Comparison of two-dimensional and three-dimensional techniques for determination of facial motion - Absolute movement in a local face frame

Background and purpose: Few studies have used motion analysis in the stud!: of facial animation. A facial animation model using an expert vision motio n analysis system was developed in this study to quantitatively evaluate ab solute movement during five facial animations. Methods: Skin markers were adhered to the face of each subject at 16 anatom ic landmarks selected to represent the functional movement of the facial mu scles. Three analysis of the 16 skin markers were used to establish a local face frame with the origin in local face frame the mid-point of the face. All of the coordinates measured in the laboratory frame facial animation we re rotated and translated to the local face frame for analysis. The traject ory of facial rotation the facial markers was evaluated from the local face frame when comparing two-dimensional (2-D) displacement of skin markers (f rontal plane) with three-dimensional (3-D) values by paired Student's t-tes t. Results: Although the correlation of 2-D and 3-D displacements of skin mark ers was high (r > 0.69), the differences between 2-D and 3-D motion were si gnificant (p < 0.001). The 2-D displacement of skin markers underestimated the 3-D facial animation in each marker and animation. There were no signif icant differences in the movement of both mouth angles or of the eyes. Thir ty repeated measurements of a subject revealed good concentration in 3-D di splacement, velocity, and angle of movement in smiling. The measurement err or was less than 0.06 mm. The normative displacement of individual anatomic landmarks was evaluated to avoid size differences of individual faces. Conclusions: We suggest that the expert vision motion analysis technique is feasible for quantitative evaluation of absolute facial movement and would be useful for further clinical evaluation of patients with facial palsy.