Search

Objective: The objective of the study was to explore the influences of seasonality, meteorological conditions, and air pollution exposure on the number of patients who visit the hospital due to seizures., Methods: Outpatient and inpatient data from the National Health Insurance Database of Taiwan from 2009 to 2013, meteorological data from the Meteorological Bureau, and air pollution exposure data from the Taiwan Air Quality Monitoring Stations were collected and integrated into daily time series data. The following data processing and analysis results are based on the mean of the 7 days' lag data of the 18 meteorological condition/air pollution exploratory factors to identify the critical meteorological conditions and air pollution exposure factors by executing univariate analysis. The average hospital visits for seizure per day by month were used as an index of observation. The effect of seasonality has also been examined., Results: The average visits per day by month had a significant association with 10 variables. Overall, the number of visits due to these factors has been estimated to be 71.529 (13.7%). The most obvious factors affecting the estimated number of visits include ambient temperature, CH 4 , and NO. Six air pollutants, namely CH 4 , NO, CO, NO 2 , PM2.5, and NMHC had a significantly positive correlation with hospital visits due to seizures. Moreover, the average daily number of hospital visits was significantly high in January and February (winter season in Taiwan) than in other months (R 2  = 0.422)., Conclusion: The prediction model obtained in this study indicates the necessity of rigorous monitoring and early warning of these air pollutants and climate changes by governments. Additionally, the study provided a firm basis for establishing prediction models to be used by other countries or for other diseases., Competing Interests: Declaration of competing interest None., (Copyright © 2020 Elsevier Inc. All rights reserved.)

This paper implements deep learning methods of recurrent neural networks and short-term memory models. Two kinds of time-series data were used: air pollutant factors, such as O3, SO2, and CO2 from 2017 to 2019, and meteorological factors such as temperature, humidity, wind direction, and wind speed. A trained model was used to predict air pollution within an eight-hour period. Correlation analysis was applied using Pearson and Spearman correlation coefficients. The KNN method was used to fill in the missing values to improve the generated model's accuracy. The average absolute error percentage value was used in the experiments to evaluate the model's performance. LSTM had the lowest RMSE value at 1.9 than the other models from the experiments. CNN had a significant RMSE value at 3.5, followed by Bi-LSTM at 2.5 and Bi-GRU at 2.7. In comparison, the RNN was slightly higher than LSTM at a 2.4 value. [ABSTRACT FROM AUTHOR]