應用於遠距醫療攜帶式生醫訊號擷取系統之EEG/ECG低雜訊前端電路及類比心電訊號動作雜訊干擾偵測器

A Low Noise EEG /ECG Signal Readout Frond-End and An ECG Motion Artifact Analog Detector for Telemedicine Mobile Biomedical Signal Acquisition Systems

黃韋智(Wei-Chih Huang)

指導教授 ： 鄭桂忠

國立清華大學/電機資訊學院/電機工程學系/碩士(2013年)

摘要

隨著目前醫療技術的進步，使社會漸漸轉型為高齡化人口結構，某些病狀或者慢性病所需要長時間並且能夠即時地監控人體各種生醫訊號的醫療儀器設備，其需求有日益增加的趨勢。這幾年來可以看到越來越多的文獻開始研究人體各類生醫訊號的特性以及各種應用於醫療上的系統電路，並且將其整合為適於人體的積體電路晶片而便於隨身攜帶的裝置。其中避免雜訊干擾而把感測到的微小振幅生醫訊號放大，是位於最前端低雜訊生醫訊號前端電路的主要功用。
本研究為擷取腦波訊號與心電訊號低雜訊前端電路及類比心電訊號動作雜訊干擾偵測器，低雜訊生醫訊號前端電路採用擁有高共模互斥比特性的全差動電流平衡式儀表放大器來做低雜訊的設計，並且利用截波穩定技術來減小電路中的閃爍雜訊，並且對於腦波訊號與心電訊號兩者不同的振幅與頻寬，利用低轉導電容濾波器與可調增益放大器來調整適當的頻寬與振幅。非侵入式的心電訊號量測容易受到量測環境中的雜訊干擾，舉凡電源線、呼吸以及動作或肌肉造成的干擾均為雜訊來源。其中最難處理的則是動作雜訊。目前最常被用來處理動作雜訊的方法是利用數位調適性濾波器演算法來移除動作雜訊的干擾。本研究提出類比心電訊號動作雜訊干擾偵測器，其目的為藉由透過類比前端偵測動作雜訊干擾的發生，來減少後端數位處理器中調適性濾波器因不必要處理動作雜訊干擾而持續運作所需要的運算與功率消耗。另外，心電訊號動作雜訊干擾偵測器也能提供其監測期間資料的可用性和監測量測電極的接觸狀況之功用。
本篇研究使用TSMC 90nm CMOS 標準製程實作。整個系統電路功率消耗為22.17 μW。低雜訊生醫訊號前端電路在EEG與ECG的頻寬範圍內，其輸入總等效雜訊分別為1.06 μVrms與1.64 μVrms，NEF為8.2，共模互斥比為107 dB，其結果顯示此電路適用於腦波與心電訊號之生醫訊號擷取系統。

關鍵字

低雜訊前端電路 ； 動作雜訊 ； 生醫訊號

並列摘要

In modern clinical practice, monitoring of biomedical signals is a crucial and important part. The biomedical signals most commonly used in medical diagnoses include EEG, ECG and EMG, etc. Therefore, there is a growing demand for small-size portable biomedical signal acquisition systems to improve the patients’ quality of life.
The low noise biomedical signal readout frond-end consists of a chopper current-balancing instrumentation amplifier (CCBIA), a small Gm-C low pass filter, and a programmable gain amplifier. The CCBIA utilizes the chopper stabilization to reduce the flicker noise and amplifiers the low amplitude biomedical signals. The common mode rejection ratio (CMRR) is also the important specification because there are Electromagnetic Interference (EMI) and Electrostatic Field Interference (EFI) when measuring biomedical signals. Behind the CCBIA, a low-pass filter is needed to reduce the out-of-band noise higher than the biomedical signal of interest, and the programmable gain amplifier is used to avoid the output signal to be saturated. The analog ECG motion artifact detector is proposed to save the power consumption of motion artifact reduction in DSP.
The article is fabricated by TSMC 90nm CMOS process. The measurement results shows that the readout frond-end achieves 107dB CMRR, the gain and the bandwidth are tunable for EEG and ECG, integrated input-referred noises are only 1.06 μVrms and 1.64 μVrms within EEG and ECG signal bandwidth. The ECG motion artifact analog detector can detect the skin-electrode impedance variation and the ECG signal variation successfully. The total power consumption is 22.17 μW.

參考文獻

[1] Hoi-Jun Yoo, Chris van Hoof, Bio-Medical CMOS ICs Integrated Circuits and Systems. Springer, 2011

[2] E. T. McAdams, Encyclopedia of Medical Devices and Instrumentation. J. G. Webster, 2002

[3] J. Ottenbacher, M. Kirst, L. Jatob´a, U. Großmann, and W. Stork, "An approach to reliable motion artifact detection for mobile long-term ECG monitoring systems using dry electrodes," ser. IFMBE Proceedings, vol. 18. Springer Berlin Heidelberg, November 2007, pp. 440–443.

[5] M. R. Nuwer et al., "IFCN standards for digital recording of clinical EEG," Electroencephalogr. Clin. Neurophysiol., vol. 106, no. 3, pp.259–261, Mar. 1998.

[7] Reid R. Harrison and Cameron Charles, ” A Low-Power Low-Noise CMOS Amplifier for Neural Recording Applications,” IEEE J. Solid-State Circuits, vol. 38, no. 6, June. 2003