Résumé
Abstract Objective: The aim of this study was to assess the accuracy of volumetric reconstruction of the pharynx by comparing the volume and minimum crosssectional area (mCSA) determined with open-source applications (ITK-Snap, www.itksnap.org ; SlicerCMF) and commercial software (Dolphin3D, 11.8, Dolphin Imaging & Management Solutions, Chatsworth, CA, USA) previously validated in the literature. Material and Methods: The sample comprised of 35 cone-beam computed tomography (CBCT) scans of patients with unilateral cleft lip and palate, with mean age of 29±15. Three-dimensional volumetric models of the pharynx were reconstructed using semi-automatic segmentation using the applications ITK-Snap (G1) and Dolphin3D (G2). Volumes and minimum cross-sectional areas were determined. Inter- and intra-observer error were calculated using ICC test. Comparison between applications was calculated using the Wilcoxon test. Results: Volumes and minimum crosssectional area were statistically similar between applications. ITK-Snap showed higher pharynx volumes, but lower mCSA. Visual assessment showed that 62.86% matched the region of mCSA in Dolphin3D and SPHARM-PDM. Conclusion: Measurements of volume and mCSA are statistically similar between applications. Therefore, open-source applications may be a viable option to assess upper airway dimensions using CBCT exams.