Pediatric EEG Reveals Stable Individual Fingerprints

Researchers applied Bayesian reduced-rank regression to EEG power spectra from 782 normally developing children aged 6 weeks to 19 years, recorded during nREM sleep, and published January 30, 2026 in PLoS Computational Biology. The learned low-dimensional representations separated individuals, generalized across N1 and N2 stages, and showed increasing stability with age. BRRR outperformed correlation-based fingerprinting, suggesting a robust tool for clinical pediatric EEG analysis.