Affects of filtering on Computer-Analysis for Detecion of Chronic Ischemic Heart Desease with Unshielded Rest Magnetocardiographyc Mapping

Recent studies have reported better sensitivity of MCG automatic analysis, as compared to ECG, in detecting myocardial ischemia in patients (pts) with Ischemic Heart Disease (IHD). The aim of our study was to evaluate if filtering modality could affect the sensitivity (S) of the algorithms used for the analysis of ventricular repolarization (VR) from MCG of chronic IHD. Method: We studied 19 (true positive) IHD pts, with angiography-documented > 70% coronary stenosis [# vessels: 12(1), 3(2), 4 (3)] and positive stress/SPECT. Rest 12-lead ECG was ischemic in 12/19 (63%). A 90 seconds rest MCG was recorded with a 36-channel system (CardioMag Imaging, -CMI) (sampling rate: 1 KHz; bandwidth DC-100 Hz). To assess VR, 3 magnetic field dynamics parameters, [i.e. +/- poles: angle (A), distance (D) and ratio (R)] during the T-wave interval, were computed from the same MCG maps: 1) after digital 20 Hz low-pass filtering (LPF), and 2) after digital 50 Hz adaptive filtering (AF). Moreover 3 quantitative scores of the T-wave (EXT, Machine learning (ML) and Q scores) were automatically calculated. For comparison Hänninen STα angle was computed. The baseline was unchanged. Results: The S of A, D, and R in detecting IHD was 47%, 74% and 63% with 50 Hz AF, and 32%, 42% and 42% with 20 Hz LPF, respectively (p<0.01, for D). Corresponding S of EXT, ML and Q (with 20 Hz LP) was 63%, 74% and 79%. STα angle sensitivity was 79% with both 20 Hz LP and 50 Hz AF. Conclusions: The S of unshielded 36-channel MCG detection of chronic IHD, with T-wave parameters (A, D and R) might be affected by LPF. The STα angle was not affected by LPF. Automatic EXT, ML and Q scores have better S than ECG.