Introduction: High incidence of skin damage over the nasal bridge and cheeks as a result of prolonged wearing of standard sized Personal Protective Equipment (PPE) including respirator masks have been reported as a serious occupational hazard for frontline Healthcare Personnel (HCP) during the COVID-19 pandemic. Respirator masks designed via anthropometric sizing methodologies have traditionally led to significant failure rates ranging from ca 10% to 90% for Quantitative Fit Test (QNFT) or Qualitative Fit Test (QLFT). Largely attributable to a high variance of facial characteristics arising from demographic differences in age, gender, ethnicity, and Body Weight Index (BMI). 3D printed respirator masks tailored to individual facial shape can potentially alleviate this problem. However, a manual and time-consuming design process to create a single custom fitted Computer-Aided Design (CAD) model add further costs and logistic burden to the already expensive Additive Manufacture (AM) process, making the mass customisation of respirator masks unrealistic. This study proposes a fully automated design pipeline to support the progress in using AM to produce custom fitted respirator masks for the post-COVID-19 era.Method: The proposed pipeline accepts 3D facial scans obtained from a mobile phone depth sensor camera and outputs a mask CAD model custom fitted to the 3D shape of the input scan. Rigid and non-rigid iterative closest point algorithms were utilised to bring a raw facial scan into dense correspondence with a universal facial template scan. Subsequently, useful topographical facial data were extracted by referencing to specific landmarks on the template scan and used to create a mask model via a CAD Application Programming Interface. 299 participants were recruited remotely via an online portal. Their facial scans and demographic information were collected to assess the robustness of the pipeline in processing facial shapes of different demographic backgrounds. 92 facial scans were excluded from the analysis due to poor quality, defects or non-neutral facial expression. The remaining 207 facial scans were processed via the pipeline. Computational time for each scan was recorded. Goodness of fit (at 1.5mm) between each mask model and the input raw scan was also evaluated via a Euclidean distance metric. Finally, chi-square tests of independence with a Confidence Level (CI) at 95% was carried out to investigate whether goodness of fit is independent of age groups (18-35, 36-55, 56 and above), gender (male or female), ethnicity (Asian, White, Others), or BMI (18.5-24.9, 25-29.9, 30.0 and above).Results: 180 out of 207 (87.0%) facial scans were successfully processed. Average computational time for processing a single scan is 3 minutes 20 seconds. Of the 180 successful subjects, 175 (97.2%) responded to the demographic survey with the majority being male (83.4%), White (79.4%) with a mean BMI at 31.1 ± 8.4 and a mean age of 47 ± 14 years. From results of chi-square test of independence, it was found that mask’s goodness of fit is independent of age, χ2 (3, N = 175) = 2.33, p = .31; BMI χ2 (3, N = 174) = 0.60, p = .74; ethnicity χ2 (3, N = 175) = 2.36, p = .31; and gender χ2 (2, N = 173) = 0.47, p = .49.Conclusion: The proposed pipeline was able to automatically create custom fitted respirator CAD model in less than 4 minutes, regardless of a person’s demographic background. This has significantly reduced the design cost, therefore making it more affordable to use AM for the mass customisation of PPE respirator masks in the future.

A Fully Automated Design Pipeline for Mass Customisation of 3D printed Respirator Masks for Post-COVID-19 Era

Paper Proceedings: https://asmedigitalcollection.asme.org/IDETC-CIE/IDETC-CIE2021/volume/85369

ASME 2021 International Design Engineering Technical Conferences & Computers and Information in Engineering Conference (IDETC/CIE2021)

Supporting the need among researchers worldwide for additional online and multimedia content, ASME now offers the ASME Conference Video Collection. This platform will collect and provide long-term access to conference presentation videos, encompassing all technology areas within ASME's technical conference and events program. About 30,000-40,000 engineers across all mechanical engineering and related disciplines participate in ASME conferences each year. This includes authors of approximately 6,000 to 8,000 technical papers published by ASME, the majority of whom also submit video presentations for our virtual conferences. The video collection will be available as an add-on to existing product packages or as a separate subscription.