We combine psychometric testing, natural language processing and electrophysiological recordings obtained using a 24 channel 24 bit fully portable EEG system (MBT Smarting) with Ag/AgCl electrodes (EasyCap) to investigate the acute effects of N,N-dimethyltryptamine (DMT) in natural settings.

We study how convolutional deep neural networks can learn phase transitions and universality classes, with the purpose of detecting critical behavior in neural systems.

We construct semi-empirical dynamical systems models of whole-brain activity, incorporating functional information (fMRI) and in vivo estimates of anatomical connectivity (diffusion tensor imaging). We use these models to rehearse manipulations transitioning between healthy and pathological brain states.

Modeling calcium imaging data obtained from Drosophilia allows us to investigate with more detail the interplay between structural and functional connectivity in neural systems.

Using a dual direct current stimulator and wireless EEG (Neuroelectrics), we investigate how external alternating currents entrain endogenous oscillations and how this entrainment modulates cognition.

We apply methods from machine learning and complex network analysis to understand the variation in chemical composition and subjective effects over >1.000 Cannabis strains.

We also use networks to model how different psychoactive compounds modify human language and underlying cognitive processes.

We develop models to predict response to neuropharmacological treatment using machine learning and functional magnetic resonance imaging data.