Predictive analysis in outpatients assisted by the Internet of Medical Things

Abstract

Abstract With the rise of the Internet of Things era, data resources can be acquired in real time through intelligent sensing technology of IoMT (Internet of Medical Things), and it is very helpful for the analysis and prediction of medical data. Through the analysis and study of the data of the department of respiration combined with the data of the air quality dimension, meteorological dimension, and time dimension, this paper studies its related features and establishes a multidimensional features prediction model based on a BP neural network. The comparative experiments prove that the model prediction has a good effect. At the same time, this paper clarifies the validity of the multidimensional prediction model by comparing the relevant research. The results of the comparative experiments in this paper show that the outpatient quantity prediction is not a simple time-series problem, and it contains a variety of nonlinear influencing factors. In the contrast experiments, the multi-dimensional prediction model including air quality feature is better than others, and the experimental results prove that air quality indicators play an important role in the prediction of respiratory clinic outpatient visits. In addition, this paper discusses the significant lag effect on the prediction of respiratory consultations, and the four-day lag prediction has the best effect in the relatively short-term study. It indirectly proves that air has a lagging effect on respiratory diseases, and provides a reference for future research, no matter medical or modeling. The research has important reference value for the management and distribution of medical resources and the formulation of medical policies as well as important significance for the formulation of policies for disease prevention and the prevention and control of environmental pollution.