Exploring aperiodic, complexity and entropic brain changes during non-ordinary states of consciousness

Victor Oswald, Karim Jerbi, Corine Sombrun, Hamza Abdelhedi, Annen Jitka, Charlotte Martial, Audrey Vanhaudenhuyse, Olivia GosseriesarXiv 2025 · DOI · arXiv

TL;DR

Researchers used machine learning and high-density EEG to decode the "Auto-Induced Cognitive Trance" (AICT)—a self-induced, drug-free non-ordinary state of consciousness. They found that trance is characterized by increased cortical excitability in frontal regions and a significant decrease in signal complexity in parietal regions, suggesting a functional "decoupling" from external sensory reality.

What Problem Does This Paper Address?

Non-ordinary states of consciousness (NOCs), such as those induced by psychedelics, meditation, or hypnosis, offer a unique window into brain plasticity and the nature of subjective experience. However, most research focuses on pharmacological inductions (drugs) or passive states (meditation).

This paper investigates Auto-Induced Cognitive Trance (AICT), a state where individuals can voluntarily enter a modified consciousness without external substances. The challenge lies in identifying the specific neural signatures that distinguish this state from normal wakefulness. Specifically, the authors wanted to know if AICT follows the "increased complexity" rule seen in psychedelic states or if it exhibits a unique pattern of neural dynamics.

Methods

The study recruited 27 highly trained AICT practitioners. Using 256-channel high-density EEG, the team recorded brain activity during a baseline "Rest" state and during "AICT."

They focused on three sophisticated EEG metrics:

  1. Aperiodic (1/f) Component: Often considered the "background noise" of the brain, the slope of this component reflects the balance between neural excitation and inhibition (E/I balance).
  2. Lempel-Ziv Complexity (LZC): A measure of how "compressed" or repetitive a signal is; higher LZC indicates a more diverse, unpredictable signal.
  3. Sample Entropy: Quantifies the regularity and complexity of time-series data.

The researchers used a Random Forest machine learning framework to classify the states and identify which brain regions (Regions of Interest, or ROIs) were most important for the transition into trance. They also employed source reconstruction to map these scalp-level signals back to specific cortical areas.

Datasets

The study utilized a specialized dataset of 27 adult participants (23 females, mean age 45). These individuals were not novices; they had an average of 28 months of AICT practice. The dataset included high-density EEG recordings across five conditions, though this paper focused specifically on the contrast between Rest and AICT.

Key Insights and Ideas

  • The "Decoupling" Signature: The most striking finding was a decrease in complexity and entropy in the left parietal cortex and temporo-parietal junction (TPJ). Since these areas are crucial for self-location and sensory integration, their "quieting" likely explains the feelings of disembodiment and immersion reported in trance.
  • Increased Excitability: The aperiodic slope flattened in the frontal regions (specifically the DLPFC) and the Posterior Cingulate Cortex (PCC). In neuroscience terms, a flatter slope suggests increased cortical excitability or arousal, indicating that the brain is highly "active" even if it is tuned out from the environment.
  • A Paradoxical Partnership: AICT appears to trigger the co-activation of the Central Executive Network (CEN) and the Default Mode Network (DMN). Usually, these networks are antagonistic (when one is on, the other is off). Their cooperation here suggests a state of "active introspection"—where the brain uses executive control to maintain a deep, vivid internal narrative.
  • Baseline Predictability: An individual's resting-state brain activity was a significant predictor of how much their brain dynamics would change during trance. This suggests that "trance-ability" might be rooted in one's underlying neural traits.

Key Results

  • Classification Accuracy: The machine learning model successfully distinguished Rest from AICT with 71% accuracy when combining all features.
  • Metric Performance: The aperiodic component was the most effective individual discriminator (65% accuracy), followed by entropy (70%) and complexity (68%).
  • Spatial Effects: The strongest effects (Cohen’s d > 0.8) were localized in the left frontal regions for aperiodic changes and the left parietal/occipital regions for entropic changes.
  • Reduced Variability: Interestingly, entering the AICT state actually reduced inter-individual differences compared to the resting state, suggesting that trance moves different brains toward a common "neural destination."

Limitations and Future Work

The authors acknowledge several hurdles:

  • Gender Imbalance: The sample was 85% female, which may limit how well these results generalize across the population.
  • Ecological Validity: EEG requires participants to remain perfectly still. However, AICT is often induced through vocalizations or movement. This "forced stillness" might have altered the natural depth or quality of the trance.
  • Expertise Levels: While all were "trained," the range of experience (9 to 216 months) was vast. Future studies could look at how "trance expertise" physically reshapes the brain over years.
  • Future Tech: The authors suggest using portable "backpack" EEG systems in the future to allow for more naturalistic, movement-based trance induction.

Summary updated

Abstract

Non-ordinary states of consciousness (NOC) provide an opportunity to experience highly intense, unique, and perceptually rich subjective states. The neural mechanisms supporting these experiences remain poorly understood. This study examined brain activity associated with a self-induced, substance-free NOC known as Auto-Induced Cognitive Trance (AICT). Twenty-seven trained participants underwent high-density electroencephalography (EEG) recordings during rest and AICT. We analyzed the aperiodic component of the power spectrum (1/f), Lempel-Ziv complexity, and sample entropy from five-minute signal segments. A machine learning approach was used to classify rest and AICT, identify discriminative features, and localize their sources. We also compared EEG metrics across conditions and assessed whether baseline activity predicted the magnitude of change during AICT. Classification analyses revealed condition-specific differences in spectral exponents, complexity, and entropy. The aperiodic component showed the strongest discriminative power, followed by entropy and complexity. Source localization highlighted frontal regions, the posterior cingulate cortex, and the left parietal cortex as key contributors to the AICT state. Baseline neural activity in frontal and parietal regions predicted individual variability in the transition from rest to AICT. These findings indicate that AICT engages brain regions implicated in rich subjective experiences and provide mechanistic insights into how self-induced trance states influence neural functioning.
BibTeX
@misc{oswald2025exploring,
  author = {Victor Oswald and Karim Jerbi and Corine Sombrun and Hamza Abdelhedi and Annen Jitka and Charlotte Martial and Audrey Vanhaudenhuyse and Olivia Gosseries},
  title = {{Exploring aperiodic, complexity and entropic brain changes during non-ordinary states of consciousness}},
  year = {2025},
  eprint = {2509.19254},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2509.19254},
  howpublished = {arXiv},
}
Figure