Enhancing Current-Source Imaging in Magnetocardiography: The Impact of Sensor Array Configuration

While magnetocardiography (MCG) is a non-invasive tool for assessing cardiac electrophysiology, its conventional planar sensor array often struggles to accurately estimate sources distant from the sensor. This limitation hinders the precise localization of deep cardiac activity. This study aims to overcome this challenge by investigating a novel cylindrical sensor array designed to improve source current estimation accuracy. We combined numerical simulations, phantom measurements, and in vivo rat experiments to compare a cylindrical sensor array with a conventional planar array. Single dipole localization was used in the simulation and phantom experiments under controlled focal source conditions, whereas both single-dipole fitting and minimum norm estimation (MNE)-based distributed source analyses were applied to the animal MCG data. The simulations and phantom measurements showed that the cylindrical array achieved more accurate single-dipole localization than the planar array under comparable sensor-source distance conditions. In the animal experiments, this advantage was not clearly observed using single-dipole fitting. The MNE-based distributed source analysis showed geometry-dependent reconstruction patterns, with the cylindrical array producing more posteriorly distributed activity and higher source map repeatability. The cylindrical sensor array provided complementary spatial information to anterior planar sampling, particularly for deeper or posterior cardiac activity. These findings suggest that surrounding sensor coverage may provide complementary information for MCG-based reconstruction, although further validation with realistic activation models remains necessary.