一种基于肌电-惯性信息的手部康复训练装置及方法

• Format: Patent
• Language: Chinese
• Published: 03.01.2023
• Subjects: ARTIFICIAL RESPIRATION; BATHING DEVICES FOR SPECIAL THERAPEUTIC OR HYGIENIC PURPOSESOR SPECIFIC PARTS OF THE BODY; DIAGNOSIS; HUMAN NECESSITIES; HYGIENE; IDENTIFICATION; MASSAGE; MEDICAL OR VETERINARY SCIENCE; PHYSICAL THERAPY APPARATUS, e.g. DEVICES FOR LOCATING ORSTIMULATING REFLEX POINTS IN THE BODY; SURGERY

本发明公开了一种基于肌电-惯性信息的手部康复训练装置及方法，本发明将肌电与惯性信息结合起来，通过采集受试者的肌电和惯性信号进行预处理、特征提取及模式识别，将肌电信号的识别结果传输到可穿戴康复机械手，同时对惯性信号进行预处理与特征提取，面向不同动作分别对训练健侧手部运动功能建立隐马尔科夫模型，运用健侧手部隐马尔科夫模型计算患侧手部运动功能对数似然概率，然后通过归一化处理得到患侧手部相对健侧手部运动功能归一化似然对数概率，从而对患者手部康复训练情况进行评估。并通过Leap Motion在场景中进行人机交互，为脑卒中患者提供一种肌电-惯性信息的手部康复训练装置。 The invention discloses a hand rehabilitation training device and method based on myoelectricity-inertia information. Myoelectricity and inertia information are combined, pretreatment, feature extraction and mode recognition are carried out by collecting myoelectricity and inertia signals of a subject, and a recognition result of the myoelectricity signals is transmitted to a wearable rehabilitation manipulator; meanwhile, pretreatment and feature extraction are carried out on the inertial signals, hidden Markov models are established for training the uninjured side hand movement function for different actions, and the uninjured side hand hidden Markov models are used for calculating the logarithm likelihood probability of the affec