Springer conference paper web page
The associated presentation is available using the link below :
Springer conference paper web page
The associated presentation is available using the link below :
The final program and schedule for the conference GSI 5th Conference on Geometric Science of Information is now available. The Conference takes place in Paris, Sorbonne University in July.
I will chair the session "Manifolds and optimization" with Bin Gao (UCLouvain)
Don't miss the talks of the Keynote Speakers :
Abstracts of GSI'21 keynote speakers are available here:
https://www.see.asso.fr/en/wiki/367468_keynote-speakers
Information on GSI'21 registration:
Our paper
Optimization of a shape metric based on information theory applied to segmentation fusion and evaluation in multimodal MRI for DIPG tumor analysis, Stéphanie Jehan-Besson (CNRS, LITO U1288 INSERM, Institut Curie), Régis Clouard (Normandie Univ, ENSICAEN), Nathalie Boddaert (Hôpital Necker Enfants Malades, AP-HP Imagine), Jacques Grill (Gustave Roussy, Inserm), Frédérique Frouin (LITO U1288 INSERM, Institut Curie).
has been accepted to the 5th conference on Geometric Science of Information in PARIS, Sorbonne University GSI 2021. As for previous editions, GSI’21 Proceedings will be published in SPRINGER LNCS. See GSI’19 Proceedings.
It takes place in the project BIOMEDE : https://www.lito-web.fr/en/projects/94-biomede-ia-2
I will present my research work the 7th april 2021 on the topic of segmentation and evaluation/fusion of ROIs in multimodal MRI. This work takes place within the BIOMEDE project (collaboration Hôpital Necker des enfants malades, Institut Gustave Roussy).
Since january 2021, I am mainly involved in the BIOMEDE project.
This project aims at better characterizing, by using machine learning or deep learning methods, the tumors of patients with diffuse intrinsic pontine gliomas (DIPG) which is a rare pediatric cancer with a very poor prognosis. The analysis is based on clinical and imaging (multi-parametric MRI) features.
https://www.lito-web.fr/en/projects/94-biomede-ia-2
Our preprint is now available on ArXiv
https://arxiv.org/abs/2102.08939
Abstract : This paper proposes the estimation of a mutual shape from a set of different segmentation results using both active contours and information theory. The mutual shape is here defined as a consensus shape estimated from a set of different segmentations of the same object. In an original manner, such a shape is defined as the minimum of a criterion that benefits from both the mutual information and the joint entropy of the input segmentations. This energy criterion is justified using similarities between information theory quantities and area measures, and presented in a continuous variational framework. In order to solve this shape optimization problem, shape derivatives are computed for each term of the criterion and interpreted as an evolution equation of an active contour. A mutual shape is then estimated together with the sensitivity and specificity of each segmentation. Some synthetic examples allow us to cast the light on the difference between the mutual shape and an average shape. The applicability of our framework has also been tested for segmentation evaluation and fusion of different types of real images (natural color images, old manuscripts, medical images).
My new web page is here : https://sites.google.com/view/jehanbesson-research/home