Artificial Intelligence (AI) cameras have been created by Scientists to help detect when hospital staff don’t wash their hands, in the hope to fight infection.
One in twenty five hospital patients will suffer from a hospital acquired infection (HAI), according to CDC. Some patients that go into hospital for treatment end up becoming more ill than when they arrived, due to lack of hygiene in the workplace.
Working to develop the technology, Stanford University research is looking to develop a non-intrusive vision-based system to track workers activity, specifically for the hand hygiene problem.
An AI camera tracking system, automatically identifying when staff and patients use hand dispensers has been developed by researchers at the University to reduce the likelihood of developing a HAI.
A trial using depth cameras and computer vision-algorithms was conducted by researchers at Stanford University and the Swiss Federal Institute of Technology, in Lausanne. With the system hospitals can track healthcare workers around the building as well as check if they are washing their hands.
After collecting data from a Friday lunchtime visiting slot, it found that of 170 people visiting patients, only 30 followed the correct hygiene system.
Research involved using top down view sensors were installed above gel dispensers, overlooking corridors, inside patient rooms, overlooking sinks and patient beds for the first part of the study. This collated imagines from the units to then assess who was using what cleaning aids.
Following the research results, the images collected were used to train algorithms for the smart technology, to detect staff, movement and hand practices.
Researchers said: “When trained from scratch, our experiments show that ResNet-152 outperforms the other model architectures, due to the high number of layers and parameters of the model.”
The old system of using ‘mystery shoppers’ to detect the effectiveness of hand hygiene only gave 63% accurate findings if visitors/workers followed the hygiene procedure.
Read More: 5 ways technology is transforming healthcare
Over a year, three hospitals will be participating in a further trial to evaluate the reduction of hospital-acquired infections.
Hospitals built with AI software are known as ‘smart hospitals’. They involve a variety of network connected products for controlling, automating and optimizing workflows in the hospitals. The technology allows hospitals to be monitored 24hours a day and deliver results and information accurately. AI software could also enable the hospital to monitor patient falls and patient condition.
CDC said: “Our goal is to build a computer vision system for automatically assessing hand hygiene compliance across an entire hospital unit.”