Comparison of Different Calculation Methods Used to Analyze Facial Soft Tissue Asymmetry: Global and Partial 3-Dimensional Quantitative Evaluation of Healthy Subjects

Abstract

An accurate assessment of range and localization of facial asymmetry is essential to restore facial symmetry. Establishment of a common and well-accepted calculation method might facilitate the interpretation of the assessment results. The aim of the present study was to analyze the global and partial asymmetry of facial soft tissues using 3 different calculation methods and investigate the relationships among them. Facial asymmetry was performed on the 3-dimensional (3D) images of 51 subjects (23 men, mean age 22 ± 2.2years; 28 women, mean age 21.1 ± 2.1years) acquired by a 3D handheld scanner. A mirror image of a facial mask was generated and superimposed on the original mask. Next, the root mean square (RMS), mean absolute deviation (MAD), and mean signed distance (MSD) values were calculated between the original and mirror images. The calculated difference between the total facial masks (amount of asymmetry) was 0.95±0.29mm for the RMS, 0.72 ± 0.22mm for MAD, and -0.04 ± 0.05mm for MSD. A very high correlation was found between the RMS and MAD (r= 0.98). In contrast, the correlation between the MSD and RMS (r= -0.26) or MAD (r= -0.25) was poor. The coefficient of variation for the MSD (133%) was significantly greater than that for the RMS (30%) and MAD (30%) (P < .05). From these results, we have concluded that both the RMS and MAD are accurate and reliable methods in a facial asymmetry assessment. The MSD is insufficient alone but can be used as a side parameter to show the direction of the asymmetry.