AI

Cancer diagnosis AI just got a green light from the FDA

Paige, a digital diagnostics startup, got a preliminary go-ahead for its prostate cancer detection software.
article cover

Paige

· less than 3 min read

Keep up with the innovative tech transforming business

Tech Brew keeps business leaders up-to-date on the latest innovations, automation advances, policy shifts, and more, so they can make informed decisions about tech.

A cancer-diagnosis AI tool received a preliminary green light from the FDA.

Last week, Paige Prostate, an AI-based software for detecting prostate cancer, became the first product of its kind to receive FDA clearance. The software was created by an NYC-based digital diagnostics startup called Paige, which has raised $220 million in funding since its 2018 founding.

Digital diagnostics

The FDA’s decision to grant De Novo clearance came in part after a clinical study in which 16 pathologists used the program to analyze tissue slides—both cancerous and benign—from more than 150 institutions in different regions.

  • Paige’s software helped improve cancer detection by more than 7%, Fierce Biotech reported—and decreased both false-negative diagnoses (by 70%) and false-positive diagnoses (by 24%).

Keep in mind: A De Novo clearance is an alternate FDA pathway to classify novel, low- or moderate-risk medical devices. That means there’s “reasonable assurance of safety and effectiveness for the intended use,” even though there’s no “legally marketed” precedent.

  • Last year, both Paige’s prostate software and breast-cancer software received CE marks (similar to US FDA approval, but in Europe).

A New York Times report from 2018 noted the company’s “exclusive” relationship with Memorial Sloan Kettering Cancer Center and some of its digital images and diagnostic data. MSKCC disputed some of the claims.

  • The cancer center itself, as well as at least one board member, hold an equity stake in Paige.

Big picture: Paige’s FDA clearance is big news for other AI companies that specialize in digital diagnostics, but it raises thorny questions, too. For one, when it comes to training algorithms with medical records, is stripping data of patient identifiers enough? And conversely, does it hold the industry back if and when companies like Paige have exclusive access to critical training data?—HF

Keep up with the innovative tech transforming business

Tech Brew keeps business leaders up-to-date on the latest innovations, automation advances, policy shifts, and more, so they can make informed decisions about tech.