坐到站時多通道表面肌電圖之即時巨幅資料縮減

Abstract

台灣急遽之人口老化，維持年長者之獨立行動能力以提高其生活品質是很重要之議題。影響人體自主活動最主要之肌群為下半身肌群，有許多動作皆需倚靠著下半身肌群來完成，其中以坐到站(Sit-to-Stand, STS)為指標性之動作，可視為獨立行動之關鍵判斷依據。坐到站之能力受到肌力及平衡之影響，本論文將針對肌力進行探討。為了增進年長者起身活動之能力，本實驗室致力於開發可程式化之坐到站起身輔助及復健系統。其中，為了後續臨床醫師定量評估復健之成效，本論文針對肌力量測開發一即時多通道表面肌電圖之記錄及分析系統，於坐到站的四個階段中顯示其肌電信號強度之變化，作為判定主要肌群施力是否適當之依據。本論文使用硬體前處理之方式，相較於傳統方法達成資料量降低至百分之一且低失真之特色，降低了對接收信號端(電腦或手機等)之規格要求。電路設計使用高共模拒斥比之儀表放大器、精準之頻寬設計，其各部分電路之串接順序及放大倍率配置至各元件之選擇考量均有詳細之闡述。配合類比開關之設計，除了短時間之量測，亦可應用於長時間之高精準度量測。此外，亦將利用LabVIEW撰寫一簡易操作者使用介面，提供肌力變化之視覺回饋。更藉由頻域及時域分析驗證本硬體前處理之可行性，此資料量大幅降低之成果更為未來能利用行動裝置(如手機)同時擷取多通道肌電信號之可能性奠定基礎。

In Taiwan, because of the aging population, the way to improve the independent mobility and quality of life of the elderly is a very important issue. The most important muscle group that affects the body’s independent mobility is the lower body muscles. Many movements need to rely on the lower body muscles to complete. Among them, Sit-to-Stand (STS) is the most important which is regarded as the key judgment for activity independence. The ability of sit-to-stand is affected by muscle strength and balance. We will focus on muscle strength in this thesis. In order to improve the ability of the elderly to get up and move, our laboratory is committed to the development of a smart STS support and quantitative evaluation system. In order to assist clinicians to evaluate the effectiveness of rehabilitation quantitatively, in this thesis, we develop a real-time multi-channel surface electromyography(EMG) recording and analysis system for muscle strength measurement, which also displays the strength of the EMG signal in the four phases of STS. In this study, we use hardware pre-processing to quantify the EMG signals. Compared with the traditional method, it reduces the data size by a hundred times with low distortion. Therefore, it reduces the computational requirements on the following processor (computer or mobile phone, etc.). In this system, we use an instrumentation amplifier with high common-mode rejection ratio, design adequate bandwidth, and arrange appropriate gain. And also, the cascade arrangement of each part of the circuit, gain configuration, and the selection of each component type are all well designed. This system, in addition to short-time measurement, can also be applied to long-term measurement with low error production. In addition, we design a friendly user interface to provide subjects with real-time visual feedback. Besides, we prove the feasibility of the hardware pre-processing of this study in frequency domain and time domain. The result of the huge data reduction greatly lays the foundation for the possibility of using mobile devices (such as mobile phones) to process multi-channel EMG signals simultaneously in the future.