AI, A New Weapon In The Fight Against Breast Cancer In France

ETVBHARAT

Villejuif (France): At France's Gustave Roussy Institute, artificial intelligence is increasingly being used to assist radiologists in analysing mammograms, offering faster and more detailed detection of suspicious areas. The AI system highlights potential anomalies almost instantly, acting as a "second look" to reduce the risk of missed signs of cancer.However, experts stress that AI is not a replacement for doctors. Radiologist Corinne Balleyguier explains that while the technology is highly sensitive, it can generate many false positives, making human interpretation essential. Reviewing AI results can sometimes even increase reading time.Professor Nathalie Lassau of the French Society of Radiology says AI's strength lies in processing vast amounts of data, supporting a more holistic view of patients. She reassures that the final diagnosis and responsibility remain firmly in human hands.

Transcript

00:00artificial intelligence is now helping doctors read mammograms at France Institute the system

00:09quickly highlights areas that may look suspicious doctors say AI works like a second set of eyes

00:16it helps make sure nothing important is missed but AI can also show problems that are not really

00:24there that is why human reading is still needed it serves a little bit of a second regard to be sure

00:32that we are not going to be close to an anomaly it is very sensitive but it can make a lot of

00:38positives so that's why the reading of a radiologist is very important because we are going to remove

00:46the positives from artificial intelligence tools radiologists carefully check the scans and then

00:53review the AI results experts say AI does not replace doctors it supports them mais en gros si on se

01:04fiat complètement à l'IA sans lecture humaine on ferait vraiment des grosses erreurs diagnostiques par

01:09exemple ces logiciels là en fait ils sont ils ont été entraînés avec des bases de données où ils

01:14sont dédiés pour détecter des anomalies subtiles ce qui est logique et donc un logiciel d'IA peut louper

01:20un cancer de 5 cm que normalement un radiologue doit pas louper globalement je pense que les patients

01:26et il y a eu aussi quelques études là dessus se sentent quand même assez rassurés que derrière ce soit un

01:33radiologue et qui soit bien habilité disons qui derrière fait la validation finale quoi

01:40by analyzing large amounts of data AI helps doctors be more accurate and consistent in cancer cure the final

01:50decision is always made by a trained radiologist

Category

Transcript

Comments

Recommended