(Deep) machine learning for the prediction of patient coma outcome based on multimodality neuroimaging

Please look at the attached pdf file for a full description of this master project.

(Deep) machine learning for the prediction of patient coma outcome based on multimodality neuroimaging

Keywords : Medical Image analysis and Modeling, Machine Learning, Deep learning, Diagnosis model

Scientific context

The IMAGINA project recently funded (oct 2018) by the french National Research Agency (ANR) targets the challenging question of providing an accurate diagnosis of patients being in acute coma. CREATIS is responsible for developing an automated diagnosis tool that will evaluate the patient coma status (degree of consciousness disorder) by combining the information provided by multimodality imaging with the most advanced machine learning methods.

Objective of the internship

The objective of this master project is to initiate the first developments in machine learning for the statistical analysis of a multimodality imaging database of coma patients and healthy subjects.

Two ways will be investigated :

  •  The first one will address the problem from the classical machine learning perspective. A series of manually engineered featured will be extracted from the different series of images and a statistical inference model will be tested to output a score referred to as the coma recovery scale.
  •  The second approach will consider and design a data driven feature extraction strategy based on the most recent advances in statistical deep learning strategies.

The experimental work will be based on a preliminary multimodality image database consisting of 40 acute comatose patients and 25 healthy subjects and including PET (Positron Emission Tomography) and multiparametric MRI (Magnetic Resonance Imaging) acquisitions .

skills

Candidate should have strong background either in machine learning and/or deep learning or image processing and some experience in both fields as well as good programming skills.

We are looking for an enthusiastic and autonomous student with strong motivation and interest in multidisciplinary research (image processing and machine learning in a medical context)

Host laboratory : Laboratoire CREATIS, 69 Villeurbanne

Supervisor : Carole Lartizien - carole.lartizien@creatis.insa-lyon.fr.

Duration : 6 months.

Starting date : february/march 2019.

Gratuity : ~560 euros/month.