US Researchers Develop Deep Learning AI Model to Predict Breast Cancer
In a significant advancement for breast cancer risk assessment, researchers have developed a deep learning model named AsymMirai. This model leverages mammographic data to predict the likelihood of developing breast cancer within one to five years.
By focusing on asymmetries between the left and right breast tissues, AsymMirai offers a more interpretable and simplified approach compared to previous models.
1 in 8 women, or approximately 13% of the female population in the U.S., will develop invasive breast cancer in their lifetime and 1 in 39 women (3%) will die from the disease, according to the American Cancer Society.
The development of AsymMirai involved analyzing over 210,000 mammograms from more than 81,000 patients. The model’s performance was found to be nearly equivalent to that of the state-of-the-art Mirai model, highlighting the clinical significance of breast tissue asymmetry as a marker for cancer risk.
The interpretability of AsymMirai’s predictions enhances its potential utility in clinical settings, providing radiologists with a valuable tool for early detection and personalized screening strategies.
“Mirai is a black box—a very large and complex neural network, similar in construction to ChatGPT—and no one knew how it made its decisions,” said Jon Donnelly, a PhD Computer Science student at Duke University in Durham, NC. “We developed an interpretable AI method that allows us to predict breast cancer from mammograms 1 to 5 years in advance. AsymMirai is much simpler and much easier to understand than Mirai.”
This innovation aligns with broader efforts in the medical community to integrate artificial intelligence into diagnostic processes. By improving the accuracy and reliability of breast cancer risk predictions, models like AsymMirai can contribute to more tailored patient care and potentially influence guidelines for mammogram frequency, ultimately aiming to improve patient outcomes through earlier intervention.
Source:
https://www.rsna.org/news/2024/march/deep-learning-for-predicting-breast-cancer
For more information, you can access the Radiology study, “AsymMirai: Interpretable Mammography-based Deep Learning Model for 1–5-year Breast Cancer Risk Prediction,” and the related editorial, “Launching a New Era of Artificial Intelligence in Health Care.”