Problem
Research questions and friction points this paper is trying to address.
myocardial perfusion
tracer-kinetic modeling
inverse problem
dynamic contrast-enhanced MRI
parameter estimation
Physics-Informed Implicit Neural Representations for Improved Myocardial Perfusion MRI Quantification
🤖 AI Summary
This work addresses the ill-posed inverse problem of quantifying myocardial perfusion MRI, which is highly sensitive to noise and acquisition variability when fitting dynamic contrast-enhanced MR data using multicompartment exchange models. For the first time, the authors integrate spatiotemporal implicit neural representations (INRs) into a physics-informed neural network (PINN) framework, modeling the MR signal as a continuous spatiotemporal function while embedding the multicompartment tracer kinetic model as a physical constraint. This approach significantly improves the accuracy, spatial smoothness, and physical consistency of estimated perfusion parameters. Evaluated on simulated myocardial perfusion MRI data, the method outperforms existing techniques and demonstrates enhanced robustness against data imperfections.
📝 Abstract
Quantifying myocardial perfusion from cardiac magnetic resonance (CMR) can be achieved by fitting tracer-kinetic models to the dynamic contrast-enhanced MR data. However, fitting the observed data with multi-compartment exchange models, which describe the evolution of the contrast agent in the tissue, to estimate perfusion parameters is a challenging inverse problem that is sensitive to noise and acquisition variability. Previously, physics-informed neural networks (PINNs) have been proposed as an alternative to conventional non-linear least squares fitting methods with promising results for quantitative perfusion CMR. In this work, we extend the previously proposed PINN framework with spatiotemporal implicit neural representations (INRs) to represent the MR signal as a continuous spatiotemporal function and to improve the accuracy, smoothness, and physical consistency of the PINN model. In realistic simulated CMR datasets, our proposed PINN with INRs demonstrates improved robustness and parameter estimation accuracy over the previously established methods. The code is available at https://github.com/q-cardIA/pinn-inr.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Physics-Informed Neural Networks
Implicit Neural Representations
Myocardial Perfusion MRI
Spatiotemporal Modeling
Tracer-Kinetic Modeling
🔎 Similar Papers
No similar papers found.
