Computational neuroscience of sensory systems

Leader

Romain
BRETTE

Research center

Institut de la Vision
17 rue Moreau
75012 Paris
José-Alain Sahel

Institution

Inserm
CNRS
ED3C
Université Pierre et Marie Curie

Laboratory

UMRS968 UMR7210 UM80
Website
Labex LIFESENSES, Label "Institut Carnot", Label CTRS / RTRS "Fondation Voir & Entendre", Qualification de « Projet Structurant » par Medicen Paris Région, Labellisation par la Foundation Fighting blindness

Keywords

computational neuroscience
sensory systems
spike initiation
neural simulation
spike timing
 

Publications

Benichoux V, Rébillat M, Brette R.On the variation of interaural time differences with frequency.J Acoust Soc Am. 2016 Apr;139(4):1810. doi: 10.1121/1.4944638.

#Laudanski J, #Zheng Y, #Brette R (2014). A structural theory of pitch. eNeuro (in press).

#Rébillat M, #Benichoux V, Otani M, Keriven R, #Brette R (2014). Estimation of the low-frequency components of the head-related transfer functions of animals from photographs. JASA, 135, 2534 (2014).

#Fontaine B, Peña JL, #Brette R (2014). Spike-threshold adaptation predicted by membrane potential dynamics in vivo. PLoS Comp Biol, 10(4): e1003560.

#Stimberg M, #Goodman DFM, #Benichoux V, #Brette R (2014).Equation-oriented specification of neural models for simulations. Frontiers Neuroinf, doi: 10.3389/fninf.2014.00006.

#Goodman DFM, #Benichoux V, #Brette R (2013). Decoding neural responses to temporal cues for sound localization. eLife2:e01312.

#Brette R (2013). Sharpness of spike initiation in neurons explained by compartmentalization. PLoS Comp Biol, doi:10.1371/journal.pcbi.1003338.

Fields of research

Computational neurosciences / neural theory

Research Theme

Our goal is to understand the neural basis of perception, using theoretical and computational models of sensory systems. These models connect the physiological level (properties of neurons) with the behavioral level. Thus theories are tested with physiological experiments (in particular electrophysiology) and behavioral experiments (psychophysics). They are also tested from a computationalperspective, by evaluating the functional performance of models in complex perceptual tasks.


Our research is organized around three broad themes:

1) Neurons

We develop predictive neuron models, i.e., models that can predict the response of a neuron (action potentials) to a sensory stimulus (in vivo) or to an injected current (in vitro). We are interested in particular in neural excitability (adaptation and plasticity) and in the spatial aspect of spike initiation (initiation in the axon).

2) Perceptual systems

We are investigating the neural basis of perception, in particular the perception of space (visual and auditory), in complex ecological environments. We try to characterize the structure of ecological environments, and we develop neural models in which selective synchronization of spikes produced by neurons reflects the detection of a structure in the sensory flow. These models are then testedby their ability to perform complex tasks in ecological environments, and by experiments (in vivo electrophysiology and psychophysics).

3) Simulation technology

We develop an open source neural network simulator, Brian (http://briansimulator.org), which was designed to allow quick development of new models, with little constraint on the type of models. We are extending this technology to fast simulation on parallel platforms.


ENP Students

Maria TELENCZUK (née KRAMAREK) Sarah GOETHALS

Lab rotation

Stability of assemblies in neuronal networks

Team leader: 

BRETTE Romain

Dates: 

September 18, 2017 - June 29, 2018

Application deadline: 

June 29, 2018

Period

~ Sept-Dec 2017

~ Jan-March 2018

~ April-June 2018


Project

Long-term memory is thought to be mediated by synaptic plasticity. However, the stability of the neuronal assemblies emerging because of synaptic changes is still an open question. Indeed, cortical networks are made of sparsely interconnected excitatory and inhibitory neurons in a so-called balanced regime, with a spontaneous activity appearing to be problematic from a memory retention point of view. In this rotation, the student will become familiar with one the most used neuronal network simulator, Brian, and use it to build neuronal networks and study the stability of neuronal assemblies in recurrent networks under several plasticity rules. The project will give a broad overview of canonical concepts in computational neuroscience by using integrate-and-fire neurons and modeling plasticity rules.

Contact

Institut de la Vision - 17, rue Moreau 75012 Paris - +33 1 53 46 25 36 - pierre.yger@inserm.fr

Supervisor: 

YGER Pierre

Computational neuroscience of unicellular organisms

Team leader: 

BRETTE Romain

Dates: 

September 18, 2017 - June 29, 2018

Application deadline: 

June 29, 2018

Period

~ Sept-Dec 2017

~ Jan-March 2018

~ April-June 2018


Project

Paramecium, as a number of other unicellular organisms, swims by alternating straight runs and rotations, and displays various taxis behaviors (chemotaxis, gravotaxis). Stimuli are transduced by ionic channels, and rotations are triggered by action potentials. The goal of this rotation is to study the computational neuroscience of this one-“neuron” system, with electrophysiological and behavioral experiments.

Contact

Institut de la Vision - 17, rue Moreau 75012 Paris - +33 1 53 46 25 36 - romain.brette@inserm.fr

Supervisor: 

BRETTE Romain

Theory of spike initiation

Team leader: 

BRETTE Romain

Dates: 

September 18, 2017 - June 29, 2018

Application deadline: 

June 29, 2018

Period

~ Sept-Dec 2017

~ Jan-March 2018

~ April-June 2018

Project

We are investigating several themes in the initiation of action potentials, including: 1) the relation between neuron geometry and excitability; 2) how a cell learns to spike (ie to properly coordinate the different channels); 3) normative constraints on the organization of the spike initiation system (eg energy). Rotation students are welcome on any of these themes, to work on theoretical problems, analysis of experimental data, and patch clamp experiments.

Contact

Institut de la Vision - 17, rue Moreau 75012 Paris - +33 1 53 46 25 36 - romain.brette@inserm.fr

Supervisor: 

BRETTE Romain

Spike sorting for large Multi-Electrodes recordings

Team leader: 

BRETTE Romain

Dates: 

September 18, 2017 - June 29, 2018

Application deadline: 

June 29, 2018

Period

~ Sept-Dec 2017

~ Jan-March 2018

~ April-June 2018

Project

To record the activity of a large population of neurons, electrophysiologists use large and dense multi-electrodes arrays (MEA) to capture the electrical signals originating from all the action potentials emitted by the neurons surrounding the electrodes. However, this source separation problem consisting in determining which neuron emitted an action potential when is very challenging, and is a crucial step toward deciphering the neuronal activities. In this rotation, the student will become familiar with data coming from MEA recordings in the retina to approach this so-called problem of  spike sorting. The project will give a broad overview of state-of-the-art signal processing techniques, data acquisition, and be a good introduction to understanding how information is represented by a population of neurons.

Contact

Institut de la Vision - 17, rue Moreau 75012 Paris - +33 1 53 46 25 36 - pierre.yger@inserm.fr

Supervisor: 

YGER Pierre