Figure 1
(A) In the Sleep (nap) MEG session, participants are delivered habituation auditory streams (HAB; regularly ascending/descending half-tone scales) while falling asleep. When stable NREM2/NREM3 sleep stages are observed, the experimenter delivers statistical (STAT) and random (RDM) 5-minutes auditory streams (counterbalanced). Participants who do not reach NREM sleep or cannot be delivered STAT streams are assigned to the No NREM/No Exposure group (N = 10); others are assigned to the NREM Sleep/Exposure group (N = 11). (B) In the subsequent MEG session at Wake, both groups are delivered 5-minutes STAT and RDM streams (2 times each, counterbalanced). Outside of the scanner, they are then informed of the presence of regularities, and administered a two alternative forced choice (2AFC) test between trained and never seen tritones. (C) Auditory streams. STAT streams feature non-melodious successions of 12 pure tones that can be statistically grouped into four randomly alternating tritones (G#CD, AC#G, FA#D# and EF#B in musical notation, A = 440 Hz). Transitional succession probability between two tones within a tritone is 1.0, and transitional probability between two tritones is 0.33, creating statistical grouping regularities. In RDM streams, the same 12 tones are randomly presented (transitional probability between tones = 0.09). Habituation (HAB) streams are constructed with the same 12 tones. Tones are delivered at a 5.505 Hz presentation rate, and tritone presentation rate is 1/3 of the tone = 1.835 Hz (see Methods).
