UC GIS Week

Expanding Care with GIS

Improving accessibility and outreach are critical in providing effective health care services to communities of all kinds. Combining spatially-aware strategies with location intelligent technology, health organizations can achieve a clearer understanding of their work at the population level and identify areas where service gaps or potential opportunities for collaboration may exist in local communities. This project, in collaboration with UCLA’s Asian Pacific American (APA) Health Care organization, aims to demonstrate how APA can adopt and implement GIS applications into their strategies and aid in improved decision-making as well as visualize client demographic, socioeconomic, and health data.

COVID Transmission

The Geospatial Agent-based COVID-19 Transmission (or GeoACT) portal is a browser-based gateway application that allows users to simulate and compare COVID mitigation strategies for K-12 schools. Scenarios with a range of pharmaceutical and non-pharmaceutical interventions are used to generate realistic simulations of viral transmission with an agent-based model that uses school floor plans and classroom layouts.to calculate maps of infection risks. While most of the risk is attributed to aerosol transmission, the uneven spread of viral load emitted by infectious student agents requires tracking these agents’ location and proximity to others, along with the duration of time spent in shared enclosed environments such as classrooms, school cafeteria, or school buses. These environments themselves are also modeled as ‘agents’, with attributes relating to the spatial dynamics of the room, including the seats in which agents are expected to spend the majority of their time. Other environments can be modeled in a similar way with minimal changes, such as public school buses or theaters. To allow for these methods to be used for large numbers of students, along with running many simulations to account for differences in initial conditions between experiments, the gateway apps are deployed to an Apache Airavata portal configured to submit simulation jobs to the Expanse supercomputer. The science gateway portal hosts several modules. The Scenario module is used to run and compare simulations for different levels of intervention parameters, such as student vaccination rates at 0, 25, 50, 75, or 100%. The Meta-analysis module enables analysis of a series of previously-generated scenarios. The portal also features a ‘San Diego Schools and COVID-19’ dashboard, which presents case counts, infection rates, and other characteristics by zip code and in the vicinity of schools in San Diego county using spatial data services from San Diego County open data portal.

Randomforest

Utilizing hospital data, we predict risk factors for ED visits of asthma in children. To demonstrate complementing multivariate cluster analysis for random forest.