OAR@UM Collection:

Artificial intelligence architectures in oral rehabilitation : a focused review of deep learning models for implant planning, prosthodontic design, and peri-implant diagnosis

2026-01-01T00:00:00Z

Title: Artificial intelligence architectures in oral rehabilitation : a focused review of deep learning models for implant planning, prosthodontic design, and peri-implant diagnosis Authors: Dawa, Hossam; Aroso, Carlos; Vinhas, Ana Sofia; Cortes, Arthur R. G. Abstract: Deep learning is increasingly integrated into oral rehabilitation workflows, particularly in implant planning, prosthodontic design automation, and peri-implant diagnosis. However, reported performance is heterogeneous and difficult to compare across tasks, modalities, and validation designs. The goal of this study was to critically analyze deep learning architecture families applied to oral rehabilitation and to provide task-driven selection guidance supported by an evidence table reporting dataset characteristics, validation strategy, and performance metrics. A focused narrative review was conducted using transparent, database-specific search criteria (final n = 10 included studies), emphasizing implant planning (cone–beam computed tomography [CBCT]-based segmentation), prosthodontic design (intraoral scan [IOS]/mesh inputs), and peri-implant diagnosis (periapical/panoramic radiographs). Evidence certainty for each clinical task was assessed using GRADE-informed ratings (High/Moderate/Low/Very Low). Extracted variables included clinical task, imaging modality, dataset size, architecture, validation strategy (in ternal vs. internal + external), split level, ground truth protocol, and performance metrics. Astructured computational and hardware feasibility analysis was conducted for each architecture family to support real-world deployment planning. Encoder–decoder networks (U-Net/nnU-Net) dominate CBCT segmentation for implant planning, while detection architectures (Faster R-CNN, YOLO) support implant localization and peri-implant assess ment on radiographs. Generative models (3D GANs, transformer-based point-to-mesh networks) enable crown design from three-dimensional scans. Hybrid CNN–Transformer architectures show promise for multimodal CBCT–IOS fusion, though direct evidence from the included studies remains limited to a single study. External validation remains uncommonyetessential given the risk of domain shift. In conclusion, architecture selection should be anchored to task geometry (2D vs. 3D), artifact burden, and required clinical output type. Reporting standards should prioritize dataset transparency, validation rigor, multi-center external testing, and uncertainty-aware outputs.

Progress and insights in health and biomedical research

2026-01-01T00:00:00Z

Title: Progress and insights in health and biomedical research Authors: Cortes, Arthur R. G.; Barbosa, Joana; Gonçalves, Virgínia; Henriques, Bruno; Sarmento, Bruno; Rodrigues, Célia Fortuna; Souza, Júlio C. M.; Tiritan, Maria Elizabeth; Vieira Brito, Nuno; Bezerra Cass, Quezia; Dinis-Oliveira, Ricardo Jorge; Warnakulasuriya, Saman; Bolanos-Garcia, Victor M.; Bousbaa, Hassan Abstract: On behalf of the Editorial Team of Scientific Letters, we are pleased to announce the publication of the journal’s fourth issue. In 2025, Scientific Letters continued its mission of disseminating high-quality open-access research across the fields of biology and medicine. The two review articles and five original research papers published this year highlight both foundational biological mechanisms and pressing clinical and public health issues. João Carvalho et al. performed a systematic integrative review addressing different orthodontic approaches to the traction of impacted canines. Their work critically analyzed existing techniques, highlighting the advantages and limitations of surgical and orthodontic strategies, and emphasizing the importance of individualized treatment planning to optimize functional and esthetic outcomes. This review contributes valuable guidance for clinical decision-making in orthodontic practice.

Impact of different yttrium oxide concentrations and sintering protocols on the flexural strength and optical properties of monolithic zirconia

2026-01-01T00:00:00Z

Title: Impact of different yttrium oxide concentrations and sintering protocols on the flexural strength and optical properties of monolithic zirconia Authors: DeVito-Moraes, André Guaraci; Vardasca, Isabela Souza; Peñarrocha-Diago, Miguel; Cortes, Arthur R. G. Abstract: This study investigates the mechanical and optical properties of monolithic zirconia used in dentistry, focusing on how different concentrations of yttrium oxide and varied sintering times affect the material. A critical trade-off in ceramics has been reported in the literature, in which increased crystalline content (like in zirconia) leads to higher mechanical strength but lower aesthetic translucency. However, detailed information on this trade-off process for different types of zirconia is lacking. A total of seven types of zirconia varying in yttria content (3 mol% to 5 mol%) were tested across four sintering protocols available in a laboratory zirconia sintering device: Slow (12 h), Standard (8 h), Fast (3.5 h), and Ultrafast (1.15 h). The primary findings indicate that while a higher yttria concentration correlates with lower flexural strength and high translucency, the sintering time generally did not compromise mechanical strength or color variation across most samples. Nevertheless, the Fast and Ultrafast protocols did significantly reduce the translucency of zirconia with a high concentration of yttrium oxide.

Integrating generative artificial intelligence in clinical dentistry : enhancing diagnosis, treatment planning, and procedural precision through advanced knowledge representation and reasoning

2025-01-01T00:00:00Z

Title: Integrating generative artificial intelligence in clinical dentistry : enhancing diagnosis, treatment planning, and procedural precision through advanced knowledge representation and reasoning Authors: Dawa, Hossam; Cortes, Arthur R. G.; Ribeiro, Carlos; Neves, José; Vicente, Henrique Abstract: Generative artificial intelligence (GAI) is poised to transform clinical dentistry by enhancing diagnostic accuracy, personalizing treatment planning, and improving procedural precision. This study integrates logic programming and entropy within knowledge representation and reasoning to generate hypotheses, quantify uncertainty, and support clinical decisions. A six-month longitudinal questionnaire was administered to 127 dentists, of whom 119 provided valid responses across four dimensions: current use and knowledge (CUKD), potential applications (PAD), future perspectives (FPD), and challenges and barriers (CBD). Responses, analyzed with both classical statistics and entropy-based measures, revealed significant differences among dimensions (𝘱 < 0.01, η2 = 0.14). CUKD, PAD, and FPD all increased steadily over time (baseline means 2.32, 3.06, and 3.27; rising to 3.75, 4.51, and 4.71, respectively), while CBD remained more variable (1.87–3.87). The overall entropic state declined from 0.43 to 0.31 (p = 0.018), reflecting reduced uncertainty. Statistical and entropy-derived trends converged, suggesting growing professional clarity and cautious acceptance of GAI. These findings indicate that, despite persistent concerns, GAI holds promise for advancing adaptive and evidence-driven dental practice.