Neuroscientists use AI to detect depression from brain scans
AI model analyzes EEG and fNIRS data to spot depressive states with 89% accuracy
Deep Dive
A pilot study of eleven healthy students establishes a framework for classifying depressive states using EEG and fNIRS brain signals, aiming toward objective, automated diagnostic tools to complement subjective clinical evaluations.
Key Points
- Researchers from the University of Tokyo and collaborators built an end-to-end ML system for detecting depression using EEG and fNIRS brain signals
- The model achieved 89% accuracy in classifying depressive states in a pilot study of 11 healthy students
- The system aims to provide objective, bias-free alternatives to traditional psychiatric evaluations
Why It Matters
Could revolutionize mental health diagnostics with objective, early detection of depression using non-invasive brain scans.