A Visual Analytics System for Syndromic Surveillance and Pandemic Readiness within a One Health Context

Abstract

AbstractEmerging infectious diseases continue to pose a significant global threat, as pandemics disrupt societies, economies, and health systems around the world. To effectively anticipate and mitigate such crises, it is essential to gain timely insights from a variety of complex data sources that encompass human, animal, and environmental health. Integrating this data within a One Health framework, supported by advanced visualization and modeling techniques, provides a promising approach to developing more effective preparedness and response strategies. This dissertation advances the use of One Health data, encompassing human, animal, and environmental domains, in conjunction with interactive visual analytics methods to enhance pandemic preparedness and response. A systematic review of One Health dashboards and visualization systems published between 2015 and 2024 has been presented. This review categorizes systems by purpose, diseases addressed, data domains, interactivity, and user groups, revealing both promising trends and significant gaps. While many studies still rely on single-domain datasets, there is a clear movement toward integrated approaches, though fully comprehensive One Health implementations remain rare. Building on these insights, the analysis examines the role of visualization during the COVID-19 pandemic, with a particular focus on dashboards that were widely used to communicate risks, policies, and projections. This analysis maps data types, visualization tasks, and representations, highlighting strengths, limitations, and opportunities for advancing visualization research to better support future pandemic response. To effectively apply these findings, the Predictive Intelligence for Pandemic Prevention (PIPP) dashboard has been introduced. This dashboard integrates various One Health datasets, such as clinical outcomes, animal health records, environmental variables, and wastewater surveillance. Additionally, the One Health Mixed-Effects Modeling Approach (OH-MEMA), which is an interactive tool, is proposed, built on a lag-aware, non-linear mixed-effects modeling framework that captures variability in both spatial and demographic factors. Predictive results are coupled with interactive visualizations, time series comparisons, regional analyses, and analytic provenance trees, enabling users to explore scenarios, evaluate predictors, and iteratively refine models. Together, these contributions demonstrate the value of combining epidemiological modeling with interactive visual analytics to provide timely, interpretable, and actionable insights. By integrating multiple data domains, this work illustrates the potential of a One Health visual analytics approach to support early detection, guide interventions, and strengthen resilience against future pandemics.