Fit prediction for filtering facepiece respirator using 3D face shape.

Area of Science:

• Biomedical Engineering
• Human Factors Engineering
• Occupational Safety

Background:

• Accurate respirator fit is crucial for protecting users from airborne hazards.
• Traditional anthropometric measurements have limitations in predicting respirator fit.
• 3D facial scanning offers a novel approach to capture detailed facial geometry.

Purpose of the Study:

• To develop a framework for predicting filtering facepiece respirator (FFR) fit using 3D face-shape elements.
• To compare the predictive power of 3D face shape elements against traditional anthropometric measurements.
• To enhance the development of predictive models for FFR fit.

Main Methods:

• Collected 3D face scans and quantitative fit factor data from 202 participants.
• Automated extraction of face shape data from 3D scans.
• Utilized Principal Component Analysis (PCA) and developed predictive models using 3D face shape elements.

Main Results:

• 3D face shape elements formed distinct and interpretable groupings via PCA.
• Specific 3D face shape elements, like lateral nose slope, were more predictive of FFR fit than traditional measurements.
• Predictive models with fewer variables, derived from 3D data, were most effective.

Conclusions:

• 3D face shape elements provide a more reliable basis for predicting FFR fit compared to traditional anthropometry.
• Geometric facial data enhances understanding of face-respirator interactions.
• This approach can lead to improved respirator fit panels, safety protocols, and design innovation.