Phase-based Epidemic Time series Analyzer(PEpiTA) is an interactive tool for public health practitioners to extract insights from time series data related to ongoing epidemic outbreaks.

Specifically, the tool will allow users to select a time series of epidemiological significance and extract categorical indicators of activity over time. These categories could either be based on level of activity (e.g., quantile of wastewater viral load) or rate of change (e.g., change in hospitalization rate). Such categorical indicators have been very useful for policymaking at state, federal, and local levels during the COVID-19 response as well as for tracking seasonal influenza over the years.

The team at UVA-BI has developed and refined various such indicators in collaboration with Virginia Department of Health and US Centers for Disease Control and Prevention for situation assessment as well as improved forecasting. Our goal is to build a pathogen-agnostic platform for public health departments to adopt for tracking and responding to various outbreaks.

Steps to be followed:
1. Upload a publicly-available time series for analysis.
2. Select the Data Preprocessing methods.
3. Choose among pre-defined rules for category extraction.
4. Customize analytical parameters such as number of bins, trend window widths.
5. Click the run button to generate the graphs.

References
1. Virginia Department of Health, District Trajectory Map, https://www.vdh.virginia.gov/coronavirus/see-the-numbers/covid-19-modeling/district-trajectory-map.
2. UVA Ensemble, FluSight experimental targets, https://github.com/cdcepi/Flusight-forecast-data/tree/master/data-experimental/UVAFluX-Ensemble.
3. Adiga et al., Enhancing COVID-19 Ensemble Forecasting Model Performance Using Auxiliary Data Sources, IEEE International Conference on Big Data, 2022.
4. Adiga et al., Phase-Informed Bayesian Ensemble Models Improve Performance of COVID-19 Forecasts, Annual Conference on Innovative Applications of Artificial Intelligence (IAAI), 2023.