Modeling the spread of the Zika virus using topological data analysis

TL;DR

This paper introduces a novel approach using topological data analysis to model and predict the spread of Zika virus in Brazil, providing a new tool for epidemiologists to assess vector-borne disease dynamics.

Contribution

It applies topological data analysis, specifically Vietoris-Rips filtration, to epidemiological data, which is a novel approach in modeling disease spread.

Findings

Topological features correlate with Zika spread patterns.

The model improves prediction accuracy over traditional methods.

Homology group analysis reveals key transmission pathways.

Abstract

Zika virus (ZIKV), a disease spread primarily through the Aedes aegypti mosquito, was identified in Brazil in 2015 and was declared a global health emergency by the World Health Organization (WHO). Epidemiologists often use common state-level attributes such as population density and temperature to determine the spread of disease. By applying techniques from topological data analysis, we believe that epidemiologists will be able to better predict how ZIKV will spread. We use the Vietoris-Rips filtration on high-density mosquito locations in Brazil to create simplicial complexes, from which we extract homology group generators. Previously epidemiologists have not relied on topological data analysis to model disease spread. Evaluating our model on ZIKV case data in the states of Brazil demonstrates the value of these techniques for the improved assessment of vector-borne diseases.