Spatiotemporal Prediction of Tropical Disease Spread Using Deep Learning and Multisource Epidemiological, Climatic, Environmental, and Demographic Data

2026-09-28T15:18:10+00:00

Tropical diseases remain a major public-health challenge because their transmission is influenced by complex interactions among historical disease incidence, climatic conditions, environmental characteristics, and demographic factors. Conventional prediction approaches often have difficulty capturing nonlinear relationships, temporal dependencies, and spatial heterogeneity. Deep learning provides an alternative approach for learning complex patterns from multidimensional epidemiological data. This study aims to develop and evaluate a deep-learning model for predicting the spatiotemporal spread of tropical diseases using epidemiological, climatic, environmental, and demographic data. The study analyzed 17,546 observations from 24 geographic areas covering January 2021 to December 2025. Historical disease surveillance data were integrated with temperature, rainfall, relative humidity, vegetation, population density, urbanization, and other relevant predictors. The data were cleaned, temporally and spatially aligned, normalized, and transformed using lagged variables before model training. A proposed spatiotemporal deep-learning architecture was evaluated against ARIMA, Random Forest, LSTM, and CNN-LSTM using MAE, RMSE, MAPE, and (R^2). The proposed model achieved the best performance, with an MAE of 1.087, RMSE of 2.291, MAPE of 14.63%, and (R^2) of 0.876. Historical disease cases, rainfall, humidity, temperature, and population density were the most influential predictors. The model also captured major outbreak patterns and demonstrated potential for supporting disease surveillance, early detection, targeted prevention, and public-health early-warning systems.

Journal Basic Science and Technology

Spatiotemporal Prediction of Tropical Disease Spread Using Deep Learning and Multisource Epidemiological, Climatic, Environmental, and Demographic Data