應用肌肉感應貼布於步態分析之實驗與分析方法開發

Abstract

目前應用於步態分析之裝置大多需在實驗室中進行，在應用環境上有許多限制，而若是使用慣性傳感器IMU，則僅能量測加速度及角速度值來間接監測人體步態的生理訊號。因此本研究旨在開發應用於步態分析之新型穿戴式裝置與其分析方法，利用具備高靈敏度、高穩定性、高生物相容性與具穿戴舒適度之肌肉感應貼布，直接量測人體活動過程中之肌肉施力的生理訊號，以進行肌肉活動狀態之監測。本研究透過人體實驗收集步態資訊，藉由跑步機調控不同速度，量測實驗參與者在不同步態模式下之小腿蹠屈肌肉之訊號變化。實驗結果顯示肌肉感應貼布訊號在頻域上之第一頻率與步頻間之相關係數達0.99，具有顯著的高度正相關，且主頻率與步頻間也呈現倍率關係，證明感測器於頻率上之準確性。除此之外，實驗結果亦顯示肌肉感應貼布具有透過量測皮膚表面形變監測肌肉收縮程度之能力，與肌電訊號具有顯著的高度正相關，相關係數r值介於0.76 ~ 0.80之間，發現在跑步模式之速度差異性判別效能高於走路模式，並能由步態訊號之曲線特徵說明此現象。本研究證明了肌肉感應貼布於步態分析上之可行性，其具備高頻率靈敏度，能夠識別走與跑之步態模式，評估肌肉在不同步行和跑步速度下之總能量大小，並進行姿態判斷，可提供新的步態資訊作為步態監測與訓練之參考依據。

Most current gait analysis methods are based on sensors that need to be carried out in laboratories, which can limit their applications. On the other hand, inertia sensors like IMU are common wearable sensors for daily activities. However, these sensors can only measure resultant accelerations and angular velocity. They cannot infer the contributions of muscles to various gait actions. Therefore, this study aims to develop a new wearable device and analysis method for gait analysis, using the muscle patch sensor (MPS) that has high sensitivity, stability, biocompatibility, and wearing comfort. The MPS can directly measure muscle force exertion during various gait activities. To verify and study the performance of the MPS, a series of human studies were conducted to collect the contraction signals of calf plantar flexor muscles during different walking and running speeds using a speed-controllable treadmill. The experimental results show that the primary frequency of the MPS signal has a high correlation with the step frequency, with a correlation coefficient of 0.99. This primary frequency also increases along with the stepping frequency, demonstrating the accuracy of this MPS device. In addition, the experimental results also show that the MPS can monitor muscle contraction has a high positive correlation with the electromyographic signal. The correlation coefficient value is between 0.76 ~ 0.80. In the running experiments, the discriminating capability is higher than that in walking experiments, and this phenomenon can be explained by the characteristic profiles of the gait signal. In summary, this study demonstrates the feasibility of using the MPS for gait analysis. It has high-frequency sensitivity that can be used to identify walking and running gait patterns, evaluate the total energy of muscles, and perform posture analysis. The MPS signals can provide new gait reference signals for gait analysis and training.