
Global Analysis of Three-Dimensional Shape Symmetry: Human Heads (Part I)
Author(s) -
Vi-Do Tran,
Tien Tuan Dao,
Tan-Nhu Nguyen
Publication year - 2022
Publication title -
technical education science/giáo dục kỹ thuật
Language(s) - English
Resource type - Journals
eISSN - 2615-9740
pISSN - 1859-1272
DOI - 10.54644/jte.68.2022.1076
Subject(s) - hausdorff distance , head (geology) , facial symmetry , mathematics , regular polygon , artificial intelligence , symmetry (geometry) , convex hull , computer science , geometry , orthodontics , medicine , biology , paleontology
Facial paralysis grading methods are mostly based on geometrical dissymmetry between left and right heads. Dissymmetry also appears even on healthy subjects. Consequently, this dissymmetry should be computed to distinguish with one on facial palsy patients. However, no studies have been quantitatively reported this characteristic. In the first part of our study, we tried to calculate this quantity by computing distance and volumetric differences between left and right heads. In particular, 329 head models were reconstructed from medical images of subjects with normal head geometries in neutral mimics. These heads were then automatically cut into left and right regions. Then, Hausdorff distances between left heads and mirrored right heads were computed. Moreover, volumetric differences between left head convex hulls and mirrored right head convex hulls were also calculated. As a result, the dissymmetry values (Mean ± SD) in mean Hausdorff distances are 1.8243 ± 0.7029 mm, and ones in volumetric are 78.1254 ± 65.7040 cm3. In perspective, in the second part of our study, we will try to analyze geometrical symmetry on human skulls. Moreover, shape symmetry will be analyzed in more detail with different local shape topologies in different facial mimics. These analyses will finally be implemented on our clinical decision-support system for facial mimic rehabilitation.