The application of the network and the intelligent terminal opens a new mode of medical treatment, in the diagnosis and treatment of mental disease, how to build a smart, efficient model is a major issue in front of scientific research. In this paper, the energy consumption of the mobile terminal is reduced based on the Q learning network and the wireless body area network routing method, improve the accuracy of data reception and the battery life. Dynamic data obtained by real-time data exchange, BP_Adaboost strong classifier algorithm is used to accurately model the psychological problems of patients.