compute_SOP

compute_SOP(EEG, Fs, varargin)

COMPUTE_SOP Compute normalized slow-oscillation power from EEG

Usage:

[SOpower_norm, SOpower_times, SOpower_stages, norm_method, ptile] = …

compute_SOP(EEG, Fs, ‘Name’, Value, …)

Inputs:

EEG : 1xN double - time series EEG data – required Fs : double - sampling frequency in Hz – required

Name-Value Pairs:

‘stage_vals’ : 1xS double - stage values (5=W,4=R,3=N1,2=N2,1=N3) (default: []) ‘stage_times’ : 1xS double - stage onset times (default: []) ‘SO_freqrange’ : 1x2 double - SO band in Hz (default: [0.3 1.5]) ‘SOpower_outlier_threshold’ : double - z-score cutoff for SOpower outliers (default: 3) ‘norm_method’ : char - ‘pNshiftS’, ‘percent’, ‘proportion’, or ‘none’ (default: ‘p2shift1234’) ‘retain_Fs’ : logical - upsample SOpower to EEG Fs (default: true) ‘tapers’ : 1x2 double - [TW, num_tapers] (default: [15 29]) ‘window_params’ : 1x2 double - [win_size, step_size] in seconds (default: [30 15]) ‘EEG_times’ : 1xN double - EEG sample times (default: (0:N-1)/Fs) ‘time_range’ : 1x2 double - restrict normalization to this window (default: [EEG_times(1), end]) ‘isexcluded’ : 1xN logical - artifact mask (default: all false)

Outputs:

SOpower_norm : 1xM double - normalized SOpower time series (row vector, possibly upsampled) SOpower_times : 1xM double - times of the SOpower samples (s) SOpower_stages : 1xM double - stage value at each SOpower sample (or true if staging not provided) norm_method : char - resolved normalization method name ptile : double or 1x2 double - percentile value(s) used for normalization (empty for ‘none’/’proportion’)

Notes

pNshiftS parses to percentile N computed across the union of stages in S (digits). Artifact samples are replaced with NaN before the multitaper estimate via compute_mtspect_power.

See also: compute_mtspect_power, nanzscore, multitaper_spectrogram_mex

∿∿∿ Prerau Laboratory MATLAB Codebase · sleepEEG.org ∿∿∿