Blocked vessels

Technical capabilities in MPflow

• Multiphase CFD modelling of blood flow in stenosed arteries
• Resolution of pressure drops and recovery across constrictions
• Cavitation modelling including bubble inception and collapse in vascular flows
• Prediction of shear stress distributions linked to plaque formation and rupture
• Support for patient-specific geometries and varying stenosis severity (50–75%+)
• Transient simulations capturing pulsatile blood flow conditions

Our CFD approach

• High-resolution simulation of flow through coronary artery stenosis
• Identification of low-pressure regions triggering cavitation
• Coupled analysis of turbulence, shear layers, and recirculation zones
• Investigation of plaque stability under varying flow conditions
• Modelling of pressure fluctuations driving bubble dynamics and collapse
• Detailed insight into flow-induced mechanisms leading to thrombotic events

Physics-based ML acceleration

• ML-enhanced prediction of pressure fields and cavitation-prone regions
• Acceleration of patient-specific vascular flow simulations
• Rapid estimation of hemodynamic indicators (WSS, oscillatory shear index)
• Hybrid CFD–ML framework ensuring physical consistency in biomedical flows
• Real-time evaluation for clinical decision support and treatment planning

Why ML-CFD matters for Blocked Vessels

• Up to 5000× faster predictions: rapid evaluation of stenosis severity and cardiovascular risk
• 5× acceleration with ML-enhanced turbulence and multiphase modelling
• High accuracy (<3% deviation) in pressure drop, wall shear stress, and cavitation onset
• Instant assessment of patient-specific conditions and disease progression
• Supports early diagnosis of plaque instability and rupture risk
• Enables digital twins for personalised cardiovascular care