Mes compétences :
Adaptabilité
Algorithmique
Analyse d'image
Anglais
Communications
EDP
imagerie
Imagerie médicale
Matlab
Projet collaboratif
Python
Veille
Veille technologique
Entreprises
CNRS / Toulouse Institute of Mathematics
- Fellow Research Engineer
2012 - maintenant
Institute of Biomedical Engineering, University of Oxford
- Research assistant
2011 - 2012Development of avanced algorithms for the automatic comparison (registration - motion tracking) of 3D medical images.
Imperial College London
- Research assistant
London2009 - 2010Development of avanced algorithms for the automatic comparison (registration - motion tracking) of 3D medical images. Until now the algorithms I developed have been mostly applied to the quantification of the cortical variability in MR mages. They have also been successfully applied to the motion tracking of coronary arteries in CT image series and the measure of the hippocampus atrophy in patients having Alzheimer's disease. My work both takes place at the Biomedical Image Analysis (BioMedIA) Group (http://biomedic.doc.ic.ac.uk/) and the Institute for Mathematical Sciences in Imperial College London.
2007 - 2008Development of strategies for the analysis the brain activity in functional MRI. My work took place in the Computer-Assisted Neuroimaging Lab (LNAO) of Neurospin - CEA Saclay (Paris suburbs). Technically speaking I participated to the development extension of an existing software (https://code.launchpad.net/~nipy-developers/nipy/pyhrf-trunk) in which the detection of the brain activity is simultaneously addressed with the estimation of impulse responses to the neuronal stimulations. My work focused in particular in the unsupervised spatial regularization of the analyses. As a result, the observed signal is analyzed with an unprecedented spatial accuracy.
Toulouse2003 - 2007Biomedical image analysis / Porous media
Analysis of the microvascular network in the gray matter of the brain. Development of algorithms dedicated to large anatomical 3D images (up to 5GB) treatment and analysis. The images were acquired using tomography at a resolution of 1.4 micron per voxel.