Purpose: To develop a time-efficient automated segmentation approach that could identify surface structures on the temporal bone for use in surgical simulation software and preoperative surgical training.
Methods: An atlas-based segmentation approach was developed to segment the tegmen, sigmoid sulcus, exterior auditory canal, interior auditory canal, and posterior canal wall in normal temporal bone CT images. This approach was tested in images of 20 cadaver bones (10 left, 10 right). The results of the automated segmentation were compared to manual segmentation using quantitative metrics of similarity, Mahalanobis distance, average Hausdorff distance, and volume similarity.
Results: The Mahalanobis distance was less than 0.232 mm for all structures. The average Hausdorff distance was less than 0.464 mm for all structures except the posterior canal wall and external auditory canal for the right bones. Volume similarity was 0.80 or greater for all structures except the sigmoid sulcus that was 0.75 for both left and right bones. Visually, the segmented structures were accurate and similar to that manually traced by an expert observer.
Conclusions: An atlas-based approach using a deformable registration of a Gaussian-smoothed temporal bone image and refinements using surface landmarks was successful in segmenting surface structures of temporal bone anatomy for use in pre-surgical planning and training.
Keywords: Atlas-based segmentation; Image registration; Pre-surgical planning; Surgical simulation; Temporal bone anatomy.