AI Cuts Missed Breast Cancers by 12% in Swedish Trial

Doctors just got powerful backup in the fight against breast cancer, and it could help catch tumors that traditional screenings miss.

A massive Swedish study involving over 100,000 women found that artificial intelligence spotted 12 percent more cancers between routine mammogram appointments compared to standard screenings. These "interval cancers" are particularly dangerous because they're typically more aggressive and harder to treat.

Here's how it worked. Half the women received Sweden's standard screening, where two radiologists examine breast X-rays for signs of cancer. The other half had their mammograms reviewed by AI software that flagged high-risk cases for two doctors while sending low-risk cases to just one physician.

The results were remarkable. About 81 percent of breast cancers in the AI group were caught during routine screenings, compared to just 74 percent in the standard group. The AI also reduced cancers that had invaded healthy tissue by 16 percent.

But the benefits extend beyond better detection. Previous research from this trial showed the AI system reduced radiologists' workload by 44 percent without missing more cancers. That's crucial as healthcare systems worldwide struggle with physician shortages.

The Ripple Effect

The timing couldn't be better. Breast cancer remains the most common cancer for women in 157 countries, with 2.3 million diagnoses worldwide in 2022 alone. Yet deaths have dropped 44 percent in the United States since 1989, thanks largely to regular screenings.

AI-assisted mammograms could accelerate that progress while easing pressure on overworked medical staff. Breast imaging specialist Jean Seely notes that catching cancers earlier also saves money. Late-stage interval cancer treatment can cost around $100,000 per patient in Canada, while early detection dramatically reduces those expenses.

The technology isn't perfect, and experts want more research before widespread rollout. "Using AI to assist in reading mammograms can be more efficient, but there's a concern that it can lead to missing some cancers," says Sowmiya Moorthie of Cancer Research UK. Since this study came from Swedish screening centers, researchers want to confirm the results hold true in other healthcare systems.

Still, study co-author Kristina Lång believes the evidence is promising: "Widely rolling out AI-supported mammography could help reduce workload pressures among radiologists, as well as helping to detect more cancers at an early stage."

For the millions of women who get routine mammograms, this technology could mean the difference between catching cancer early or facing a much harder fight later.