Radiomic imaging features extracted from digital mammography associated with breast cancer subtypes

Researchers used machine learning techniques to confirm that radiomic imaging features of breast tumors extracted from digital mammography are associated with breast cancer subtypes, according to a new study published in Academic Radiology. Previous research has demonstrated that radiomic imaging features derived from breast MRI examinations are associated with breast cancer subtypes, but digital mammography represents a cheaper, more widely available option.

The authors used imaging data from more than 300 women diagnosed with invasive breast cancer from August 2015 to October 2015. All cancers were detected using full-field digital mammography and more than 76 percent were palpable. A breast imaging specialist manually outlined the tumor in each image and automated computer programs determined a set of 39 radiomic features. Machine learning was used for classification.

Overall, the team had the best results classifying subtypes by using a combination of mediolateral oblique view and craniocaudal view images. Accuracy ranged from more than 74 percent to more than 79 percent.

Since breast cancer subtyping helps healthcare providers determine the best plan for treatment—luminal tumors respond well to endocrine therapy, for instance, while other tumors respond better to targeted antibody therapy—these findings could make a significant impact on patient care, the authors explained.

“Mammogram images are the most commonly available examination for breast cancer screening and diagnosis,” wrote lead authors Shandong Wu, PhD, with the departments of radiology, biomedical informatics, and bioengineering at the University of Pittsburgh, and Peifang Liu, PhD, at the Tianjin Medical University Cancer Institute and Hospital in Tianjin, China, and colleagues. “If the automated radiomic features like we identified in this study are validated to be predictive of the molecular subtypes, it can provide further information from the images to aid radiologists in mammographic reading and to better inform clinical diagnosis and decision-making. This would have important additional value too for patients who do not have a breast MRI scan available.”