A groundbreaking study shows a new AI system surpassing radiologists in identifying breast cancer from mammograms.
Early detection dramatically improves breast cancer survival rates, which is why regular mammograms are recommended for women. While highly effective, mammography misses about one in five breast cancers and leads to one in two women receiving a false positive over a decade.
Researchers worldwide have spent years refining AI to boost mammography accuracy. Google Health's advanced system, trained on 91,000 mammograms, was rigorously tested on two datasets totaling 28,000 more scans—one from U.S. women and one from the U.K.
In the U.S. dataset, the AI reduced 5.7% fewer false positives and 9.4% fewer false negatives than physicians, catching more cancers they overlooked.
For the U.K. data, reviewed by two radiologists to minimize human error, the AI still excelled with 1.2% fewer false positives and 2.7% fewer false negatives.
In a head-to-head challenge, six U.S. radiologists analyzed 500 random mammograms against the AI. The system spotted cancers missed by all experts, though it missed one case they caught—highlighting its impressive, though not flawless, performance.
The goal isn't to replace radiologists but to empower them, speeding up diagnoses for better patient outcomes.
"This demonstrates how AI can support human experts," says co-author Dominic King. "It could highlight potential abnormalities, prompting doctors for a closer review."
Medical professionals largely welcome AI: An MIT Technology Review survey of 900 respondents found 80% agree it reduces burnout, freeing doctors to focus more on patients.
Source
Related articles:
