Data regarding the mechanical properties of three-dimensionally (3D) printed materials for occlusal splint manufacturing are scarce. The aim of the present study was to evaluate the flexural strength and surface hardness of modern 3D-printed occlusal splint materials and compare them with two control groups, namely, milled and conventional cold-polymerized occlusal splint materials. A total of 140 rectangular specimens were manufactured for the present study. The specimens were prepared in accordance with the International Organization for Standardization standards (ISO 20795-1:2013). Five 3D-printed (NextDent Ortho Rigid, Dental LT Clear, Dentona Flexisplint, Cosmos Bite Splint, and ProArt Print Splint), one milled (ProArt CAD Splint), and one cold-polymerized (ProBase Cold) occlusal splint materials were used to determine flexural strength and surface hardness values. The three-point flexure test was used for the determination of flexural strength values, while Vickers hardness was measured to determine surface hardness. Ten specimens (n = 10) of each material were tested using these procedures. One-way ANOVA and Tukey’s post-hoc test were used to analyze the obtained results (α = 0.05). The values of flexural strength ranged from 46.1 ± 8.2 MPa to 106 ± 8.3 MPa. The Vickers hardness values ranged from 4.9 ± 0.5 VHN to 20.6 ± 1.3 VHN. Significant differences were found among the tested materials (p < 0.0001). The milled and cold-polymerized materials yielded higher values for both flexural strength (only one 3D-printed resin had comparable results to cold-polymerized acrylics) and surface hardness. There are differences in the mechanical properties of the various tested occlusal splint materials. The flexural strength of most of the 3D-printed materials and their surface hardness values are still inferior when compared to the milled or cold-polymerized materials.

Objective: Estimating biological sex from orthopantomograms (OPGs) is a challenging and time-consuming task in forensic dentistry. This study aims to validate the predictive ability of a custom DASE-AI machine learning algorithm, leveraging multiple deep learning architectures, for determining patient sex from OPGs. Materials and Methods: The DASE-AI algorithm was trained and fine-tuned using DenseNet201, InceptionResNetV2, ResNet50, VGG16, VGG19, and Xception feature extractors on a dataset of 4,000 OPGs. The two best-performing models were subsequently validated on a large, independent cohort of 18,780 anonymized OPGs. Performance was assessed by calculating accuracy, sensitivity, specificity, and the area under the receiver operating characteristic curve (AU-ROC) for the entire cohort and across different age groups. The algorithms were compared for significant differences using the DeLong test. Results: The validated algorithms demonstrated high overall performance, with AU-ROC values of 0.896 and 0.899 across the entire cohort. Predictive ability was highest in the 20-30 years age group (AU-ROC: 0.925/0.924) and lowest in patients over 80 years (AU-ROC: 0.837/0.843). The DeLong test indicated no significant differences between the two top-performing algorithms. Conclusion: Machine learning analysis of OPGs is a viable and effective tool for determining patient sex. While a clear correlation exists between dataset size and algorithm accuracy, future model refinement should focus on accounting for changes in dental status acquired during training to further enhance performance. This technology has significant potential to facilitate and accelerate sex estimation in clinical and forensic contexts.