Research On Intelligent Gait Rehabilitation Based On Digital Human Movement Model

With the increasing health awareness of Chinese residents and the increasingly prominent problem of population aging,the role of rehabilitation medicine in the entire medical system is becoming more and more obvious,and it is particularly important to further promote the development of rehabilitation medicine.Elderly people,chronic diseases,sub-healthy people and disabled people all have different degrees of physical motor dysfunction.Society and families need to spend a great deal of money to treat and care for such groups.In addition,the late start of China’s rehabilitation medicine,the shortage of medical staff in the industry,the uneven distribution of medical resources and the insufficient service capacity make the social and family burden heavier.Starting from the common gait disorders,this paper proposes to combine the digital human body movement model with intelligent algorithm to achieve intelligent gait rehabilitation assessment and assist the clinician’s diagnosis and treatment,in an attempt to explore the possibility of this new mode to solve some of the above problems.Firstly,the digital motion model of human body is established by the bone critical point detection technology based on depth image and the bone muscle simulation technology based on the principle of reverse dynamics.The model can realize the quantitative rehabilitation evaluation of kinematics and dynamics.Then,the digital human movement model is combined with the machine learning algorithm,the gait characteristics derived from the digital human movement model are used as input and the corresponding gait pattern as output,and the mapping relation is learned by the integrated learning algorithm,so as to realize the intelligent gait rehabilitation evaluation.Considering the diversity and complexity of gait analysis parameters,information redundancy and attribute duplication may exist in a large number of parameters.Meanwhile,the performance of integrated learning is greatly affected by specific parameters.The random forest algorithm selected in this paper is affected by the depth and number of decision trees.Therefore,a new heuristic algorithm,whale optimization algorithm,is introduced to automatically adjust the parameters of the classification and recognition algorithm.Finally,taking the automatic recognition and analysis of hemiplegic gait as an example,the feasibility of intelligent gait rehabilitation evaluation by combining digital human motion model with intelligent algorithm is verified by experimental method.There are two innovative points in this paper: first,non-invasive gait evaluation parameters are obtained by constructing digital human movement model;Secondly,by combining digital human motion model with intelligent algorithm,automatic recognition and analysis of gait disorders are realized,which can avoid misjudgment caused by doctors’ inexperience to some extent.Under the background of high rehabilitation medical cost,insufficient service capacity and shortage of rehabilitation doctors in China,it is of certain practical significance and innovation to use the new and non-invasive technology of digital human movement model to carry out rehabilitation assessment.The combination of digital mannequin and intelligent algorithm to realize intelligent assessment of gait rehabilitation conforms to the guidance of national policy and social development,and can promote the development of telemedicine,mobile medical treatment and intelligent medical treatment to a certain extent.The intelligent rehabilitation evaluation with low cost and easy operation is easy to be popularized in clinic and serve patients.