"We know starting screening mammography at age 40 saves lives," Lewin explained. "It reduced breast cancer mortality by about 30%. How can we move the needle higher up? One thing is to get everyone to screening."

One of the barriers to increasing screening is that some ethnic and racial minorities and rural populations have limited access to mammography screening programs, which has led to healthcare disparities.
  
Another way to boost the screening rate and save more lives is to identify people who have risk factors that increase their chances of getting cancer and target them for additional screening. This includes women with family history of breast cancer, women with genetic markers associated with increased breast cancer risk and women with dense breast tissue.

For women with dense breasts, it is difficult to detect cancers because the dense tissue appears as white on mammograms, and so do cancers. He said the dense tissue can mask smaller cancers, so these women may need supplemental breast screening with breast ultrasound, contrast enhanced mammography or breast MRI. 

"Using screening ultrasound in addition to mammography, we will find two additional cancers per 1,000 women on average, and that is on top of the three to four we typically find from mammography alone," Lewin explained. "That is fine, but if you add contrast breast MRI, you increase that number to 15-19 more cancers. Most of these additional cancers will be invasive and identified because of the cancer's increased blood supply. And with MRI, it is not really finding more cancers, we are just finding cancers earlier when they are smaller and more easily treated." 

While MRI is very effective, it is also a longer exam time and more expensive, and it is not always covered by insurance.

"We know that if we screened everyone using MRI we would save more lives, but we have never proven this in a randomized trial. So what we do instead is to screen women at higher risk using MRI, and we know anecdotally that we do find cancers that we would not have found for years using standard methods," Lewin explained. 

There have been questions in the breast imaging field over when women should start screening and how often they need to get a screening. Lewin said women at higher risk should screen earlier and more often, while some women may be fine with screenings every other year.

"There will be women at the very low end of the risk spectrum who truthfully do not need a yearly mammogram, but there are women at the middle and high ends who need more than a yearly mammogram," Lewin said. "This is what we call personalized screening and that is a place we are going to go one way or another. We now have more tools available to assess a women's risk, including genetic tools and AI analysis of their mammograms."

Lewin also explains some of the issues associated with artificial intelligence (AI), but adds that AI is already being used across radiology and in breast imaging and it will get better with time. He believes we will see a lot more use of AI in the coming years, especially for first pass views to flag exams that radiologists need to spend more time on. FDA-cleared AI already exists to help identify suspected cancers and flag them to outlines and a percentage of diagnostic certainty, or a color coded heat map to show areas of interest on the image.