Modelling, Learning and Image Computing

Aims

Computer-aided diagnosis, prognosis and therapy from biomedical images and genomic data

Leader name(s)

Pr N. Ayache
Pr M. Barlaud

General description

Modelling:

  • biophysical models of tumour growth at various scales
  • biophysical models of brain atrophy at various scales

Image Computing:

  • Personalization of biophysical models from time series of medical images
  • Extraction of biomarkers from images based on biophysical models

Machine Learning:

  • Statistical Risk prediction
  • Supervised Classification (Proximal Methods)
  • Control of false positives or negatives (Neyman-Pearson method)
Task Title
Lung / Head & Neck Tumour growth quantification by Image Computing (micro-macro scales)
Task Description
  • Collection of images/data of patients with Lung and Head and Neck tumours
  • Modelling tumour growth from micro to macroscopic scales
  • Personalize tumour growth models from medical images and data to assess tumour progression and predict therapy effect
Task Improvement
  • Improve assessment of the disease progression
  • Improve therapy planning (surgery & chemo- & radio-& targeted therapies)
  • Reachable in 3 years
Task Title
Brain atrophy Quantification by Image Computing (micro-macro scales)
Task Description
  • Collection of Images/data of patients with neurodegenerative diseases
  • Modelling brain atrophy from micro to macroscopic scales
  • Personalize atrophy models from medical images and data for early detection of disease and to quantify the effect of therapy
Task Improvement
  • Improve early detection of degenerative diseases and quantify effect of treatment on disease progression thanks to image biomarkers
  • Reachable in 2 years
Task Title
Machine Learning on Lung Tumour and COPD (nano)
Task Description
  • Collection of large scale Genomic data
  • Statistical COPD prognosis
  • Statistical Relapse Risk prediction
  • Statistical Treatment response prediction
Task Improvement
  • Improve COPD prognosis, relapse risk prediction and prediction of treatment response
  • Improve therapy planning
  • Reachable in 5 years
Task Title
Modelling Head and Neck Lung Tumour (nano-micro)
Task Description
  • Collection of Genomic and images of tumour cells after drug exposure
  • Estimation of cell heterogeneity (genomic and non-genomic) by time lapse imaging and drug testing
Task Improvement
  • Improve early prediction of treatment response and help for optimal drug combination design (product, concentration)
  • Reachable in 1 years