Abstract
The subjective nature of analyzing outcomes following facial plastic surgery poses a great challenge to providing objective evaluations of patients. Currently, there is a lack of accurate quantitative tools to monitor facial appearance and function. We introduce a novel quantitative facial analysis tool dubbed numeriFACE, which utilizes the recent advances in mobile camera technology and machine learning to provide precise and reliable automated facial mapping and scoring. numeriFACE allows for both retrospective analysis of existing 2D images, as well as more advanced prospective analysis utilizing depth information. numeriFACE was used in its two study arms to harvest six key facial measurements: intercanthal distance, mouth width, pronasale to menton, alar base width, mid-face height, and lower-face height. These were then compared to standard-of-care caliper measurements showing a strong degree of correlation overall. numeriFACE provides a reliable and repeatable point-by-point registration of human facial features. It has the potential to be used in a vast array of facial characterization most specifically analyzing mid-face symmetry. Future studies are aimed at utilizing the software in the fields of reconstructive as well as aesthetic surgery.
