Semi-automatic segmentation of computed tomographic images in volumetric estimation of nasal airway

Abstract The objective of this study was to determine nasal cavity volumes and cross-sectional profiles from segmented coronal high-resolution computed tomography (HRCT) images. Pathological mucosal changes and congenital sinonasal variants were quantitated and three-dimensional (3D) images for determining sinonasal airway diseases evaluated by using the new semiautomatic segmentation software, Anatomatic. Anterior to posterior cross-sectional profiles of the sinonasal airway were obtained from acoustic rhinometry and segmented coronal HRCT images and compared in five patients having complaints of nasal obstruction and chronic sinusitis. Results showed that accurate volumes of air spaces in the nasal cavity and paranasal sinuses were obtained. When compared, the cross-sectional profiles of the nasal cavities obtained from acoustic rhinometry and the segmentation technique were similar in the anterior portion, but differed in the posterior portion. The results obtained by coronal HRCT and segmentation were more reliable than those produced with acoustic rhinometry. 3D images acquired from segmented images were found to help make a good pre-operative assessment of the whole sinonasal compartment. Segmentation and volumetric analysis using the Anatomatic technique also proved to be well suited to the evaluation of the nasal cavity and paranasal sinus geometry in patients with sinonasal diseases.