Cerebral Infarction Rehabilitation Evaliiation with Posture Analyses

摘要

A cerebral infarction is a brain illness caused by a blockage in or narrowing of the arteries that supply blood and oxygen to the brain.The restricted amount of oxygen to the brain results in varying levels of disorder in the limb function of patients,severely affecting their normal lives.For cerebral infarction patients,physical rehabilitation is crucial in the early stage of their illness.The correct instruction of rehabilitation exercises can effectively restore the patient's limb function,reduce the chance of reoccurrence and improve the patient's daily life.In order to provide cerebral infarction patients effective early treatment,this project strives to develop a motion evaluation model based on deep learning.In order to enhance the instruction of rehabilitation exercises,the project's model defines 6 standard exercises as training input.It then employs the python Openpose framework to extract the coordinates of 18 joints of the human body,and traces the trajectory of these 18 joints when a person carries out one of the standard exercises.The trajectory of the joints during the exercise is used to train the LSTM network.The final model can be used as a guiding model for rehabilitation training.