Description du projet
Prévention du cancer: le rôle de l’intelligence artificielle
La coloscopie constitue l’approche conventionnelle de détection et de retrait des polypes précancéreux. Bien que l’intelligence artificielle (IA) soit considérée comme une aide à la détection de cellules malignes, il n’existe aucune preuve directe de son impact sur la prévention du cancer colorectal. L’objectif principal du projet ACCEPT, financé par l’UE, est de mettre en place l’infrastructure de recherche permettant de comparer directement la coloscopie standard et la coloscopie assistée par l’IA. Les chercheurs utiliseront des données réelles ainsi que des informations de suivi issues des programmes nationaux de dépistage du cancer colorectal en Europe. Les résultats fourniront des preuves fiables sur le rôle de l’IA dans la prévention du cancer et contribueront à la diffusion de cette technologie en Europe et dans le reste du monde.
Objectif
Removal precancerous polyps (so-called adenomas) during colonoscopy reduces colorectal cancer incidence and mortality. We have recently developed artificial intelligence (AI) systems which optimise colonoscopy quality by aiming at increasing the detection of adenomas. However, it is unknown if this benefit of AI translates into improved cancer prevention.
To establish the role of AI in colorectal cancer prevention, we here propose a three-step research portfolio targeting individuals in the national colorectal cancer screening programmes in Norway and Poland. 1) A 1-year observational study to clarify the superiority of AI in adenoma detection in the average-risk population. 2) A cost-effectiveness analysis for cancer prevention, using real-world data obtained from the part 1 study. 3) Establishment of an infrastructure for long-term (10 years) follow-up of 40,000 individuals to quantify the effect of AI on colorectal cancer incidence.
We use propensity score matching to balance the background characteristics of the two comparing arms; AI-assisted colonoscopy and standard colonoscopy. Participants are followed through national cancer registries for ten years after cancer screening.
Our project is the first of its kind and made possible by the unique collaboration of world-class environments in clinical epidemiology and medical device assessment with world-leading AI developers. The project will enable evidence-based implementation of AI technologies into nationwide cancer screening programmes in Europe and the world.
The project’s achievement will be maximised through detailed plans for training, monitoring, management, exploitation, dissemination, and communication. While adding essential competence to the host institution, this project will increase the applicant’s skill, especially in epidemiologic aspects, helping him obtain an independent academic position in translational research of AI for medicine.
Champ scientifique
Mots‑clés
Programme(s)
Régime de financement
MSCA-IF - Marie Skłodowska-Curie Individual Fellowships (IF)Coordinateur
0313 Oslo
Norvège