Bridging integrated information theory and the free-energy principle in living neuronal networks
The University of Tokyo（东京大学）；International Research Center for Neurointelligence (WPI-IRCN), The University of Tokyo Institutes for Advanced Study (UTIAS)（东京大学高等研究院神经智能国际研究中心（WPI-IRCN））
Teruki Mayama, Dai Akita, Sota Shimizu, Yuki Takano, Hirokazu Takahashi
Integrated Information Theory (IIT) links consciousness to integrated causal structure, whereas the Free-Energy Principle (FEP) explains self-organization through variational free-energy minimization. Their relationship in living neural systems remains unclear. We analyzed dissociated neuronal cultures learning to infer hidden signal sources. Across repeated stimulation, variational free energy decreased, while inference accuracy and Bayesian surprise, defined as the divergence between prior and posterior beliefs, increased. An IIT-inspired integrated-information proxy and main-complex size followed a non-monotonic, hill-shaped trajectory. The proxy correlated most strongly with Bayesian surprise and more weakly with accuracy and variational free energy. An Ising-model analysis indicated that Bayesian surprise and integrated information can be jointly amplified near shared positive critical modes and suggested how early connectivity development followed by response stabilization could generate the observed trajectory. These results link belief updating to integrated information in living neuronal networks and provide an empirical point of contact between IIT and the FEP.
Neuronal networks in dissociated culture combined with cell engineering technology offer a pivotal platform to constructively explore the relationship between structure and function in living neuronal networks. Here, we fabricated defined neuronal networks possessing a modular architecture on high-density microelectrode arrays (HD-MEAs), a state-of-the-art electrophysiological tool for recording neural activity with high spatial and temporal resolutions. We first established a surface coating protocol using a cell-permissive hydrogel to stably attach polydimethylsiloxane microfluidic film on t…
Neuronal networks provide living organisms with the ability to process information. They are also characterized by abundant recurrent connections, which give rise to strong feedback that dictates their dynamics and endows them with fading (short-term) memory. The role of recurrence in long-term memory, on the other hand, is still unclear. Here we use the neuronal network of the roundworm C. elegans to show that recurrent architectures in living organisms can exhibit long-term memory without relying on specific hard-wired modules. A genetic algorithm reveals that the experimentally observed dyn…
Identify the directions of signal flows in neural networks is one of the most important stages for understanding the intricate information dynamics of a living brain. Using a dataset of 213 projection neurons distributed in different regions of Drosophila brain, we develop a powerful machine learning algorithm: node-based polarity identifier of neurons (NPIN). The proposed model is trained by nodal information only and includes both Soma Features (which contain spatial information from a given node to a soma) and Local Features (which contain morphological information of a given node). After i…
Identifying the brain's neuronal cluster size to be presented as nodes in a network computation is critical to both neuroscience and artificial intelligence, as these define the cognitive blocks required for building intelligent computation. Experiments support many forms and sizes of neural clustering, while neural mass models (NMM) assume scale-invariant functionality. Here, we use computational simulations with brain-derived fMRI network to show that not only brain network stays structurally self-similar continuously across scales, but also neuron-like signal integration functionality is pr…
We investigate the dynamics of two models of biological networks with purely suppressive interactions between the units; species interacting via niche competition and neurons via inhibitory synaptic coupling. In both of these cases, power-law scaling of the density of states with probability arises without any fine-tuning of the model parameters. These results argue against the increasingly popular notion that non-equilibrium living systems operate at special critical points, driven by there by evolution so as to enable adaptive processing of input data.
Improvements in technique in conjunction with an evolution of the theoretical and conceptual approach to neuronal networks provide a new perspective on living neurons in culture. Organization and connectivity are being measured quantitatively along with other physical quantities such as information, and are being related to function. In this review we first discuss some of these advances, which enable elucidation of structural aspects. We then discuss two recent experimental models that yield some conceptual simplicity. A one-dimensional network enables precise quantitative comparison to analy…
We apply the approach of the Google matrix, used in computer science and World Wide Web, to description of properties of neuronal networks. The Google matrix ${\bf G}$ is constructed on the basis of neuronal network of a brain model discussed in PNAS {\bf 105}, 3593 (2008). We show that the spectrum of eigenvalues of ${\bf G}$ has a gapless structure with long living relaxation modes. The PageRank of the network becomes delocalized for certain values of the Google damping factor $α$. The properties of other eigenstates are also analyzed. We discuss further parallels and similarities between th…
The dynamic range measures the capacity of a system to discriminate the intensity of an external stimulus. Such an ability is fundamental for living beings to survive: to leverage resources and to avoid danger. Consequently, the larger is the dynamic range, the greater is the probability of survival. We investigate how the integration of different input signals affects the dynamic range, and in general the collective behavior of a network of excitable units. By means of numerical simulations and a mean-field approach, we explore the nonequilibrium phase transition in the presence of integratio…
In vitro primary cultures of dissociated invertebrate neurons from locust ganglia are used to experimentally investigate the morphological evolution of assemblies of living neurons, as they self-organize from collections of separated cells into elaborated, clustered, networks. At all the different stages of the culture's development, identification of neurons' and neurites' location by means of a dedicated software allows to ultimately extract an adjacency matrix from each image of the culture. In turn, a systematic statistical analysis of a group of topological observables grants us the possi…
The aim of this paper is to leverage the free-energy principle and its corollary process theory, active inference, to develop a generic, generalizable model of the representational capacities of living creatures; that is, a theory of phenotypic representation. Given their ubiquity, we are concerned with distributed forms of representation (e.g., population codes), whereby patterns of ensemble activity in living tissue come to represent the causes of sensory input or data. The active inference framework rests on the Markov blanket formalism, which allows us to partition systems of interest, suc…
The spontaneous activity pattern of cortical neurons in dissociated culture is characterized by burst firing that is highly synchronized among a wide population of cells. The degree of synchrony, however, is excessively higher than that in cortical tissues. Here, we employed polydimethylsiloxane (PDMS) elastomers to establish a novel system for culturing neurons on a scaffold with an elastic modulus resembling brain tissue, and investigated the effect of the scaffold's elasticity on network activity patterns in cultured rat cortical neurons. Using whole-cell patch clamp to assess the scaffold …
An inverse procedure is proposed and tested which aims at recovering the a priori unknown functional and structural information from global signals of living brains activity. To this end we consider a Leaky-Integrate and Fire (LIF) model with short term plasticity neurons, coupled via a directed network. Neurons are assigned a specific current value, which is heterogenous across the sample, and sets the firing regime in which the neuron is operating in. The aim of the method is to recover the distribution of incoming network degrees, as well as the distribution of the assigned currents, from g…
