How AI is transforming breast cancer detection: a game-changing tool for early diagnosis and personalized care

If you had an assistant who helped you do your work faster, more efficiently, and more precisely, would you fear it?

With artificial intelligence (AI) at the forefront of virtually everything today, two of the most common questions I hear in my daily practice are, “Do you use AI?” and “What does AI do for you?” For some, these questions stem from a desire to access cutting-edge tools, while others express concerns about its rapid advancements.
The evolving capabilities of AI raise concerns across many fields. But as a breast cancer radiologist who has used AI since 2019, I can confidently say that AI is improving breast cancer detection and care every day.

Mammography technology has come a long way in recent decades, and with advancements like 3D imaging, mammograms remain a necessary and important tool in cancer detection. Yet, while some cancers are identifiable on mammograms despite dense tissue, others are not. As health care experts, we would be remiss in not utilizing a tool capable of closing gaps in diagnostic care. This tool is AI.

Detection. The gold standard of breast cancer prevention, mammograms are vital to a woman’s health, but like all imaging and screening processes throughout health care, nothing is foolproof. AI is a tremendous addition to mammography, helping radiologists identify breast cancer that might otherwise go undetected, especially in early stages or in people with dense breasts. For example, breast density can impact the timely detection of tumors—so much so that in September 2024, the federal government passed a law requiring U.S. mammogram providers to notify all patients of their density types. Normal fatty breast tissue tends to appear dark in a mammogram image, while normal dense fibro-glandular tissue appears white, making cancers harder to spot. Every woman has both types of tissue, but with varying proportions and patterns. AI helps serve as a “second set of eyes” and safety net for radiologists interpreting these images.

Improving image quality. AI is also becoming more helpful in creating clearer mammogram images. We are now using advanced algorithms to enhance the resolution and clarity of mammograms, as well as their interpretation. By reducing “background shadows,” for example, AI makes images easier to read. This means an increased ability to detect breast cancer at its earliest stages and added confidence in evaluating normal mammograms.

Predicting risk. Studies have shown that AI has the ability to augment and even surpass our current ability to predict a person’s long-term risk of breast cancer based on information gathered from the pixels in a mammogram image. This can help identify people who might benefit from additional screening or testing while minimizing unnecessary testing for others.

Characterizing and assessing tumors. AI can help identify tumor subtypes and assess how breast cancer responds to chemotherapy. This work complements the expertise of radiologists and oncologists and can help specialists plan targeted therapy and follow-up care.

When a radiologist reviews a mammogram, AI acts as an assistant by highlighting suspicious areas, providing clarity and direction. For images without such findings, it reinforces confidence in a normal result. AI does not replace the duties performed by a radiologist. It enhances them.

As the impact of AI in health care continues to evolve, radiology is on track to become more efficient, precise, and comprehensive. From my vantage point, the future looks bright.

January Lopez is Medical Director of Breast Imaging, Hoag Memorial Hospital Presbyterian, Newport Beach, CA.

Hoag is a nonprofit, regional health care delivery network in Orange County, California, that treats more than 30,000 inpatients and 480,000 outpatients annually. Hoag consists of two acute-care hospitals – Hoag Hospital Newport Beach, which opened in 1952, and Hoag Hospital Irvine, which opened in 2010 – in addition to nine health centers and 13 urgent care centers.