Application of machine learning and complex network measures to an EEG dataset from ayahuasca experiments

This preprint (2022) used an EEG dataset from a study involving ayahuasca to investigate the ability of machine learning and complex network measurements to automatically detect changes in brain activity. After applying machine learning at three different levels of data abstraction, machine learning proved to be consistent with the current literature and showed the highest accuracy in detecting the correlation of the EEG time series.

Abstract

“Ayahuasca is made from a mixture of Amazonian herbs and has been used for a few hundred years by the people of this region for traditional medicine. In addition, this plant has been shown to be a potential treatment for various neurological and psychiatric disorders. EEG experiments have found specific brain regions that changed significantly due to ayahuasca. Here, we used an EEG dataset to investigate the ability to automatically detect changes in brain activity using machine learning and complex networks. Machine learning was applied at three different levels of data abstraction: (A) the raw EEG time series, (B) the correlation of the EEG time series, and (C) the complex network measures calculated from (B). As a result, the machine learning method was able to automatically detect changes in brain activity, with case (B) showing the highest accuracy (92%), followed by (A) (88%) and (C) (83%), indicating that connectivity changes between brain regions are more important than connectivity changes within brain regions. The most activated areas were the frontal and temporal lobe, which is consistent with the literature. In terms of brain connections, the correlation between F3 and PO4 was the most important. This connection may point to a cognitive process similar to face recognition in individuals during ayahuasca-mediated visual hallucinations. Furthermore, closeness centrality and assortativity were the most important complex network measures. These two measures are also associated with diseases such as Alzheimer’s disease, indicating a possible therapeutic mechanism. Overall, our results showed that machine learning methods were able to automatically detect changes in brain activity during ayahuasca consumption. The results also suggest that the application of machine learning and complex network measurements are useful methods to study the effects of ayahuasca on brain activity and medical use.

Authors: Caroline L. Alves, Rubens G. Cury, Kirstin Roster, Aruane M. Pineda, Francisco A. Rodrigues, Christiane Thielemann & Manuel Ciba

Become a psychedelic insider!

With a free Blossom membership you will always be in the know.

📰 Weekly newsletter about the psychedelic research

✔️ Unlimited access to our database and original articles

🖊️ Add (private) notes and comments to each page

Make an account

Study details

Topics studied
Neuroscience Technology

Study characteristics
Theory Building Bio/Neuro

Participants
0 Humans

Institutes

Institutes associated with this publication

University of São Paulo
The University of São Paulo has been conducting research with psychedelics for many years, with a focus on ayahuasca given its traditional use in Brazil.

0 Comments
Inline Feedbacks
View all comments