Researchers Use Machine Learning to Improve Breast Cancer Screening Techniques

Imagine enduring a painful, expensive and scar-inducing surgery—only to find out afterwards that it wasn’t necessary.

This is the situation for many women with high-risk breast lesions—areas of tissue that appear suspicious on a mammogram and have abnormal but not cancerous cells when tested by needle biopsy. Following surgical removal, 90% of these lesions end up being benign.

A change in the standard of care could be on the horizon thanks to researchers at Massachusetts General Hospital and MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) who have found a more precise and less invasive way to separate harmful lesions from benign ones.

MIT-AI-Cancer-Detection-01_0.jpg
From left: Manisha Bahl, director of the Massachusetts General Hospital Breast Imaging Fellowship Program; MIT Professor Regina Barzilay; and Constance Lehman, chief of the Breast Imaging Division at MGH’s Department of Radiology. Photo courtesy of MIT News

“The decision about whether or not to proceed to surgery is challenging, and the tendency is to aggressively treat these lesions [and remove them],” said Manisha Bahl, MD, Director of the Breast Imaging Fellowship Program at Mass General, in a recent interview.

Bahl, along with a team of researchers, have harnessed the power of artificial intelligence (AI) to develop a more accurate and less invasive screening method for high-risk lesions. When tested, the machine correctly diagnosed 97 percent of 335 high-risk breast lesions as malignant and reduced the number of benign surgeries by more than 30 percent compared to existing approaches. These results were recently published in Radiology.

The team developed an AI system that uses machine learning to distinguish between high-risk lesions that need to be surgically removed from those that should just be watched over time. They created this model by feeding it data on over 600 high-risk lesions, including information on the patient’s demographics and pathology reports, and then tasked it to identify patterns among the different data elements.

Through a process called deep learning, the machine uses the data to create an algorithm that can be used to predict which high-risk lesions should be surgically removed. This process differs from traditional software programming in that the researchers did not give the machine the formula for diagnosis, but rather let it analyze the data and identify patterns on its own.

“To our knowledge, this is the first study to apply machine learning to the task of distinguishing high-risk lesions that need surgery from those that don’t,” said collaborator Constance Lehman, MD, PhD, chief of the Breast Imaging Division at Mass General’s Department of Radiology, in a recent interview. “We believe this could support women to make more informed decisions about their treatment and that we could provide more targeted approaches to health care in general.”

Lehman says Mass General radiologists will begin incorporating the model into their clinical practice over the next year.

Specialized Screening Protocol May Improve Detection of Ovarian Cancer in High-Risk Women

First, let’s define a few key words:

  • CA125: A protein found on the surface of many ovarian cancer cells. Most women with ovarian cancer have CA125 levels in the blood that are over 35.
  • ROCA: An abbreviation for Risk of Ovarian Cancer Algorithm, a test that assesses risk for developing ovarian cancer. This test tracks CA125 levels over time to identify significant elevations above each patient’s baseline levels.

What’s the issue researchers wanted to address? The standard advice for women at high risk of ovarian cancer, due to either family history or inherited gene mutations, is to have their ovaries and fallopian tubes surgically removed as a preventative measure once they are done having children. Some women choose to postpone this surgery and there’s a need for a screening test for these women that will detect the development of tumors while the cancer is still in a treatable stage.

How did they address this issue? Researchers from the Massachusetts General Hospital (MGH) Cancer Center and the Biostatistics Unit, and the National Cancer Institute and Anderson Cancer Center enrolled more than 3,800 women at elevated risk for ovarian cancer in two screening trials. Participants had blood tests every three months to establish their baseline levels of CA125 and to monitor for any changes using the ROCA assessment tool. Women at intermediate ROCA risk were referred for an ultrasound examination, while those at an elevated ROCA risk received both ultrasound and clinical evaluation.

What did they find? 19 malignant tumors of the ovaries or fallopian tubes were identified during the study periods. Ten cases were diagnosed during screening, and nine were diagnosed by preventive surgery. This protocol increased the proportion of tumors detected at early stages from 10 percent – which is typically seen in high-risk women who are not screened – to 50 percent.

What do these results mean? The combined results of the two screening trials suggest that a protocol involving quarterly blood testing to identify significant increases above each patient’s personal baseline in levels of CA125, followed by ultrasound examination when such elevations are detected, could improve the chance that tumors are detected at early stages when they are easier to treat.

What do researchers have to say about their findings? Researchers caution that surgery is the primary and best option for reducing the risk of ovarian cancer, and ROCA should only be considered as a promising but unproven option for patients who decide, against medical advice, to postpone their surgery. They plan to conduct more research to identify other ovarian cancer indicators and improved imaging technologies that may help to detect even more tumors at even earlier stages.

Steven Skates, PhD, of the Massachusetts General Hospital (MGH) Cancer Center and the Biostatistics Unit, is co-lead and corresponding author of the report. Learn more here.

Fascinating Findings from Massachusetts General Hospital: Recent Research on Alzheimer’s, Patient Navigators and the Wage Gap

By Milo Goodman
Research Institute Intern

Over the past few weeks, I’ve had the chance to read about some extremely interesting research studies that were recently conducted by investigators at Mass General.

Out of the dozens I’ve encountered, three have stood out to me in particular: research on sex-based income disparities among physicians, a study on the development of a polygenic risk score to determine one’s chances of developing Alzheimer’s and dementia, and research on the importance of patient assistance for cancer screening rates.

Continue reading “Fascinating Findings from Massachusetts General Hospital: Recent Research on Alzheimer’s, Patient Navigators and the Wage Gap”