MSxplain
On going project
In Spring 2023, the EPFL+ECAL Lab joined the consortium MSxplain investigating how to better explain AI decisions in personalized healthcare. This collaboration, focusing on Multiple Sclerosis, aims to incorporate deep learning into diagnosis and treatment planning. Key partners include the Medical Image Analysis Laboratory (mial, UNIL, Lausanne) specializing in medical image analysis, the Translational Imaging in Neurology group (ThINk, University of Basel, Basel) focusing on neurology imaging, and MedGIFT from HES-SO in Sion, bridging computer science and medicine. This consortium is pioneering in adopting a transdisciplinary approach to AI decisions in personalized healthcare.
Multiple Sclerosis (MS) is a widespread, debilitating condition, impacting over 2 million individuals globally and representing the most common neurological disorder in young adults. The progression of MS significantly deteriorates the quality of life of those affected and imposes considerable financial strains on patients, their families, and society.
While deep learning (DL) methods have recently advanced in monitoring MS patients, integrating these techniques into clinical neuro-radiology workflows remains challenging, impeding their broad validation. A key obstacle is the perceived opacity of these tools by physicians, who are the primary users. EPFL+ECAL Lab is dedicated to comprehending and addressing the hurdles in adopting Explainable Artificial Intelligence (XAI), with a particular focus on how its presentation impacts understanding and trust within the healthcare industry.
The results of the project will be released in 2024 and 2026.
PROJECT MANAGEMENT & ENGINEERING LEAD
UX Assistant
Clara Evans
Principal Investigator
Dr MER Merixtell Bach Cuadra, group head mial
Investigators
Pr Cristina Granziera (ThINK, University of Basel)
Pr Henning Müller and Prof. Adrien Depeursinge (medGIFT, HES-SO Valais)
PhD candidates Ms Nataliia Molchanova (UNIL, HES-SO)
PhD candidates Mr Federico Spagnolo (Basel, HES-SO)
Part-time software engineer Mr Roger Schaer (HES-SO)
Dr Mara Graziani (HES-SO)