Image-Based Web Application for Respirator Sizing: Contactless Mask-Fitting during a Pandemic

ABSTRACT

At the beginning of the COVID-19 pandemic, many hospitals and healthcare institutions lacked an adequate supply of masks and other personal protective equipment. Moreover, protocols that were in place to ensure healthcare workers had appropriately sized masks consumed precious time and resources. Any determination of a user's correct respirator size demanded an in person assessment and had the potential to waste multiple respirators. Here we introduce IBARS (Image-based Application for Respirator Sizing), a novel tool which provides respirator size recommendations based on a facial image and basic user demographics. This solution obviates the need for an in person assessment, providing an accurate size recommendation within seconds. The application has the potential to reduce time-per-worker respirator fitting, reduce overall respirator usage, and increase safety by providing hospitals with a non-contact option for sizing. Furthermore, future applications may assist healthcare institutions optimize supply chains by providing rapid assessments and reassessments of appropriate respirator sizes used by their workers. Early testing indicated accuracy of 71.3% for the software (N=16), and further testing is underway at Houston Methodist Hospital.