Artificial intelligence (AI) is becoming part of many areas of healthcare, but one of its most promising uses may be in breast cancer screening. Every year, radiologists examine millions of mammograms looking for tiny signs of cancer. Even with years of training, some cancers can be difficult to detect, especially in dense breast tissue or when changes are very subtle.
A large new clinical trial examined whether AI could help radiologists find more breast cancers during routine mammography screening. The results suggest that AI can improve cancer detection without increasing false alarms, showing that it may become a valuable tool for supporting doctors rather than replacing them.
How Is AI Used in Mammograms?
Artificial intelligence is designed to analyze mammogram images and identify areas that may look suspicious for cancer.
Instead of making the diagnosis itself, AI acts as a second reader, highlighting areas that deserve a closer look by a radiologist.
The goal is to:
- Detect cancers that might otherwise be missed.
- Reduce the workload for radiologists.
- Improve the accuracy of breast cancer screening.
- Help diagnose cancers at an earlier stage.
The final decision about whether a mammogram is normal or abnormal is still made by a healthcare professional.
What Did the Study Look At?
Researchers conducted the MASAI randomized clinical trial, one of the largest studies to evaluate AI-assisted breast cancer screening.
Women attending routine mammography screening were randomly assigned to one of two groups:
- Standard screening with radiologists.
- Screening that included AI support.
The researchers compared how many cancers were detected, how many women were called back for additional testing, and how much work was required from radiologists.
What Did the Researchers Find?
The study found that adding AI helped detect more breast cancers than standard screening alone.
Importantly, this improvement came without increasing the number of women called back for unnecessary follow-up testing, one of the biggest concerns about introducing AI into screening programs.
The researchers also found that AI reduced the amount of work required from radiologists by helping prioritize which mammograms needed the closest review.
These findings suggest AI could improve screening efficiency while maintaining high-quality care.
Why Could This Matter?
Finding breast cancer earlier often means:
- More treatment options.
- Smaller tumors at diagnosis.
- Less extensive surgery.
- Better long-term outcomes.
By helping radiologists identify subtle cancers that might otherwise be overlooked, AI could improve early detection for thousands of women each year.
At the same time, keeping false-positive results low is important because unnecessary follow-up tests and biopsies can cause anxiety, increase healthcare costs, and expose patients to procedures they ultimately do not need.
The MASAI trial suggests AI may help improve one side of this balance without making the other worse.
Will AI Replace Radiologists?
AI will not be replacing radiologists. The researchers emphasize that AI is designed to assist, not replace, trained radiologists.
Radiologists still evaluate the images, consider a patient’s medical history, compare previous mammograms, and decide whether additional testing is needed.
AI simply provides another layer of analysis that may help doctors identify cancers they might not otherwise notice.
As AI technology continues to improve, it may become a routine part of breast cancer screening, working alongside radiologists to improve accuracy and efficiency.
What This Means for You
This large clinical trial suggests that artificial intelligence may become an important tool for breast cancer screening. By helping radiologists detect more cancers without increasing unnecessary follow-up testing, AI has the potential to improve early diagnosis while making screening programs more efficient.
If AI-assisted mammography becomes available where you receive care, it should be viewed as an additional layer of support—not a replacement for the expertise of your radiologist. Regular mammograms remain one of the best ways to detect breast cancer early, when treatment is often most successful.
Source
- Hernström E, et al. AI-Assisted Mammography Screening: Results From the MASAI Randomized Clinical Trial. Lancet Digit Health. 2025. Available from: https://pubmed.ncbi.nlm.nih.gov/39904652/. Accessed by: https://pubmed.ncbi.nlm.nih.gov/39904652/
