Three-dimensional reconstruction of individual cervical vertebrae from cone-beam computed-tomography images

The visualization of cervical vertebral morphology has potential in skeletal age assessment; however, thus far, it has only been described in planar images. The objective of this article is to present a novel segmentation algorithm for automatic 3-dimensional (3D) reconstruction of individual cervical vertebrae from cone-beam computed-tomography (CBCT) volumetric data sets.

CBCT data sets of 3 subjects representing different skeletal age groups with no potential health influences on cervical anatomy were identified from a larger subject sample. A visualization toolkit was used to demonstrate the surface topologic shape of cervical vertebrae C1 through C3. The cervical vertebrae were segmented by using a custom algorithm based on individual voxel intensity distribution analysis and propagation from a densitometric start point to generate the whole vertebra. The segmentation algorithm was combined with toolkit visualization to render and save the cervical vertebra in 3D space.

The developed segmentation algorithm separated individual cervical vertebrae successfully. It was robust and efficient. Observed 3D cervical vertebral morphologic features from the 3 examples matched the known 2-dimensional sagittal shape changes of the cervical vertebra with respect to subject age and skeletal maturation.

Segmentation of individual vertebrae proved possible from CBCT volumetric data sets. This provides a 3D approach to the biologic aging of orthodontic patients by using images of the cervical spine. It also has potential in studying disease processes such as spinal fractures consequent to osteoporosis.