Related Experiment Video
Updated: Aug 17, 2025
08:51
Statistical Modelling of Cortical Connectivity Using Non-invasive Electroencephalograms
Published on: November 1, 2019
5.7K
Application of machine learning and complex network measures to an EEG dataset from ayahuasca experiments
Caroline L Alves1,2, Rubens Gisbert Cury3, Kirstin Roster2
1BioMEMS Lab, Aschaffenburg University of Applied Sciences (UAS), Aschaffenburg, Germany.
Plos One
|December 16, 2022
Summary
Ayahuasca use alters brain activity, detectable by machine learning analyzing EEG data. Connectivity changes between brain regions, particularly F3 and PO4, were key indicators, offering insights into psychedelic mechanisms.
Area of Science:
- Neuroscience
- Computational Psychiatry
- Pharmacology
Background:
- Ayahuasca, an Amazonian plant blend, has a long history in traditional medicine.
- It shows potential for treating neurological and mental health conditions.
- Previous EEG studies identified brain region changes associated with ayahuasca.
Purpose of the Study:
- To investigate the automatic detection of brain activity changes induced by ayahuasca using machine learning.
- To analyze brain activity at different levels of data abstraction: raw EEG, time series correlation, and complex network measures.
- To develop novel complex network measures for enhanced analysis.
Main Methods:
- Applied machine learning to raw EEG time series, EEG time series correlations, and complex network measures derived from correlations.
- Calculated complex network measures, including novel community detection metrics.
- Utilized a dataset of EEG recordings from individuals consuming ayahuasca.
Main Results:
- Machine learning successfully detected ayahuasca-induced brain activity changes.
- Correlation analysis (92% accuracy) outperformed raw EEG (88%) and complex network measures (83%).
- Identified frontal and temporal lobes as most affected, with F3-PO4 connections being crucial, potentially linked to visual hallucinations and face recognition.
Conclusions:
- Brain connectivity changes are highly indicative of ayahuasca's effects.
- Closeness centrality and assortativity were significant network measures, potentially relevant to neurodegenerative diseases.
- Ayahuasca may slow information dissemination in functional brain networks, suggesting altered cognitive processing and providing insights into psychedelic mechanisms of action.
