Researchers at the University of South Australia have developed a computer vision system that can automatically detect a baby’s face in a hospital bed. This enables health experts to remotely monitor vital signs from a digital camera, and it demonstrates the same accuracy rate as an electrocardiogram machine.
While artificial intelligence (AI) software and facial recognition technologies are often used to detect adult human faces, this new system is the first developed to detect a premature baby’s face and skin. It can do this despite the various tubes and clothing surrounding the baby.
Remotely Monitoring Heart and Respiratory Rates
Engineering researchers worked with a neonatal care specialist from UniSA to remotely monitor the heart and respiratory rates of seven infants in the Neonatal Intensive Care Unit (NICU) at Flinders Medical Centre in Adelaide with the use of a digital camera.
UniSA Professor Javaan Chahl is one of the lead researchers.
“Babies in neonatal intensive care can be extra difficult for computers to recognise because their faces and bodies are obscured by tubes and other medical equipment,” says Chahl.
“Many premature babies are being treated with phototherapy for jaundice, so they are under bright blue lights, which also makes it challenging for computer vision systems,” he continued.
Developing the System
The technology was developed using a dataset of videos of babies in the NICU, which enabled the system to detect their skin tone and faces accurately.
The research demonstrated that the system’s vital sign readings are on par with those of an electrocardiogram (ECG). It even outperformed the conventional electrodes in some cases.
The study is part of a larger UniSA project that is working to replace contact-based electrical sensors with non-contact video cameras, which can help avoid skin tearing and infections caused by adhesive pads. The latter can happen due to the fragile nature of babies’ skin.
High-resolution cameras were used to film the infants at close range while vital psychological data was extracted using advanced signal processing techniques that can detect subtle colour changes from heartbeats and body movements. These cannot be detected by the human eye, which is another key factor of the new system.
According to UniSA neonatal critical care specialist Kim Gibson, neural networks for detecting the faces of babies is a major breakthrough in non-contact monitoring.
“In the NICU setting it is very challenging to record clear videos of premature babies. There are many obstructions, and the lighting can also vary, so getting accurate results can be difficult. However, the detection model has performed beyond our expectations.”
“Worldwide, more than 10 percent of babies are born prematurely and due to their vulnerability, their vital signs need to be monitored continuously. Traditionally, this has been done with adhesive electrodes placed on the skin that can be problematic, and we believe non-contact monitoring is the way forward,” Gibson says.
The COVID-19 pandemic means these results are even more important, professor Chahl says. With physical distancing, technologies such as this could play an increasingly important role in hospitals.