C. Lartizien                CNRS Research Associate

Ma photo

Research interest

I am a CNRS Research Associate (section 7). I am conducting research at the CREATIS laboratory in Lyon, France, whose aim is to develop image processing methods for medical imaging. I am the co-leader of research team Images and Models

My current research interests focus on machine learning methods for medical image analysis. This includes kernel learning approaches (e. g. SVM) for multiclass classification or outlier detection as well as unsupervised deep feature learning. The clinical applications concern the prototyping of computer aided diagnosis systems (CAD) for cancer and brain pathology screening based on multimodality medical imaging.

My previous research projects were in the field of signal detection theory for medical imaging as well as simulation of realistic nuclear imaging databases.


Job offers

09/10/17 : We are currently hiring two master students for internships starting early 2018.

The first subject in the field of deep learning for multiparametric brain pathology screening. Please download it here

The second subject in the field of multitask learning applied to prostate cancer mapping in multiparametric MRI. Please download it here.

New publications

July 2017 : Our paper on the generalization of the SVDD algorithm has just been published in Neurocomputing. Please have a look here. This publication proposes a generalization of the SVDD outlier detection algorithm allowing to estimate a probability of belonging to the normal class in contrast to the classical model which provides unbounded scores. This probabilistic information is very important for making inter and inter-individual comparisons.