Ph.D. or Master Project Proposals
Not for the moment
Ph.D.
October 2010 - October 2013
Active contour and statistical learning for automatic structure segmentation:
Application to Computer Aided Diagnosis in MRI
(Contours actifs et apprentissage pour la détection automatique de
structures dans les images : application à l’aide au diagnostic en IRM)
Funding: Doctoral allocation from University of Cergy-Pontoise
Complete Description:
Download
Ph.D. student: Leila Meziou
Supervising: Aymeric Histace (aymeric.histace_at_u-cergy.fr), Frédéric Precioso (frederic.precioso_at_ensea.fr)
Master Projects
Avril 2012 - September 2012 (Master STIC from University of Cergy-Pontoise)
Hardware implementation of active contour based image segmentation algorithms for embedded analysis of wireless videocapsule images
MSc student: Juan Silva da Quintero
Supervising: Aymeric Histace (aymeric.histace_at_u-cergy.fr), Olivier Romain (olivier.romain_at_u-cergy.fr)
December 2010 - September 2011 (Master STIC from University of Cergy-Pontoise)
3D Confocal Microscopic images segmnetation for analysis of cell cytoskeleton changes due to radiotherapy treatment
MSc student: Mickael Garnier
Supervising: Aymeric Histace (aymeric.histace_at_u-cergy.fr)
December 2009 - March 2010 (Master STIC from University of Cergy-Pontoise)
Boosting approach for medical image segmentation: application to tagged cardiac MRI
Complete Description: Aim of this project is to prospect adaptability of classical Adaboost algorithm to automatic ROI segmentation in medical images.
Most precisely, we show that Haar filter can provide some interesting results but can be efficiently replaced by more adapted weak classifiers that integrate some prior
informations on the type of scan used for imaging. Application on tagged MRI segmentation is proposed.
MSc student: Leila Meziou
Supervising: Aymeric Histace (aymeric.histace_at_u-cergy.fr), Frédéric Precioso (frederic.precioso_at_ensea.fr)
Histogram based segmentation using active contours in a paramteric framework : application to prostate MRI segmentation
Complete Description: Aim of this project is to investigate how the histogram based segmentation using active contour can be adapted to MRI segmentation.
Most precisely, by taking into account the particular acquisition noise of MRI scans (Rician distribution), we propose to compare performances of different classical distances
like Kullback-Leibler divergence, Hellinger distance (related to the Bhatacharyya divergence), the classical Ki2 distance and the Wasserstein distance. Application to bladder
segmentation in prostate MR imaging is proposed.
MSc student: Laetitia Lamard
Supervising: Aymeric Histace (aymeric.histace_at_u-cergy.fr), Frédéric Precioso (frederic.precioso_at_ensea.fr)
December 2008 - March 2009 (Master STIC from University of Cergy-Pontoise)
Hybrid level set approach for medical image segmentation: application to extraocular muscle segmentation in MRI for early diagnosis of exophtalmia
Complete Description: Aim of this project is to developp a segmentation tool using level set approach for segmentation of extraocular muscles in head MRI.
Such a tool has become of primary importance for fast segmentation of extraocular muscles which morphological characteristics (diameter, muscle to fat ratio,...) can make
an early diagnosis of exophtalmia pathology possible.
MSc student: Marine Breuilly
Supervising: Aymeric Histace (aymeric.histace_at_u-cergy.fr)
December 2007 - March 2008 (Master STIC from University of Cergy-Pontoise)
Fractionnal PDE based restoration approach for image denoising
Complete Description: Aim of this project is to investigate what fractional PDE can bring to the PDE based restoration approach for image denoising.
As a prospectove study, we propose to compare classical Heat equation diffusion and its fractionnal version but also classical Perona-Malik's PDE and its fractionnal version.
MSc student: Marina Dagaudez
Supervising: Aymeric Histace (aymeric.histace_at_u-cergy.fr)