2 innovative software programs predict the impact of MS

Improving the treatment of multiple sclerosis using artificial intelligence? “A reality in France from 2025! “, affirms Professor Gilles Edan, scientific and technical manager of the RHU* Primus program promoted by the university hospital center (CHU) of Rennes. How ? Thanks to a unique medical decision support service, developed as part of the Primus project (projection in multiple sclerosis or projection in MS), soon to be evaluated by the FCRIN4MS clinical research network. His credo? “Choose the right treatment, for the right patient, at the right time. » The one who defines himself as “one of the oldest expert neurologists in MS”, with 50 years of practice under his belt, participated in the creation of this innovative program. It sets out its challenges and its strengths, ahead of the World Day dedicated to this autoimmune disease of the central nervous system on May 30, 2024.

Handicap.fr: How is multiple sclerosis diagnosed today?
Professor Gilles Edan : Its diagnosis was refined in 2024. Until then, it exclusively concerned patients who had presented at least one suggestive episode (sensory disturbance, reduction in visual acuity, etc.). At the instigation of Doctor Christine Lebrun from Nice University Hospital, we are now moving towards the possibility of making a diagnosis before the first attack, on the basis of additional examinations which combine MRI and lumbar puncture.

H.fr: A concrete example?
GE : After a shock to the brain during a football or rugby match, a teenager seems groggy, he is given an MRI. We can then discover a signature which suggests MS, which will lead to the start of treatment. These new criteria could therefore make it possible to make a diagnosis of MS following examinations carried out in a context which is not that of the clinical expression of the disease.

H.fr: What is at stake?
GE : It is above all to prevent the disease in order to better control its impact. Progress made in the treatment of MS mainly concerns patients who have benefited from early diagnosis and treatment. We are much less equipped to face diseases when they are well established and cause disabilities.

H.fr: How is artificial intelligence an asset for improving diagnosis?
GE : In MS, focal lesions accumulate in the brain and spinal cord tissue. MRI is an exceptional spy for detecting new lesions; this is what has revolutionized the confidence placed in treatments.

To analyze their evolution precisely, our ambition is to provide radiologists and neurologists with tools that have learned to recognize the changes that occur via artificial intelligence techniques. Thus, the first technological tool developed as part of the Primus project reads and interprets patients’ MRI and then detects each change in a much more operational manner. Objective ? Offer all patients a scholarly, precise and certain reading of their examinations in order, ultimately, to prescribe the appropriate treatment.

H.fr: Concretely, this automatic reading tool will make it possible to identify possible new spots in the brain?
GE : Exactly, but also of the spinal cord, which are particularly difficult to distinguish from artifacts (editor’s note: disturbance of the image due to the pulsatile flow of blood or cerebrospinal fluid which can be wrongly interpreted by the doctor as a lesion).

H.fr: How was this tool developed?
GE : To ensure he identified real lesions, he had to practice. For this, we fed the software “ground truths” (MRIs authenticated by neuroradiologists) so that it can integrate the difference with an artifact.

H.fr: But are errors possible?
GE : This automatic reading must be validated by a radiologist but it is a valuable aid to support the interpretation of this examination and saves time in reading for the radiologist and the neurologist.

H.fr: What is the other function of the Primus project?
GE : Make software available to all neurologists allowing them to observe the impact of different treatments on a large panel of patients from a specific database. Concretely, it makes it possible to compare the data of each patient with other patients of the same sex, same age group, same age at the onset of the disease and same clinical and MRI characteristics, but having received different treatments. The challenge is to observe the differences in the number of relapses, lesions and the progression of the disability according to the different treatments. Data from approximately 10,000 patients, followed over several years, were recorded. To summarize, the challenge is that all the knowledge accumulated over time is accessible in a few clicks. It’s a real revolution!

H.fr: When will these two technological tools be implemented?
GE : With the help of FCRIN4MS, they will be tested, together, in 28 French neurology departments in 2025; half are expert centers (CHU) and the rest are neurology departments of general hospitals. Nearly 500 patients will participate in this study coordinated by the Nantes University Hospital.

H.fr: When will there be a generalization?
GE : That takes time. In five years, we should be able to make these two tools available to treating neurologists.

H.fr: Based on your work, do you think AI will revolutionize the medical sector and the way we support patients?
GE : Yes, provided that it is used intelligently and thoughtfully and that the data used is authenticated ” high quality “. This is the case for those we use in the Primus program. The data comes from therapeutic trials carried out to obtain marketing authorization for drugs and from the French Multiple Sclerosis Observatory (OFSEP). AI will not make the decision, it will provide knowledge that enriches the doctor’s ability to make the right decision. A doctor will perhaps have in memory, during his consultation, five or ten patients with the same characteristics while with our tool, the AI ​​can store thousands.

Want to know more ?

To find out more about the F-CRIN network of excellence and two other promising studies on MS (early treatment and remyelination), continue reading in our second article: Multiple sclerosis: 3 innovative studies made in France!..

* University hospital research

© Stocklib Sergey Nivens