AI-Powered MRI: Guillaume, ESILV alumnus, performs Research on Deep Learning to improve medical diagnosis

As A.I., Robotics and Data are redrawing the healthcare landscape, researchers from all over the world started to study and to publish about deep learning-based magnetic resonance imagining (MRI). One of them is Guillaume Fradet, ESILV alumnus, class of 2019, who is a data scientist specialized in e-health, machine learning, and computer vision.

Guillaume is a deep learning researcher at AZmed, an e-health start-up based in Paris that deploys Artificial Intelligence and machine learning to help radiologists and emergency physicians make more precise and rapid diagnoses.

“As a researcher at AZmed, I have to improve the performance of our algorithms, mostly so that they better generalize over various medical centres. That’s a well-known problem in the machine learning environment, called Domain Adaptation.” (Guillaume Fradet, class of 2020, e-health data scientist)

In a recent research report on domain adaptation in medical imaging, Guillaume presents the benefits of various technologies used to train A.I. so it can adapt to different medical environments.

• Transfer Learning: this methodology involves leveraging (“adapting”) data from the source domain (step one) so that it can be used in the target domain (step two or second round of training)
• Image enhancing/editing: the use of CLAHE ( Contrast Limited Adaptive Histogram Equalization) technology
• Artificial neural networks: the Batch Normalization method is ideal for domain adaptation of the A.I. medical solutions

As a team member at AZmed, Guillaume’s mission is “to optimize the workflow of radiologists allowing faster and more precise detections”, and help them “conserve time and energy for intricate cases”. He describes itself as a supporter of the #AIforGood movement. 

ESILV Graduate School of Engineering believes in the power of A.I. to improve people’s lives. The Health Engineering & Biotechnology major focuses on data, connected health, mechatronics, A.I. and human knowledge to train engineers to evolve in the multidisciplinary ecosystem of technology for health. By combining life sciences expertise with technologies, I.T., Big Data & Artificial Intelligence, ESILV aims to inspire new ways to impact modern medicine positively.