Three-Dimensional Quantification of Facial Morphology and Movements Using a Wearable Helmet

Abstract

This work proposes a 3D normative database of facial ranges of motion in adults free from facial disorders. Ten facial movements were analyzed, each targeting the activity of specific muscle groups innervated by the facial nerve. The experimental protocol included a test-retest reliability positioning procedure of 25 skin markers based on clinical expertise in facial morphology. Three maximal voluntary contractions were recorded for each facial movement studied, using a 3D facial motion capture helmet. We included 53 adults free from facial disorders (26 men; age 43 ± 14 ), evaluated twice one week apart. The reliability of marker positioning was expressed as absolute measurement errors. The range of motion vectors of all markers from the best rest to the maximal voluntary contraction was calculated for each muscle group. Primary, secondary, and tertiary markers were extracted for each facial movement. 3D Procruste and asymmetry indices were developed. This allowed the identification of common thresholds of 10% for the asymmetry index and of 6 mm for the Procruste index, beyond which facial motions would be considered abnormally asymmetric. The normative database quantifies facial motions and allows assessment of the degree of clinical disorders by comparison. This protocol is currently being investigated in patients with chronic unilateral peripheral facial paresis.

Guided Self-rehabilitation Contract vs conventional therapy in chronic peripheral facial paresis: VISAGE, a multicenter randomized controlled trial

“…Regarding the secondary objectives, it should be noted that this sample size will be comparable to other reliability and validity studies for the FaCE scale ( n = 67 to 122 patients; [ 47 , 48 , 59 – 61 ] and larger than the initial Sunnybrook validation study ( n = 19; 12). Regarding reliability parameters such as the intraclass correlation coefficients (ICC) and for ICC values expected to be > 70%, the suggested sample size will allow estimating ICCs with an accuracy of ± 15% for two measures and ± 10% for three measures and more [ 63 ].…”

“… Objective 3D facial motor function quantification through 3D wearable (helmet) facial motion analysis through the Cara™ system. Asymmetry ratios and Procuste analysis will be performed at rest (during a 30-s period), during twelve maximal movement efforts and during spontaneous smile [ 59 ]. Measurement of the amount of synkinesis during voluntary movement, of spasms and dystonia at rest will also be performed using Procuste distances: synkinesis (quality, frequency and severity) will be assessed during each maximal voluntary movement (12 movements in total) and during one spontaneous smile; spasms will be assessed during a 30-s period of rest, as well as dystonia.…”