"Useful AI in healthcare is AI that addresses a real need, with applications designed by and for healthcare professionals"

Interview with Naoual Bakrin, surgeon,
and Thomas Guyet, computer scientist

Read the article

The QuickRare project

In France, more than 3 million people suffer from a rare disease (a disease with a prevalence of less than 1 in 2,000), and despite the progress achieved through the 3ème PNMR (PNMR), the patient journey to diagnosis remains long and complex.

The QuickRare project aims to reduce the diagnostic odyssey associated with rare diseases by developing an innovative digital tool for primary care physicians, including general practitioners.

The two objectives of this project will be:

  • to explore recent advances in text analysis using artificial intelligence techniques to identify markers in medical documents associated with specific types of rare diseases.
  • to validate the feasibility of such a solution with the primary care physician community, addressing the ethical issues raised by the implementation of such an AI tool and its acceptability among the physicians involved, while also considering the practical aspects of its use.

The integration of digital technology into healthcare cannot be successfully achieved without taking into account several success factors, including:

  1. reliable, secure, and user-friendly digital tools;
  2. health data management—both within the regulatory framework and regarding data utilization;
  3. ethics;
  4. the societal dimension, ensuring professional buy-in for this digital health evolution, particularly concerning priority public health issues.

    This project incorporates most of these components, requiring a high degree of transdisciplinarity.

    An interdisciplinary consortium

    QuickRare draws on an interdisciplinary consortium—bringing together experts from within and beyond the SHAPE-Med@Lyon community—specializing in rare disease treatment, the humanities and social sciences (sociology and philosophy of technology), and computer science. The project aims to combine these areas of expertise to explore the feasibility and implementation of a technological solution for early medical referral recommendations for patients facing diagnostic delays, while also addressing societal issues associated with the use of such technologies in healthcare.

    Project leaders:

    Justine BACCHETTA

    Justine Bacchetta is a Professor of Pediatric Nephrology at  Lyon 1 Université, specializing in bone disorders caused by rare kidney diseases. As a researcher at LYOS, she studies these genetic diseases affecting children, whose daily lives are severely impacted. The QuickRare project will focus on patients with suspected nephrological conditions seen at the rare disease reference center hosted within Professor Bacchetta’s department.

    Naoual BAKRIN

    Naoual Bakrin is a physician in the digestive and oncological surgery department, a collaborator with the Cicly laboratory, and the head of the mission-driven company R2D (Rare Disease Diagnosis), which contributes to research aimed at improving the care of patients with rare diseases—specifically regarding their diagnosis and management.

    Élodie GIROUX

    Elodie Giroux is a Professor of Philosophy of Science and Medicine at Université Jean Moulin Lyon 3. Her areas of specialization include the philosophy of medicine, as well as the philosophy and history of epidemiology and public health. She has focused in particular on the concept of disease and its representations.

    Thomas GUYET

    Thomas Guyet is an Inria researcher with the AIstroSight team. He conducts research on artificial intelligence applied to healthcare, specifically developing tools to analyze care trajectories. He has also led several projects at the intersection of computer science, the humanities, and healthcare. The work carried out by the AIstroSight team is conducted in collaboration with the HCL (Hospices Civils de Lyon), which hosts part of the team at the Bron hospital.

    Muriel SALLE

    Muriel Salle is a historian and an associate professor at Lyon 1 Université. A specialist in women’s history, she focuses her research on medical discourses and health. She teaches at the medical school and at Sciences-Po Lyon. As part of the QuickRare project, she aims specifically to analyze and understand the differences in diagnostic delays observed between girls and boys. A key objective of the project is to ensure that the tools developed correct these gender biases.

    They have joined us:

    Aoitif EL-MOKHLESSE

    Aoitif El-Mokhlesse has been recruited as a Clinical Research Associate (CRA) for the QuickRare project (with additional funding from R2D). With a background in quality management, she is responsible for compiling the model’s training dataset in collaboration with clinicians.

    Anne FENOY

    Anne Fenoy was recruited as a postdoctoral researcher for the QuickRare project to support ethical and epistemological reflection on the design of an artificial intelligence tool for diagnostic recommendations. She holds a PhD in the philosophy of medicine from Sorbonne University, where her research focused on the representation of amyotrophic lateral sclerosis (ALS), a rare disease.

    Antoine RICHARD

    Antoine Richard has been recruited as a research engineer for the QuickRare project to develop artificial intelligence tools based on natural language processing and machine learning techniques. After completing a PhD in medical decision-support systems, he led various IT projects at HCL, particularly in the field of natural language processing.

    Partners

    FOLLOW US

    Linkedin Twitter Youtube Instagram