Your last mammogram required a radiologist’s trained eye to spot potential cancer—but NYC’s hospital chief wants AI reading those images instead. Mitchell Katz, who runs America’s largest public hospital system, just made a bold proposal that could reshape how millions get medical imaging.
The $224 Million Bet on Machine Vision
NYC Health + Hospitals pushes AI-first diagnostics to cut costs and expand access.
Katz oversees 11 hospitals serving over a million New Yorkers annually through NYC Health + Hospitals, and he’s ready to make a dramatic shift. “We could replace a great deal of radiologists with AI… if we are ready to do the regulatory challenge,” he declared at a March 25 Crain’s panel. His system already invested $224 million in GE imaging upgrades—now he wants AI handling initial reads on low-risk screenings like mammograms, with radiologists stepping in only for abnormal cases.
The Numbers Look Compelling
Early AI performance data suggests machines might outperform humans at spotting cancer.
David Lubarsky from Westchester Medical Center dropped some eye-opening statistics: their AI misses breast cancers in only about 3 out of 10,000 low-risk negative cases. “AI… actually better than human beings,” he reported. Sandra Scott from One Brooklyn Health called AI-first reads a “game-changer” for safety-net hospitals squeezed by tight margins. The math seems straightforward—fewer radiologist salaries, faster throughput, broader screening access.
Doctors Push Back Hard
Radiologists warn that cost-cutting could compromise patient safety.
Not everyone’s buying the AI revolution pitch. Mohammed Suhail from North Coast Imaging called Katz’s view “naive” and dangerous, claiming AI-only reads would cause “patient harm and death” due to current technology limitations. He accused hospital administrators of prioritizing costs over safety—a familiar tension in healthcare’s ongoing automation debates. The pushback mirrors broader professional anxiety about AI displacement, from legal research to creative work.
Regulatory Reality Check
New York’s rules currently require radiologist oversight, but that could change soon.
The real battle isn’t technological—it’s regulatory. Current New York state rules mandate radiologist review of imaging studies, but Katz and his peers are pushing for exceptions that would allow AI-only reads on normal cases. This echoes recent claims from tech leaders about AI’s radiology potential, though critics argue such statements overstate current capabilities. If NYC succeeds in changing regulations, expect other states to follow quickly.
Whether you trust machines with your cancer screening depends partly on whether you think hospital accountants or radiologists better understand acceptable risk.
