Yale Child Study Center is launching a two-year research project funded by Bukhman Philanthropies designed to harness Artificial Intelligence (AI) to advance child mental health.
With the aim to build capacity and train the next generation of clinician-scientists, a core part of the initiative will train and support faculty, trainees, and research staff through workshops, bootcamps, online modules, and embedded consultation. The programme will focus on practical skills for child mental health research, including working with unstructured clinical data, responsible use of electronic health records, reproducible research, and applied machine learning.
A second focus of the initiative will be on two pilot research studies, one focused on predicting ADHD in children. ADHD affects an estimated 5–7% of children and is often under-identified. Many children are recognized only after problems have already grown. To move beyond this “wait-to-fail” pattern, the research team will develop and test a multimodal prediction approach, combining multiple types of data—genetic (common and rare variation), clinical, and environmental information from health records and surveys. Using large-scale datasets and machine learning methods, they will test which data sources improve prediction and whether combined models outperform simpler ones.
A second study will examine how compulsive digital media use relates to youth suicidality. While public discussion often centers on total screen time, recent findings suggest addictive patterns of use may be more closely linked to suicidal thoughts and behavior. The team will test whether compulsive use itself increases risk or whether shared genetic factors contribute to both addictive behaviors and psychiatric vulnerability. Using genotype and behavioral data from the ABCD Study, researchers will calculate genetic risk scores and evaluate whether the association between addictive screen use and suicidality remains after accounting for genetic liability, and whether genetic risk changes the strength of that relationship.