Automatic extraction of mandibular nerve and bone from cone-beam CT data

Area of Science:

• Medical Imaging
• Computational Anatomy
• Oral and Maxillofacial Surgery

Background:

• Accurate localization of the mandibular nerve is crucial for dental implantology and maxillofacial surgery.
• Cone Beam Computed Tomography (CBCT) is widely used in dental imaging, but offers limited soft tissue contrast.
• Identifying small structures like alveolar nerves in CBCT data is challenging.

Purpose of the Study:

• To develop and validate a fully automatic method for reconstructing the 3D bone surface and mandibular nerve course from CBCT data.
• To improve the visualization and localization of the mandibular nerve for surgical planning.

Main Methods:

• A combined statistical shape model of the nerve and bone was employed.
• A Dijkstra-based optimization procedure was utilized for reconstruction.
• The method was validated on 106 clinical Cone Beam Computed Tomography (CBCT) datasets.

Main Results:

• The automated method achieved an average bone reconstruction error of 0.5 +/- 0.1 mm.
• The mandibular nerve was reconstructed with an average error of 1.0 +/- 0.6 mm.
• The method demonstrated accurate 3D reconstruction of both bone and nerve structures.

Conclusions:

• Fully automatic 3D reconstruction of the mandibular nerve and bone surface from CBCT is feasible.
• This method offers a valuable tool for enhancing precision in dental implantology and maxillofacial procedures.
• The validated accuracy supports the clinical utility of this automated approach.