News

Industry leading news about artificial intelligence in healthcare and diagnostics.

New AI Technology Shows Promise in Decreasing Unnecessary Mammograms

A new AI tool from the Massachusetts Institute of Technology (MIT) may someday improve how quickly breast cancer is diagnosed while reducing the number of unnecessary mammograms.

The AI-designed exams personalize recommendations based on individual characteristics of each patient, providing a precision medicine approach to breast cancer screening. This new development may lead to improved diagnostics and more efficient care delivery, according to MIT.

In a study published on Jan. 13 in Nature Medicine, MIT researchers describe their newly-developed AI program, called Tempo. Tempo is designed to review the risk of breast cancer for individual patients based on previous mammograms, age, family history, hormonal factors, and other considerations. Tempo will then recommend when the patient should next receive a breast cancer screening. That next exam might be in just months, or it could be years away, depending on the AI-based recommendation. For some women, this new approach could reduce unnecessary mammograms.

Current breast cancer screening exams are normally scheduled on the one-size-fits-all basis that much of medicine still uses. Women ages 45 to 54 typically receive mammograms annually, then once every two years after that. One significant problem with this method is that breast cancer that develops a few months after a screening exam can go undetected for over a year and a half in patients 55 years or older.

Yala_Adam, PhD student at MIT

“Current guidelines divide the population into a few large groups, like younger or older than 55, and recommend the same screening frequency to all the members of a cohort,” said Adam Yala in an MIT news release. 

"The development of AI-based risk models that operate over raw patient data gives us an opportunity to transform screening, giving more frequent screens to those who need it and sparing the rest,” added Yala, who is a PhD student in electrical engineering and computer science and the lead researcher for this study.

Personalized Breast Cancer Screenings Can Help Better Allocate Resources

Tempo uses reinforcement learning, a machine learning method widely known for success in chess, to develop an approach that predicts a follow-up recommendation for each patient. 

Using Tempo will provide patients who have an increased risk of developing breast cancers with additional screening. It will also allow for more time between screening exams for those who are less likely to develop breast cancer. This efficiently allocates resources while personalizing screening in a way that results in better diagnostics.

“By tailoring the screening to the patient’s individual risk, we can improve patient outcomes, reduce overtreatment, and eliminate health disparities,” Yala explained. “Given the massive scale of breast cancer screening, with tens of millions of women getting mammograms every year, improvements to our guidelines are immensely important.”

One key benefit to Tempo is that it does not just recommend a certain screening pattern for an individual; rather, it modifies individualized screening patterns based on changes in their health or history.

“A key aspect of these models is that their predictions can evolve over time as a patients raw data changes, suggesting that screening policies need to be attuned to changes in risk and be optimized over long periods of patient data,” Yala said.

AI Framework is not Just for Unnecessary Mammograms

The underlying AI principles used by the MIT researchers can be applied to a variety of other conditions.

“Our framework is flexible and can be readily utilized for other diseases, other forms of risk models, and other definitions of early detection benefit or screening cost,” Yala explained. “We expect the utility of Tempo to continue to improve as risk models and outcome metrics are further refined. We’re excited to work with hospital partners to prospectively study this technology and help us further improve personalized cancer screening.”

In the future, Tempo may allow healthcare leaders to run leaner operations while still providing optimal care for breast cancer and other illnesses. By personalizing screenings, hospital and oncology leaders will be able to improve the effectiveness of screening while ensuring resources go where they are most needed.

—Caleb Williams

Related Resources:

Optimizing risk-based breast cancer screening policies with reinforcement learning

CDC Breast Cancer Screening Guidelines for Women

Seeing into the future: Personalized cancer screening with artificial intelligence

To Fight Cancer, Machine Learning Identifies DNA Repair Proteins

New Test Under Development That Detects Breast Cancer within One Hour with 100% Accuracy Has Potential to Help Pathologists Deliver More Value