There are more than 3.5 million breast cancer survivors in the United States, with 268,600 new cases expected to be diagnosed this year alone. With an estimated 90% of United States adults online and 81% owning smartphones, digital health technologies are uniquely situated to bridge the gap in breast cancer care through detection, intervention, and management. Here are some of the most promising digital health solutions coming from Massachusetts to promote early breast cancer detection and improve patient care for those undergoing treatment or in remission.
Imagine Elise, a 40-year-old mother who discovers a lump on her breast, makes an appointment with her physician who takes a sample and sends it out to a pathologist. The pathologist must make a critical decision: Does Elise have breast cancer or not? Path AI, founded in Boston, MA, is utilizing artificial intelligence and machine learning technologies to improve both the accuracy and speed of pathologist diagnoses and ensure patients get the right diagnosis and the most effective treatment.
Breast cancer surveillance data reports that conventional mammography accurately identifies roughly 87% of women who truly have breast cancer. But what about the other 13%? These women are either devastated by a false positive—often leading to overdiagnoses and overtreatment when it is not warranted—or blindsided by a false negative—delaying treatment until a later stage, significantly increasing treatment severity and decreasing survival rates. With innovations like PathAI, these misdiagnoses—both false positives and false negatives—can be lessened, improving the lives of the 13% one diagnosis at a time.
Did you know the Massachusetts Institute of Technology’s Computer Science and Artificial Intelligence Lab developed a new deep learning-based AI prediction model that can anticipate the development of breast cancer up to five years in advance? This innovative technology, trained on over 90,000 mammograms and 600,000 patient outcomes, can accurately predict over 30% of all cancer patients in the highest-risk category compared to the 18% detected by current models.
In developing its technique, MIT sought to address disparities in detection inequality among minorities. Black women are 42% more likely than white women to die from breast cancer, a statistic driven largely by the lack of minority representation in current early detection techniques. By creating a model that assesses health risks more accurately for minorities, who are often not well represented in the development of deep learning models, MIT is bridging that gap and working to save all women.
Newton-based Mammosphere has created a breast imaging and cancer prevention application that makes it easy for patients to take control of their health records and transfer them to providers with the click of a button. One in four patients fail to gather prior records in a timely manner, increasing the risk of being called back for additional testing or receiving a false positive, a problem that is compounded when breast health records aren't always readily available across health care systems.
Mammograms conducted without prior images for comparison lead to thousands of false positives every year in the U.S. driving up patients’ chances of receiving an accurate diagnosis and increasing health care costs. In honor of Breast Cancer Awareness Month, Mammosphere is offering this service completely free of charge for Massachusetts women.