A Bidimensional System of Facial Movement Analysis Conception and Reliability in Adults

Objective. To design a bidimensional facial movement measuring tool and study its reliability. Methods. We utilized the free video-analysis software Kinovea that can track preselected points during movements and measure two-point distances off-line. Three raters positioned facial markers on 10 healthy individuals and video-taped them during maximal bilateral contractions of frontalis, corrugator, orbicularis oculi, zygomaticus, orbicularis oris, and buccinator, on two occasions. Each rater also analyzed the first video twice, one week apart. For each muscle, intrarater reliability was measured by percent agreements (PA) and intraclass correlation coefficients (ICC) between two assessments of the same video one week apart and between assessments of two videos collected one week apart. Interrater reliability was measured by PA, ICC, and coefficients of variation (CV) between assessments of the first video-recording by the three raters. Results. Intrarater and interrater reliabilities were good to excellent for frontalis (PA and ICC > 70%; CV < 15%), moderate for orbicularis oculi, zygomaticus, and orbicularis oris, and poor for corrugator and buccinators. Discussion. Without formal prior training, the proposed method was reliable for frontalis in healthy subjects. Improved marker selection, training sessions, and testing reliability in patients with facial paresis may enhance reliability for orbicularis oculi, zygomaticus, and orbicularis oris.