Annonce

Les commentaires sont clos.

Proposition d'une thèse en imagerie médicale

26 Octobre 2022


Catégorie : Doctorant


Sujet: :
Deep neural network-based prediction of brain tumor treatment outcomes from MRI images

 

Glioma is the most common primary malignant tumor of the central nervous system. The typical treatment is the resection of the tumor lesion, followed by radiotherapy on the operating region coupled with chemotherapy. Despite this standard treatment, the prognosis of glioma (especially the high-grade ones) is poor, featured by a very high risk of tumor recurrence, commonly called cancer progression. Although glioma progression systematically happens within the central nervous system, the topography of the relapse and recurrence time differ significantly across different patients. Timely predict the recurrence and the topography of relapse at the time of initial diagnosis is of great clinical value towards the personalized treatment of glioma, e.g., to adapt treatment and follow-up strategies for individual patients. The goal of this thesis is to develop a set of cutting-edge AI methods to predict the topography and time to relapse for patients with newly diagnosed glioma and to follow up its progression during the treatment.

The thesis will be carried out in the Quantif team of the LITIS laboratory, University of Rouen Normandy and the University of Bourgogne.

Profil: Master student in image processing, statistical analysis, applied mathematics with a wide background in deep-learning, statistical inference.

Required computer language: Python (libraries Keras or Pytorch).

Duration : 3 years starting from December 15 2022.

Supervisors:

-Su Ruan su.ruan@univ-rouen.fr

-TRUC Gilles PU-PH