Artificial intelligence in clinical genetics: Current practice and attitudes among the clinical genetics workforce.

Abstract

Because clinical genetics artificial intelligence (AI) applications and capabilities are on a rapid rise, this study sought to assess workforce readiness. To assess the workforce's use, knowledge, and attitudes about medical AI applications, we conducted a survey of 215 US-based genetics clinicians and trainees. More than half (51.2%) of participants reported little to no knowledge of AI in clinical genetics; 64.3% reported no formal training. Formal training correlated with greater self-reported knowledge of AI in clinical genetics: 69.3% of respondents with formal training (vs 37.5% without) reported intermediate to extensive knowledge of AI. Most participants reported insufficient knowledge of clinical AI (83.4%), desired more education (97.6%), and would take available training (89.3%). The majority (51.6%) of clinician participants initially reported not using AI applications in the clinic. However, after a tutorial describing clinical AI applications, 75.8% reported some use. When asked about specific applications, the majority of clinician participants used facial diagnostic applications (54.9%) and AI-generated genomic testing results (62.1%); other applications, such as chatbots, large language models, pedigree or medical summary generators, and risk assessment, were less commonly used (11.1%-12.5%). Further education is both desired and needed to optimally use AI applications in clinical genetics.