Long-term meditators self-induce high-amplitude gamma synchrony during mental practice

Lutz et al. 10.1073/pnas.0407401101.

Supporting Information

Files in this Data Supplement:

Supporting Methods
Supporting Figure 4
Supporting Figure 5
Supporting Figure 6
Supporting Figure 7
Supporting Figure 8
Supporting Figure 9

Supporting Figure 4

Fig. 4. Difference between meditation and muscle activity. (a) Raw electroencephalographic signals. At t = 45 s, practitioner S4 started generating a state of "nonreferential compassion" or mimicked and exaggerated the possible muscle activity during this meditative state. (b) Different power spectral distributions were found during these two states. (c) Meditation exhibits a different linear relation than muscle activity between gamma activity and 80-120 Hz activity. This difference was revealed by an interaction between states and covariate (80-120 Hz activity) in a separate slope analysis of covarience (ANCOVA) design [F(2,286) = 166, P < 0.0001]. Points correspond to average power in a 2-s epoch during four blocks, with slopes equal to 0.89 and 4.01 and intercepts equal to 25 and 74.8 for muscle and meditation, respectively (SE equals 0.14 for slope and 23 for intercept).

Supporting Figure 5

Fig. 5. Modulation of the event-related potential (ERP) by the ongoing gamma oscillatory activity. Subjects were passively listening to auditory pulses (65 dB for 50 ms at 1,000 or 1,200 Hz) during a concentrative meditation on a visual object, during reading, and during another objectless meditative state called open presence (rig pa’i cog bgzag in Tibetan). This objectless meditation is similar to the pure compassion state but without its affective components. In particular, this state requires a complete relaxation of all muscles, including facial muscles. EEG signals showing eye movements or muscular artifacts 1 s before or after a pulse were manually excluded from the study. For subjects S2 and S4 there were, respectively, 80 and 1,080 trials during open presence, 88 and 414 trials during concentration on a visual object, and 410 trials for subject S4 during reading. (a--c) High-gamma oscillation was still found during the open presence state for subjects S2 and S4. (a) Subject S4; raw signal, 25-42 Hz (F3); 250 auditory pulses during the open presence state. (b) For subject S4, we collected enough data to show that the amplitude of the gamma oscillations measured before the external stimulation predicts the amplitude of the high oscillatory responses evoked by the pulses. The 1,080 trials were ranked depending on the global prestimulation gamma activity across electrodes. The ERP was computed on F3 across 12 clusters of 90 trials each. A significant correlation (r = 0.85, P < 0.005) was found between the ranked prestimulation gamma activity and the evoked gamma power (25-42 Hz). (c) For subjects S2 and S4, the frequencies in the 25- to 42-Hz band showing a peak of energy during the 1-s prestimulation were the frequencies for which the oscillatory evoked responses increased the most compared with the prestimulation baseline noise. For the controls, the evoked gamma response was negligible compared with those for subjects S2 and S4 and was correlated with low gamma activity before the stimulation. Only the two blocks ranked by subject S4 as the best were used in c (439 trials). The ERP was computed by averaging across trials the band-passed electroencephalogram (EEG) signal over channel F3. The time–frequency power emission, 128-sampled-points time window of the ERP (0-250 ms) was z-transformed by using the mean and standard deviation of the energy between –600 and 0 ms. Because the evoked activity is relatively independent of muscle activity, the relationship between the prestimulation fast-frequency oscillation and the evoked activity suggests that these high-amplitude gamma rhythms are not muscle artifacts.

Supporting Figure 6

Fig. 6. Electroencephalogram (EEG) raw signal and event-related potential (ERP) after independent component analysis (ICA) correction (1). High-amplitude gamma oscillations (25-42 Hz) (a) and high-amplitude, fast-frequency ERP (15-30 Hz) (b) remain after ICA extraction of muscle activity.

1. Makeig, S., Jung, T. P., Bell, A. J., Ghahremani, D. & Sejnowski, T. J. (1997) Proc. Natl. Acad. Sci. USA 94, 10979-10984.

Supporting Figure 7

Fig. 7. Dipole modeling of oscillatory evoked responses (BESA). Two deep and two shallow dipole sources model 64% of the variance from the high-amplitude oscillatory evoked responses for subject S4 (15-30 Hz) from 30 to 190 ms (≈400 trials during open presence meditation).

Supporting Figure 8

Fig. 8. Independent component analysis (ICA) of event-related potential (ERP) data (1). ICA identifies in the electroencephalogram (EEG) signals (subject S4, open presence) independent sources over temporalis muscles. Dipole modeling of the contribution of these sources to the ERP reveal two shallow dipole sources over temporal electrodes (57% of total variance) and suggest a muscle origin for these sources.

1. Makeig, S., Jung, T. P., Bell, A. J., Ghahremani, D. & Sejnowski, T. J. (1997) Proc. Natl. Acad. Sci. USA 94, 10979-10984.

Supporting Figure 9

Fig. 9. Brain origin of the high-amplitude, fast-frequency event-related potential (ERP). The independent component analysis (ICA) sources corresponding to muscle activity were removed from the electroencephalogram (EEG) signals (subject S4, open presence). Dipole modeling of the ERP revealed one deep central dipole and one dipole over the right frontal cortex, which explains 28% and 22% of the total variance, respectively, for brain model (a) and the brain anatomy of subject S4 (b). The exact location of the dipoles cannot be fully resolved because of the imprecision of the method [on the order of 1-2 cm (1)]. The localization within the brain of the dipole sources of these fast-frequency-evoked oscillations further supports the brain origin of these high-amplitude gamma oscillations.

1. Baillet, S., Riera, J. J., Marin, G., Mangin, J. F., Aubert, J. & Garnero, L. (2001) Phys. Med. Biol. 46, 77–96.