Application of Artificial Neural Networks for Dengue Fever Outbreak Predictions in the Northwest Coast of Yucatan, Mexico and San Juan, Puerto Rico

Abstract

Modelling dengue fever in endemic areas is important to mitigate and improve vector-borne disease control to reduce outbreaks. This study applied artificial neural networks (ANNs) to predict dengue fever outbreak occurrences in San Juan, Puerto Rico (USA), and in several coastal municipalities of the state of Yucatan, Mexico, based on specific thresholds. The models were trained with 19 years of dengue fever data for Puerto Rico and six years for Mexico. Environmental and demographic data included in the predictive models were sea surface temperature (SST), precipitation, air temperature (i.e., minimum, maximum, and average), humidity, previous dengue cases, and population size. Two models were applied for each study area. One predicted dengue incidence rates based on population at risk (i.e., numbers of people younger than 24 years), and the other on the size of the vulnerable population (i.e., number of people younger than five years and older than 65 years). The predictive power was above 70% for all four model runs. The ANNs were able to successfully model dengue fever outbreak occurrences in both study areas. The variables with the most influence on predicting dengue fever outbreak occurrences for San Juan, Puerto Rico, included population size, previous dengue cases, maximum air temperature, and date. In Yucatan, Mexico, the most important variables were population size, previous dengue cases, minimum air temperature, and date. These models have predictive skills and should help dengue fever mitigation and management to aid specific population segments in the Caribbean region and around the Gulf of Mexico.

Weathering the pandemic: How the Caribbean Basin can use viral and environmental patterns to predict, prepare, and respond to COVID‐19

“…Public Health Agency has already put forth 64 that heavy showers are expected in May, with likely floods for the Greater Antilles and Guianas, and short-term droughts in other areas. The rains will increase the incidence of endemic vector-borne diseases including Dengue, [65][66][67] Zika, 68 and Chikungunya. 69 Consequently, jurisdictions should prepare for potential impacts in food production, water availability, and wildfires, in addition to preparing for epidemics.…”

Review on Nowcasting using Least Absolute Shrinkage Selector Operator (LASSO) to Predict Dengue Occurrence in San Juan and Iquitos as Part of Disease Surveillance System

“…Laureano-Rosario [27] made use of the climatic and population variables from National Oceanic and Atmospheric Administration (NOAA) to forecast the weekly dengue incident rate (per 100,000) for two population subgroups. Confirmed daily cases specific to the population at risk (those younger than 24 years old) and vulnerable population (small children and old folks) were obtained from two medical sources.…”

“…The covariates underwent extensive feature engineering which entailed the combination of autoregressive (climatic lags), trend (cumulative, average, smoothing, simple linear regression), trigonometric (sine and cosine functions to capture seasonality) and transformation (logarithmic and squared) elements. Monte Carlo simulations were run to generate the forecasts followed by evaluation against the other top entries from the competitions plus [27]. The hybrid hetGP was among the top three models for San Juan while its performance was less consistent for Iquitos.…”

“…The hybrid hetGP was among the top three models for San Juan while its performance was less consistent for Iquitos. The authors noticed that the parsimonious F2 model from [27] shared the advantageous historical seasonal pattern memorization and matching capability as their hetGP. [9] performed massive feature engineering with the creation of 103 new variables from Mosquito lifecycle related lagged effects, up to 32 weeks, interaction between climatic variables and forecasted dengue cases from decomposition-based time series models.…”

A dynamic neural network model for predicting risk of Zika in real-time

“…The aims of our work are to 1) present recurrent neural networks for time ahead predictive modelling as a highly flexible tool for outbreak prediction, and 2) implement and evaluate the model performance for the Zika epidemic in the Americas. The application of neural networks for epidemic risk forecasting has previously been applied to dengue forecasting and risk classification [45][46][47][48][49][50], detection of mosquito presence [51], temporal modeling of the oviposition of Aedes aegypti mosquito [52], Aedes larva identification [53], and epidemiologic time-series modeling through fusion of neural networks, fuzzy systems and genetic algorithms [54]. Recently, Jian et al [55] performed a comparison of different machine learning models to map the probability of Zika epidemic outbreak using publically available global Zika case data and other known covariates of transmission risk.…”