| Abstract [eng] |
Currently, the surface quality of scans generated by intraoral scanners is typically evaluated using best-fit global alignment techniques, most commonly iterative closest point (ICP) algorithms. However, these methods often encounter difficulties when dealing with the complex morphology of individual dental surfaces, thereby limiting the reliability of accuracy assessments in clinical research. Meanwhile, augmented reality (AR) headsets—particularly Microsoft HoloLens 2—have been successfully implemented in various medical fields, yet their application in dentistry, especially in implantology, remains insufficiently explored, notably in terms of accuracy. Evaluating the performance of the HoloLens 2 in computer-guided dynamic navigation systems is crucial, as even minimal deviations can substantially affect procedural precision and, consequently, treatment outcomes. In this doctoral thesis, a set of guidelines was developed for the comprehensive accuracy assessment of intraoral scanners. The proposed framework encompasses reference object creation, scanning procedures, and reverse engineering techniques to analyze surface deviations and structural discrepancies. This workflow effectively differentiates the scan quality of intraoral and laboratory scanner systems, revealing distinct accuracy levels across various surface types. Additionally, the potential of augmented reality for dynamic navigation systems was examined using a reference-free evaluation approach. While integration proved feasible, accuracy deviations exceeding 1 mm in registration and visual perception restrict its clinical applicability for implant placement procedures, thereby limiting its practical utility. |
