Talk description
Medicinal cannabis use is widespread but poorly captured in electronic health records and rarely structured enough to support research. This talk presents our efforts to address that gap by applying large language models to millions of clinical notes, extracting structured data on cannabis use patterns, indications, and outcomes. The focus is on the translational potential of this approach: how AI tools can convert an underutilized clinical data source into a research-ready dataset capable of supporting observational studies, and to ultimately inform cannabis policy and improve patient care.
Johannes Thrul
Johannes Thrul, PhD, is an Associate Professor in the Department of Mental Health at the Johns Hopkins Bloomberg School of Public Health and has a joint appointment at the Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins. He received his PhD in Psychology from the Friedrich-Alexander University Erlangen-Nuremberg and completed postdoctoral training at UCSF. His research focuses on substance use and addiction, as well as digital and mobile health research, including the use of smartphone-based ecological momentary assessment and just-in-time interventions. He leads multiple NIH-funded studies and co-directs a national longitudinal research registry to examine the health effects of medicinal cannabis. He has received numerous teaching awards at Johns Hopkins and serves in editorial and scientific advisory roles across the addiction and digital health fields.
