Will artificial intelligence (AI) replace doctors? To this question, which fuels many fears, Charlotte Alliod opposes a « non » firm and frank. But she has no doubt: “It will allow them to provide better and faster care!”
For more than four years, this 37-year-old researcher, doctor of health sciences and AI specialist, has been interested in a complex pathology: endometriosis. This very painful inflammatory gynecological disease (1) affects nearly 10% of women. But since its causes are poorly understood and its forms are very diverse, it is difficult to identify by professionals. At the heart of the “Research & Innovation” department of CapGeminini Engineering, Charlotte Alliod’s team hopes to reduce the time and risk of diagnostic errors by leveraging the power of algorithms.
“When I think about what you can do with these calculations, it’s dizzying,” she says with a big smile. Her wide blue eyes and her slightly rushed speech suggest an unfailing enthusiasm. When I was younger, I wasn’t very good at science subjects and my family didn’t have the talent: it takes passion and motivation to force destiny! »
From pharmacy to artificial intelligence
Originally from the region of Gex, in Ain, this curious woman has always been « fan » laboratories but hesitated for a long time to choose those that would suit him best. After a first year in pharmacy, ” by chance “, she decides to try studies in biochemistry, but she is bored. The student finally believes she has achieved her dream by joining a scientific police school in Switzerland. But faced with the difficulty of finding work, she turned back.
Back in biochemistry where she discovered, this time, a kind of vocation, Charlotte Alliod began a thesis devoted to “explosives of tomorrow”, which is slowly shifting her towards biotechnology, a discipline that combines life sciences, computer science and mathematics. Her mission? To create an algorithm capable of predicting whether future molecules developed could be less toxic for humans and the planet.
This first contact with artificial intelligence was a revelation. So much so that in the fall of 2019, she was recruited as a project manager at Altran. – since become CapGemini Engineering – to develop research projects combining AI and health. The mission excites him.
His energy “in the right place”
The subject of endometriosis emerged during the first brainstorming session, when one of her colleagues stuck this post-it on the wall: “I have a friend who has it and who suffers a lot, I would like someone to be able to help her.” After three months spent “map” With this disease, Charlotte Alliod discovers its share of problems: diagnostic wandering, complex medical journey, difficulties in treating pain… For her, it is obvious: she and her colleagues put their energy ” at a good place “.
His team approached the Croix-Rousse hospital, which helped him refine the research protocol and provided motivated professionals. Above all, the establishment allowed him to access 200 files of patients already diagnosed and treated, from medical reports to surgical assessments, prescriptions and blood tests.
Faster than a doctor
Once anonymized, this valuable data was used to train a program that was continuously tested and adjusted. The objective: to create an artificial intelligence system capable, simply by reading a file, of detecting the disease. “Today, it is estimated that our AI is as effective as a doctor, but faster!” summarizes Charlotte Alliod. The researcher therefore imagines, ultimately, transforming it into a tool.
“Professionals tell us that they spend half of their consultation trying to understand their patients’ journey, which is often chaotic: by saving this time, they could be more available to listen to them and support them in relieving their pain.”
While a first scientific article is being written, two other projects have already been launched: Endoview to better identify lesions on MRI images and Endocare, to assess the psychological state of patients.
(1) Endometriosis is defined by the presence and development of endometrial-like tissue (uterine lining) outside the uterus.
——
His compass. Collective emulation
“While we hardly talked about it a few years ago, the subject of endometriosis is finally gaining momentum. We are interested in it, we are studying it, and we are taking action on all sides. In Lyon, the Hospices Civils are opening dedicated day hospitals and the start-up Ziwig has developed a saliva test. In Bordeaux, the Tivoli clinic is constantly innovating. There is real emulation! Knowing that we are not alone and that we are not doing this for nothing guides and motivates me. Especially when I think of all my friends who have been diagnosed and who I see suffering on a daily basis.”
