Title: Unsupervised and Weakly Supervised Deep Learning for Medical Image Analysis
​​​​​​
Time: Wednesday June 12, 2024 - 13:00-14:00

Location: Room H-0.09​​​​​​

Abstract: 
Machine Learning and Deep Learning are powerful tools that have become ubiquitous in many fields when it comes to image processing. Medical imaging is no exception, and these AI methods are often used to help diagnose diseases or their progression. However, Deep Learning algorithms are also notorious for their huge requirements in terms of high-quality annotated data. As it turns out, such high volumes of annotated data are rarely available for rare diseases and new medical problems, thus making it impossible to use mainstream and high performing Deep learning methods in many cases. In this presentation, we explore 3 case studies of unsupervised and weakly supervised deep learning architectures applied to medical image analysis for diseases as diverse as age-related macular degeneration, glaucoma and amyotrophic lateral sclerosis.

Prof. Dr. Jérémie Sublime got his PhD in Applied Computer Science from Paris-Saclay University in 2016. Since then he has been an Associate Professor in the Departement of Data Science of ISEP school of digital Engineers in Paris. His research activities include machine Learning as well as clustering and image processing, with a particular focus on unsupervised learning and its applications in the fields of remote sensing and medicine.
 

Prof. Pierre-Alexandre Murena