Stockholm News: An international research endeavor has demonstrated that when it comes to disease diagnostic accuracy, utilizing AI assistants to conduct mental health assessment interviews outperforms the standardized scales that are widely employed in the present healthcare system. This finding opens up a fresh avenue for digital assessment in the realm of mental health.

A research team spearheaded by Lund University in Sweden created an AI assistant called Alba. Alba conducted interviews with 303 participants. This group encompassed patients diagnosed with nine different types of mental disorders, including depression and anxiety, as well as healthy individuals.

During the interviews, the AI assistant began by asking 15 to 20 open-ended questions. Subsequently, it offered recommendations based on the diagnostic manual for mental disorders. Meanwhile, the participants also completed standardized scales.

The results were quite telling. The AI assistant attained a higher diagnostic accuracy for eight types of mental disorders. It was especially adept at differentiating between mental disorders that exhibit overlapping symptoms. Moreover, the majority of participants had a positive experience during the AI interview. They characterized the AI assistant as empathetic and regarded its questions as highly pertinent.