Three-dimensional motion analysis - an exploratory study. Part 2: Reproducibility of facial movement

摘要: Authors – Popat H, Richmond S, Playle R, Marshall D, Rosin PL, Cosker D Objectives To investigate the reproducibility of using spoken word as a measure facial movement. Design Experimental study. Setting and Sample Population Department Dental Health Biological Sciences, University Hospital, Cardiff. 22 normal subjects. Variable Subjects were asked to say ‘puppy’ in normal, relaxed manner twice within 10-second time interval. The sequence was recorded non-invasive, three-dimensional motion analysis image capture system (3DMDface™ Dynamic System) at 48 frames per second. Outcome Corresponding between two utterances aligned three-dimensionally best fit on non-moveable points upper half face. Reproducibility measured percentage point deviation ± 0.5 mm corresponding frames. Results Mean intrasessional (SD) for group 86.2% (5.8). ranged from minimum 66.8% maximum 97.5%. When utterance split into its separate viseme segments (/p//u//p/ /p//y/), second part seen be more reproducible than first. male female group. Conclusion Intrasessional shows high intra- intersubject variability this subjects therefore further research needs conducted before being able confidently use