基于患者疼痛感的康复机器人系统控制方法研究
发布时间：2018-07-28 10:09
【摘要】：全世界每年大约有1500万人因脑卒中等心脑血管疾病,导致永久性的肢体瘫痪,这给患者日常生活带来极大不便,也给家庭和社会带来沉重的精神与经济负担。越来越多的脑卒中患者需要接受康复治疗,以重获肢体运动功能。目前,国内外研究者开发出多种康复机器人辅助脑卒中患者进行康复训练。康复机器人的应用有望解决康复师人工训练中存在的问题和缓解康复师资源紧张状况。但是大多数康复机器人执行机械式地辅助运动,不能有效重塑患者受损的神经通路,并且缺乏患者主动运动意图,难以调动患者参与康复训练的积极性;此外,康复训练中缺乏对患者疼痛、疲劳等主观感受的监测,容易对患者造成二次伤害。针对这些问题,本文研究了基于患者疼痛感的康复机器人系统控制方法,通过检测患者的生物反馈信号识别运动意图并量化疼痛等级,以期建立有效安全的康复系统。首先,本文介绍了疼痛评估方法、脑-机接口(Brain-Computer Interface,BCI)技术、功能性电刺激(Functional Electrical Stimulation,FES)在康复机器人领域应用的国内外研究现状,提出本文的主要研究内容和工作。其次,开展了基于多生理信号的疼痛强度识别方法研究。针对原始特征中含有大量无关或冗余的特征,导致疼痛强度识别率下降问题,设计基于遗传算法的特征选择技术,寻找与疼痛有关的特征组合,建立优化的疼痛强度识别模型。再次,研究基于脑电信号识别患者主动意图的方法。设计12HZ、15HZ、20HZ频率的视觉刺激方案,提取相应的稳态视觉诱发电位信号,通过信号处理过程,建立有效的脑-机接口,识别患者的主动意图。然后,引入现代控制方法,研究FES应用于上肢康复训练中的最优控制策略。设计基于PD反馈的迭代学习控制算法,优化FES的电刺激控制序列,完成轨迹跟踪的康复任务,实现神经通路重塑和运动功能恢复。最后,进行基于疼痛反馈的脑-控康复系统的初步实验研究。利用BCI辨识患者主动运动意图作为上层康复指令,集成FES与康复机器人的康复系统执行具体的康复训练策略,监测的疼痛信息作为反馈参数,调节康复训练。
[Abstract]:Every year, about 15 million people in the world suffer from stroke and other cardiovascular and cerebrovascular diseases, resulting in permanent paralysis of limbs, which brings great inconvenience to patients' daily life, and also brings heavy mental and economic burden to family and society. More and more stroke patients need rehabilitation to regain limb motor function. At present, researchers at home and abroad have developed a variety of rehabilitation robots to assist stroke patients for rehabilitation training. The application of rehabilitation robot is expected to solve the problems existing in the artificial training of rehabilitators and relieve the shortage of resources of rehabilitators. However, most rehabilitation robots perform mechanically assisted exercise, which can not effectively reshape the injured nerve pathway, and lack the initiative motion intention of the patients, which makes it difficult to motivate the patients to participate in the rehabilitation training. The lack of monitoring of subjective feelings such as pain and fatigue in rehabilitation training can easily cause secondary injury to patients. Aiming at these problems, this paper studies the control method of rehabilitation robot system based on patient's pain feeling. By detecting the patient's biofeedback signal, we can recognize the motion intention and quantify the pain grade in order to establish an effective and safe rehabilitation system. Firstly, this paper introduces the methods of pain assessment, Brain-Computer interface (BCI) technology, and the application of functional electrical stimulation (Functional Electrical stimulation) in the field of rehabilitation robot, and puts forward the main research content and work of this paper. Secondly, the method of pain intensity recognition based on multiple physiological signals is studied. Aiming at the problem of reducing the recognition rate of pain intensity caused by a large number of unrelated or redundant features in the original features, a feature selection technique based on genetic algorithm is designed to search for the combination of features related to pain, and an optimized pain intensity recognition model is established. Thirdly, the method of recognizing patient's active intention based on EEG signal is studied. A visual stimulation scheme with 12HZ ~ 15HZ ~ 20HZ frequency was designed to extract the corresponding steady-state visual evoked potential (VEP) signal. Through signal processing, an effective brain-computer interface was established to recognize the active intention of the patient. Then, the optimal control strategy of FES in upper limb rehabilitation training is studied by introducing modern control method. An iterative learning control algorithm based on PD feedback was designed to optimize the electrical stimulation control sequence of FES to complete the rehabilitation task of track tracking and to achieve neural pathway remodeling and motor function recovery. Finally, a preliminary experimental study of brain-controlled rehabilitation system based on pain feedback was carried out. The active motion intention of patients was identified by BCI as the upper rehabilitation instruction, the rehabilitation system of FES and rehabilitation robot was integrated to carry out specific rehabilitation training strategy, and the monitored pain information was used as feedback parameter to adjust rehabilitation training.
【学位授予单位】：沈阳理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：R49;TP242
本文编号：2149749
[Abstract]:Every year, about 15 million people in the world suffer from stroke and other cardiovascular and cerebrovascular diseases, resulting in permanent paralysis of limbs, which brings great inconvenience to patients' daily life, and also brings heavy mental and economic burden to family and society. More and more stroke patients need rehabilitation to regain limb motor function. At present, researchers at home and abroad have developed a variety of rehabilitation robots to assist stroke patients for rehabilitation training. The application of rehabilitation robot is expected to solve the problems existing in the artificial training of rehabilitators and relieve the shortage of resources of rehabilitators. However, most rehabilitation robots perform mechanically assisted exercise, which can not effectively reshape the injured nerve pathway, and lack the initiative motion intention of the patients, which makes it difficult to motivate the patients to participate in the rehabilitation training. The lack of monitoring of subjective feelings such as pain and fatigue in rehabilitation training can easily cause secondary injury to patients. Aiming at these problems, this paper studies the control method of rehabilitation robot system based on patient's pain feeling. By detecting the patient's biofeedback signal, we can recognize the motion intention and quantify the pain grade in order to establish an effective and safe rehabilitation system. Firstly, this paper introduces the methods of pain assessment, Brain-Computer interface (BCI) technology, and the application of functional electrical stimulation (Functional Electrical stimulation) in the field of rehabilitation robot, and puts forward the main research content and work of this paper. Secondly, the method of pain intensity recognition based on multiple physiological signals is studied. Aiming at the problem of reducing the recognition rate of pain intensity caused by a large number of unrelated or redundant features in the original features, a feature selection technique based on genetic algorithm is designed to search for the combination of features related to pain, and an optimized pain intensity recognition model is established. Thirdly, the method of recognizing patient's active intention based on EEG signal is studied. A visual stimulation scheme with 12HZ ~ 15HZ ~ 20HZ frequency was designed to extract the corresponding steady-state visual evoked potential (VEP) signal. Through signal processing, an effective brain-computer interface was established to recognize the active intention of the patient. Then, the optimal control strategy of FES in upper limb rehabilitation training is studied by introducing modern control method. An iterative learning control algorithm based on PD feedback was designed to optimize the electrical stimulation control sequence of FES to complete the rehabilitation task of track tracking and to achieve neural pathway remodeling and motor function recovery. Finally, a preliminary experimental study of brain-controlled rehabilitation system based on pain feedback was carried out. The active motion intention of patients was identified by BCI as the upper rehabilitation instruction, the rehabilitation system of FES and rehabilitation robot was integrated to carry out specific rehabilitation training strategy, and the monitored pain information was used as feedback parameter to adjust rehabilitation training.
【学位授予单位】：沈阳理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：R49;TP242
【参考文献】
相关期刊论文 前3条
1 燕铁斌;;神经肌肉电刺激及其临床应用[J];继续医学教育;2006年30期
2 吴毅;刘罡;;神经系统可塑性的理论研究与实践[J];中华物理医学与康复杂志;2007年04期
3 黄敬,梅元武,童萼塘;脑卒中后脑的可朔性与康复[J];中国康复;2004年01期
相关博士学位论文 前1条
1 周鹏;基于运动想象的脑机接口的研究[D];天津大学;2007年
，本文编号：2149749
本文链接：https://www.wllwen.com/huliyixuelunwen/2149749.html
最近更新
教材专著
