Skip to main content
. 2020 Jul 28;9:e55092. doi: 10.7554/eLife.55092

Author response image 3. The influence of different fit algorithms on spectral slope estimation in general anesthesia with propofol.

Author response image 3.

a, 1/f spectral slope in wakefulness (Wake, red) and under anesthesia (Ana, blue) using the original polyfit (Orig., orange), FOOOF (lblue) or eBOSC (purple) algorithm (n = 9, averaged across electrodes). Paired t-tests (uncorrected): Orig.-FOOOF: p = 0.082, t8 = 1.99, d = 0.18; Orig.- eBOSC: p = 0.369, t8 = -0.95, d = -0.03, FOOOF - eBOSC: p = 0.118, t8 = -1.75, d = -0.21. n.s. – not significant. b, At Fz ,1/f slope from original polyfit (Orig., orange) and FOOOF (blue; left panel: r = 0.98, p < 0.001) and original and eBOSC (purple; right panel: r = 1.00, p < 0.001) are strongly correlated.c, Goodness of fit (R2) of the different slope estimates (Orig. - original polyfit (orange); FOOOF (light blue); eBOSC (purple)) to the power spectra in wakefulness and under propofol anesthesia at electrode Fz (n = 9). Permutation t-tests: Orig. - fooof: p = 0.538, Orig. - eBOSC: p = 0.726, fooof - eBOSC: p = 0.690. n.s. – not significant.