AI provides accurate classification of breast density
OAK BROOK, Ill. (March 17, 2022) – An artificial intelligence (AI) tool can accurately and consistently classify breast density on mammograms, according to a study conducted in Radiology: Artificial Intelligence.
Breast density reflects the amount of fibroglandular tissue in the breast commonly seen on mammograms. High breast density is an independent risk factor for breast cancer, and its masking effect of underlying lesions reduces the sensitivity of mammography. Therefore, many US states have laws requiring women with dense breasts to be notified after a mammogram, so they can choose to undergo additional testing to improve cancer detection.
In clinical practice, breast density is assessed visually on two-view mammograms, most commonly with the American College of Radiology Breast Imaging-Reporting and Data System (BI-RADS) four-category scale, ranging from Category A for almost entirely fatty breasts to Category D for extremely dense. The system has limitations, because visual classification is subject to inter-observer variability, or differences in ratings between two or more people, and intra-observer variability, or differences that appear in repeated ratings by the same person.
To overcome this variability, Italian researchers have developed breast density classification software based on a sophisticated type of AI called deep learning with convolutional neural networks, a sophisticated type of AI capable of discerning subtle patterns in images beyond the capabilities of the human eye. . The researchers trained the software, known as TRACE4BDensity, under the supervision of seven experienced radiologists who independently visually assessed 760 mammogram images.
External validation of the tool was carried out by the three radiologists closest to the consensus on a dataset of 384 mammographic images obtained in a different center.
TRACE4BDensity showed 89% accuracy in distinguishing between low density (BI-RADS categories A and B) and high density (BI-RADS categories C and D) breast tissue, with 90% agreement between the tool and the three drives. All disagreements were in adjacent BI-RADS categories.
“The particular value of this tool is the ability to overcome the suboptimal reproducibility of visual human density classification that limits its practical usefulness,” said study co-author Sergio Papa, MD, of Centro Diagnostico Italiano from Milan, Italy. “Having a robust tool that offers density assignment in a standardized way can help a lot in decision-making.”
Such a tool would be particularly valuable, the researchers said, as breast cancer screening becomes more personalized, with density assessment an important factor in risk stratification.
“A tool such as TRACE4BDensity can help us advise women with dense breasts to undergo, after a negative mammogram, further screening with ultrasound, MRI or contrast-enhanced mammography,” said study co-author Francesco Sardanelli. , MD, from IRCCS Policlinico San Donato in San Donato, Italy.
The researchers are planning additional studies to better understand the full capabilities of the software.
“We would like to further evaluate the TRACE4BDensity AI tool, especially in countries where female density regulation is not active, by evaluating the usefulness of such a tool for radiologists and patients” , said study co-author Christian Salvatore, Ph.D., Principal Investigator, IUSS Pavia University School of Advanced Studies and co-founder and CEO of DeepTrace Technologies.
“Development and validation of an AI-based mammographic breast density classification tool based on the consensus of radiologists.” In collaboration with Drs. Papa, Sardanelli and Salvatore were Veronica Magni, MD, Matteo Interlenghi, M.Sc., Andrea Cozzi, MD, Marco Alì, Ph.D., Alcide A. Azzena, MD, Davide Capra, MD, Serena Carriero, MD, Gianmarco Della Pepa, MD, Deborah Fazzini, MD, Giuseppe Granata, MD, Caterina B. Monti, MD, Ph.D., Giulia Muscogiuri, MD, Giuseppe Pellegrino, MD, Simone Schiaffino, MD and Isabella Castiglioni, M.Sc., MBA
Radiology is edited by David A. Bluemke, MD, Ph.D., University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin, and owned and published by the Radiological Society of North America, Inc. (https://pubs.rsna.org/journal/radiology)
The RSNA is an association of radiologists, radiation oncologists, medical physicists and related scientists promoting excellence in patient care and healthcare delivery through education, research and technological innovation. The Company is based in Oak Brook, Illinois. (RSNA.org)
For user-friendly information on breast imaging, visit RadiologyInfo.org.
Artificial Intelligence Radiology
The title of the article
Development and validation of a mammographic tool for classifying breast density based on the consensus of radiologists