Computational geometry assessment for morphometric analysis of the mandible

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Anthropology

Background:

• Accurate geometric assessment of the human mandible is essential for various applications.
• Existing methods for mandible analysis may lack automation or statistical rigor.
• Understanding population-level mandibular variations is key for clinical and research purposes.

Purpose of the Study:

• To develop and validate a fully automated algorithm for mandible geometry assessment.
• To establish a statistically representative mean mandible shape with defined geometrical variations.
• To compare the geometric accuracy of existing mandible models with the derived population mean.

Main Methods:

• Development of a fully automated algorithm for detecting anatomical landmarks on mandible CT-scans.
• Statistical evaluation of detected landmarks and distances using principal component analysis (PCA).
• Generation of a mean mandible shape and its variations from a dataset of 497 human mandible CT-scans.

Main Results:

• Reliable detection of anatomical landmarks and statistically validated distance measurements.
• Creation of a mean mandible shape representing geometrical variations within the studied population.
• Demonstration that commercially available mandible replicas significantly differ from the population's mean shape.

Conclusions:

• The automated algorithm provides a robust method for mandible geometry assessment.
• The generated mean mandible shape and variation data are valuable for bioengineering and surgical planning.
• Significant discrepancies between current mandible models and population data highlight the need for improved anatomical accuracy.