The integration of computational intelligence and augmented reality has become increasingly prevalent in dental practices. Three-dimensional (3D) printing techniques have now become routine in orthodontics, prosthetics, and maxillofacial surgery.
The objective of this study is to assess the effectiveness of stereolithography (SLA) printed models compared to traditional plaster casts over a three-year period. The experimental dataset consisted of 36 orthodontic patients, each with SLA printed models and plaster casts for both the upper and lower jaws, resulting in a total of 72 scans in the form of Standard Template Library (STL) files and 72 traditional impressions.
The upper dental arch models were constructed using an SLA 3D printer, employing a blue 405 nm laser beam to solidify a liquid polymer. The classical plaster casts were prepared in a dental laboratory.
The models were retained for long-term orthodontic therapy control. To evaluate the differences between the dental models, a laboratory scanner was used to generate virtual casts.
The meshes obtained were adjusted and pre-aligned using the best-fit algorithm. Subsequently, registration of the models was performed using the iterative closest point (ICP) algorithm.
Distances between the point clouds and meshes for each point of the printed model were calculated by determining the nearest triangle on the reference mesh (cast). Additionally, the model surfaces were assessed using a scanning electron microscope and a stereomicroscope.
The results from 22 experimental datasets indicated a high level of agreement between the plaster casts and virtual surfaces, with a mean absolute difference of 0.018 mm and a standard deviation of 0.17 mm. These values were obtained by averaging 231,178 data points for each comparison.
Overall, this study demonstrates the comparability and accuracy of SLA printed models in relation to traditional plaster casts, supporting their potential as reliable alternatives in dental practice.