Classification Diagnosis Model of Recovery Status for Hemiplegic Patients Based on XGBoost Algorithm

The lower limb exoskeleton robot has been widely used in the rehabilitation treatment of stroke patients. However, the rehabilitation status of patients is usually evaluated manually by physicians, which is not only labor-intensive, but also susceptible to the physicians’ personal subjective experience. To objectively evaluate the recovery status of patients during the rehabilitation process, a rehabilitation diagnosis model that integrates long short-term memory (LSTM) network and eXtreme Gradient Boosting (XGBoost) algorithm is proposed in this paper. The six-minute walk distance (6MWD) value of the patient was predicted by LSTM. Then the predicted 6MWD value and the patient’s gait parameters were utilized to train the XGBoost model. The experimental results show that compared with other classification models, the proposed model can improve the classification accuracy with less computational complexity, and obtain the best hemiplegia diagnosis results.