Research and current projects
The lab is following several directions of research
Modeling and mathematical asymptotics, Simulations to study interactions from subcellular, cellular level to Brain areas
- 1 Network modeling approaches, coarse-grained n-dimensional Ornstein-Ulhenbeck processes, revealing key parameters for brain prediction.
- 2 Modeling and asymptotic expansion of solution of partial differential equation to model current-voltage (I-V) conversion in nanodomains using electrodiffusion theory.
- 3-We develop a theory of rare events in cell biology : theory of extreme statistics and application to cellular microdomains, trafficking in the endoplasmic reticulum.
- 4 We develop model of phase separated domains (PSDs) that are ubiquitous in cell biology, representing nanoregions of high molecular concentration. PSDs appear at diverse cellular domains, such as neuronal synapses but also in eukaryotic cell nucleus, limiting the access of transcription factors and thus preventing gene expression.
- 5 We have interested in organization and dynamics of chromatin in the cell nucleus. We specialized our recent efforts in two ensembles of data : stochastic single particle trajectories (SPTs) and the Hi-C distribution of contact frequencies across cell populations : We are developing algorithms to recovered the geometrical organization of chromatin from these two ensemble of data.
- 6 We develop novel statistical methods and clustering methods assisted by Machine-Learning to study cell proliferation (Microglia) during foetal development.
Developing Algorithms based on imaging data and Machine-Learning
- 1 To reconstruct Glial and Neuronal network.
- 2 To segment EEG in real-time
- 3 Modeling EEG using high dimensional stochatic processes coupled to wavelet decomposition
- 4 Algorithms to simulate fast diffusion events that we applied to model neuronal navigation in the brain and develop a theory of combinatorix to predict brain wiring.
- 5 We develop statistical approaches and Machine-Learning algo. to classify patient state during anesthesia. This approach allows to generate a state-chart representation of the brain and to quantify the transitions between brain profound states using control theory and viterbi algorithm.
Predictive medecine and modeling-statistics of electrophysiology time series such as EEG, Ca GCAMP, LFP,etc...
- 1-We develop model and algorithm to detect tremor, predict coma output and control general anesthesia.
