Evaluation of Sensor and Analysis Area in the Signal Source Estimation by Spatial Filter for Magnetocardiography

Abstract

Magnetocardiography (MCG) has been extensively investigated and evaluated in various clinical studies for its potential in detecting the early stages of cardiac diseases. In particular, the establishment of a technology that can provide detailed information on the lesion parts of the heart is expected. The estimation result of the current spatial filter method, such as the exact low-resolution brain electromagnetic tomography (eLORETA), contains errors and has insufficient estimation accuracy for clinical applications. The information of sensors located at the ends cannot be utilized, and the amount of information that affects the estimation result is different for each sensor, which causes the derivation of a biased estimated solution. Therefore, in this study, we considered that the amount of information of each sensor can be used equally by setting an analysis area larger than the sensor area and proposed a method focusing on the relationship between the analysis area of signal source estimation and the sensor plane. We simulated the most efficient and optimized analysis area to be used and performed the source signal estimation by eLORETA using the QRS-complex peak data of the actual MCGs used in adjusting the 51ch sensor directly above the xiphoid process. We compared the following three sizes of the analysis area: <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$x =120$ </tex-math></inline-formula> mm, <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$y =120$ </tex-math></inline-formula> mm, and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$z =100$ </tex-math></inline-formula> mm (a); <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$x =180$ </tex-math></inline-formula> mm, <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$y =180$ </tex-math></inline-formula> mm, and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$z =100$ </tex-math></inline-formula> mm (b); and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$x =240$ </tex-math></inline-formula> mm, <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$y =240$ </tex-math></inline-formula> mm, <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$z =100$ </tex-math></inline-formula> mm (c). Thus, because the signals from the heart are originally considered to be a single block, the estimated solution is also expected to be a single block. The solution was estimated as a single block, except case with analysis area (a), which corresponds to the conventional method. In addition, the position and size of the estimated solution are similar to those of the heart obtained from the computed tomography (CT) images. The goodness of fit of each analysis area was approximately 0.99 or more, which indicated that there was no significant difference. Thus, we demonstrated that it is possible to visualize the movement of the myocardium by estimating the time-series waveform of the MCG.