Welcome to my web-page!


I am a CNRS Research scientist at the INS, Aix-Marseille University (Theoretical Neuroscience Group), with an interdisciplinary background in computational and systems neuroscience, physics of complex systems/networks and theoretical computer science.


Present address Institute for Systems Neuroscience, INSERM UMR 1106,    Aix-Marseille University - 27, bd Jean Moulin - 13005 Marseille, France


Phone +33 6 35 95 33 94

E-mail

Dynamics of Functional Connectivity


Structural connectivity constrains brain dynamics but does not fully determine it. Even very simple structural motifs can give rise to a multiplicity of dynamical states, each one of them associated to different patterns of functional interaction between system's components. The sampling of these states thus gives rise to a characteristic switching of functional connectivity over time. We characterize Functional Connectivity Dynamics (FCD) from multichannel recordings in rodents and primates to human brain imaging. We develop FCD descriptors which may lead to new biomarkers, e.g. in aging. We also reproduce FCD in computational models informed by realistic connectivity, linking it to the hidden brain “dynome”.  Ongoing projects include:


  1. BulletWhole brain mean field modeling of resting state Functional Connectivity Dynamics [a poster]


  1. BulletFunctional Connectivity Dynamics through human aging [preprint]

Oscillations and information


Oscillations of neural activity are ubiquitous and yet their possible

relevance for information processing has not been proved. We use computational and theoretical neuroscience approaches to investigate and probe the potential of oscillatory activity and its dynamic self-organization for the state-dependent routing of information. We use information theory and machine learning to quantitatively measure in actual data-sets the involvement of oscillations in information storage, transfer and integration, primitive components of cognitive functions such as attention and working memory. We also model the emergence of dynamic coherence networks in multi-scale brain circuits with realistic anatomy.


  1. BulletInformation routing with stochastic metastable oscillatory bursts “at the edge of synchrony” [a poster]

  2. BulletComplex oscillatory dynamics of multi-layer cortex [a preprint]

Algorithms for (neural) data analysis


We develop computational methods for different types of neural data analyses (or other complex datasets as well, like

peer-reviewing social networks, revealing gender bias).

We are keen on experimenting with new promising tools (e.g. topological data analysis), to benchmark their potential for neuroscience.

Applications have included so far:


  1. BulletInference of connectivity from calcium imaging of neuronal cultures: a crowdsourcing challenge [a poster]


  1. BulletModel-free characterization of irregular tuned responses and of their modulation by attention [an article]


  1. BulletSoft clustering approaches for the unsupervised classification of neuronal types (and even primate vocalizations!)

Email Me

Demian Battaglia’s Home Page