Description of a Novel Technique for Three-Dimensional Fit Assessment of Dental Restorations

Objective: Precision of fit of CAD/CAM generated dental implant superstructures is crucial to reduce mechanical complications and ensure longevity. Numerous studies based on finite element analyses (FEA) have been published predicting impact of component accuracy on bone-implant interfacial and superstructure loading conditions. A new protocol for three-dimensional (3D) fit assessment of dental restorations was investigated to obtain ensure correct virtual alignment and data acquisition of component fit. Material and Methods: Fit of ten three-unit screw-retained implant superstructures manufactured in high-noble alloy was investigated. To obtain 3D information on component fit a new triple-scan protocol was applied utilizing an industrial non-contact scanner (ATOS SO, GOM mbH, Braunschweig, Germany). Following separate digitization of the objects and the master-casts a third scan with the object screw retained in definitive position on the master-cast was acquired. Following established manual and automated alignment processes the third scan was used as a reference for transformation alignment ensuring that object and master-model were virtually aligned in correct 3D position. Fit assessment was performed using surface deviation statistics. A correlation coefficient was calculated based on five repeated measurements per framework. Results: Statistic analysis resulted in an intraclass correlation of 0.991 (95%-CI between 0.978 und 0.998) and therefore a statistically significant repeatability of measurments (p<0.001, F = 112.95). Discussion: The triple-scan protocol is a highly reliable approach to obtain true 3D information of component fit for conventional and implant retained restorations. Significant shortcomings of currently applied two-dimensional fit assessment methods of dental restorations can be eliminated. The most eminent one being the non-destructive nature of the approach allowing for repeated measurement of components and extensive virtual analytical options. In addition the true 3D information on fit provides is valuable for numerical models (e.g. FEA) reducing the number of assumptions for biomechanical research.