A key symptom of COVID-19 – oxygen saturation – is now being estimated remotely from a camera, thanks to research from University of South Australia engineers Professor Javaan Chahl, Dr Ali Al-Naji and their team of graduate students.

AI health monitoring software developed by the team this year already detects temperature, heart and breathing rates from drones and fixed cameras.

Technology developed under the research project is being used widely in the US to monitor for COVID-19 symptoms and ensure social distancing compliance in a range of environments.

The contactless sensing technology takes just 15 seconds to measure heart rate, breathing rate and now blood oxygen levels from webcam video of a person’s face, while also measuring core temperature using a thermal camera.

Also known as SpO2, blood oxygen saturation is a measure of how well the lungs absorb oxygen and the circulatory system transports oxygenated blood. 

“An individual’s SpO2 level is usually measured using connected sensors that project light through a finger or earlobe and, until now, there has been no real means to make a non-contact assessment,” Prof. Chahl says.

“In this pandemic, a low SpO2 level has risen to prominence as an important symptom of individuals with COVID-19.”

Dr Al-Naji and Prof Chahl have previously demonstrated the value of monitoring vital health signs from fixed cameras via contactless monitoring to reduce the risk of infection in neonatal wards, replacing adhesive electrodes.

Their work to develop specialised sensor and computer vision systems to remotely monitor temperature, heart and respiratory rates was originally envisaged for war zones and natural disasters to detect signs of life.

The technology has also been successfully trialled to undertake basic health checks of exotic wildlife using a digital camera, saving them the stress of an anaesthetic.

In future do you think AI technology will replace the need to ever see a doctor in person?