Online Lecture by Jennifer Goldman

Bridging scales from single neurons to whole brain dynamics for a formal understanding of brain states

This work aims at connecting empirical knowledge across scales from microscopic (nanometers to micrometers—molecules to whole neurons) to macroscopic brain activity (centimeters to meters—brain areas to individual subjects’ brains) with theoretical tools (Goldman et al. 2019). Not unlike how microscopic interactions between molecules underlie structures formed in macroscopic states of matter, using statistical physics, the dynamics of microscopic neural phenomena can be linked to macroscopic brain dynamics through mesoscopic scales. While statistical physics has been utile to describe patterns of pair-wise and population-wide correlations at the microscopic scale in brain research (Nghiem, Telenczuk, et al. 2018; Schneidman et al. 2006; Zanoci, Dehghani, and Tegmark 2019), formal mesoscopic methods have yet been used to bridge scales microscopic to macroscopic scales. Here, methods for the macroscopic quantification of brain states, inspired by statistical physics, are introduced and used to classify global brain dynamics in states spanning deep sleep to resting and active wakefulness in human MEG and EEG data. Multi-scale simulations were then built to bridge single neuron properties to whole brain activity through mean-field models (El Boustani and Destexhe, 2009; di Volo et al. 2019) connected using human tractography data and implemented in The Virtual Brain (Sanz-Leon et al. 2015). The results indicate that macroscopically observed high synchrony, low complexity brain signals recorded during unconscious states may be accounted for by an increased coupling in the system’s components, behaving more like a solid (Peyrache et al. 2012; Le Van Quyen et al. 2016; Olcese et al. 2016; Nghiem, Lina, et al. 2018). In contrast, conscious brain states may be described as less ordered, higher complexity (Sitt et al. 2014; Engemann et al. 2018; Nghiem, Lina, et al. 2018), more liquid-like.