NYC's Largest Hospital System Proposes AI-Only Cancer Screenings

Hospital chief wants to replace radiologists with AI to cut costs, but doctors warn of patient safety risks.

Apr. 1, 2026 at 2:35pm

Got story updates? Submit your updates here. ›

Mitchell Katz, who runs America's largest public hospital system NYC Health + Hospitals, has proposed using AI to handle initial reads on low-risk medical imaging like mammograms, with radiologists only stepping in for abnormal cases. Katz says this could save millions in radiologist salaries and expand screening access, but doctors warn that current AI technology is not advanced enough to safely replace human radiologists, risking patient harm and death.

Why it matters

This proposal highlights the growing tension between using AI to cut healthcare costs and the medical community's concerns about patient safety. As hospitals face tight budgets, the push to automate diagnostic imaging could have major implications for the radiology profession and access to quality cancer screening.

The details

Katz's system has already invested $224 million in GE imaging upgrades and now wants to leverage AI to handle initial reads on low-risk screenings like mammograms. Early AI performance data suggests machines may outperform humans at spotting cancer, with AI missing breast cancers in only about 3 out of 10,000 low-risk negative cases. However, radiologists warn that AI-only reads could compromise patient safety due to current technology limitations.

  • On March 25, Katz made the proposal at a Crain's panel discussion.
  • NYC Health + Hospitals has already invested $224 million in GE imaging upgrades.

The players

Mitchell Katz

The chief executive officer of NYC Health + Hospitals, which is America's largest public hospital system serving over 1 million New Yorkers annually.

David Lubarsky

A radiologist from Westchester Medical Center who shared data showing AI misses breast cancers in only about 3 out of 10,000 low-risk negative cases.

Sandra Scott

A representative from One Brooklyn Health who called AI-first reads a 'game-changer' for safety-net hospitals facing tight budgets.

Mohammed Suhail

A radiologist from North Coast Imaging who called Katz's proposal 'naive' and 'dangerous', claiming AI-only reads would cause 'patient harm and death' due to current technology limitations.

Got photos? Submit your photos here. ›

What they’re saying

“We could replace a great deal of radiologists with AI… if we are ready to do the regulatory challenge.”

— Mitchell Katz, CEO, NYC Health + Hospitals

“AI… actually better than human beings.”

— David Lubarsky, Radiologist, Westchester Medical Center

“AI-first reads [are] a 'game-changer' for safety-net hospitals squeezed by tight margins.”

— Sandra Scott, Representative, One Brooklyn Health

“Katz's view is 'naive' and dangerous, claiming AI-only reads would cause 'patient harm and death' due to current technology limitations.”

— Mohammed Suhail, Radiologist, North Coast Imaging

What’s next

New York state regulators will need to decide whether to grant exceptions to the current rules requiring radiologist oversight of medical imaging studies, which could pave the way for AI-only reads on low-risk screenings.

The takeaway

This proposal highlights the growing tension between using AI to cut healthcare costs and the medical community's concerns about patient safety. As hospitals face tight budgets, the push to automate diagnostic imaging could have major implications for the radiology profession and access to quality cancer screening.