AI changes future of breast cancer detection

This paid piece is sponsored by Sanford Health.

In the past few months, it has been nearly impossible to avoid talking about artificial intelligence. Whether it’s writing papers for college students, generating art for public murals or just keeping someone company, AI is weaving its way into everyday life.

For some, it’s exciting. For others, it’s uncomfortable.

But when it comes to scientific advances, there seems to be less hesitation.

It’s hard to argue that lifesaving technology developed through AI is anything but positive. And while AI might feel new for the general population, at Sanford Health, AI has been assisting medical professionals for years.

Developing the algorithm

In early 2020, Dr. Chris Johansen, a radiologist at the Edith Sanford Breast Center, and his team kept running into a discouraging situation: Women younger than age 40, the recommended age to start breast cancer screenings, would find a lump in their breast, come in for testing and be diagnosed with advanced cancer. And while their medical records sometimes indicated that they might be at an elevated risk of developing breast cancer, there was no process in place for connecting those dots.

It would have been nearly impossible to have humans sort through the files, and Johansen knew this. So he reached out to Austin Amdahl, an advanced business intelligence analyst at Sanford Health, for help developing a system of filters, or an algorithm, that could sift through information from previous doctors’ visits to look for keywords that would flag the appropriate indicators for someone at elevated risk.

Larissa Risty, a certified genetic counselor, was instrumental in assisting Johansen with outlining which risk indicators to include.  If a patient has a mother who was diagnosed with breast cancer at a young age, assessing their risk is straightforward. But it gets more complicated if the relative is a sister or an aunt. And then, it’s even more complicated when age of diagnosis, lifestyle choices, genetic traits and dense breast tissue are factored in.

But the AI tool Amdahl, Risty and Johansen were building was programmed to consider all of that and more.

Complementary approach

It does take more than just an algorithm to make a difference.

AI would not be nearly as impactful if there weren’t highly trained professionals to receive the information and take action.

When the Edith Sanford Breast Center team uses the AI algorithm and identifies a patient who may be at elevated risk, staff connect with the primary care provider and notify the patient. The patient is encouraged to schedule an appointment at the Breast Specialty Clinic inside the Edith Sanford Breast Center, which offers comprehensive, preventive care to people with an elevated risk of developing breast cancer.

Appointments at the Breast Specialty Clinic include a consultation with a specialized breast health provider and a certified genetic counselor to talk through preventive care, including earlier or more frequent screenings.

The clinic providers coordinate with fellowship-trained breast radiologists to establish an individualized breast imaging care plan. This collaborative effort, paired with the data generated by AI, allows the specialized team to improve breast cancer screening for each patient.

Improved outcomes

Technology advances in health care are exciting. Anything that can minimize illness and potentially keep loved ones around longer should be considered a major win.

When it comes to breast cancer care, AI is just one part of this. But it’s not a coincidence that in the years since this algorithm was developed, outcomes have improved.

In 2022, over 80 percent of patients diagnosed with breast cancer across Sanford Health had their cancer detected in the earliest stages. That number outpaces national averages by 15 percent.

With years of evidence to draw from, one thing is clear: AI paired with skilled health care professionals is saving lives in our community.