New York: Artificial Intelligence (AI) algorithms can outperform the standard clinical risk model for predicting the five-year risk for breast cancer according to a large study of thousands of mammograms.
A woman’s risk of breast cancer is typically calculated using clinical models such as the Breast Cancer Surveillance Consortium (BCSC) risk model, which uses self-reported and other information on the patient — including age, family history of the disease, whether she has given birth, and whether she has dense breasts — to calculate a risk score.
The study, published in the journal Radiology, showed that AI could make strong predictive performance over a five-year period.
AI was also able to identify both missed cancers and breast tissue features that help predict future cancer development.
“Something in mammograms allows us to track breast cancer risk. This is the ‘black box’ of AI,” said lead researcher Vignesh A. Arasu, a research scientist and practising radiologist at Kaiser Permanente Northern California.
When evaluating women with the highest 10 percent risk as an example, AI predicted up to 28 percent of cancers compared to 21 percent predicted by BCSC.
“We’re looking for an accurate, efficient, and scalable means of understanding a women’s breast cancer risk,” Dr. Arasu said.
“Mammography-based AI risk models provide practical advantages over traditional clinical risk models because they use a single data source: the mammogram itself.”
For the study, Dr. Arasu included 13,628 women with negative (showing no visible evidence of cancer) screening on 2D mammograms in 2016.
In addition, 4,584 patients who were diagnosed with cancer within five years were also studied. All the women were followed until 2021.
Using the 2016 screening mammograms, risk scores for breast cancer over the five-year period were generated by five AI algorithms, including two academic algorithms used by researchers and three commercially available algorithms.
The risk scores were then compared to each other and to the BCSC clinical risk score.
“All five AI algorithms performed better than the BCSC risk model for predicting breast cancer risk at 0 to 5 years,” Arasu said.
Some of the AI algorithms also excelled at predicting patients at high risk of interval cancer, which is often aggressive and may require a second reading of mammograms, supplementary screening or short-interval follow-up imaging.
Even AI algorithms trained for short time horizons (as low as 3 months) were able to predict the future risk of cancer up to five years when no cancer was clinically detected by screening mammography.
When used in combination, the AI and BCSC risk models further improved cancer prediction.
A person’s future risk score, which takes seconds for AI to generate, could be integrated into the radiology report shared with the patient and their physician.
“AI for cancer risk prediction offers us the opportunity to individualise every woman’s care, which isn’t systematically available,” Arasu said. “It’s a tool that could help us provide personalised, precision medicine on a national level.”