PLoS ONE
December 16, 2022
Caroline L. Alves, Rubens Gisbert Cury, Kirstin Roster et al.
13 citations
Ayahuasca, an Amazonian plant blend used in traditional medicine for centuries, is a promising therapy for neurological and mental diseases. Using an EEG dataset, machine learning and complex network analysis automatically detected changes in brain activity at three data abstraction levels. Connectivity changes between brain regions (correlation of EEG time series) yielded the highest accuracy (92%), followed by raw EEG (88%) and complex network measures (83%). The frontal and temporal lobes were most activated, consistent with prior work. A novel finding identified F3 and PO4 as the most important brain connections, possibly linked to face-recognition-like cognitive processes during visual hallucinations.
bioRxiv (Cold Spring Harbor Laboratory)
December 11, 2025
Caroline L. Alves, Fernanda Palhano-Fontes, Thaise G. L. de O. Toutain et al.
Ayahuasca alters conscious experience, and this study identifies EEG markers of its network-level effects using machine learning and complex-network analysis. In a randomized, double-blind, placebo-controlled trial with naïve ayahuasca users, resting-state EEG was recorded before dosing, 2 hours after, and 4 hours after. Connectivity was estimated with sliding windows; optimal classification performance occurred at 60–70 seconds (AUC and accuracy = 0.93). Network analysis revealed a bilateral decrease in eigenvector centrality (weaker hub influence), increased degree heterogeneity in the right hemisphere, and reduced global efficiency in the left. Posterior-left connections weakened acutely, while right temporal–central coupling transiently strengthened. The findings suggest that hub-centric shortcuts weaken, routing communication through more distributed, less efficient pathways with right-lateralized expression.
May 5, 2023
Caroline L. Alves
Predictive models using complex network analysis of EEG and fMRI data can automatically detect brain changes associated with schizophrenia, Alzheimer's disease, and autism, and also identify functional changes induced by psychedelics such as ayahuasca and N,N-dimethyltryptamine. The models outperformed those previously reported in the literature. Interpreting the models and network metrics provided new insights into how these substances work. New metrics developed by the author were especially relevant for autism and psychedelic studies. The approach is flexible and likely applicable to other conditions and substances, like antidepressants.
medRxiv
June 16, 2022
Caroline L. Alves, Thaise G. L. de O. Toutain, Joel Augusto Moura Porto et al.
preprint
A machine-learning method using support vector machines classified EEG data from volunteers before and after inhaling the psychedelic DMT. Complex network measures derived from brain connectivity achieved 89% AUC, outperforming raw connectivity matrices. Key distinguishing features included connections between temporal and central cortex regions (TP8-C3) linked to finger movements, and between precentral gyrus and lateral occipital cortex (FC5-P8) potentially related to emotional and mystical experiences. Closeness centrality was the most important network measure. DMT increased community size and average path length, disrupting the balance between functional segregation and integration, supporting the idea that cortical activity becomes more entropic under psychedelics.