🔒New Haven biotech DeepLook Medical’s AI-driven imaging software targets better breast cancer detection

This number is hard to ignore: 99% of women who receive an early breast cancer diagnosis survive the disease.

That’s what makes a new federal rule — set by the U.S. Food and Drug Administration (FDA) — that took effect on Sept. 10 so noteworthy. It requires healthcare providers to tell patients during mammography screenings how dense their breasts are, because women with dense tissue may need additional screenings to detect cancer.

While that’s a key step in the fight against the disease, the cost of those additional screenings can be prohibitive.

That concerned U.S. Rep. Rosa DeLauro (D-CT), who in May 2023 introduced a bill in Congress to address the issue. The Find It Early Act would require all health insurance plans to cover screenings and diagnostic breast imaging with no out-of-pocket costs for women with dense breasts, or who are at higher risk of breast cancer.

In September, DeLauro said the bill, which still hasn’t passed Congress, is needed “so that women do not face exorbitant out-of-pocket costs when it comes to receiving these additional screenings, or forgo them entirely because they cannot afford them.”

According to the American Cancer Society, nearly 311,000 women are diagnosed with invasive breast cancer each year, with an estimated 43,250 deaths annually.

Given those numbers, the bill has many supporters. Yet, there is another way to help women with dense breasts save money on additional screenings — by eliminating them.

Enter New Haven-based medical technology startup DeepLook Medical.

Since its founding in December 2018, DeepLook, based at 470 James St., has developed and secured two patents and FDA clearance for its medical imaging analysis software, called DL Precise.

The company, which has raised millions of dollars in venture funding, says its shape-recognition software significantly improves the accuracy and efficiency of medical imaging, and is specifically designed to detect cancer in dense breast tissue.

When it was granted its second patent in April, DeepLook stated that its software can leverage medical imaging data to isolate and display the underlying structure of masses in breasts that previously were “obscured, hidden or masked from human vision.”

CEO Marissa K. Fayer said the artificial intelligence-driven software can be used in a variety of ways, but company officials identified breast cancer imaging as a “huge niche.”

“All women about 40 (years old and up) have annual mammograms, and 45% of them have dense breasts,” she said.

Fayer called the second patent, which protects the unique radiomic data that DL Precise generates, a “paradigm shift in decision-support technologies — especially for breast imaging.”

In other words, the software makes it easier to detect breast cancer earlier, reducing or eliminating the need for more tests.

In general, Fayer said, systems like magnetic resonance imaging (MRI) or CT scans highlight an area of the breast for further study.

“They’re just circling an area,” she said. “They’re putting a bull’s-eye on it. … That’s super helpful, … but they’re not looking at the actual mass. We highlight the actual mass.”

‘It’s just genetics’

Fayer, who was named DeepLook’s CEO in September 2022, is a former executive at Hologic Inc., one of largest medical imaging companies in the U.S.

In addition to serving as CEO of a biotechnology company that has just six full-time employees — relying instead on retired industry executives and occasional contract employees nationwide — Fayer also is founder of HERHeatlhEQ, a nonprofit that raises awareness and works for equitable access to health care for women worldwide.

The issues of healthcare access, and dense breasts requiring additional scanning, actually intersect, she says, because there are “great technologies out there, but there needs to be add-ons and adjuncts to supplement that technology, because it’s not ‘one-size-fits-all’ for everyone, and especially for women with dense breasts.”

Fayer said dense breast tissue has nothing to do with body composition, meaning it doesn’t matter whether a woman is thin, physically fit or obese.

“It’s just genetics,” she said. “It’s also an equity issue, because Black women, Asian women and Jewish women just generally have dense breasts.”

Testing dense breasts for signs of cancer is like searching for an airplane in a cloud, she said. “You can’t see it, so how is a radiologist supposed to be able to visualize something that they can’t see?”

She continued, “That’s very much what we focus on, because when you look at the details on a computer basis, as opposed to just with your eyes, you’re able to show a lot more.”

Sales ramping up

DeepLook is selling its DL Precise software to hospital systems and medical imaging service providers worldwide, Fayer said, noting that it already is used by White Plains Hospital in New York’s Westchester County.

Other deals are in the works, she said.

“We’d love this to be in 500 places,” Fayer said. “It’s not an exclusive license that is only for certain hospitals.”

With sales ramping up this year, DeepLook Medical is a “pre-revenue company,” Fayer said, adding the startup expects to generate revenue in the current quarter.

That doesn’t mean it lacks funding. In three rounds of fundraising, including a Series A financing round that closed in September, the company has raised approximately $6.9 million overall.

Investors include Connecticut Innovations, the state’s quasi-public venture investment arm, and Tidal River Investments, a Stamford-based angel investment firm that supports early-stage companies with a focus on diversity and equity.

Fayer said there are plenty of incentives for health systems to use the software, because eliminating duplicative and unneeded tests produces significant savings for patients, providers and payers, and results in a faster diagnosis.

“Even if we could reduce additional imaging by just 1% — and we have initial studies that show a 12% reduction — we can save payers $302 million a year,” she said.

While that would make any insurance company CFO happy, it’s even better news for patients, Fayer said.

According to her, 80% of breast biopsies discover benign tissue, which means most of those procedures are not needed.

“There’s a ridiculous amount of imaging that happens, and so many steps that a patient has to pay for,” she said. “So, even if they have insurance, most of them are still paying a copay, and that adds up.”

In fact, the average copay is $1,400, and for those without insurance, a biopsy averages $17,000, Fayer said.

“There are new studies that have just come out that found if anything’s over $2,000, most people are just not going to do it,” she said.

42M mammograms

Fayer cites one other benefit of the DL Precise software — its ability to take some pressure off radiologists.

According to the FDA, 42.4 million mammograms were performed in the U.S. this year through Oct. 1. That’s 4.2 million per month.

Radiologists, Fayer says, are exhausted.

“They need help,” she said. “I mean, they want help and this is another tool that helps them.”

She adds that what DL Precise doesn’t do is make medical decisions.

“They’re still physicians,” she said of radiologists, “and one thing that we’ll never do is take the decisions away from them. They are the doctors. They are specially trained for this.”

As for the future, DeepLook is working to expand the types of cancers its imaging software can be used for, including lung, thyroid and liver cancers.

The company is also working on a new addition to its existing software that is expected to debut in early 2025, Fayer said.