EEG Paradox — Vigilance (VIGALL-inspired)

Global result

Slow-dominant (V5)
VIGALL Stage Slow-dominant Level V5 Index 23.3 (0–100) Gunkelman Microsleep Stability 97% Transitions 1 Duration 31.0 s

Stage distribution: B1 ×1 B2_3 ×30

Timeline (epoch-by-epoch)

0 A1 A2 A3 B1 B2_3 C

Region breakdown (subnet)

RegionDominant stateMean indexDistribution
CentralLight Sleep21.50Deep Sleep:2, Light Sleep:15, Microsleep:14
FrontalMicrosleep21.90Deep Sleep:3, Light Sleep:13, Microsleep:15
PosteriorMicrosleep24.20Deep Sleep:1, Light Sleep:10, Microsleep:20
TemporalMicrosleep26.80Light Sleep:10, Microsleep:21

Per-site assessment (endpoint)

SiteRegionGunkelmanVIGALLIndexLevelSFIHFTIPAIα/θCmplx
C3centralLight SleepB2_320V53.250.120.530.330.05
C4centralMicrosleepB2_323.70V42.270.110.390.290.06
CZcentralLight SleepB2_320.90V42.870.100.490.310.05
F3frontalMicrosleepB2_323.80V42.250.120.400.300.06
F4frontalMicrosleepB2_322V42.470.120.420.320.05
F7otherMicrosleepB2_326V41.990.130.300.270.07
F8otherLight SleepB2_321.80V42.520.100.300.260.05
FP1frontalLight SleepB2_321.50V43.180.100.300.280.04
FP2frontalMicrosleepB2_321.20V43.040.100.320.290.04
FZfrontalMicrosleepB2_323.60V42.250.110.450.330.05
O1posteriorMicrosleepB2_326.60V41.930.140.450.350.07
O2posteriorMicrosleepB2_328.10V41.810.130.490.410.07
P3posteriorMicrosleepB2_323.90V42.170.120.390.300.06
P4posteriorMicrosleepB2_324.90V42.230.110.410.320.06
P7temporalMicrosleepB2_329.30V41.570.120.300.280.07
P8temporalMicrosleepB2_327.50V41.890.120.350.300.06
PZposteriorMicrosleepB2_323.20V42.400.110.480.350.05
T7temporalMicrosleepB2_324.30V42.140.100.350.290.05
T8temporalMicrosleepB2_327.30V41.740.090.230.230.05

SFI = Slow-Fast Index (delta+theta)/(alpha+beta); HFTI = High-Freq Tension Index beta/(alpha+theta); PAI = Posterior Alpha Index site_alpha/frontal_alpha_median; α/θ = alpha/theta ratio; Cmplx = Complexity beta/(alpha+theta+delta)

Instability metrics

Transitions/min 1.94 Lability 0.13 (0–1) Index σ 6.60

Frontal–posterior coherence

Divergent epochs 0% Mean divergence 4.7 pts Max divergence 11.4 pts

Dwell times

StageTime (s)%
B2_360.098%
B11.02%

Transitions

#FromToTime (s)
1B2_3B130.0

All stages (reference)

LevelStageIndex rangeDescription
V1080–100Active wakefulness — desynchronized, high beta
V2A165–80Relaxed wakefulness — posterior-dominant alpha
V3A2/A345–65Alpha anteriorization — drowsy-onset
V4B125–45Drowsiness — low amplitude, desynchronized
V5B2/30–25Deep drowsiness — delta/theta dominant

Computation log

adaptive_alpha_threshold: 8.004541837735692e-13

epoch_duration_s: 2.0

Method

Per-epoch: Welch PSD → band power per channel → SFI/HFTI/PAI/alpha-theta features per site → sigmoid evidence → continuous index (0-100) → Gunkelman 7-state mapping per site → region median → weighted global

n_epochs: 31

Scalp-level adaptation of VIGALL 2.1 (no LORETA). Each electrode gets a continuous vigilance index plus Gunkelman state (Alert/Relaxed/Drowsy/Microsleep/Light Sleep/Deep Sleep/REM Sleep). Regions aggregate via median; global via weighted mean.

overlap: 0.5

regions_available: ['posterior', 'central', 'temporal', 'frontal']

sites_available: ['C3', 'C4', 'CZ', 'F3', 'F4', 'FP1', 'FP2', 'FZ', 'O1', 'O2', 'P3', 'P4', 'P7', 'P8', 'PZ', 'T7', 'T8']

threshold_rule: max(alpha_based, 0.2*posterior_dt). A1: alpha>beta*1.0. B2_3: dt>threshold*3 and dt>alpha*1.5. Stage0: beta>alpha*0.8, beta>dt*0.5, dt<=alpha*1.5. B1: else.

Interpretation

Vigilance is classified per 2-second epoch using a VIGALL-inspired algorithm with three-level hierarchy: Site (endpoint) — each electrode gets a continuous vigilance index (0–100) from SFI/HFTI/PAI features plus a Gunkelman 7-state label; Region (subnet) — median of member-site indices, dominant state by vote; Global (network) — weighted mean of region indices. Alpha topography drives A-substages; delta+theta dominance → B-stages; high beta ratio → Stage 0 (active). Not a diagnosis — a descriptive EEG arousal proxy.

Subject: EO | Session: 001