Skip to main content

Using cross-correlation networks to identify and visualize patterns in disease transmission

Description

Syndromic surveillance data such as the incidence of influenza-like illness (ILI) is broadly monitored to provide awareness of respiratory disease epidemiology. Diverse algorithms have been employed to find geospatial trends in surveillance data, however, these methods often do not point to a route of transmission. We seek to use correlations between regions in time series data to identify patterns that point to transmission trends and routes. Toward this aim, we employ network analysis to summarize the correlation structure between regions, whereas also providing an interpretation based on infectious disease transmission. Cross-correlation has been used to quantify associations between climate variables and disease transmission. The related method of autocorrelation has been widely used to identify patterns in time series surveillance data. This research seeks to improve interpretation of time series data and shed light on the spatial–temporal transmission of respiratory infections based on cross-correlation of ILI case rates.

Objective

Time series of influenza-like illness (ILI) events are often used to depict case rates in different regions. We explore the suitability of network visualization to highlight geographic patterns in this data on the basis of cross-correlation of the time series data.

Submitted by teresa.hamby@d… on