PhysIQ software for smart watches has been created that recognizes the first symptoms of coronavirus based on the owner's biometric data.
A team of researchers from Purdue University, in West Lafayette (Indiana, USA), has created a program for smart watches based on artificial intelligence PhysIQ, which is able to warn the owner about the need to be tested for COVID-19.
When developing an algorithm for detecting the first signs of coronavirus from biometric data, we relied on previous research, said Craig Gergen, associate Professor of biomedical engineering. Doctors have found that viral infections increase the heart rate and breathing at rest, as well as reduce heart rate variability even before the patient's temperature rises. Our app cannot detect that you are actually infected with COVID-19, but now the watch is able to warn the owner: "you may have symptoms of a viral disease. You should be tested for infection". This can be called proactive support, the scientist explained.
About a hundred participants of the experiment – students and University teachers – received a standard Samsung Galaxy smart watch with the PhysIQ app installed. The gadget is complemented by several miniature biosensors placed in a strip of sticky patch attached to the chest. These sensors record a wide range of biometric data of the owner: take a cardiogram, determine body temperature, blood pressure, respiratory rate and pulse. In real time, the biosensor data stream is continuously transmitted to the cloud server. The PhysIQ program compares changes in the performance of sensors that are unique to humans, "learning" from large amounts of data that is continuously collected when a huge number of people wear watches.
We consider this work a serious step towards expanded personalized monitoring of people's health, said Stefan Wegerich, chief researcher at the Department of electrical and computer engineering. A smart watch with the PhysIQ app installed will be another tool to help your doctor make an accurate diagnosis.
At the next stage, the study will be expanded to include several categories of people at high risk of COVID-19 infection. The new technology for reliable and accurate detection of signs of infection at an early stage has been approved by experts from the food and drug administration (FDA). It will soon be widely used in the field of digital medical services and in clinical practice.