Related Experiment Video
Updated: Aug 16, 2025
07:21
Magnetic Resonance Derived Myocardial Strain Assessment Using Feature Tracking
Published on: February 12, 2011
14.4K
Localization of magnetocardiographic sources for myocardial infarction cases using deterministic and Bayesian
Vikas R Bhat1, Basudha Pal2, H Anitha2
1Department of Biomedical Engineering, Manipal Institute of Technology, Manipal Academy of Higher Education (MAHE), Manipal, India.
Scientific Reports
|December 21, 2022
Summary
This study solves cardiac inverse problems using Tikhonov regularization and Bayesian methods for Magnetocardiographic (MCG) signals. Probabilistic models, particularly Variational Bayesian inference, offer improved accuracy in localizing heart sources, even with noisy data.
Area of Science:
- Biophysics
- Computational Electrophysiology
- Medical Imaging
Background:
- Cardiac source localization from Magnetocardiography (MCG) signals presents an inverse problem.
- Estimating heart surface potentials from under-determined MCG data is challenging, especially with measurement noise.
Purpose of the Study:
- To solve cardiac inverse problems using both analytical and probabilistic methods.
- To compare deterministic Tikhonov regularization with advanced Bayesian approaches for MCG source reconstruction.
- To apply and evaluate Variational Bayesian inference for improved heart source localization.
Main Methods:
- Tikhonov regularization for initial estimation of heart surface potentials.
- Hierarchical Bayesian modeling with fixed Gaussian priors for uncertainty quantification.
- Novel application of Variational Bayesian inference with non-stationary priors.
- Performance evaluation using Root Mean Square Error (RMSE) and correlation coefficient.
Main Results:
- Deterministic methods show sensitivity to measurement noise.
- Probabilistic models, especially Variational Bayesian inference, provide more robust source reconstructions.
- Bayesian solutions successfully localized MCG sources in simulated Myocardial infarction cases.
Conclusions:
- Variational Bayesian inference is a promising approach for accurate and robust cardiac source localization from MCG.
- Probabilistic methods offer superior performance over deterministic techniques in handling noisy MCG data.
- The developed Bayesian framework aids in understanding and localizing cardiac abnormalities like myocardial infarction.
