The home page of the FACETS project can be found at http://www.kip.uni-heidelberg.de/facets or http://www.facets-project.org. The home page of FACETS-ITN is at http://facets.kip.uni-heidelberg.de/ITN/index.html.
The goal of the FACETS project is to create a theoretical and
experimental foundation for the realisation of novel computing
paradigms which exploit the concepts experimentally observed in
biological nervous systems. The research will be carried out by an
interdisciplinary consortium, involving 16 groups of neuroscientists,
computer scientists and physicists. The institutions involved
represent a major fraction of the European groups working in the
relevant fields. The three major lines of research will be: (a)
experimental characterisation of cortical cells and networks in-vivo
and in-vitro; (b) study of theoretical and computer based models of
cells and networks; (c) design, construction and operation of VLSI
circuits emulating the biological example. Each of the 3 lines
involves studies on the level of individual computing elements
(neurons) and on the network level. The continuous interaction and
scientific exchange between biological experiments, computer
modelling and hardware emulations within the project provides a
unique research infrastructure that will in turn provide an improved
insight into the computing principles of the brain. This insight may
potentially contribute to an improved understanding of mental
disorders in the human brain and help to develop remedies.
The home page of the BrainScaleS project can be found at http://brainscales.kip.uni-heidelberg.de/.
The BrainScaleS project aims at understanding function and interaction of multiple spatial and temporal scales in brain information processing. The fundamentally new approach of BrainScaleS lies in the in-vivo biological experimentation and computational analysis. Spatial scales range from individual neurons over larger neuron populations to entire functional brain areas. Temporal scales range from milliseconds relevant for event based plasticity mechanisms to hours or days relevant for learning and development. In the project generic theoretical principles will be extracted to enable an artificial synthesis of cortical-like cognitive skills. Both, numerical simulations on petaflop supercomputers and a fundamentally different non-von Neumann hardware architecture will be employed for this purpose.
Neurobiological data from the early perceptual visual and somatosensory systems will be combined with data from specifically targeted higher cortical areas. Functional databases as well as novel project-specific experimental tools and protocols will be developed and used. New theoretical concepts and methods will be developed for understanding the computational role of the complex multi-scale dynamics of neural systems in-vivo. Innovative in-vivo experiments will be carried out to guide this analytical understanding.
Multiscale architectures will be synthesized into a non-von Neumann computing device realised in custom designed electronic hardware. The proposed Hybrid Multiscale Computing Facility (HMF) combines microscopic neuromorphic physical model circuits with numerically calculated mesoscopic and macroscopic functional units and a virtual environment providing sensory, decision-making and motor interfaces. The project also plans to employ petaflop supercomputing to obtain new insights into the specific properties of the different hardware architectures. A set of demonstration experiments will link multiscale analysis of biological systems with functionally and architecturally equivalent synthetic systems and offer the possibility for quantitative statements on the validity of theories bridging multiple scales. The demonstration experiments will also explore non-von Neumann computing outside the realm of brain-science.
The BrainScaleS consortium has several PhD and post-doc
fellowships available and is currently looking for filling these
positions. Please contact the different laboratories or see the BrainScaleS webpage.
Summary of the project:
One of the characteristics of neural activity in neocortical networks is that there is a considerable level of self-sustained ongoing activity, which exhibits highly complex but structured spatiotemporal patterns of action potentials and whose irregularity in time is often interpreted as "noise". In view of the recurrent nature of the network, where most links between neural units are achieved through distributed reverberating loops, it is therefore impossible to apply the classic paradigm of distinguishing the "signal" from the "noise".
In this project, we would like to depart from this paradigm and rather consider that information is potentially present in ongoing activity (for instance as internally stored memories) and that external inputs, carrying stimulus-driven information, are interacting non-linearly with the ongoing activity. We aim at characterizing network states using diverse methods applied at different scales of spatial integration, from the microscopic (conductance and single-neuron level), up to large populations of neurons measured mesoscopically (multiple recordings and voltage sensitive dye imaging). We will provide a characterization of the correlation state of network activity, as seen through the measurement techniques associated to each scale:
I.At the single-neuron level, intracellular recordings will be used to resolve subthreshold membrane potential fluctuations, of synaptic origin, and analyze a dynamic multiscale "image" of the activity of the effective afferent network in which the cell is embedded at any point in time.
II.At the level of populations, the recording of many neurons simultaneously will be confronted with theories inspired from Ising models and provide a characterization of the network state through pairwise correlations.
III.At more macroscopic levels, we will use local field potential recordings using array of electrodes, and voltage-sensitive dye imaging to record simultaneously larger assemblies and extract possible relations between correlation patterns and the context of functional cortical maps.
By using theoretical approaches adapted to such scales, such as electrodynamics or mean-field models, we also hope to provide characterizations of network states during ongoing activity, and during visual inputs of different dimensionalities. In all cases, the in vivo experiments and modeling will be done in parallel with in vitro experiments to determine key aspects and correlation measures in controlled conditions, as well as to validate or test some of the assumptions of the models. These different studies should allow us to study the interdependency between different levels of neural-based processing in neocortical networks, and address experimentally the concepts of âemergenceâ (micro to macro) and âimmergenceâ (macro to micro) characteristic of complex dynamic systems
By this interdisciplinary approach, we hope to provide decisive
data, tools and concepts on the different network states involved
in visual processing, with possible future applications as diverse as
artificial vision, "mind reading" in brain imaging, brain and
machine interface in the field of Neuroprosthetics and Medicine,
and life-inspired computing architectures in the field of
Information and Technology.
The Blue Brain Project will consist in gathering knowledge from different fields such as neuroanatomy, neurophysiology and computational neuroscience. The goal is to investigate cortical computations using an accurate software replica of neocortical microcircuits ("the Blue Column"). This sophisticate model will be run on a 8K-processor Blue Gene supercomputer build by IBM. The Blue Column will be composed of 104 morphologically complex neurons, which will be reconstructed from in vitro experiments and matched to models in order to capture their main electrical properties. The neurons will be interconnected in a 3-dimensional (3D) space with 107 -108 dynamic synapses, directly derived from morphological measurements.
The home page of the MANDy project can be found at href=http://www.proba.jussieu.fr/pageperso/thieullen/MANDy/accueil.html.
The activated cerebral cortex displays "high-conductance states" characterized intracellularly by intense subthreshold fluctuations, which are due to the high level of activity in the local surrounding network. Present intracellular methods to characterize this activity are limited in resolution due to the bias introduced by recording electrodes. In the present project, we plan to address these limitations by proposing a new recording paradigm based on a computer-contolled feedback with the cell. Developing and implementing this paradigm will require a tight association between mathematics, computer science, computational neuroscience and intracellular electrophysiology (in vivo and in vitro). We aim at both the conception of novel methodologies, their testing in real neurons (essentially in vitro), as well as applying these methods to intracellular recordings in primary visual cortex in vivo.
The project combines different expertises, such as mathematics,
computer science, computational neuroscience and intracellular
electrophysiology (in vitro and in vivo), to yield accurate and
reliable methods to properly characterize high-conductance states in
neurons. We plan to address several of the caveats of present
recording techniques, namely 1) the impossibility to perform reliable
high-resolution dynamic-clamp with sharp electrodes, which is the
intracellular technique mostly used in vivo; 2) the unreliability and
low time resolution of single-electrode voltage-clamp recordings in
vivo; 3) the impossibility of extracting single-trial conductances
from Vm activity in vivo. We propose to address these caveats with
the following goals: 1. Obtain high-resolution recordings applicable
to any type of electrode (sharp and patch), any type of protocol
(current-clamp, voltage-clamp, dynamic-clamp) and different
preparations (in vivo, in vitro, dendritic patch recordings). 2.
Obtain methods to reliably extract single-trial conductances from Vm
activity, as well as to "probe" the intrinsic conductances in
cortical neurons. These methods will be applied to intracellular
recordings during visual responses in cat V1 in vivo. 3. Obtain
methods to extract correlations from Vm activity and apply these
methods to intracellular recordings in vivo to measure changes in
correlation in afferent activity. 4. Obtain methods to estimate
spike-triggered averages from Vm activity and obtain estimates of the
optimal patterns of conductances that trigger spikes in vivo. These
results will be integrated into computational models to test
mechanisms for selectivity. These methods will be based on a
real-time feedback between a computer and the recorded neuron. This
real-time feedback will be used not only to improve existing
techniques, but also to extract essential information to better
understand spike selectivity of cortical neurons in vivo.
In this HFSP project, our plan was to evaluate how the rhythmic activity of the brain during slow wave sleep influences synaptic transmission and plasticity at a cellular and network level in the cerebral cortex. We have addressed this general question by combining three approaches: (i) Extra-, intracellular and optical recordings in mouse primary visual and somatosensory cortex in vivo; (ii) Extra-, intracellular and optical recordings in ferret primary visual and somatosensory cortex in vitro; (iii) Computational models of morphologicallyreconstructed cortical neurons and network simulations. Optical recordings were done using voltage sensitive dye fluorescence as well as calcium indicators and a fast CCD camera. Our working hypothesis was that periods of electroencephalogram (EEG) activation (wakefulness, REM) provide ideal conditions for "priming" cortical synapses and that these synapses are later subject to long-term changes during slow-wave sleep. We have tested this hypothesis by using various paradigms in vivo, in vitro, and in models. The project provided data essential to the long-term goal of establishing firm evidence for a role of slow wave sleep in memory consolidation.
This research project led to several publications (see Publication list). In particular, we published a review article in Science.
This research project led to several publications (see Publication list).
This project (entitled "Impact of synaptic bombardment on neocortical neurons") proposed to combine computational models and intracellular recordings of neocortical pyramidal cells in vivo to quantify the total amount of sustained synaptic activity and to study how it affects the integrative properties of pyramidal cells. The long-term objective of this project was to build better representations of neuronal networks of the neocortex during intense network activity similar to the waking state. This is of great benefit for investigating information processing paradigms that involve cortical networks. Experiments and models estimated the amount of synaptic conductances tonically activated in soma and dendrites, and how this could potentially affect the basic integrative and response properties of pyramidal cells.
This research project led to numerous publications and review articles (see Publication list).
This research project led to numerous publications and one monograph (see Publication list).
Unité de Neurosciences, Information & Complexité
(UNIC)
CNRS
UPR-3293, Bat 33,
1 Avenue de la Terrasse,
91198 Gif-sur-Yvette, France.
Tel: +33-1-69-82-34-35
Fax: +33-1-69-82-34-27