A series of studies, starting as a steady drip and quickening to a deluge, has reported the same core finding amid the global spread of Covid-19: Artificial intelligence could analyze chest images to accurately detect the disease in legions of untested patients.

The results promised a ready solution to the shortage of diagnostic testing in the U.S. and some other countries and triggered splashy press releases and a cascade of hopeful headlines. But in recent days, the initial burst of optimism has given way to an intensifying debate over the plausibility of building AI systems during an unprecedented public health emergency.

Unlock this article by subscribing to STAT Plus and enjoy your first 30 days free!

GET STARTED

What is it?

STAT Plus is STAT's premium subscription service for in-depth biotech, pharma, policy, and life science coverage and analysis. Our award-winning team covers news on Wall Street, policy developments in Washington, early science breakthroughs and clinical trial results, and health care disruption in Silicon Valley and beyond.

What's included?

  • Daily reporting and analysis
  • The most comprehensive industry coverage from a powerhouse team of reporters
  • Subscriber-only newsletters
  • Daily newsletters to brief you on the most important industry news of the day
  • Online intelligence briefings
  • Frequent opportunities to engage with veteran beat reporters and industry experts
  • Exclusive industry events
  • Premium access to subscriber-only networking events around the country
  • The best reporters in the industry
  • The most trusted and well-connected newsroom in the health care industry
  • And much more
  • Exclusive interviews with industry leaders, profiles, and premium tools, like our CRISPR Trackr.
  • As a CT technologist, I can say with a high degree of confidence that relying heavily on radiology for this disease will likely increase the spread. We can clean and clean and clean, but inevitably someone will cross paths with this pathogen who would not have otherwise. When a stroke patient comes in to my facility by EMS, they’re basically pounding at the doors with ED and Neuro staff before the patient is even taken off the gurney. The same thing happens when a stroke patient comes down to us from the Neuro ICU for their daily head scan follow up. It feels rather foolish to send these types of patients to the same room minutes apart from one another. Scanning vented patients is also very time consuming for floor nurses, Respiratory staff, and CT staff. CT could be viable as a study tool on a limited basis, but most CT departments are not big enough to handle scanning an increase that this article appears to be suggesting, at least not while sustaining the other exams we have coming to us. We cannot order more scanners the way that hospitals order more ventilators. A new vent rolls in the front door. A new CT scanner requires a construction crew.

  • The machine learning experts will know if their system is any good or not. The methods for validation are very well defined at this point.

  • The idea that CT-scans should be widely used to diagnose Covid-19 is ridiculous. The increased risks from radiation would not justify such a use, especially when the primary goal is the health of the patient. One chest CT-scan is equivalent to about 1,000 chest x-rays. The technology is overused throughout the US health care system, and probably elsewhere.

  • I’m not in any way knowledgeable in this area, but initially it seems like the x-ray route would be far more reasonable than anything with CT scans – a cheaper test and with less radiation exposure.

Comments are closed.

A roundup of STAT’s top stories of the day in science and medicine

Privacy Policy