flowCare uses patient-specific computational fluid dynamics (CFD) to help clinicians plan treatment before they treat. Our current focus is cerebral aneurysms: weakened, bulging areas of brain arteries that can rupture and cause life-threatening bleeding. Starting from 3D models of each patient's cerebral arteries, we simulate blood flow before and after surgical clipping and endovascular coil embolization, and compare how each option changes velocity, pressure, secondary flow, and wall shear stress in and around the aneurysm, including the transitional and turbulent flow common at the middle cerebral artery bifurcation.

Because image-based CFD is computationally expensive, we also explore how deep learning can make these simulations fast enough for clinical use. With human factors collaborators, we are mapping the steps of endovascular coiling to design a mixed reality and AI system for training clinicians in the procedure.

Objective

Develop personalized treatments for cardiovascular diseases based on patient-specific numerical models

Skillsets

• Patient-specific 3D modeling of cerebral arteries from medical image data
• Computational fluid dynamics modeling of blood flow, including turbulence modeling (low-Reynolds k–ω SST)
• Virtual treatment modeling: surgical clipping and porous-medium models of coil embolization
• Physiological boundary conditions (three-element Windkessel model)
• Mesh independence and verification (grid convergence index)
• Hemodynamic analysis: velocity, pressure, secondary flow, and wall shear stress
• Deep learning for computational hemodynamics
• Hierarchical task analysis for designing clinical training systems

Researchers

Dr. Amirtahà Taebi

Principal Investigator