Multiscale Modeling and Machine Learning of Thrombosis, Vascular Flow, and Patient-Specific Anticoagulant Therapy

Thrombosis arises from a tightly orchestrated interplay of vascular flow, platelet activation, and coagulation biochemistry. Despite decades of research in hemodynamics and hematology, accurately predicting thrombus formation and dissolution remains a major challenge due to the nonlinear, multiscale nature of clotting processes that couple molecular kinetics, cell–cell interactions, and macroscopic fluid mechanics. Traditional mechanistic models have clarified individual pathways, from fibrin polymerization to platelet adhesion, but extending these frameworks to clinically actionable, patient-specific predictions is constrained by computational demands and biological variability. Emerging approaches that integrate machine learning with physics-based modeling now enable faster and more adaptive simulations, offering new routes toward predictive and personalized thrombosis modeling.This Research Topic aims to advance integrative computational frameworks that combine multiscale modeling, data-driven inference, and clinical translation in thrombosis research. The goal is to link molecular- and cellular-level mechanisms of clotting with organ- and patient-level hemodynamics, enabling predictive simulation of thromboembolic risk and therapeutic outcomes. By uniting efforts in computational biomechanics, systems hematology, and clinical data modeling, this collection seeks to promote hybrid methods that accelerate model calibration, incorporate patient heterogeneity, and inform optimized anticoagulant strategies.To gather further insights into the mechanistic and data-driven modeling of thrombus dynamics and vascular flow, we welcome articles addressing, but not limited to, the following themes:• Multiscale and multiphase continuum models of thrombus initiation, platelet activation, and fibrin network formation under flow• Patient-specific modeling of coagulation kinetics and shear-induced clotting thresholds• Hybrid frameworks integrating mechanistic fluid–structure interaction models with machine learning surrogates• Computational modeling of anticoagulant and antiplatelet drug mechanisms, resistance, and dose–response dynamics• In silico trials and virtual patient populations for assessing cardiovascular and procedural thrombotic riskWe welcome Original Research, Methods, Review, Perspective, and Mini Review articles. Topic editors: Haipeng Liu, RAJESH KUMAR TRIPATHY, Francisco Miguel Sanchez Margallo, Cheng Zhang Source: https://www.frontiersin.org/research-topics/84742/multiscale-modeling-and-machine-learning-of-thrombosis-vascular-flow-and-patient-specific-anticoagulant-therapy