Patient-Specific Preoperative Flow Analysis of Mitral Valve Regurgitation Through Experimental Measurements and Numerical Simulations

1. Report on the Work Performed
Supported by the YIN Start-up grant, we established a fruitful collaboration between the group Artificial Intelligence in Cardiovascular Medicine (AICM), at the University Hospital Heidelberg (UKHD) and the Institute of Fluid Mechanics (ISTM), KIT to investigate the limitations of the flow convergence method in mitral regurgitation (MR) quantification. This collaboration resulted in a joint publication [P1] and the supervision of several master’s [T1, T2] and bachelor’s theses [T3]. The study focused on the method’s effectiveness using particle image velocimetry (PIV) measurements in a controlled in-vitro environment (Fig. 1a), with particular attention to its inaccuracies for large regurgitation orifices, where MR severity tends to be underestimated. PIV measurements were carried out by an experienced PostDoc, Dr. Robin Leister, who was funded for two months through the YIN grant.

Our comparative analysis of ultrasound measurements and PIV data across nine artificial valve geometries of varying shapes and sizes revealed medium inter-observer variability but significant systematic underestimation of regurgitation volume for large orifice areas due to violations of flow convergence assumptions. Computational fluid dynamics (CFD) analysis demonstrated that increased aliasing velocities amplify uncertainty in MR assessment, especially for large apertures. This confirms that reliable outcomes using the Proximal Isovelocity Surface Area (PISA) method are primarily achievable for small circular apertures. The study emphasizes the importance of selecting appropriate aliasing velocities in clinical echocardiography and recommends adjustments based on observed isovelocity contours to improve the accuracy of MR severity assessment. The results have been presented at a medical conference [C2] and a bioengineering conference [C1].

Additionally, as part of this collaborative effort, a digital twin of the in-vitro dynamic simulator, originally conceptualized and built at AICM, UKHD, was developed at ISTM, KIT by a student research assistant, B.Sc. Jakob Hoffmann, funded through the YIN grant. Using the Siemens StarCCM+ CFD software package, the digital twin facilitated a comprehensive analysis of the 3D dynamics in pulsatile flow and complex geometries (Fig. 1b). Comparative analysis of the temporal development of flow fields in the simulation and time-resolved PIV measurements for a specific stiff orifice geometry demonstrated good agreement (Fig. 1c). This digital twin now serves as a foundational model for simulating flow in complex mitral valve geometries suggested in the proposal, verifiable with the in-vitro simulator, and will be further developed. The numerical study is now being continued as part of an external Master’s thesis conducted at the Institute of Industrial Science (IIS), University of Tokyo. This work leverages the simulation database to develop a predictive tool based on a physics-informed neural network.

Building on our preliminary study, we submitted several proposals during the project period, though they were unfortunately not successful:
• S. Engelhardt, A. Stroh, HEiKA project proposal ”Characterization of Regurgitation Jets through Particle Imaging Velocimetry and Numerical Simulations”, 2023
• S. Engelhardt, A. Stroh, DFG proposal in the framework of SPP 2311 Continuum Biomechanics ”Physics-Informed Neural Networks (PINNs) to Accelerate Coupled Fluid StructureInteraction for Heart Valve Simulation: Evaluation Against a Physical Simulator”, 2024
• S. Engelhardt, A. Stroh, W. Liebig, G. Romano, U. Scheppers, HEiKA-STAR proposal ”HEiKAValveTech: Advancing Personalized Heart Valve Therapies through Digital Twins and Bioprinting”, 2024

Nevertheless, based on the preliminary work, Dr. Robin Leister, the PostDoc involved in the execution of PIV measurements has been accepted into the Baden-Württemberg Foundation’s Postdoctoral fellowship for leading early career researchers. With his project proposal “Preventing heart failure - quantification of mitral valve regurgitation flow”, he was successful in a two-stage review process and is now a scholarship holder of the Elite Program Year 2025. The aim of the research project, which involves a cooperation with AICM at UKHD is to set up an in-vitro test facility and measure realistic heart valves using state-of-the-art, high-resolution optical laser measurement technology.

2. Project-Related Publication List
Reviewed publications
[P1] R. Leister, R. Karl, L. Stroh, D. Mereles, M. Eden, L. Neff, R. de Simone, G. Romano, J. Kriegseis, M. Karck, C. Lichtenstern, N. Frey, B. Frohnapfel, A. Stroh, and S. Engelhardt. Investigating the shortcomings of the flow convergence method for quantification of mitral regurgitation in a pulsatile in-vitro environment and with computational fluid dynamics. Cardiovascular Engineering and Technology, In Print. doi:10.1007/s13239-024-00763-w. Conference contributions
[C1] A. Stroh, R. Leister, R. Karl, L. Stroh, D. Mereles, M. Eden, L. Neff, R. de Simone, G. Romano, M. Karck, C. Lichtenstern, N. Frey, J. Kriegseis, B. Frohnapfel, and S. Engelhardt. Evaluating the flow convergence method in mitral regurgitation analysis: Insights from computational fluid dynamics and pulsatile in-vitro studies. In Virtual Physiological Human Conference 2024, Stuttgart, Germany, 2024.
[C2] L. Stroh, R. Karl, R. Leister, M. Derliz, R. de Simone, G. Romano, A. Stroh, J. Kriegseis, M. Karck, N. Frey, M. Weigand, C. Lichtenstern, B. Frohnapfel, and S. Engelhardt. Comparison of transesophageal echocardiography and particle image velocimetry to quantify mitral regurgitation in a high-fidelity environment. In DGAI Wissenschaftlicher Arbeitskreis Kardioanästhesie, 35. Herbsttreffen, Fulda, Germany, 2023. Student theses
[T1] L. Neff. Characterization of flow dynamics in a heart simulator by means of PIV. Master’s thesis, Insititute of Fluid Mechanics, Karlsruhe Institute of Technlogy, Karlsruhe, Germany, 11 2023. doi:10.5445/IR/1000164471.
[T2] J. Saal. Characterization of flow dynamics in a heart simulator by means of CFD. Master’s thesis, Insititute of Fluid Mechanics, Karlsruhe Institute of Technlogy, Karlsruhe, Germany, 4 2024.
[T3] Y. Wang. Simplified CFD for mimicking of PISA characterisation method. Bachelor’s thesis, Insititute of Fluid Mechanics, Karlsruhe Institute of Technlogy, Karlsruhe, Germany, 8 2024.