Background/Objectives: Accurate diagnosis and staging of periodontitis rely on clinical measurements and radiographic assessment of alveolar bone loss. Methods: Studies published between 1 January 2020 and 31 October 2025 were searched in the Web of Science and PubMed databases in accordance with the PRISMA guidelines. Original research articles that evaluated periodontal pathology on radiographic images using fractal analysis and/or artificial intelligence approaches, with clearly defined methodologies, were included. Due to methodological heterogeneity, a quantitative meta-analysis was not performed, and the findings were summarized using a narrative synthesis approach. Results: Of 346 records, 80 studies (9 fractal, 71 AI) met the inclusion criteria. Fractal analysis studies predominantly calculated the fractal dimension on panoramic or periapical radiographs using the box-counting method. In artificial intelligence studies, the task types mainly comprised classification, segmentation, detection, and hybrid approaches (multi-stage models or models combining multiple tasks). Panoramic and intraoral radiographs were the predominant imaging modalities. Performance metrics were reported across wide ranges (sensitivity 0.23–1.00; accuracy 0.506–1.00; specificity 0.41–0.99; F1 score 0.15–0.99; AUC 0.75–0.99), and in some studies, these metrics were only partially reported. Conclusions: Fractal analysis and artificial intelligence approaches offer objective and reproducible assessment of periodontal bone loss; however, methodological and reporting heterogeneity limit comparability and generalizability. Standardization of ROI definitions, datasets, study designs, and performance reporting is needed to improve clinical applicability. Future research should also explore hybrid models that combine the quantitative microstructural insights of fractal analysis with the automated detection capabilities of artificial intelligence to enhance diagnostic precision.

Osseointegration is defined as the structural and functional integration between alveolar bone and a dental implant. Photo-functionalization (PF) refers to ultraviolet (UV)-induced surface modifications of titanium implants, including changes in physicochemical properties and biological responsiveness. The aim of this study was to evaluate the effects of PF on early osteoblast responses related to osseointegration on titanium dental implants in vitro. We hypothesized that PF applied to titanium implants enhances early osteoblast responses related to osseointegration. Sixteen titanium dental implants were divided into two equal groups of eight: untreated (PF−) and PF-treated (PF+). PF+ implants were exposed to UV light at 172 nm for 10 s. An additional cell-only control group was incubated without an implant. All groups were cultured in vitro with SAOS-2 human osteoblast-like cells. Cell proliferation and viability were assessed using standard in vitro assays, and DNA damage was evaluated using the Terminal deoxynucleotidyl transferase [TdT] dUTP Nick End Labeling (TUNEL) assay. Early cellular responses related to osseointegration were assessed by evaluating adhesion-related vinculin levels and alkaline phosphatase (ALP) activity. After 24 h of incubation, cell proliferation was comparable between the groups, whereas after 48 h, cell number was significantly lower in the PF− group compared with the PF+ group and the control group (p = 0.039). Osteoblast viability was significantly lower in the PF− group than in the control group (53% vs. 94%, p = 0.002), while the PF+ group showed a numerically higher viability value than the PF− group (70% vs. 53%). TUNEL assay showed no statistically significant difference in DNA damage among the groups, although the PF− group showed a slightly higher TUNEL-positive cell ratio (p = 0.563). Vinculin levels were significantly higher in the PF+ group at both 24 and 48 h compared with the PF− group and control group (p < 0.0001). ALP activity increased significantly over time in cells incubated with PF-treated implants (p < 0.0001). Within the limitations of this exploratory pilot in vitro study, UV photo-functionalization was associated with more favorable early osteoblast-like cell responses on titanium dental implants, particularly in terms of proliferation, adhesion-related vinculin levels, and ALP response. PF did not increase TUNEL-positive cell ratios compared with untreated implants under the present experimental conditions. These findings should be interpreted as preliminary biological evidence and require confirmation through larger experimental designs, detailed physicochemical surface characterization, and in vivo validation.

Background and Aim: To provide the first bibliometric analysisfocused exclusively on artificial intelligence (AI) in periodontology,mapping publication trends, contributors, collaboration networks,and thematic hotspots.Materials and Methods: The Web of Science Core Collectionwas searched on May 8, 2025 using AI- and periodontologyrelatedterms. Only original research articles were included.Bibliographic data from 206 eligible records were analyzed inVOSviewer (v1.6.20) to generate networks of co-authorship,keyword co-occurrence, country- and institution-levelcollaboration, citation patterns, and source journals; minimumthresholds were applied for robust clustering.Results: Relevant literature began in 2005, with a sharp riseafter 2018 and a peak in 2024. The United States and Chinawere the most productive countries, and Türkiye ranked third.In the co-authorship network, Kaan Orhan occupied the mostcentral position. Dominant keywords were “periodontitis,”“machine learning,” and “deep learning,” indicating an emphasison diagnostic/classification tasks. Temporal overlays showed ashift from early algorithm development toward clinically orienteddiagnostic and decision-support applications. Citation mappingsuggested that earlier influential papers were often isolated,whereas recent studies were more integrated. The Journal ofDentistry was the most productive and well-connected source.Conclusion: AI research in periodontology is expanding rapidlywith increasing clinical orientation. Strengthening internationalcollaboration and extending work beyond diagnosis totherapeutic and prognostic applications could enhance clinicalimpact and global integration.

Large language models (LLMs) are increasingly used in dental education; however, their performance on high-stakes national examinations remains insufficiently explored. The Dentistry Specialization Examination (DUS) in Türkiye is a standardized gateway to postgraduate dental training and includes a dedicated periodontology component. No previous study has evaluated contemporary LLM accuracy on DUS periodontology questions or examined whether Bloom's Revised Taxonomy predicts model performance. A total of 128 validated periodontology single-best-answer multiple-choice questions from publicly released DUS booklets (2012-2021) were analyzed. Each question was presented in Turkish to ChatGPT-5, Gemini 2.5 Pro, Claude Sonnet 4.5, and DeepSeek v3.2 using a standardized single-answer prompt. Responses were scored against the official key. Two periodontology specialists independently classified items according to Bloom's Revised Taxonomy and eight periodontal content domains. Accuracy differences were assessed using Cochran's Q test and Bonferroni-adjusted McNemar tests. The association between Bloom level and accuracy was examined using binary logistic regression. Accuracy rates ranged from 80.47% to 91.41% across the four large language models. Incorrect responses were distributed across content domains, with a greater number observed in the non-surgical and surgical periodontal therapy domains. Regarding cognitive demand, inter-rater reliability for Bloom classification was high (κ = 0.948); however, Bloom level was not a statistically significant predictor of model accuracy (p > 0.05). Differences in performance across models were observed; a statistically significant difference was identified only between Gemini 2.5 Pro and DeepSeek v3.2 (p = 0.0081). Contemporary LLMs showed high accuracy on released DUS periodontology single-best-answer multiple-choice items. While performance was strong in this structured, text-based examination context, the ability of these models to replicate higher-order clinical reasoning remains uncertain. Within the context of the analyzed DUS periodontology dataset, these findings contribute to ongoing discussions regarding the design and potential vulnerabilities of high-stakes single-best-answer MCQ-based assessments in the era of artificial intelligence. Consequently, it may be important for educators to understand the potential role of LLMs in students' preparation for assessments, suggesting that future research may explore modifications in test development to better capture student understanding.