Published on
It seems that artificial intelligence (AI) models can detect early acute systemic inflammation in response to viral respiratory tract infections (VRTIs) through biometric data collected from wearable devices, according to a small study published in Lancet Digital Health. In the prospective study, 55 healthy adults aged 18–59 years were monitored before and after receiving a live attenuated influenza vaccine, which simulated real-world infection. Participants continuously wore a smart ring, smart watch, and smart shirt, which together collected 2 billion physiological data points over 12 days, including heart rate, heart rate variability, body temperature, respiratory rate, blood pressure, physical activity, and sleep quality. The data was paired with blood-based inflammatory biomarkers and self-reported symptoms. The top-performing AI model among all the models the researchers developed collected nighttime data from the smart ring and achieved nearly 90% sensitivity using a 24-hour detection window. The authors say this AI model leveraging wearables may be able to provide objective early warning of systemic inflammatory events due to viral respiratory infections, and they found it to be superior to a model that was based on symptoms alone.
Nice bonus: What was surprising in the study was that the AI models even detected systemic inflammation in 4 participants with incidental SARS-CoV-2 infections before symptom onset or testing confirmation. If models like this enter everyday consumer health markets, urgent care would likely rely on point-of-care testing for confirmation of any health issue the wearable device suggests.