Adaptive Group Testing Strategy for Infectious Disease Using Social Contact Graph Partition
Speaker: Lenwood Heath and Jingyi Zhang, Virginia Tech
Abstract: During epidemics, mass testing stands as a frontline defense, enabling timely identification and isolation of infected individuals while safeguarding the community. The traditional Dorfma’s group testing technique, while reducing test counts, neglects factors like infection dynamics, testing capacity challenges, and community-based transmission patterns in a person-to-person contact network. To address these challenges, we propose the adaptive group testing (AGT) strategy, based on graph partitioning. In comparative evaluations against six synthetic and seven real social contact networks, AGT consistently outperformed Dorfman’s method, requiring fewer tests and better-controlling disease spread. AGT emerges as a robust, strategic blueprint, providing testing efficiency and invaluable guidance to policymakers.
Bio: Jingyi Zhang is currently pursuing her Ph.D. in Computer Science at Virginia Tech, with a keen focus on computational biology and computational epidemiology. Before this, she earned her M.S. in Information Science from the University of Pittsburgh. Her research mainly focuses on leveraging computational methodologies to understand biological systems.
Lenwood S. Heath is a professor in the Department of Computer Science at Virginia Tech with research interests including theoretical computer science, algorithms, graph theory, computational biology and bioinformatics, computational genomics, complex networks, and computational epidemiology. He completed a Ph.D. in Computer Science (1985) at the University of North Carolina, Chapel Hill, an M.S. in Mathematics (1976) at the University of Chicago, and a B.S. in Mathematics (1975) at the University of North Carolina, Chapel Hill.