Prediction of outpatient visits for allergic rhinitis using an artificial intelligence LSTM model - a study in Eastern China.

backgroundAllergic rhinitis is a common disease that can affect the health of patients and bring huge social and economic burdens. In this study, we developed a model to predict the incidence rate of allergic rhinitis so as to provide accurate information for the treatment, prevention, and control of allergic rhinitis.

methodsWe developed a Long Short-Term Memory model for effectively predicting the daily outpatient visits of allergic rhinitis patients based on air pollution and meteorological data. We collected the outpatient data from the departments of otolaryngology, emergency medicine, pediatrics, and respiratory medicine at the Affiliated Hospital of Hangzhou Normal University, from January 2022 to August 2024. The data were stratified by gender and age and were separately input into the model for evaluation. A total of 25,425 outpatient data samples were assessed in this study.

resultsBased on the data obtained from males (n = 13,943), females (n = 11,482), adults (n = 17,473), and minors (n = 7,952), the normalized mean squared errors of the Long Short-Term Memory model were 0.4674976, 0.3812502, 0.418301, and 0.4322124, respectively. By comparing the NMSE prediction results of ARIMA and LSTM models on this dataset, the LSTM model was found to outperform the ARIMA model in terms of stability and accuracy.

conclusionsThe model presented here could effectively predict the daily outpatient visits for allergic rhinitis patients based on air pollution and meteorological data, thereby offering valuable data-driven support for hospital management and for potentially improving societal management and prevention of allergic rhinitis.