Facebook uses AI to predict if COVID-19 patients will need more care

Facebook is harnessing the power of artificial intelligence to help doctors predict whether they will need more resources, such as extra oxygen to care for COVID-19 patients in hospitals.

The social network said Friday it developed two AI models, one based on a single chest X-ray, and another from a series X-rays, that could help forecast if a patient infected by the coronavirus is likely to get worse. A third model predicts the amount of extra oxygen a COVID-19 patient might need. Facebook's AI models generally did a better job than a human when it came to forecasting up to four days in advance if a patient will need more intensive care resources.

"These predictions could help doctors avoid sending at-risk patients home too soon, and help hospitals better predict demand for supplemental oxygen and other limited resources," Facebook employees said in a blog post.

Partnering with with New York University Langone Health's Predictive Analytics Unit and Department of Radiology, Facebook's AI research is another example of how tech companies are trying to help the health industry battle COVID-19. Some people infected by the novel coronavirus have trouble breathing or experience shortness of breath. Hospitals in some areas, including Los Angeles, are reportedly running low on oxygen needed to help COVID-19 patients. 

Meanwhile, COVID-19 cases continue to climb in the US as essential workers and older adults get vaccinated. Since January 2020, coronavirus cases topped roughly 30 million in the US, according to data on Thursday from the Centers for Disease Control and Prevention. More than 383,300 people have died from the respiratory illness caused by the virus.

Facebook's models rely on a technique in which AI learns on its own rather than depending on data labelled by humans, which can be a time-consuming process. The social network and NYU are publishing their research and open sourcing the AI models.