Early detection is critical when it comes to fighting cancer. A new tool developed by MIT can anticipate the development of breast cancer in women up to five years in advance.
In a recent blog post, MIT announced the collaboration between MIT’s Computer Science and Artificial Intelligence Lab and Massachusetts General Hospital. The team has worked together to develop a new deep-learning model that can predict from a mammogram if a patient is likely to develop breast cancer as many as five years in the future.
This new diagnostic tool is based on over 60,000 patients’ mammograms and known outcomes, discovering subtle patterns in breast tissue that are precursors to malignant tumors. It has already placed 31 percent of cancer patients in the high-risk category, compared to 18 percent from traditional modes of diagnosing.
Because this tool is based on actual hard data, rather than suggestive behaviors, it allows for a more comprehensive approach that is free of bias. This means that it is equally successful in predicting a diagnosis in women of color as in caucasian women—and it is also risk-based and not age-based as in previous models.
MIT and MGH hope that this new tool can lead to accurate, early detection of breast cancer for more women and that it might be applied to other types of diseases that face similar detection hurdles.
Featured photo: DarkoStojanovic via Pixabay