互补型自适应滤波器在心磁信号处理中的应用

将心磁信号从干扰噪声中加以提取并有效地消除噪声干扰是心磁信号处理中尤为重要的环节 .从改进算法的角度出发，提出互补型自适应滤波器结构以实现心磁信号的消噪处理.该滤波器针对心磁这类非平稳信号进行设计，有效地解决了常规自适应滤波器应用于心磁信号处理时收敛速度和稳态误差的矛盾.通过仿真实验和心磁实验结果表明，该算法能有效地消除心磁信号的背景噪声和工频干扰噪声.同时该算法也可用于其他非平稳信号的消噪处理. 关键词： 自适应滤波 心磁图 最小均方误差     

A SQUID gradiometer module with large junction shunt resistors

A dual-washer superconducting quantum interference device(SQUID) with a loop inductance of 350 pH and two onwasher integrated input coils is designed according to conventional niobium technology. In order to obtain a large SQUID flux-to-voltage transfer coefficient, the junction shunt resistance is selected to be 33 ?. A vertical SQUID gradiometer module with a baseline of 100 mm is constructed by utilizing such a SQUID and a first-order niobium wire-wound antenna.The sensitivity of this module reaches about 0.2 fT/(cm·Hz1/2) in the white noise range using a direct readout scheme, i.e.,the SQUID is directly connected to an operational amplifier, in a magnetically shielded room. Some magnetocardiography(MCG) measurements with a sufficiently high signal-to-noise ratio(SNR) are demonstrated.     

Magnetic-susceptibility plethysmography

Summary Magnetic-susceptibility plethysmography (MSPG) is a new method proposed for monitoring cardiac-volume changes. The measurement principle and recent work by other investigators are reviewed. The effect of the cardiac-related motion of the anterior chest wall on the MSPG signal is studied. The results suggest that the motion of the chest-wall-air boundary may contribute significantly (∼75%) to the MSPG signal when recorded near the heart. Paper presented at the ?IV International Workshop on Biomagnetism?, held in Rome, September 14–16, 1982.     

一种互补型自适应滤波最小均方法在SQUID测量的心磁信号中的应用

提出了一种互补型自适应滤波最小均方算法在SQUID测量的心磁信号中的应用，快的收敛速度和很好的跟踪能力通过根据最小均方原则在两个滤波器之间来回转换加权参量来实现     

Atrial activity during thePR segment of the MCG

Summary Sloping signals were observed in thePR segment of normal magnetocardiograms recorded from five dogs. Larger ramps were observed after first-degree AV nodal blockade was induced pharmacologically, and complete atrial repolarization was observed following isolated atrial contractions in two animals with second-degree block. These ramps, ranging up to 6.7 pT in size and arising from the atria, may complicate magnetic detection of His-Purkinje activity. This work was supported by the National Institute of Health under Grant No. HL 23184.     

Magnetic sensors with good signal-to-interference discrimination

Summary Principles are established for the design of coil arrays which generate distant magnetic fields diminishing as a high inverse power of the distance. By a principle of reciprocity these arrays, when used as sensors in induction and SQUID magnetometers, have an inherently low sensitivity to distant sources of magnetic field. Paper presented at the ?IV International Workshop on Biomagnetism?, held in Rome, September 14–16, 1982.     

Slow magnetic fields of the brain preceding movement and speech

Summary Magnetic fields of the brain were recorded preceding selfinitiated finger, foot, toe movements, stimulus-triggered finger and toe movements, and speech. Slow magnetic-field shifts were found with morphology similar to coincident electrical potentials, but with very different distributions. Inferences from the data about localization of cerebral generators were discussed. These studies were supported by grants from the B.C. Research Council and N.S.E.R.C. of Canada.     

Measurement of biomagnetic fields using multichannel superconducting-magnetometer techniques

Summary We have constructed a three-channel differential SQUID magnetometer for the simultaneous measurement of the three orthogonal components of the magnetic field in the same location. The structure of the device is described and MCG data measured are shown. Interference effects between the channels are discussed. A preliminary report of this work has been presented     

A digital filtering scheme for SQUID based magnetocardiography

Considering the properties of slow change and quasi-periodicity of magnetocardiography (MCG) signal, we use an integrated technique of adaptive and low-pass filtering in dealing with two-channel MCG data measured by high $T_{\rm c}$ SQUIDs. The adaptive filter in the time domain is based on a noise feedback normalized least-mean-square (NLMS) algorithm, and the low-pass filter with a cutoff at 100Hz in the frequency domain characterized by Gaussian functions is combined with a notch at the power line frequency. In this way, both relevant and irrelevant noises in original MCG data are largely eliminated. The method may also be useful for other slowly varying quasi-periodical signals.     

Singular value decomposition with normalized period for magnetocaridiography signal processing

In this paper, we have developed an algorithm based on singular value decomposition (SVD) for matrix. And the novel SVD algorithm with normalized period of cardiac cycles is presented. The results from real magnetocardiography (MCG) data processing show that the new algorithm is better than the standard one not only in suppressing noises, but also in providing high-fidelity MCG signals.