The researchers found that the overall health monitoring tool, combined with a computer algorithm, correctly identified 68% of people who tested positive for Covid two days before their symptoms appeared.
An international research team, which includes the University of Basel (Switzerland) and Imperial College London, has shown that although the PCR swab test remains the gold standard for confirmation of Covid-19 infection, “the findings of We show that a wearable informed machine learning algorithm can serve as a promising tool for detecting symptomatic or asymptomatic Covid-19.”
Wearable AI tracker to detect Covid-19
The team conducted a test on an AVA bracelet and included 1,163 participants, all under the age of 51 who wore the tracker at night. The device saves data every 10 seconds and requires at least 4 hours of relatively continuous sleep. The bracelets have been synced with a free smartphone app upon waking.
All participants performed routine rapid antibody tests for Covid infections. People with indicated symptoms also have a PCR swab test.
The algorithm was ‘trained’ using 70% of the data from day 10 to day 2 before the onset of symptoms within 40 days of continuously monitoring 66 people who tested positive with SARS-CoV-2. It is then tested on the remaining 30% of the data.
Approximately 73% of laboratory-confirmed positives were selected in the training kit and 68% in the test kit, up to 2 days before the onset of symptoms.
The researchers acknowledge that their results may not be more widely applicable.
But “wearable sensor technology is an easy-to-use, low-cost method to enable individuals to monitor their health and wellness during a pandemic,” they wrote in the paper. .
Furthermore, “these devices, in partnership with artificial intelligence, can push the boundaries of personalized medicine and disease detection before (symptoms appear), potentially reducing the spread of disease.” transmission of the virus in the community”.