Lucas M Ritschl1, Florian D Grill2, Fabienna Mittermeier2, Daniel Lonic3, Klaus-Dietrich Wolff2, Maximilian Roth4, Denys J Loeffelbein4. 1. Department of Oral and Maxillofacial Surgery, Klinikum rechts der Isar, Technische Universität München, Germany. Electronic address: lucas.ritschl@tum.de. 2. Department of Oral and Maxillofacial Surgery, Klinikum rechts der Isar, Technische Universität München, Germany. 3. Department of Plastic and Reconstructive Surgery, Helios Klinikum München West, Germany. 4. Department of Oral and Maxillofacial Surgery, Klinikum rechts der Isar, Technische Universität München, Germany; Department of Oral and Maxillofacial Surgery, Helios Klinikum München West, Germany.
Abstract
BACKGROUND: Three-dimensional (3D) photogrammetry has reached high standards and accuracy but is mainly conducted with stationary and expensive systems. The purpose of this study was to evaluate the accuracy of a low-budget portable system with special regard to the gracile and challenging nasal region. MATERIAL AND METHODS: 3D models of the perinasal area were acquired by impression-taking and the scanning of the generated plaster models (3Shape D500) or with a portable low-budget 3D stereophotogrammetry (FUEL3D® SCANIFY®) system. Four examiners analysed defined landmarks of the generated Standard Tessellation Language files with regard to accuracy and interobserver reliability by using 3dMDvultus™ software. A semi-automatic 3D best-fit analysis of both models was performed by using Geomagic® and the Root Mean Squared (RMS) errors were calculated. RESULTS: 41 volunteers were included, with 22 perinasal and perioral landmarks, 15 3D distances and eight 3D angles being analysed per data set. In a point-based analysis the mean spreads were partially smaller in the plaster model scans. Most measurements showed very high (>0.8) to excellent (>0.9) intraclass correlation coefficients, the lowest being found for columella length (0.686) and left nostril width (0.636). Overall, the mean RMS error between the superimposed surfaces was 0.89 ± 0.22 mm in the best-fit analysis. CONCLUSIONS: The corresponding software program was operator-friendly. The findings indicate that the analysed, affordable and portable system is a feasible solution for 3D image acquisition with comparable accuracy reported in the literature. Further studies will analyse the feasibility in neonates.
BACKGROUND: Three-dimensional (3D) photogrammetry has reached high standards and accuracy but is mainly conducted with stationary and expensive systems. The purpose of this study was to evaluate the accuracy of a low-budget portable system with special regard to the gracile and challenging nasal region. MATERIAL AND METHODS: 3D models of the perinasal area were acquired by impression-taking and the scanning of the generated plaster models (3Shape D500) or with a portable low-budget 3D stereophotogrammetry (FUEL3D® SCANIFY®) system. Four examiners analysed defined landmarks of the generated Standard Tessellation Language files with regard to accuracy and interobserver reliability by using 3dMDvultus™ software. A semi-automatic 3D best-fit analysis of both models was performed by using Geomagic® and the Root Mean Squared (RMS) errors were calculated. RESULTS: 41 volunteers were included, with 22 perinasal and perioral landmarks, 15 3D distances and eight 3D angles being analysed per data set. In a point-based analysis the mean spreads were partially smaller in the plaster model scans. Most measurements showed very high (>0.8) to excellent (>0.9) intraclass correlation coefficients, the lowest being found for columella length (0.686) and left nostril width (0.636). Overall, the mean RMS error between the superimposed surfaces was 0.89 ± 0.22 mm in the best-fit analysis. CONCLUSIONS: The corresponding software program was operator-friendly. The findings indicate that the analysed, affordable and portable system is a feasible solution for 3D image acquisition with comparable accuracy reported in the literature. Further studies will analyse the feasibility in neonates.