Automatic Sleep Staging

Classifies EEG epochs into sleep stages (Wake, N1, N2, N3, REM) using spectral power analysis based on simplified AASM criteria. Each epoch is scored by computing spectral band powers (delta, theta, alpha, sigma, beta) and applying rule-based classification.

Usage

eegSleepStage(
  x,
  method = c("spectral", "rule_based"),
  epoch_sec = 30,
  assay_name = NULL
)

Arguments

x

A PhysioExperiment object with EEG data.

method

Staging method: "spectral" (FFT power-based) or "rule_based" (synonym, same algorithm).

epoch_sec

Epoch duration in seconds (default: 30, AASM standard).

assay_name

Input assay name. If NULL, uses the default assay.

Value

A data.frame with columns:

epoch

Integer epoch number.

stage

Character sleep stage: "W", "N1", "N2", "N3", or "REM".

start_sample

Integer start sample of the epoch.

end_sample

Integer end sample of the epoch.

delta_power

Numeric delta band power (0.5-4 Hz).

theta_power

Numeric theta band power (4-8 Hz).

alpha_power

Numeric alpha band power (8-13 Hz).

sigma_power

Numeric sigma band power (12-16 Hz).

beta_power

Numeric beta band power (16-30 Hz).

The result is also stored in metadata(x)$sleep_stages.

References

Berry, R. B., et al. (2017). AASM Scoring Manual Updates for 2017. Journal of Clinical Sleep Medicine, 13(5), 665-666.

See also

eegSpindleDetect(), eegKcomplexDetect(), eegSlowWaveDetect(), eegSleepMetrics(), eegPlotHypnogram()

Examples

if (FALSE) { # \dontrun{
pe <- make_eeg_sleep(n_time = 90000, n_channels = 2, sr = 500)
stages <- eegSleepStage(pe, epoch_sec = 30)
head(stages)
} # }