M2 Computational Neurosciences and Neuroengineering

No prerequisites

no prerequisites

Anatomy and Physiology of the mammalian auditory system, Perception of artificial and natural stimuli such as communication sounds in mammals.

Acoustic communication in songbirds, Sensorimotor learning in juvenile and adult birds.

Anatomy and physiology of the mammalian visual system: from retinal light capture to complex representation of visual scenes

Active vision: Non-canonical modulation of visual cortex

Interactions between somatosensory and motor systems: bases of neuroprosthetic devices

Multisensory processing study with large scale two-photon imaging

Olfaction in vertebrates and insects

This teaching unit aims at providing detailed knowledge on the anatomical organization and the functional properties of neurons in all sensory modalities. Students will learn the foundations of sensory physiology underlying stable perception despite the flow of fluctuating information coming from the peripheral receptors. Results from electrophysiological recordings and calcium imaging will be used to visualize and quantify the responses of individual neurons, or of large cell assemblies, to artificial and to natural stimuli. We will then determine how behaving animals use these stimuli to master new tasks and reach optimal behavioral performance. Multimodal processing will also be presented with large scale neuronal recordings, and examples of the role of corticofugal feedbacks will presented. A large part of the courses will be based on recent findings from leading labs in all sensory modalities.

Master and explain the way sensory systems process stimuli of the external world to guide the animal’s behavior.

Analyze papers and interpret findings in the different sensory modalities.

Conceive an experimental project using appropriate methods to study the perception of specific stimuli.

background in signal processing
programming skills

Regression and classification (Ridge Regression, Generalized Linear Model -including logistic regression, Support Vector Machines). Extensions to core methods.

Generalized Linear Model and Selection of Variables (Sparsity Constraints).

Aggregation methods (Boosting, Random Forrest)

Statistical tools for the analysis of large data.

Model of mixtures

Dimension reduction (Analysis in Principal Components).

K-means / Hierarchical clustering.

Neuroscience is a field where very large data are generated ("Big Data", several tens of GB per session in fMRI, EEG, high temporal and spatial resolution optical recordings, or multi-electrode electrophysiology). The analysis of these data requires very specific analytical methods, which are themselves the subject of advanced research. This course presents an overview of machine learning methods as well as examples of the application of the different approaches developed.

CM et TP, challenge

At the end of this course, students will be able to define, understand and choose a machine learning method and to implement it in line with  the specific problem.

no prerequisites

Issues and questions in neuroscience

 Development of the nervous system

Nervous system organization: from neuron to network, transmission of the nervous signal

Synaptic transmission

Synaptic plasticity, functional plasticity

This course will focus on the study of the brain, its functions and its adaptability (brain plasticity) to the environment. The latest knowledge on the molecular and physiological basis of synapse and neuron function will be presented. At a more macroscopic level, the anatomy and structure of the brain will be addressed. The neural processes responsible for processing sensory information and perception, as well as their short- and long-term regulation through interaction with the environment, will be described.

CM

At the end of this course, the student will be able to:

- Understand neural communication processes at the physiological level.

- Have a global vision of the complexity of the nervous system and the interactions between the different components of the brain.

- Be able to identify the neural processes responsible for processing sensory and motor information.

no prerequisites

A supervised scientific project will complete the students' training during the semester one. This project is focused on a scientific issue including a bibliographic and/or experimental and/or simulation works.

A scientifc project is proposed to student with supervision of a project leader

no prerequisites

The course is divided into lectures given by research experts. Each researcher will introduce some of the different approaches used in closed-loop neuroscience including experimental, computational, theoretical approaches. One (possibly two) practical classes will be organized in order to practice some algorithmic and analysis skills. Each teacher will provide their course slides after the class.

This UE addresses the recent interest in closed-loop experiments in neuroscience. Closed loops arise from the bidirectional interaction between a biological system and a technological device. In the case of brain-machine interfaces, for example, it is an interaction between a brain structure and a computer or robot. To this end, this UE presents various techniques for recording and influencing neural activity in real time. It discusses sensory neuroprostheses and brain-machine interface techniques based on interpreting user intentions from EEG signals or invasive recordings, as well as more recent techniques exploiting or studying brain plasticity. Finally, the UE presents neuro-robotics techniques, focusing in particular on neuroinspired learning algorithms (in which the robot is programmed based on knowledge of brain function) and on robotic prostheses or ortheses.

The course takes place mostly during one week in December (second or third week, to be confirmed).

On successful completion of this course, students should be able to:       

•    Understand the techniques used for artificially influencing neuronal activity, for therapeutic or experimental purposes;

•    Understand the different components of a brain-machine interface device and how to tailor those components to the specific aims;

•    Read the latest scientific publications in the field and critically evaluate the performance and limitations of the different approaches in relation to their goals;

•    Present a recent publication in an oral session and place it in the context of recent literature.

No prerequisites

Chapter 1: Introduction to the different scales of neural activity recording and stimulation, what are the main issues and challenges to be addressed? Chapter 2: Targeted neural stimulation/inhibition methods Chapter 3: Biophysical models of brain signals 

Chapter 4: Good practices for data acquisition and analysis, statistical tools for the analysis of large data sets 

Chapter 5: Non-invasive neural recording methods used in humans. 

Chapter 6: Electrophysiology 

Chapter 7: Optical recording of neuronal network dynamics

Chapter 8: Functional Ultrasound Imaging

The aim of this UE is to present methods for recording and stimulating neuronal activity as a means to access the organization and functioning of the brain. It will present methods to record or stimulate neural activity at different spatial and temporal scales. The principles and main applications of cellular (unit or multiple activity) and network (EEG, ECoG, LFP) electrophysiological approaches as well as invasive (calcium or voltage-sensitive imaging in conventional or multi-photonic microscopy) and non-invasive (fMRI) brain imaging approaches will be described. Data analysis methods specific to each modality will be detailed, in particular through practical sessions where actual data sets will be analyzed. 

We will see how these methods, applied to either human subjects or animal models make it possible to establish causal links between patterns of neural activity and cognitive functions.

13h CM, 12h TP

At the end of this course, students will be able to:

- Have a comprehensive overview of the technological arsenal available nowadays to neuroscientists. 

- Assess the relevance of using one method over another based on the scientific questions addressed.

- Identify the limitations of the different methods for a well-versed interpretation of the experimental results described in the literature of the field.

linear algebra tools
programming skills

Chapter 1: Mathematical models of neurons

This chapter presents well-known neuron models. It introduces conductance-based models through the famous Hodgkin-Huxley model and underlines its electronic analogy. It then presents simplified models, such as integrate & fire or FitzHugh-Nagumo models, as well as simple models of synapses and neuronal plasticity. Numerical simulation of these models is also introduced.

Chapter 2: Analysis of neuron models

This chapter presents mathematical tools to study neuronal behavior. It introduces the notion of phase diagram and bifurcation. These notions are first given for one-dimensional systems, and then for planar systems. The chapter establishes a link between these bifurcations and the qualitative behavior of the neuron. It also shows how to simplify a neuron model based on singular perturbation theory. It finally introduces the notion of “phase response curve”, which helps simplifying complex oscillatory behaviors to the mere time evolution of its phase, and shows how to use this technique in practice for the prediction of neuronal synchronization.

Chapter 3: Biophysical neuron model

This chapter presents a realistic neuron model based on dendritic modeling.

Chapter 4: Neuronal populations “Mean field model”

This chapter addresses the dynamics of a whole population of neurons or a cerebral structure. It presents simplified models of the activity of a population based on a mean field theory.

Chapter 5: Neuronal populations “Neural field model”

This chapter addresses the dynamics of a whole population of neurons or a cerebral structure. It presents simplified models of the activity of a population, such as the Wilson-Cowan model or neural fields. It shows how to predict the behavior of such models by stability or bifurcation analysis. It also provides tools to identify such models based on experimental data.

Chapter 6: Synapse modeling and plasticity

This chapter focuses on synaptic modeling and the exploration of synaptic plasticity.

Chapter 7: Neuromorphic computing

This chapter focuses on computing units and processors inspired by the processing of information by the brain.

This course constitutes an introduction to tools for the analysis of dynamical processes involved in brain functioning. Despite their huge complexity, brain functions are indeed based on elementary dynamics, some of which can be apprehended by a mathematical approach. Mastering these techniques is fundamental to progress in our understanding of brain functioning, to optimize instrumentation for brain activity measurements (brain imaging, electrophysiological recordings…), to improve brain machine interfaces, to build up neuro-inspired computational units, and to understand the mechanisms involved in some brain diseases and thus improve their treatment.

19h CM et 6h TD

By the end of this course, students will be able to:

• Understand neuroscience fundamentals, for possible interaction with professionals of the field (neurosurgeons, computational neuroscientists, experimenters).
• Model the activity of a neuron or a whole neuronal population.
• Predict their behavior both analytically and numerically.

- Dynamical systems in Neuroscience (Eugene M. Izhikevitch)

- Theoretical Neuroscience (Peter Dayan, LF Abbott)

The objectives of the module is to show students how to assess cognitive functions in humans and animal models. Students will visit imaging centers for animal and human research in order to learn how to measure brain activity using various cutting edge technics. They will also learn how to test animal and human brain functions and their neural and physiological correlates during practical trainings.

Learning outcomes (OAVs):

By the end of this unit, students will be able to: 

1: understand how to investigate, measure and analyse human and animal brain functions

2: understand the principle of brain imaging in humans and rodent models

3: understand how to study physiological correlates of brain functions

The course consists of practical assessments of cognitive functions in humans and rodents and of their neurophysiological correlates, and of visits of brain imaging centers dedicated to human and animals to learn the cutting edge current methods

This course explores the neural underpinnings of human cognitive development. It draws parallels between typical and atypical developmental trajectories and integrates insights from adult cognitive neuroscience to better understand developmental mechanisms. Students will be introduced to the main cognitive functions, core neuroscientific methods, and major developmental disorders.

By the end of this unit, students will be able to: 

• OAV 1: understand the main concepts of developmental cognitive neuroscience. 
• OAV 2: develop critic lecture of the experimental approach deployed to test typical and atypical developments of cognitive functions. 
• OAV 3: Extract and comment relevant information from the literature

This teaching unit includes 16 hours of lectures delivered by research experts, illustrating the foundations of developmental cognitive neuroscience, the typical development of cognitive functions and educational aspects in healthy and pathological conditions in children. In addition 4 hours of supervised sessions, during which students will work and discuss together with researchers to design some specific studies.

The objectives of the module is to show students how to assess cognitive functions in humans and animal models. Students will visit imaging centers for animal and human research in order to learn how to measure brain activity using various cutting edge technics. They will also learn how to test animal and human brain functions and their neural and physiological correlates during practical trainings.

Learning outcomes (OAVs):

By the end of this unit, students will be able to: 

1: understand how to investigate, measure and analyse human and animal brain functions

2: understand the principle of brain imaging in humans and rodent models

3: understand how to study physiological correlates of brain functions

The course consists of practical assessments of cognitive functions in humans and rodents and of their neurophysiological correlates, and of visits of brain imaging centers dedicated to human and animals to learn the cutting edge current methods

This course explores the neural underpinnings of human cognitive development. It draws parallels between typical and atypical developmental trajectories and integrates insights from adult cognitive neuroscience to better understand developmental mechanisms. Students will be introduced to the main cognitive functions, core neuroscientific methods, and major developmental disorders.

By the end of this unit, students will be able to: 

• OAV 1: understand the main concepts of developmental cognitive neuroscience. 
• OAV 2: develop critic lecture of the experimental approach deployed to test typical and atypical developments of cognitive functions. 
• OAV 3: Extract and comment relevant information from the literature

This teaching unit includes 16 hours of lectures delivered by research experts, illustrating the foundations of developmental cognitive neuroscience, the typical development of cognitive functions and educational aspects in healthy and pathological conditions in children. In addition 4 hours of supervised sessions, during which students will work and discuss together with researchers to design some specific studies.

Réservé aux étudiants de la mention HSB de CentraleSupelec

Acquérir des notions de traitement des images médicales et en particulier des images IRM