The primary validation domain for QNFM. Scalp EEG seizure detection using the A-Gate architecture on IBM quantum hardware.
The CHB-MIT Scalp EEG Database (PhysioNet) contains continuous EEG recordings from 23 pediatric patients with intractable epilepsy. Recordings use the international 10-20 system at 256 Hz sampling rate.
Our experiments use a 7-patient subset for hardware validation: chb01, chb03, chb05, chb07, chb11, chb14, chb21. Note: chb07 (not chb08) — selected for sufficient seizure events, consistent recording montage, and coverage of both polarity populations.
Full 22-patient simulation results inform the calibration tier analysis. Leave-One-Subject-Out (LOSO) cross-validation throughout — no within-patient training.
The Mark IV headset (18-channel, OpenBCI Cyton + Daisy) is built and awaiting first recording. Mark IV data will provide a second independent dataset beyond CHB-MIT, with controlled recording conditions and known electrode placement.