AI could speed up psychosis assessment process, says study

D-Keine/Getty Images

By Olivia Bowthorpe

Artificial intelligence could help clinicians identify people at high risk of developing psychosis, by analysing transcripts of psychiatric interviews, researchers say.

The study by the University of Oxford and King’s College London tested 11 large language models (LLMs) and found that some were able to assess symptoms and identify people at risk with around 80% accuracy.

"These findings demonstrate that open-weight LLMs have the potential to assess psychosis risk from psychometric interview transcripts, supporting scalable, human-in-the-loop approaches to early detection," the team reported in npj Digital Medicine.

Patients at a high risk of psychosis often face long waiting lists for assessment, with almost nine out of 10 not identified in time to prevent onset.

Early detection of those at high risk is currently constrained due to the need for specialist interpretation of narrative interviews, said lead author Dr Taiyu Zhu, lecturer in Large Language Models for Healthcare at King’s College London, adding the use of AI could help speed up assessments and access to care, leading to better outcomes for patients.