Automatic classification of patient groups using magnetocardiography in the diagnosis of the atria lesions in patients with chronic obstructive pulmonary disease and coronary heart disease

The work is devoted to study the possibilities of automatic classification of patient groups using magnetocardiography (MCG) in the diagnosis of lesions of the atria. Analyzed magnetocardiographic data for the three groups of patients. The first group included 31 ΜCG record healthy volunteers. The second group - 45 MCG records of patients with chronic obstructive pulmonary disease (COPD). The third group - 58 ΜG records of patients with coronary heart disease (CHD). Statistical analysis showed that the decision rule for classification of patients with COPD and IHD can have four information parameter: 1 - the rate of variation of the magnetic field (RVMF), which characterizes the percentage of exceeding a given level of magnetic field in the whole time interval (104 msec), points on the plane dimensions of each magnetic card. 2 - parameter inversion (PI), which is defined on the interval 2-40 msec. 3 - parameter changes integrated map currents (ICT) characterizing changes in the structure (the values of vectors; directions of vectors; spatial distribution within the boundaries of the field measurements) distribution maps of vector current density during msec (on P wave in this case). 4 - parameter integral of the minimum magnetic field (PIMF).