Artificial intelligence (AI) is proving to be a powerful tool in the early detection of breast cancer, with new research showing that it can help radiologists concentrate more effectively on areas of concern during mammogram evaluations.

A recent study published in the journal Radiology revealed that radiologists using AI-assisted technology spent more time examining regions of mammograms that contained suspicious lesions—without increasing the overall time required to review the scans. This focused attention led to improved accuracy in identifying breast cancer.

Eye-Tracking Reveals Radiologist Behavior

The study, conducted by researchers at Radboud University Medical Center in the Netherlands, employed eye-tracking technology to monitor how radiologists examined mammograms with and without AI assistance. Using a device equipped with infrared lights and a central camera, researchers captured precise data on where radiologists looked on the screen and how long they focused on specific areas.

“With this eye-tracking data, we could map out the radiologists’ reading patterns and understand how AI influenced their attention,” said Jessie Gommers, the study’s co-author.

The study involved 12 radiologists analyzing mammograms from 150 women—half of whom had confirmed breast cancer. The findings showed that with AI support, radiologists spent significantly more time on lesion-containing regions, boosting cancer detection rates.

Smarter Focus, Same Efficiency

Interestingly, while the radiologists zeroed in more on potential cancer sites when aided by AI, their overall reading time didn’t increase. This suggests that AI helped optimize their workflow by signaling which cases deserved more scrutiny and which were likely benign.

“When AI rated a mammogram as low risk, radiologists moved through the case more quickly,” Gommers explained. “In contrast, high AI suspicion scores prompted a more detailed review of that specific area, especially when the cancer signs were subtle or challenging to detect.”

A Collaborative Future for AI and Radiologists

While AI clearly improved performance, the researchers cautioned against overreliance on the technology. Educating radiologists on how to interpret AI-generated data is crucial to prevent false positives, unnecessary imaging, or overlooked cancers.

“Radiologists must remain critical thinkers,” Gommers emphasized. “AI is a valuable tool, but the final responsibility should always rest with the medical professional.”

The team is now working on ways to gauge when AI is uncertain in its assessments, which could further guide radiologists in their decision-making process.

Looking Ahead

This study highlights AI’s growing potential to refine diagnostic accuracy and efficiency in breast cancer screening. By helping radiologists focus on the most relevant parts of mammograms, AI not only improves detection but also streamlines workflow, making routine screenings more effective.

As Gommers puts it: “AI is not replacing radiologists—it’s making them even better.”