Amid a scientist shortage, AI is being used to scan for diseases

In this article:

The U.S. is facing a doctor shortage that is getting worse as an aging population of Baby Boomers lives longer, increasing the demand for medical professionals. But doctors are not the only ones feeling the pinch. Pathologists, clinical professionals who study disease, have also been hit hard, with an overall decline in professionals from 2007 to 2017.

“With many senior pathologists expected to retire in the coming years, a ‘pathologist gap’ is likely to increase through 2030,” according to a 2018 study by the National Center for Biotechnology Information.

David West, co-founder and CEO of digital startup Proscia, said his company is hoping to help pathologists use their time more efficiently.

“The pathologist workforce is declining by 12% every decade, while biopsy volume is increasing. So the pressure on the laboratories is very intense,” resulting in increased turnaround time for lab results, West told Yahoo Finance.

The Philadelphia-based company recently teamed up with Johns Hopkins School of Medicine to develop applications that incorporate artificial intelligence (AI) to advance the practice of pathology for multiple diseases.

The platform helps pathologists scan samples faster and evaluate those that need a scientist’s attention, versus those with negative results, West said. With AI, the technology can learn what to scan for, find areas in need of attention, and flag them for the pathologist.

“Most of what the pathologist is doing is spending time looking for that needle-in-a-haystack instance of tumor in a case that’s otherwise benign,” West said.

Johns Hopkins director of pathology informatics, Dr. Alexander Baras, told Yahoo Finance the partnership is groundbreaking.

“We’re generating new data that doesn’t exist through this collaboration,” he said. “Right now, the lion’s share of diagnostic knowledge is done with microscope and slides, not by digitizing the slides.

AI (Artificial Intelligence) concept.
AI (Artificial Intelligence) concept.

Target market

Founded in 2014, Proscia has primarily focused on, similar to IBM’s Watson, finding more efficient ways to treat cancer, but has also pivoted to work on other specialties.

The Philadelphia-based company’s target market right now is Europe and the Middle East. West said that is because those areas are at least two years ahead of the U.S. in terms of digital pathology options.

“A lot of people are surprised about that. But the crisis in pathology shortage is much more intense in those geographies,” West said.

According to a study from London’s Royal College of Pathologists, only 3% of U.K. pathology labs have enough pathologists.

“We’ll start in Europe because it’s the most mature of all these geographies,” West said. “We’re seeing Asia grow very quickly too.”

The company plans to eventually join the health care economy back home in the U.S., where West hopes to pitch insurance companies to reimburse it for the reads, similar to the way pathologists are paid today.

Failures of the past

AI’s role in health care has been heavily criticized in recent years, stemming from the lack of breakthroughs from more established companies.

But health providers are still working with the promising technology. Besides Johns Hopkins, other big players, like Geisinger Health, are also working with AI products.

Memorial Sloan Kettering also recently announced its spinoff, Paige, which it also works exclusively with, raised $45 million in Series B funding.

Building out an AI platform is just one step, West says, because the data sets that are used can create an unintentional bias, even in the machines. Paying attention to the quality of the data is important, and ensuring enough variety to provide accurate results is key.

“It’s where AI has always failed,” West said. “The scientific community and industry is just now beginning to think about the scalability challenges. So that’s where we’ve focused almost all of our work.”

He said Proscia is working to improve on the failures of the past in order to stand up to the regulatory barriers, when the time comes.

“It’s crucial to make sure that AI actually does translate to improved diagnostics and improve economics and better patient outcomes,” West said. “We need that because we don’t have enough pathologists to keep up with the volume today.”

Despite the setbacks of the past, the digital health industry overall continues to explore ways to use AI, and West believes the trend will only continue,

“We’re sort of at the tip of the iceberg with seeing how these actually manifest, how AI applications can manifest in better patient outcomes,” he said.

“And this is a trend that we’re going to see for the next five, 10 years. We’re just beginning that from an economic standpoint.”

Anjalee Khemlani is a reporter at Yahoo Finance. Follow her on Twitter: @AnjKhem

Read the latest financial and business news from Yahoo Finance

Follow Yahoo Finance on Twitter, Facebook, Instagram, Flipboard, SmartNews, LinkedIn, YouTube, and reddit.

Read more:

Express Scripts launches new digital formulary aimed at lowering health costs, increasing access

Study: The health sector is splurging on digital, but not reaping the rewards just yet

Advertisement