Researchers have introduced a new, interpretable model designed to predict five-year breast cancer risk from mammograms, according to a study recently published in Radiology.

Breakthroughs in AI are reshaping cancer detection and treatment. Scientists report that AI technology can spot breast cancer up to five years before it would typically be diagnosed, leading to earlier detection and more personalized care.

While traditional mammography has its limitations, AI systems like INSIGHT MMG can scan mammograms and detect future cancer risks, even when no visible signs are present. A study from Norway showed that AI could accurately forecast cancers developing as far as six years later.

In Germany, the use of AI boosted detection rates by 17.6% without increasing the number of false positives. AI is also driving advances in predictive medicine, helping physicians better anticipate cancer risks and how patients may respond to treatments.

At Weill Cornell Medicine, researchers found that AI-assisted liquid biopsies could detect cancer recurrence months or even years before traditional testing methods. Despite ongoing challenges such as potential biases and regulatory issues, AI’s impact on cancer care is significant.

AI is helping to usher in an era of precision oncology, including innovations like personalized cancer vaccines. Although obstacles remain, AI continues to offer hope for earlier diagnoses and better outcomes for breast cancer patients.

Study Findings Could Influence Mammogram Schedules For this study, researchers examined 210,067 mammograms from 81,824 patients in the Emory BrEast imaging Dataset (EMBED), collected between January 2013 and December 2020, using both the Mirai and AsymMirai models. They found that their simplified deep learning model performed nearly as well as the advanced Mirai model when predicting one- to five-year breast cancer risk.

Their findings also emphasized the clinical significance of breast asymmetry, suggesting that bilateral dissimilarity could become a key imaging marker for breast cancer risk in the future.

Because AsymMirai’s predictions are easy to interpret, it could serve as a useful companion to radiologists in diagnosing breast cancer and assessing risk, said Donnelly.

“We can, with surprisingly high accuracy, predict whether a woman will develop cancer in the next 1 to 5 years based solely on localized differences between her left and right breast tissue,” he explained. “This could have public impact because it could, in the not-too-distant future, affect how often women receive mammograms.”