Which AI tools for mental healthcare are being invented, which ones reach the market, and what does it take for them to be used in practice? We bring together evidence from patents, regulated products, and clinicians’ perspectives.
Six focus groups with therapists and patients revealed 16 facilitators, 11 barriers, and 9 context-dependent factors for adopting AI in psychotherapy – from useful features and cost coverage to the fear of losing human contact.
Almost half of the 392 psychotherapists, psychiatrists, and clinical psychologists we surveyed in Germany and the US had never heard of AI-enabled technologies for their field. Learning intentions were higher than use intentions – and highest for practice management tools.
Out of 84 candidate products, only seven regulated AI-enabled clinical decision support systems for mental healthcare met our inclusion criteria – and the scientific evidence behind them is thin. Our product review in Artificial Intelligence in Medicine.
At the DGPs Congress in Vienna, I chaired a symposium on trust in AI and presented two studies: on the landscape of AI-enabled clinical decision support systems in mental healthcare, and on how performance and trust shape the use of generative AI in academia.
Our review of AI-enabled precision psychiatry patents published between 2015 and mid-October 2022, now published in Computers in Human Behavior Reports. Multiple analytic approaches, such as graphic network analysis and topic modeling, are used to analyze the scope, content, and trends of the retained patents. The tools described aim to provide diagnosis, prediction of treatment responses, and prognosis of mental disorder symptoms. Additionally, about one-third of the tools suggest treatment options related to selection, adjustment, and management. The complexity of technology combinations has increased over the years. This review highlights the potential of AI-enabled precision psychiatry tools for adoption in practice.