Stanford researchers, who have previously witnessed artificial intelligence (AI) performing on par with board-certified dermatologists, are turning to computer vision to ensure patient safety and improve physician hygiene.
“We may be approaching the limits of what is achievable through improvements in clinical processes, culture and narrowly focused technological assistance,” lead author Serena Yeung, MS, and colleagues wrote in the New England Journal of Medicine. “Expectations that fatigued clinicians will reliably execute each behavioral step of complex hospital treatments ignore evidence from cognitive science that humans usually operate in error-prone ‘fast thinking’ mode.”
Computer vision is “no longer science fiction,” Yeung et al. said. Deep learning has allowed Google to develop self-driving cars, so why can’t it detect deviations in important clinical behaviors and duties?
“If successfully developed and deployed, ambient computer vision carries the potential to discern diverse bedside clinician and patient behaviors at superhuman performance levels and send user-designed prompts in real time,” the authors wrote.
The group of Stanford, Intermountain LDS Hospital and Lucile Packard Children’s Hospital researchers applied the idea to something rudimentary but essential to clinical care: hand hygiene. They trained a computer, using AI, to produce heatmaps of patient rooms based on camera footage of staff using available sanitation stations.
Because of concerns about staff and patient privacy, the method utilizes thermal and depth sensors to create images, the authors said. Despite its lack of explicit clarity, it offers structural advantages over current bedside-assessment systems.
Unlike “secret shopper” observations or clandestine nurse-led inspections, ambient computer vision is “ceaseless and fatigue free,” Yeung and co-authors said, while maintaining a low overall cost. The technique could also be integrated with electronic health record systems to take further documentation and data-entry tasks off physicians’ shoulders.
“Although safe hospital care presents unique challenges, if productivity gains seen in other industries are an indication, computer vision may contribute significantly to clinical quality and efficiency while freeing clinicians to focus on nuanced decision-making, engaging with patients and delivering empathetic care,” the authors wrote. “Given its rapid pace of improvement in accuracy and affordability in other industries, computer vision may soon bring us closer to resolving a seemingly intractable mismatch between the growing complexity of intended clinician behaviors and human vulnerability to error.”