Data-driven Neural Networks for Windkessel Parameter Calibration

2025-09-25

AI Summary

Calibrating Windkessel (WK) parameters in 1D-0D coupled hemodynamic models remains challenging—particularly when clinical measurement locations are uncertain and arterial pressure waveforms are corrupted by noise, hindering accurate parameter inversion. Method: We propose a data-driven neural network approach trained on synthetically generated left brachial artery pulse pressure waveforms. The network reconstructs full-domain pulse pressure waveforms with high accuracy and low computational cost. To enable robust real-world deployment, we introduce an “extended dummy neuron” mechanism and perform targeted retraining to facilitate rapid, noise-resilient WK parameter inversion from measured photoplethysmographic (PPG) or tonometric waveforms. Results: Experiments under diverse noise conditions and sensor misplacement scenarios demonstrate mean relative errors <3% in WK parameter estimation, with inference time <0.1 s per waveform—substantially outperforming conventional optimization-based methods. The framework exhibits strong robustness and promising clinical translatability.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Planning, Routing, and Scheduling: Activity and Plan Recognition

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

In this work, we propose a novel method for calibrating Windkessel (WK) parameters in a dimensionally reduced 1D-0D coupled blood flow model. To this end, we design a data-driven neural network (NN)trained on simulated blood pressures in the left brachial artery. Once trained, the NN emulates the pressure pulse waves across the entire simulated domain, i.e., over time, space and varying WK parameters, with negligible error and computational effort. To calibrate the WK parameters on a measured pulse wave, the NN is extended by dummy neurons and retrained only on these. The main objective of this work is to assess the effectiveness of the method in various scenarios -- particularly, when the exact measurement location is unknown or the data are affected by noise.

Problem

• Calibrating Windkessel parameters in 1D-0D blood flow models
• Emulating pressure pulse waves using a data-driven neural network
• Assessing method effectiveness with unknown locations and noisy data

Innovation

• Data-driven neural network for parameter calibration
• Emulates pressure pulse waves with negligible error
• Extended with dummy neurons for retraining on measurements