| Summary: | 碩士 === 國立中央大學 === 電機工程學系 === 102 === In recent years, our westernized eating habits have led to the growth of stroke patient numbers. Therefore, the stroke rehabilitation has become one of the main important facilities in the hospital. The rehabilitation of a stroke patient can be presented by quantitative assessment index. Thus, it is desired to obtain the patient's rehabilitation condition by measuring the electromyography (EMG) of the patients. Further, through the analysis of EMG, we can provide a quantitative indicator of muscle strength which is intended for the physical therapist. In this study, the EMG measurement system constructed by NI instruments was first used to measure a total of twenty-seven stroke patients' EMG signals which were obtained before and after rehabilitation. Sixteen of them conducted the rehabilitation by using virtual reality (VR). Then we compared the EMG analysis result with the inertial measurement results and also with the Wolf motor function test (WMFT) to find the same trend in order to verify the reliability of EMG indicators. The experimental results showed that the patients who have more than six months to conduct virtual reality rehabilitation have excellent results. The 18 out of 27 patients presented the same trend in the EMG results and the Wolf motor function test. This study used XBee module and STM32 system board to develop a small-sized and multi-channel wireless EMG measurement system with a LabVIEW-based software interface to monitor, display and analyze the signal. The new system (6cm × 2.5cm × 3cm, 30.6g) has significantly reduced size and weight compared to the old system (10.5cm × 5cm × 6cm, 101.3g). The interface can display the muscle strength with bars and analyze the real-time co-contraction index to inform the users whether they used the correct muscles with the green light. In order to examine the signal quality of the sensors, EMG signals were first measured from five healthy subjects (5 males, age 23-25) with three different wireless EMG measurement systems (i.e., New system, Old system and Shimmer system) in terms of signal-to-noise ratio (SNR). The results showed that both the new system (28.69dB) and the Shimmer system (28.10dB) were better than the old system (28.94dB). For comparison purpose, the same sinusoidal signal was selected as input for three different systems (i.e., New system, Old system and Shimmer system) to obtain 21.39dB, 19.41dB, and 22.65dB, respectively. The spectral fatigue indices of three systems also indicated obvious appearance of muscle fatigue. Before performing the multi-channel EMG measurement, the signal quality of the sensors had been examined whether they were identical or not. The results showed that the sensors’ signal SNR were about 28dB with 0.63% variation. We then conducted the multi-channel EMG measurement on the same five healthy subjects and five stroke patients (4 male and 1 female, age 56, 28, 64, 45 and 28) with three different movements (i.e., shoulder flexion, shoulder abduction and elbow flexion) and compared the differences between the healthy subjects and stroke patients. We also compared the CI index of EMG measured by the new system and the NI-based system. Finally, the EMG signals of five patients were measured with the new system and NI-based system and analyzed with the Paired-Samples T Test. The result showed that there was no significant difference (p > 0.05) between the data measured with the two systems. This means that the new system could streamline the movement and mobility of the body without affecting the experimental results. Through the real-time RMS, CI index, and green light indicator on the interface, the subjects can immediately know the situation of their muscles. The above results showed that EMG signals were not only able to be used to assess the degree of muscle strength by RMS and muscle fatigue by using median frequency, but also able to be used to observe the contribution mode of muscles through multi-channel EMG measurement.
|
|---|
