Emotion-focused AI spots early signs of psychosis in youths from short speech samples
MedicinePsychologyNeuroscience
THE AI ANGLE
Extracting emotional cues from speech transcripts to predict psychosis riskResearchers created an AI tool to detect psychosis risk in youths from 45-second transcripts of casual speech. The system maps subtle emotional cues like fear and joy, reaching up to 93 percent detection sensitivity. This approach gives clinicians explainable assessments for patients who do not openly disclose their symptoms.
THE TEACHING ANGLE
Direct model prompts failed on casual conversation, highlighting the difficulty of spotting clinical risk when patients do not explicitly mention symptoms.Read the original at medicalxpress.com Generate teaching or study materials
More in Medicine
- AI companion startup Character.ai files for Chapter 11 bankruptcy, citing costs and litigationCNBC — Technology · October 14, 2026
- Can AI feel safer than human advice for consumers who face discrimination?The Conversation — AI · October 14, 2026
- Wearable cyborg training linked to longer walking distances in people with ALSMedical Xpress · October 14, 2026
- The crucial responsibilities doctors must never offshore to AIMedical Xpress · October 14, 2026
- Why Agentic AI Works in Hospitals When You Strip It DownMedCity News · October 14, 2026