| TITLE |
| Heterogeneous and global itinerancy in brain activity illuminated by Hodge decomposition |
| OBJECT |
| The aim of the research is to characterize the dynamic information itinerancy in the brain by using the Hodge decomposition. Here we utilize the discrete graph version of the Hodge decomposition published in 2011. This study complements conventional, homogeneous and local methods such as tree approximations in complex network theory with global and heterogeneous detection of recurrent flows and layers in neural circuits.
|
| |
NAME |
AFFILIATION |
SPECIALTY |
ROLE |
| Leader |
MIURA, Keiji |
Tohoku University |
Collaborative Mathematics |
Research Director |
| TITLE |
| Proposal of a New Brain Model based on Internet Architecture and Communication |
| OBJECT |
| In this project, we focus on two features, namely, the default mode (activity patterns underlying the cerebral cortex at rest) and episodic memory of the brain. We model these two features of the brain based on the algorithms and infrastructures developed recently to Web search companies such as Google to support their services. Internet is now often used as a social network these days and is flooded with a large amount of episodic memory. Recent study of brain science shows that the default mode in the brain is involved in the recall of episodic memory. We expect our model to reveal modes and memory structures in the brain necessary for communication. In addition, we verify predictions from the model by using experiments on "perceptual crossing" and clarify the functions the default mode and the episodic memory in communication.
|
| |
NAME |
AFFILIATION |
SPECIALTY |
ROLE |
| Leader |
IKEGAMI, Takashi |
University of Tokyo |
Complex Systems |
Principal Investigator |
| Collaborator |
OKA, Mizuki |
University of Tokyo |
Computer Scientist |
Data Analysis and Experiments |
| Collaborator |
FRIESE, Tom |
University of Tokyo |
Cognitive Scientist |
Perceptual Crossing Experiments |
| TITLE |
| Theoretical study toward understanding hierarchy and
instability in decision-making dynamics |
| OBJECT |
There exist hierarchical structures behind decision-making process. A large network of heterogeneous elements generates decisions in each person. The network is considered to have a module structure and can effectively be regarded as a network of subnetworks. Such networks are further interacted via, e.g., conversation, finally leading to a decision.
Decision making processes is considered to posses some kind of instability. In this study, we deal with network systems with instability to construct a fundamental theory for understanding complex dynamics in large, heterogeneous networks.
|
| |
NAME |
AFFILIATION |
SPECIALTY |
ROLE |
| Leader |
KORI, Hiroshi |
Ochanomizu University |
Nonlinear Science |
Modeling and numerical simulations |
| Collaborator |
TOENJES, Ralf |
Potsdam University |
Nonlinear Science |
Analysis of chaotic dynamics |
| Collaborator |
KOBAYASHI, Yasuaki |
Hokkaido University |
Nonlinear Science |
Analysis of noisy oscillators |
| TITLE |
| Dynamics of network plasticity during communication between heterogeneous neural cell assemblies |
| OBJECT |
| The adaptivity of neural network is based on synaptic plasticity, through which the neuronal activity induces changes in the network connections. On the other hand, the underlying synaptic network significantly affects the neuronal activity. The purpose of this research is to understand this interplay between neurons and synapses during communication between heterogeneous neural cell assemblies. We construct a theoretical model of oscillatory neurons coupled with activity-dependent synaptic connections and analyze the effects of communication on network plasticity.
|
| |
NAME |
AFFILIATION |
SPECIALTY |
ROLE |
| Leader |
AOKI, Takaaki |
Kagawa University |
Nonlinear Physics, Computational Neuroscience |
General planning and theoretical analysis |
| TITLE |
| Clarification of high-functioning neural information processing by heterogeneous frequency oscillators |
| OBJECT |
| Although we have long been familiar with the advanced processing capabilities of the human brain, complicated issues remain. Heterogeneous frequencies in the EEG band, known as θ waves and α waves have been recorded within the brain, and these waves are understood to influence each other in processing information. Further, in recent years, it has come to be understood that unicellular organisms such as the amoeba or paramecium also experience periodic heterogeneous oscillation phenomena and that these primitive life forms exhibit intelligence by making use of resonance phenomena. By using a mathematical model, we have studied the emergent intelligence inherent in these heterogeneous rhythms. This research has adopted a two-pronged bottom-up (intelligence from primitive intelligence) and top-down (mathematical modeling and analysis of measurement results regarding human-brain activity when carrying out tasks) method to clarify the processing capabilities of the human brain.
|
| |
NAME |
AFFILIATION |
SPECIALTY |
ROLE |
| Leader |
TERO, Atsushi |
Kyushu University |
Applied Mathematics |
Mathematical modeling and mathematical analysis |
| Collaborator |
AKIYAMA, Masakazu |
Kyushu University |
Applied Mathematics |
Simulation and data analysis |
| TITLE |
| New developments of learning theory that enables organizing, understanding, and generating various dynamical systems |
| OBJECT |
| The purpose of this work is to establish the learning theory of multiple dynamical systems. Thus it enables (i) learning various dynamical systems from a set of observed time series data, (ii) estimating the latent rule and the latent variables which govern the system class, and (iii) generating new, as-yet-unseen dynamical systems and time series. In this project, the following themes are focused on especially. (a) Self-organization of self-understanding through mutual interactions, (b) Latent state estimation of brains, (c) Neuron-based representation of this theory, and (d) Further development of the theoretical insight of this theory.
|
| |
NAME |
AFFILIATION |
SPECIALTY |
ROLE |
| Leader |
FURUKAKWA, Tetsuo |
Kyushu Institute of Technology |
Neurocomputing |
Establishing the theory, Simulations, and Modeling |
| TITLE |
| Neuronal mechanism of long-range EEG synchronization based on geometrically structured neuronal projections |
| OBJECT |
| Long-range EEG synchronization is an important phenomenon for understanding information transfer between cortical regions. However, its relationship with ensemble neuronal activities, that are key elements for information processing in the brain, is still unclear. In this study we hypothesize a geometrically structured neuronal projection between cortical regions and elucidate the relationship between neuronal activities and EEG by using computer simulations. Through constructing a whole brain network model, the principle organizing cortical areas will be discussed.
|
| |
NAME |
AFFILIATION |
SPECIALTY |
ROLE |
| Leader |
SATO, Naoyuki |
Future University Hakodate |
Computational Neuroscience |
Computational modeling of EEG |