Automatic landmark annotation in 3D surface scans of skulls: Methodological proposal and reliability study

Background and Objectives: Craniometric landmarks are essential in many biomedical applications, such as morphometric analysis or forensic identification. The process of locating is usually a manual slow task, highly influenced by fatigue, skills the experience practitioner. Localization errors propagated magnified subsequent steps, which can result incorrect measurements assumptions. Thereby, standardization, reliability reproducibility lay foundations for necessary accuracy anatomical analysis. In this paper, we present an automatic method to annotate 3D surface skull models taking into account geometrical features. Methods: proposed follows hybrid structure where deformable template used initialize landmark positions. Then, refinement stage applied using prior knowledge ensure correct placement. Our proposal validated over thirty scans male Caucasians, acquired hand-held scanning, set 58 craniometric landmarks. A statistical was carried out analyze inter- intra-observer variability annotations results, along with visual assessment final results. Results: Inter-observer show significant differences, reflected expert consensus reference. average localization error 2.19±1.5 mm when comparing reference location. confirmed most Conclusions: Repeated high depending on both expertise observer, landmarks’ location characteristics. contrast, provides accurate, robust reproducible alternative tedious error-prone task landmarking.