Research On Signal Processing Methods Based On SQUID Cardiac Magnetic Field Measurement Technique

As societal aging and lifestyle changes increase cardiovascular disease incidence,early diagnosis becomes a crucial social challenge.To address this issue,medical facilities are continuously updated and improved.Among the diagnostic methods for heart disease,magnetocardiography technology,characterized by its non-invasive,passive measurement,non-contact,and high-precision features,is a hot research topic in the field of new medical technology.Magnetocardiography technology can measure the weak magnetic field generated by the electrophysiological activity of the human heart,offering broad application prospects and profound research value.The magnetic signals generated during cardiac electrical stimulation show better specificity and richer information than electrical signals.By using magnetic sensors to collect these signals,we can reconstruct the position and structure of the current conduction in the heart,thereby determining the state of the heart.As the peak of the cardiac magnetic signal is only100 p T(picoteslas),sensors with extremely high sensitivity are required for detection.The magnetic field detection sensitivity of the Superconducting Quantum Interference Device(SQUID)can reach 10-15 T,which can well meet this requirement.The SQUID sensor,together with backend circuits and algorithms,constitutes a complex system.Due to the very weak signal,it poses high requirements for processing methods and algorithms.Therefore,this paper has conducted detailed research on the signal processing methods of unshielded magnetocardiography measurement equipment based on multi-channel SQUID,and proposed improvements.The specific research contents are as follows:(1)Various analyses have been conducted on the magnetocardiogram data collected by the SQUID second-order gradiometer in an unshielded environment.In response to the baseline drift problem that exists,simulation methods such as wavelet transform,empirical mode decomposition,independent component analysis have been used,and a baseline correction method based on Savitzky-Golay filtering was employed.For power frequency interference,IIR notch filtering was used,and for other high-frequency noise,low-pass filtering was adopted.After preprocessing,power frequency interference and baseline drift were essentially eliminated,laying the foundation for subsequent work.(2)In response to the issue of discontinuous time and difficulty in locating waveforms in multi-channel magnetocardiogram signal processing,this paper proposes a magnetocardiogram R-wave positioning method based on the improved Pan-Tompkins method and a local outlier factor algorithm based on density for identifying local outliers in high-dimensional data sets.This method uses local neighborhood information to calculate the degree of abnormality of each data point and filter it out,and outliers are removed.Then,the paper uses the radial basis function interpolation algorithm for image regularization and drawing of magnetograms,using the Gaussian radial basis function as the interpolation basis function,transforming the values of scattered data points into a continuous function,thus achieving data interpolation and approximation.(3)To further enhance the usable value of magnetocardiographic data,this paper analyzes the magnetic forward problem based on current elements and magnetic materials and the uniqueness problem of magnetic inversion in two and three dimensions.Comparing various magnetic inversion methods,simulation verification was performed on the H-C transformation,proving that it can reflect the two-dimensional current density distribution in the measured plane under near-source conditions quite well,and pseudo-current density vector graphs were generated from the isomagnetic diagrams of the actual magnetocardiographic data,vigorously advancing the clinical application process of magnetocardiography.This paper,in response to the current needs of cardiac diagnosis development,explores the signal processing methods obtained from unshielded magnetocardiography measurement equipment based on multi-channel SQUID,conducts detailed research on the shape recognition and peak positioning technology of magnetocardiographic signals,achieves the detection and processing of magnetocardiographic signals in a non-magnetic shielding environment,and discusses the application prospects of magnetocardiography technology and its profound research value.These studies have important and far-reaching implications for the diagnosis and early screening of heart disease.