fNIRS-Based Awareness Detection in Disorders of Consciousness
A clinical brain-computer interface using functional near-infrared spectroscopy to detect covert awareness in patients with disorders of consciousness. The system employs individualized Riemannian classification combined with permutation-based statistical analysis for bedside diagnostic assessment.
The portable, ICU-deployable system achieves 100% sensitivity and 89% specificity for awareness detection using just 2-second encoding windows — from portable bedside cap acquisition through covariance feature construction and Riemannian classification to permutation-based awareness decisions.
Published in Neurophotonics 12(4), 2025. Winner of the IEEE Brain Discovery Workshop Poster Prize (1st place).