Adolescent Violence and Injury Detection System
Beginning in January 2020, the Johns Hopkins International Injury Research Unit, along with project partners in Vietnam and India, began Phase 1 of an innovative research project that includes e-consultation workshops, qualitative sub-studies, systematic reviews, secondary data analysis, and the development of an AI algorithm to flag at-risk adolescents experiencing violence and injuries.
Impact and innovation
The machine learning algorithm for predicting the risk of adolescent injuries would be the first application of AI in injury research and prevention.
Project partners
- Hanoi University of Public Health
- Indian Institute of Health Management Research
- Indian Institute of Public Health, Hyderabad
Goal
- The overall goal of this project is to develop a tool using artificial intelligence (AI) to detect adolescents at risk of experiencing violence and injuries.
Objectives
- Identify the main predictors of injuries and violence in the literature
- Develop an AI prediction model to detect adolescents at risk of violence and injuries, and compare its performance with the logistic model
- Conduct formative qualitative work (interviews and FGDs) to understand key contextual issues around adolescent injury and violence and inform development of a mobile and web-based platform, including the review of regulatory and built-capacity for integration and sustainability.