rapidly from research into clinical practice, while scientific machine learning, including physics-informed neural networks and operator-learning methods, is transforming these computations into fast, reproducible, patient-specific pipelines. Despite this progress, such methods remain difficult to reproduce. They combine many steps (segmentation, geometric reconstruction, mesh generation, computational fluid dynamics, and model training), each with implementation details that written papers rarely capture fully, which slows validation and adoption across laboratories.



This Topical Collection curates demonstrable, end-to-end methodologies for cardiovascular image analysis and hemodynamic modeling. Its video-article format is uniquely suited to communicating these multi-step pipelines, allowing other groups to reproduce and extend them directly. We welcome contributions covering vessel and chamber segmentation from CTA, ICA, MRI, and echocardiography; three-dimensional patient-specific reconstruction; CFD-based hemodynamic simulation; physics-informed and data-driven functional assessment (FFR, wall shear stress, perfusion); cardiovascular digital twins; and validation against clinical reference standards. By assembling rigorous, reproducible methods from computational modelers, imaging scientists, and clinical researchers, the collection aims to establish a shared methodological foundation and accelerate the translation of computational cardiovascular imaging into clinical practice.