High-precision cardiac magnetic field modeling and multi-source spatial reconstruction based on OPM-MCG: experimental validation

Ischemic chronic heart failure (ICHF) is a common cardiac condition. Cardiac electrical activity is traditionally captured by electrocardiography (ECG) using surface electrodes. However, due to its limited spatial resolution, ECG struggles to accurately detect deep cardiac electrophysiological signals. Magnetocardiogram (MCG), an ultra-sensitive, non-invasive, high-resolution cardiac magnetic field detection method, enables advanced cardiac source imaging. However, the lack of a high-fidelity model capable of effectively decoupling intrinsic cardiac sources from volumetric conduction effects in the torso limits its clinical diagnostic accuracy. An MCG forward modeling approach based on a three-dimensional heart-body surface torso model is proposed in this study. It aims to elucidate the mechanism of magnetic field generation by accurately simulating the dynamic evolution of spatial current density in the myocardium. A three-dimensional reference heart-torso geometry was reconstructed from CT images of a 26-year-old healthy male. The torso geometry, global cardiac position, and sensor locations were held constant across simulations, whereas the ICHF model incorporated disease-related electrophysiological changes and predefined ventricular remodeling, including ventricular dilatation and modeled ischemic scar regions. Experimental validation is performed using a 64-channel OPM atomic magnetometer system. Quantitative evaluation shows strong alignment between simulation and measured data. Specifically, the dimensionless structural similarity index between measured and simulated two-dimensional magnetic-field maps ( B s ± B v ) was 0.9335 and 0.9763 at the normal R- and T-wave peaks, respectively. In the ICHF group, the model captures a 70°–85° shift in the by component during the R-wave peak, revealing reduced maximum current vector (MCV) amplitude and disordered total current vector angles attributed to functional myocardial loss. Logistic regression analysis confirms statistical consistency between simulated and measured data. These findings demonstrate the capacity of this electrophysiological model to distinguish ICHF from normal states, identifying it as a promising tool for early diagnosis.