Abstract
BackgroundFoodborne disease is a common threat to human health worldwide, leading to millions of deaths every year. Thus, the accurate prediction foodborne disease risk is very urgent and of great importance for public health management.ObjectiveWe aimed to design a spatial–temporal risk prediction model suitable for predicting foodborne disease risks in various regions, to provide guidance for the prevention and control of foodborne diseases.MethodsWe designed a novel end-to-end framework to predict foodborne disease risk by using a multigraph structural long short-term memory neural network, which can utilize an encoder–decoder to achieve multistep prediction. In particular, to capture multiple spatial correlations, we divided regions by administrative area and constructed adjacent graphs with metrics that included region proximity, historical data similarity, regional function similarity, and exposure food similarity. We also integrated an attention mechanism in both spatial and temporal dimensions, as well as external factors, to refine prediction accuracy. We validated our model with a long-term real-world foodborne disease data set, comprising data from 2015 to 2019 from multiple provinces in China.ResultsOur model can achieve F1 scores of 0.822, 0.679, 0.709, and 0.720 for single-month forecasts for the provinces of Beijing, Zhejiang, Shanxi and Hebei, respectively, and the highest F1 score was 20% higher than the best results of the other models. The experimental results clearly demonstrated that our approach can outperform other state-of-the-art models, with a margin.ConclusionsThe spatial–temporal risk prediction model can take into account the spatial–temporal characteristics of foodborne disease data and accurately determine future disease spatial–temporal risks, thereby providing support for the prevention and risk assessment of foodborne disease.
Highlights
Foodborne disease is caused by pathogenic bacteria that enter the body due to ingestion of contaminated food, resulting in symptoms such as diarrhea and abdominal pain [1]
In this paper, we propose the use of a multigraph structural long short-term memory (LSTM) based spatial–temporal prediction model to determine the risk of foodborne disease in different regions in future periods, which considers various spatial dependencies and uses a dynamic fusion method, with multistep prediction using a encoder–decoder structure, to support future disease prevention and control, and with attention mechanisms in spatial and temporal dimensions, as well as external features, to further improve performance
We propose a multistep spatial–temporal data prediction model based on encoder–decoder structure and composed entirely of LSTM modules, to address the problem of spatial–temporal foodborne disease risk prediction; we propose a dynamic fusion method to fuse region proximity, historical trend similarity, regional function similarity and food exposure similarity, with a spatial–temporal attention mechanism and external feature embedding; and we validated our model with extensive experiments on a long-term real-world foodborne disease data set, with data from 2015 to 2019 in multiple provinces of China; experimental results clearly demonstrated that our approach can outperform other state-of-the-art methods, with a margin
Summary
Foodborne disease is caused by pathogenic bacteria that enter the body due to ingestion of contaminated food, resulting in symptoms such as diarrhea and abdominal pain [1]. While considerable efforts have been made, an open challenge remains—accurately predicting foodborne disease risk by mining spatial–temporal patterns in historical disease records, using similar methods to those used for flu prediction [16,17,18], which is of great significance for public health management. By providing estimates of the trends of foodborne disease in future periods, accurate foodborne disease risk prediction can support effective guidance for government epidemic prevention policies. Objective: We aimed to design a spatial–temporal risk prediction model suitable for predicting foodborne disease risks in various regions, to provide guidance for the prevention and control of foodborne diseases. Conclusions: The spatial–temporal risk prediction model can take into account the spatial–temporal characteristics of foodborne disease data and accurately determine future disease spatial–temporal risks, thereby providing support for the prevention and risk assessment of foodborne disease
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
