Statistical modelling of lip movement in the clinical context
Authors: Popat H, Zhurov AI, Toma AM, Richmond S, Marshall D, Rosin PL.
DOI: 10.1111/j.1601-6343.2011.01539.x.
Abstract:
Objective – To establish three-dimensional (3D) reference data on average lip movement in normal healthy subjects using statistical shape analysis techniques.
Setting and Sample Population – School of Dentistry and Cardiff School of Computer Science, Cardiff University, United Kingdom. One hundred and fifteen white subjects.
Material and Methods – Subjects performed four reproducible verbal gestures (puppy, rope, baby and bob) in a normal relaxed manner, which were captured using a non-invasive, 3D motion scanner (3dMDFace™ Dynamic System). Six landmarks were manually placed around the lips of the 3D facial shells showing maximum lip displacement. Generalized procrustes analysis followed by principal component analysis was applied to the landmark coordinates to characterize lip movement for each word.
Results – The first four principal components (PCs) describe the majority of variation in lip movement for the four words involving a complex interaction of lip movements in three dimensions. Bilateral landmarks were paired within PCs showing that movement was largely symmetrical. Female resting lip shape was narrower and shorter in height than males. During motion, females preferred a more protrusive articulation than males.
Conclusion – Statistical shape analysis techniques can be used to characterize lip movement during articulation. Data from this study can act as a reference for average lip movement to compare similar population groups.
