Applications of artificial intelligence in predicting dengue outbreaks in the face of climate change: a case study along coastal India

Abstract

The climate crisis will have an increasingly profound effect on the global distribution and burden of infectious diseases. Climate-sensitive diseases can serve as critical case studies for assessing public health priorities in the face of epidemics. Preliminary results denote that machine learning-based predictive modeling measures can be successfully applied to understanding environmental disease transmission dynamics. Ultimately, machine learning models can be trained to detect climate-sensitive diseases early, diseases which might represent serious threats to human health, food safety, and economies. We explore how machine learning can serve as a tool for better understanding climate-sensitive diseases, taking dengue dynamics along the Godavari River of coastal India as our case study. We hypothesize that a climate-driven predictive model with controlled calibration can help us understand several of the most critical relationships and climate characteristics of climate-sensitive disease dynamics.

Assessing Eelgrass Wasting Disease Dynamics in California Estuaries in the Context of Climate Change

“…On an individual leaf level, disease prevalence was determined based on the presence or absence of lesions, assessed through visual examination. To evaluate disease severity, we employed a machine learning imaging model to measure the lesion area and total leaf area (Tallam et al 2023). Disease severity was then calculated as the proportion of leaf area covered by lesions.…”

“…As for the linear mixed effects model, the following was the final "best fit model" for our purposes (Figure 1). However, the strength and direction of the relationships had to be explored further and so we developed a structural equation model (SEM) to further interrogate these relationships (Figure 8) (Pourtois, Tallam, et al 2023). To further assess the distribution of our dataset, we compared between both upper and lower estuary spatial gradients at each site, as well as comparatively across the three sites, we can visualize the symmetry, skew, variance, and outliers in our dataset overall [Figure 2A; 2B; 2C].…”

“…This package allowed us to account for indirect pathways and interrelationships among the explanatory variables. The approach was adapted from Pourtois, Tallam, et al (2023). The structural equation modeling (SEM) analysis was conducted using aggregated data across multiple years, with temporal means calculated for each site.…”

“…Specific to global coastal and estuarine ecosystems, climate is also altering these sensitive environments with severe fluctuations from decadal averages (Colombano et al, 2021), directional changes of environmental variables (i.e., sea surface temperatures, pH levels, and rising sea levels) (Cabral et al, 2019), as well as less predictable oscillations (e.g., the Pacific and Atlantic decadal oscillations) and more extreme weather events such as coastal flooding and marine heatwaves (Babcock et al, 2019;Tallam and Quang, 2023). Shallow and intertidal coastal habitats including mangroves, wetlands, saltmarshes, tidal flats, and estuaries are particularly exposed to extreme weather events such as extreme heat or cold waves, and this has in turn begun to increase the thermal stress that many of these environments are facing globally (Malhi et al, 2020).…”

Environmental Analysis of Hemorrhagic Fever in Iraq Using Machine Learning

“…Hemorrhagic fevers are serious viral illnesses; Crimean-Congo Hemorrhagic Fever (CCHF) spreads via infected ticks or contact with contaminated animal substances [1][2] [3]. This study explores the correlation between Hemorrhagic Fever and regional factors in Iraq to understand the disease's spread and prevention strategies [4] [5]. The World Health Organization reported 1,085 suspected, 287 confirmed cases, and 52 deaths due to CCHF in Iraq up to August 2022 [6].…”