AI-assisted screening boosts breast cancer detection: Study

STUDY EVALUATED 105,000 WOMEN IN SWEDEN BETWEEN APR 2021 AND DEC 2022, MAKING IT THE LARGEST TRIAL OF A.I. IN MAMMOGRAPHY SCREENING SO FAR

- Rhythma Kaul

NEW DELHI: A landmark clinical trial has demonstrated that artificial intelligence can dramatically improve breast cancer screening, achieving a 29% increase in early cancer detection while reducing radiologists' workload by nearly half—findings that could transform how mammograms are evaluated worldwide and underscore the potential of machine learning.

The randomised controlled study, published on Tuesday in The Lancet Digital Health, evaluated over 105,000 women in Sweden between April 2021 and December 2022, making it the largest and most comprehensive trial of AI in mammography screening to date.

"There was a significant 29% increase in cancer detection when AI-supported screening was used compared with standard of care," the researchers reported. "There were non-significant increases in the recall rate (8%) and false-positive rate (1%). The findings suggest AI contributes to the early detection of clinically relevant breast cancer and reduces screen-reading workload without increasing false positives."

AI-supported screening detected 338 cancers compared to 262 in the control group. Crucially, the technology excelled at finding small, early-stage cancers that hadn't spread to lymph nodes—detecting 58 more small tumours and 46 more lymph node-negative cancers than traditional screening methods.