Wearable devices have new features coming out constantly, but perhaps the most useful are those related to health. A recent Stat News article discussed a new study aimed at determining if wearables can be used to detect presymptomatic and asymptomatic infections. Subjects in the study were given Empatica’s E4 wristbands, which record heart rate, skin temperature, movement, and electrodermal activity. They were then exposed to viruses, and were instructed to report daily symptoms to researchers so they could quantify their viral shedding.
A machine learning algorithm compared this data to the subjects’ baseline biometric data, and predicted the presence of infection. After the data was gathered, the algorithm was found to predict infection 12 hours after exposure with 78% accuracy, when the median time of symptom onset was 48 hours for flu, and 36 hours for rhinovirus. That number increased to 92% for flu alone after 24 hours, and 88% for rhinovirus after 36 hours. The results were promising enough to give hope for a public health tool that could lead to early detection of diseases like COVID-19.
