Yuk Sun-Woo, Hwang In-Ho, Cho Hyeon-Rae, Park Sang-Geon
Korea Orthopedics and Rehabilitation Engineering Center, 26 Gyeongin-ro 10beon-gil, Bupyeong-gu, Incheon 21417, Korea.
Department of Medical Device Industry; Dongguk University; Goyang-si, Gyeonggi-do 10326, Korea.
Micromachines (Basel). 2018 Oct 29;9(11):555. doi: 10.3390/mi9110555.
The form of the collection of bio-signals is becoming increasingly integrated and smart to meet the demands of the age of smart healthcare and the Fourth Industrial Revolution. In addition, the movement patterns of human muscles are also becoming more complex due to diversification of the human living environment. An analysis of the movement patterns of normal people's muscles contracting with age and that of patients who are being treated in a hospital, including the disabled, will help improve life patterns, medical treatment patterns, and quality of life. In this study, the researchers developed a smart electromyogram (EMG) sensor which can improve human life patterns through EMG range and pattern recognition, which is beyond the conventional simple EMG measurement level. The developed sensor has a high gain of 10,000 times or more, noise of 500 uVrms or less, and common mode rejection ratio (CMRR) of 100 dB or more for EMG level and pattern recognition. The pattern recognition time of the sensor is 30 s. All the circuits developed in this study have a phase margin of 75 degrees or more for stability. Standard 0.25 μm complementary metal oxide semiconductor (CMOS) technology was used for the integrated circuit design. The system error rate was confirmed to be 1% or less through a clinical trial conducted on five males in their 40s and three females in their 30s for the past two years. The muscle activities of all subjects of the clinical trial were improved by about 21% by changing their life patterns based on EMG pattern recognition.
生物信号采集形式正日益走向集成化与智能化，以满足智能医疗时代和第四次工业革命的需求。此外，由于人类生活环境的多样化，人体肌肉的运动模式也变得更加复杂。分析正常人肌肉随年龄收缩的运动模式以及包括残疾人在内的正在医院接受治疗的患者的运动模式，将有助于改善生活方式、医疗模式和生活质量。在本研究中，研究人员开发了一种智能肌电图（EMG）传感器，它可以通过肌电图范围和模式识别来改善人类生活方式，这超越了传统的简单肌电图测量水平。所开发的传感器对于肌电图水平和模式识别具有10000倍或更高的高增益、500 μVrms或更低的噪声以及100 dB或更高的共模抑制比（CMRR）。该传感器的模式识别时间为30秒。本研究中开发的所有电路为确保稳定性都具有75度或更高的相位裕度。集成电路设计采用标准的0.25μm互补金属氧化物半导体（CMOS）技术。通过在过去两年对五名40多岁男性和三名30多岁女性进行的临床试验，确认系统错误率为1%或更低。通过基于肌电图模式识别改变生活方式，临床试验所有受试者的肌肉活动提高了约21%。
