Tracking COVID-19 with artificial intelligence (July 22)
Artificial intelligence (AI) is becoming a powerful tool for tracking and treating COVID-19 in the U.S. and abroad.
Several U.S. institutions are developing new AI technology or using preexisting technology to monitor and treat the new coronavirus.
HealthMap, an AI application run by Boston Children’s Hospital, was launched in 2006. It was among the first tracking mechanisms to detect the COVID-19 outbreak in China.
“HealthMap data has been used for research studies of infectious disease events,” said Kara Sewalk of Boston Children’s Hospital, “and can even be used by the general public to receive real-time information on disease events in their community” or where they may travel.
HealthMap’s algorithm collects online data about infectious-disease events around the world from news organizations and social media in 15 languages. The system then uses machine learning and natural language processing technologies to track outbreaks, Sewalk said.
Before the pandemic, researchers at the University of California, San Diego (UCSD) worked for 18 months to create an AI system to detect pneumonia and lung damage in patients.
Once COVID-19 cases started to arrive at UCSD-affiliated hospitals, doctors used their technology to help diagnose and track COVID-19 in over 6,000 chest X-rays, according to media reports.
A recent study by researchers in the U.S. and China found AI correctly diagnosed COVID-19 in 68 percent of a sample of patients who had normal chest scans and had been classified as negative for COVID-19 by radiologists.
This U.S. innovation follows on “$12 billion allocated by agencies and departments across the U.S. government to benefit the global response, including vaccine and therapeutics development, preparedness efforts and humanitarian assistance,” Secretary of State Michael R. Pompeo said in a statement.