A hammer in search of a nail
🤔 Why is Artificial Intelligence of COVID-19 Imaging considered ‘A Hammer in Search of a Nail’ by a radiologist of the Radiological Society of North America ?
📰 The current situation :
- On January 15 2021, Facebook AI open sourced new algorithms to predict the need of oxygen based on a single X-ray using NYU Langone Health database
- There are more than 500 such manuscripts on arXiv, showing a clear appetite for the subject despite its repetitiveness and unclear path to clinical utility
- However, chest CT (and even less a chest X-ray) is not recommended as a first-line test by the American College of Radiology
- We are also beginning to understand some of the non-pulmonary manifestations of COVID-19 that can lead also to death
🏥 “It is time to move beyond studies showing that AI can detect opacities at CT or chest radiography—this is now well established. Instead, there is a great need for AI systems, based on a combination of imaging, laboratory, and clinical information, that provide actionable predictions otherwise unavailable or less accurate without AI.”
💻 For what it’s worth, in the FB AI article, they introduce a new technique that allows researchers to train image classifiers with a small number of labeled examples.
 
 
- Radiological Society of North America article : https://pubs.rsna.org/doi/10.1148/radiol.2020204226
- New AI research to help predict COVID-19 resource needs from a series of X-rays