NOV 30, 2020 3:07 PM PST

Using deep learning to identify breast cancer biomarkers in mammograms

A team from Massachusetts General Hospital (MGH) has developed a deep learning model capable of predicting a patient’s risk of getting breast cancer. The team will present their findings at the annual meeting of the Radiological Society of North America (RSNA).

The model identifies the risk level by detecting imaging biomarkers on screening mammograms. According to the researchers, it shows greater accuracy than current risk assessment tools, which incorporate information such as breast biopsies, family history, and breast density, but lack other integrative data.  

“Traditional risk assessment models do not leverage the level of detail that is contained within a mammogram," said Leslie Lamb, M.D., M.Sc., breast radiologist at MGH. "Even the best existing traditional risk models may separate sub-groups of patients but are not as precise on the individual level."

"Why should we limit ourselves to only breast density when there is such rich digital data embedded in every woman's mammogram?" questioned senior author Constance D. Lehman, M.D., Ph.D., division chief of breast imaging at MGH. "Every woman's mammogram is unique to her just like her thumbprint. It contains imaging biomarkers that are highly predictive of future cancer risk, but until we had the tools of deep learning, we were not able to extract this information to improve patient care."

Photo: Pexels

To integrate all of the information provided in a mammogram, the team trained the deep learning algorithm utilizing data from five MGH breast cancer screening sites. The dataset was comprised of 245,753 consecutive 2D digital bilateral screening mammograms performed in 80,818 patients between 2009 and 2016. After training, the deep learning model scored a predictive rate of 0.71, higher than that of the traditional risk model, which scored a rate of 0.61.

"Our deep learning model is able to translate the full diversity of subtle imaging biomarkers in the mammogram that can predict a woman's future risk for breast cancer," Dr. Lamb said. "Traditional risk models can be time-consuming to acquire and rely on inconsistent or missing data. A deep learning image-only risk model can provide increased access to more accurate, less costly risk assessment and help deliver on the promise of precision medicine."

Sources: Eureka Alert

About the Author
  • Kathryn is a curious world-traveller interested in the intersection between nature, culture, history, and people. She has worked for environmental education non-profits and is a Spanish/English interpreter.
You May Also Like
NOV 18, 2020
Genetics & Genomics
Choosing an NGS workflow: What are you looking for?
NOV 18, 2020
Choosing an NGS workflow: What are you looking for?
  It’s easy to be overwhelmed by the number of workflows that are available for NGS. How do you choose? While ...
NOV 23, 2020
Cancer
Women show higher survival rates than men post lung cancer surgery
NOV 23, 2020
Women show higher survival rates than men post lung cancer surgery
New research published in the journal Chest shows that women fare better than men following lung cancer surgery. Th ...
NOV 30, 2020
Cancer
Developing Handheld Pulse Lasers to Destroy Cancer Tissue
NOV 30, 2020
Developing Handheld Pulse Lasers to Destroy Cancer Tissue
Many take modern surgeons and surgical methods for granted. In the grand scheme of things, it wasn’t too long ago ...
DEC 02, 2020
Cancer
Telomere shortening doesn't only indicate ageing...
DEC 02, 2020
Telomere shortening doesn't only indicate ageing...
While telomere shortening has been thought to be an adverse sign of aging, new research suggests that the shortening of ...
DEC 23, 2020
Cancer
Using AI to identify new measures of mammography
DEC 23, 2020
Using AI to identify new measures of mammography
New findings published in the International Journal of Cancer from the University of Melbourne report that new mamm ...
JAN 14, 2021
Clinical & Molecular DX
Tip of the Iceberg: Inaccuracies in Prostate Cancer Diagnostics
JAN 14, 2021
Tip of the Iceberg: Inaccuracies in Prostate Cancer Diagnostics
  Only 10 percent of icebergs are visible on the surface of the water; the remaining 90 percent remains submerged. ...
Loading Comments...