The role of artificial intelligence (AI) is transforming breast cancer screening and methods to lower risks, as highlighted in a recent review published by Cell Press on December 12 in the journal *Trends in Cancer*.
AI is influencing the future of breast cancer screening and preventative strategies, as discussed in a recent review article released by Cell Press on December 12 in the journal Trends in Cancer.
Erik Thompson, a senior study author from the Queensland University of Technology in Brisbane, Australia, states, “We explore the latest developments in AI-based breast cancer risk evaluation, the implications for upcoming screening and prevention efforts, and the essential research required to transition mammography findings from the lab into practical use.”
In mammograms, breast tissue that appears white is known as dense, while dark tissue is classified as non-dense. It’s broadly recognized that women with greater mammographic density—considering their age and body mass index—face an elevated risk of breast cancer. Moreover, higher density can obscure breast cancer detection during mammograms, a phenomenon often called the “masking effect.”
Global advocacy groups are urging that women be informed about their mammographic density, leading to policy updates in countries like the U.S., Canada, and Australia. In certain regions, mammographic density is influencing the adoption of supplementary imaging methods, such as ultrasound and MRI, which have shown to enhance cancer detection rates for individuals with very dense breast tissue. However, the complexity of the masking effect, the breast cancer risk linked to mammographic density, and the best ways to modify clinical practices remain challenging for researchers and healthcare providers.
To forecast possible future breast cancer cases, advanced techniques including deep learning are being employed to examine mammographic images. Notably, AI is revealing specific mammographic characteristics that may serve as stronger indicators of breast cancer risk than any previously known factors. These characteristics could also clarify much of the relationship between mammographic density and breast cancer risk. The identification of these AI-developed risk-predictive mammographic features opens up new avenues for determining which women are at a heightened risk of developing breast cancer and distinguishing them from those at risk of missed diagnoses due to the masking effect.
“A woman displaying mammographic features that indicate a high risk of breast cancer detection might benefit from more regular screenings or risk-lowering medications,” explains Thompson. “Conversely, a woman assessed with a low probability of a breast cancer diagnosis over the next five years could have extended intervals between screenings. Furthermore, a woman with dense breast tissue who does not exhibit high-risk features could potentially benefit from additional imaging procedures like MRI or ultrasound.”
Research indicates that some AI-identified mammographic traits may signify early-stage malignancies that traditional radiologist-reviewed mammograms can overlook, while others could point to benign conditions that nonetheless increase breast cancer risk. However, the classification of certain AI-generated mammographic features, which do not clearly indicate cancer or a benign state, remains uncertain.
Thompson emphasizes, “It is vital to uncover the pathobiological links associated with mammographic features and the fundamental mechanisms connecting them to breast cancer development. This will be crucial for assessing their significance for both short- and long-term breast cancer risk and forthcoming strategies to minimize that risk.”
