These digital tools, sometimes referred to as “symptom checkers,” could potentially be used to determine whether a patient has COVID-19 without the individual needing to come into a physicians’ office or emergency room where they could spread the infection. These tools are built by companies that use machine learning computer algorithms, which can learn to diagnose diseases.
Some digital companies said they are “wary of updating their algorithms to incorporate questions about the new coronavirus strain,” because they lack enough data about the disease to train their detecting algorithms, according to the report. AI systems built on machine learning adapt based on new data they receive, which makes them more difficult to alter for a new virus, Hugh Harvey, co-founder of digital healthcare consulting firm Hardian, told the WSJ. He said these AI systems need more accurate training data before being widely implemented.
Your.MD, a London-based symptom checking app, recently added a “coronavirus checker” that prompts users to answer a series of questions about their symptoms, according to the report. Instead of machine learning, the symptom checker is based on a decision tree process to provide online advice for the health issue.
“We made a decision not to do it through the AI because we haven’t got the underlying science,” Maureen Baker, CMO at Your.MD, told the WSJ. Ms. Baker said it could take six months to a year before peer-reviewed scientific research is published and can be used to help design algorithms in today’s symptom checkers.
Babylon Healthcare Services, a digital health firm whose AI chatbot is used by the National Health Services in the U.K., said it refrained from adding coronavirus screening to its symptom checker technology and instead recommends users to call NHS’ hotline if they develop symptoms of the virus.
Despite AI’s opportunity to enhance healthcare, “In a situation like this, the tools aren’t ready,” said Keith Grimes, clinical innovation director at Babylon.
More articles on artificial intelligence:
Cleveland Clinic’s AI center spearheads 20 research projects
Mount Sinai AI ‘reads’ physician notes to determine patient pain levels
AI chatbot startup Babylon Health attacks physician for ‘2,400 Twitter troll tests’
