中华口腔医学杂志最新文献

{"title":"[Role and mechanism of methyltransferase-like 3 in promoting macrophage NLRP3 inflammatory responses in medication-related osteonecrosis of the jaw].","authors":"Z S Jiang, C Z Zhao, Z J Pan, L R Hu, N Zhang, J Pan","doi":"10.3760/cma.j.cn112144-20260425-00259","DOIUrl":"10.3760/cma.j.cn112144-20260425-00259","url":null,"abstract":"<p><p><b>Objective:</b> To investigate the role and mechanism of methyltransferase-like 3 (METTL3) in regulating macrophage inflammatory responses and the development of medication-related osteonecrosis of the jaw (MRONJ) at the single-cell level. <b>Methods:</b> Single-cell RNA sequencing was used to construct an immune atlas of zoledronic acid (ZA)-induced bisphosphonate-related osteonecrosis of the jaw(BRONJ)-like lesions in mice. Key epigenetic regulators were screened by bioinformatic analysis, and key genes were predicted using virtual knockdown analysis. METTL3 expression <i>in vivo</i> and <i>in vitro</i> was validated by immunohistochemical staining and Western blotting. <i>In vitro</i> knockdown and overexpression experiments were performed to clarify the regulatory effect of METTL3 on NLRP3 inflammasome activation in macrophages. <b>Results:</b> The single-cell atlas showed that ZA treatment significantly induced the recruitment of specific pro-inflammatory macrophage subsets in extraction sockets, accompanied by marked activation of inflammation-related pathways. Bioinformatic screening indicated that NLRP3 inflammasome-related genes served as a central bridge linking inflammatory responses to RNA processing and modification. Pseudotime trajectory and virtual knockdown analyses further suggested that METTL3 functioned as a key node in maintaining the pro-inflammatory state of macrophages. In vivo and in vitro experiments confirmed that ZA significantly induced METTL3 upregulation at both tissue and cellular levels, accompanied by an increase in global m6A levels. Functional assays showed that METTL3 knockdown markedly suppressed NLRP3 inflammasome activation, whereas METTL3 overexpression exerted the opposite effect. <b>Conclusions:</b> This study reveals the critical role of METTL3 as a positive regulator in promoting macrophage NLRP3 inflammatory responses in MRONJ pathogenesis. METTL3 is expected to provide clues for targeted therapy research on modulating local immune dysregulation and promoting bone repair in MRONJ.</p>","PeriodicalId":23965,"journal":{"name":"中华口腔医学杂志","volume":"61 8","pages":"1166-1177"},"PeriodicalIF":0.0,"publicationDate":"2026-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148631854","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"[Evaluation of nasal prosthesis automatic construction integrating a flexible point cloud deformation matching algorithm and generative artificial intelligence].","authors":"S W Guo, Q Z Qin, Y L Li, H Y Wu, A N Wen, Y J Zhu, W B Li, Y J Zhao, Y Wang","doi":"10.3760/cma.j.cn112144-20260207-00106","DOIUrl":"10.3760/cma.j.cn112144-20260207-00106","url":null,"abstract":"&lt;p&gt;&lt;p&gt;&lt;b&gt;Objective:&lt;/b&gt; To develop a point cloud non-rigid deformation algorithm, local projection deformation match (LPD-Match), for optimizing the edge transition of the nasal prosthesis complementing the generative artificial intelligence-based facial mesh generation network (FMGen-Net) model previously developed by the research team to automate the construction of target reference data for the nasal prosthesis. &lt;b&gt;Methods:&lt;/b&gt; The LPD-Match algorithm was developed based on the local projection deformation strategy. Three-dimensional facial scans from 20 subjects without significant facial deformities who first visited the Department of Prosthodontics, Peking University School and Hospital of Stomatology, between February 2023 and November 2023 were collected, and full nasal defect data were simulated on these scans. To evaluate the effectiveness of LPD-Match algorithm in edge transition optimization, a paired experiment was designed. For each model, two methods for constructing target reference data were compared. The control group utilized FMGen-Net alone, while the experimental group integrated FMGen-Net with the LPD-Match algorithm. The \"edge fitness\" metric was defined to evaluate the edge transition of the prosthesis by calculating the three-dimensional (3D) curve deviation and maximum curve deviation between the defect boundary curve and the prosthesis boundary curve. Additionally, the \"morphological similarity\" metric was introduced to assess the morphology restoration of the target reference data.Using the nasal region data from the original three-dimensional facial scans as the gold standard, the root mean square error of the three-dimensional morphological deviation and the maximum morphological deviation between the prosthesis design data and the gold standard were calculated for both groups. &lt;b&gt;Results:&lt;/b&gt; Analysis of the \"edge fitness\" metric showed that the root mean square error of the 3D curve deviation in the control group averaged (0.62±0.23) mm, whereas in the experimental group, it was (0.16±0.07) mm. The maximum curve deviation in the control group averaged (1.57±0.64) mm, while the experimental group showed a reduced value of (0.52±0.30) mm. Statistical analysis showed significant differences between the two groups (&lt;i&gt;t=&lt;/i&gt;11.33, &lt;i&gt;P&lt;/i&gt;0.001; &lt;i&gt;t=&lt;/i&gt;10.17, &lt;i&gt;P&lt;/i&gt;0.001), with the experimental group showing superior edge transition. For the \"morphological similarity\" analysis, the root mean square error of the 3D morphology deviation was 1.30 (1.00, 1.83) mm in the control group and 1.23 (0.90, 1.63) mm in the experimental group. The maximum morphology deviation in the control group was 4.26 (2.96, 5.02) mm, while the experimental group demonstrated 3.85 (2.92, 4.56) mm. Statistical testing showed significant differences between the two groups (&lt;i&gt;Z=&lt;/i&gt;-3.40, &lt;i&gt;P&lt;/i&gt;0.001; &lt;i&gt;Z=&lt;/i&gt;-2.09, &lt;i&gt;P=&lt;/i&gt;0.037), indicating the superior morphological restoration effect of the experimental group. &lt;b&gt;Conclusions:&lt;/b&gt; This","PeriodicalId":23965,"journal":{"name":"中华口腔医学杂志","volume":"61 7","pages":"996-1003"},"PeriodicalIF":0.0,"publicationDate":"2026-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148377293","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"[Selection of restorative strategies for extensive tooth defects in young permanent molars].","authors":"M Wan, X Xu, X D Bao, L W Zheng","doi":"10.3760/cma.j.cn112144-20251025-00429","DOIUrl":"10.3760/cma.j.cn112144-20251025-00429","url":null,"abstract":"<p><p>Restoration of extensive defects in young permanent molars is more complex than that in mature permanent teeth. Young permanent teeth are characterized by insufficient enamel mineralization, large pulp chambers, short occlusogingival distance, and unstable gingival margins, limiting both mechanical retention and bonding conditions. In addition, individual factors such as patient cooperation, caries risk, and orthodontic needs further complicate treatment planning. Based on current clinical evidence, this review summarizes the indications, advantages, and limitations of direct adhesive restorations, preformed crowns, provisional crowns, inlays/onlays/overlays, endocrowns, and computer-aided design and computer-aided manufacturing (CAD/CAM) indirect restorations, proposing a clinical decision-making framework for extensive defects in young permanent molars. The proposed pathway follows a \"three-level assessment and integrated decision-making\" approach, in which the physiological characteristics of young permanent teeth, the objective condition of the affected tooth, and patient-and treatment-related factors are comprehensively evaluated to develop an individualized restorative plan. When moisture control is achievable and tooth conditions are favorable, adhesive restorations with cusp coverage can be selected. When moisture control is difficult or rapid restoration of coronal integrity is required, preformed metal crowns may serve as interim restorations. For non-vital teeth with short occlusogingival distance and insufficient retention, endocrowns may be considered. At present, CAD/CAM resin-ceramic indirect adhesive restorations have shown favorable short-to medium-term outcomes over 18-24 months; however, the long-term performance still requires further validation through multicenter studies with larger sample sizes.</p>","PeriodicalId":23965,"journal":{"name":"中华口腔医学杂志","volume":"61 7","pages":"937-945"},"PeriodicalIF":0.0,"publicationDate":"2026-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148377304","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"[Factors influencing accuracy and corresponding strategies for intraoral scanning in full-arch implant rehabilitation].","authors":"Y P Guo, C X Jin, C Huang","doi":"10.3760/cma.j.cn112144-20260204-00095","DOIUrl":"10.3760/cma.j.cn112144-20260204-00095","url":null,"abstract":"<p><p>The rapid development of digital dentistry has established intraoral scanning as a core technology for acquiring digital impressions in full-arch implant rehabilitation for edentulous patients, and its accuracy plays a decisive role in clinical restorative outcomes. This article reviews the major factors influencing intraoral scanning accuracy in edentulous implant cases, including scanner-related factors, scanning procedure-related factors, and factors associated with scan bodies and auxiliary devices. Corresponding solutions and optimization strategies for improving scanning accuracy are also discussed, aiming to provide clinicians with more precise and reliable approaches for digital impression acquisition.</p>","PeriodicalId":23965,"journal":{"name":"中华口腔医学杂志","volume":"61 7","pages":"954-962"},"PeriodicalIF":0.0,"publicationDate":"2026-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148377237","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"[Artificial intelligence empowers prosthodontics: opportunities and challenges of technological integration and disciplinary innovation].","authors":"X Q Jiang","doi":"10.3760/cma.j.cn112144-20260324-00188","DOIUrl":"10.3760/cma.j.cn112144-20260324-00188","url":null,"abstract":"<p><p>The rapid development of artificial intelligence (AI) technology has injected new impetus into the digital transformation and high-quality development of prosthodontics. Its integration with core technologies such as deep learning, image recognition, and big data analytics has permeated the entire clinical chain of prosthodontics, including preoperative diagnosis and planning, intraoperative precise operation, and postoperative evaluation and follow-up. This has driven the discipline to transform from a traditional \"experience-driven\" model to a \"data-driven+precision\" model. This paper systematically reviews the clinical application scenarios and core innovative value of AI in the field of prosthodontics, deeply integrates international cutting-edge research results and domestic expert consensus, and analyzes the core bottlenecks faced in current technological development, clinical translation, ethical privacy, and other aspects. Combined with the industry development trend, targeted solutions and future prospects are proposed, aiming to provide academic references for the deep integration and standardized development of AI and prosthodontics, and promote the discipline to achieve the development goals of personalization, intelligence, and universalization.</p>","PeriodicalId":23965,"journal":{"name":"中华口腔医学杂志","volume":"61 7","pages":"923-928"},"PeriodicalIF":0.0,"publicationDate":"2026-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148377066","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"[Mechanism of low-concentration IL-6 in promoting osteogenic differentiation of human stem cells from apical papilla via the JAK2-STAT3 signaling pathway under inflammatory condition].","authors":"T F Du, W F Zhao, L D Wu, G Hu, K Gan, K Song, J F Zhu","doi":"10.3760/cma.j.cn112144-20260421-00244","DOIUrl":"10.3760/cma.j.cn112144-20260421-00244","url":null,"abstract":"&lt;p&gt;&lt;p&gt;&lt;b&gt;Objective:&lt;/b&gt; To investigate the effects and underlying mechanisms of low-concentration interleukin-6 (IL-6) on the proliferation, migration and osteogenic differentiation of human stem cells from apical papilla (hSCAPs). &lt;b&gt;Methods:&lt;/b&gt; hSCAPs were isolated from the apical papilla of third molars extracted in the Department of Oral Maxillofacial Surgery, the First Affiliated Hospital of Zhengzhou University, and identified by flow cytometry and immunofluorescence. Cell counting kit-8 (CCK-8) and Transwell migration assay were used to detect the effects of IL-6 in different concentrations (10, 20, 50 μg/L) on the proliferative and migration ability of hSCAPs. Alizarin red staining, cetyl pyridine chloride quantification, alkaline phosphatase (ALP) staining and ALP activity assay were performed to evaluate the osteogenic differentiation capacity of hSCAPs under inflammatory conditions induced by different concentrations of IL-6, so as to determine the optimal IL-6 concentration to simulate the inflammatory microenvironment of hSCAPs osteogenic differentiation. Real-time fluorescence quantitative PCR (RT-qPCR) and Western blotting were used to detect the osteogenic differentiation ability of hSCAPs treated with low-concentration (10 μg/L) IL-6. Western blotting, Transwell migration assay, ALP staining, ALP activity assay and RT-qPCR were applied to explore the mechanism by which low-concentration IL-6 promotes the migration and osteogenic differentiation of hSCAPs via the Janus kinase 2-signal transducer and activator of transcription 3 (JAK2-STAT3) signaling axis. &lt;b&gt;Results:&lt;/b&gt; The isolated and cultured hSCAPs were positive for surface markers CD24 and vimentin, while negative for CD45 and cytokeratin. Different concentrations of IL-6 exerted no significant effect on the proliferation of hSCAPs (&lt;i&gt;P&lt;/i&gt;0.05). The number of migrated cells (11 463.33±622.70), cetyl pyridine chloride quantification (2.087±0.121) and ALP activity (94.11±4.06) U/g of hSCAPs in 10 μg/L IL-6 treatment were all significantly increased than those in 0 μg/L IL-6 control group, respectively [8 246.00±674.02, 1.330±0.096, and (53.87±3.70) U/g] (all &lt;i&gt;P&lt;/i&gt;0.05). Moreover, the cell migration number and ALP activity in the 10 μg/L IL-6 group were more remarkably increased than in 20 μg/L IL-6 group [9 503.33±275.03, (74.44±3.55) U/g] (both &lt;i&gt;P&lt;/i&gt;0.01). After osteogenic mineralization induction of hSCAPs under low-concentration IL-6 intervention, the mRNA expression levels of osteogenic-related factors Osterix (OSX) and Runt-related transcription factor 2 (Runx2) at day 3 (2.033±0.308, 1.433±0.215) and day 7 (3.497±0.153, 3.360±0.061) were significantly upregulated compared to the control group (all &lt;i&gt;P&lt;/i&gt;0.05), while the mRNA level of osteocalcin (OCN) was significantly elevated at day 7 (2.950±0.315) (&lt;i&gt;P&lt;/i&gt;0.05). The protein expression levels of OCN were significantly increased at day 3 (2.077±0.447) and day 7 (5.023±0.537) than those in the control group ","PeriodicalId":23965,"journal":{"name":"中华口腔医学杂志","volume":"61 7","pages":"1039-1049"},"PeriodicalIF":0.0,"publicationDate":"2026-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148377167","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"[A retrospective study on changes in buccal bone thickness following immediate implant placement after dental trauma in the esthetic zone].","authors":"H R Zhou, Y W Wu, X H Lin, H D Wang, X Y Li, L Jiang","doi":"10.3760/cma.j.cn112144-20260129-00063","DOIUrl":"10.3760/cma.j.cn112144-20260129-00063","url":null,"abstract":"<p><p><b>Objective:</b> To evaluate the effects of flap and flapless surgical approaches on changes in buccal bone thickness around implants in patients undergoing immediate implant placement after dental trauma under a fiber-reinforced composite (FRC) splint-supported immediate provisionalization protocol, and to provide clinical evidence for surgical decision-making in the esthetic zone following dental trauma. <b>Methods:</b> This retrospective cohort study included patients who presented to the Department of General Dentistry, Shanghai Ninth People's Hospital between December 2022 and December 2025 with traumatic maxillary anterior tooth loss requiring immediate implant placement and who had completed definitive prosthetic rehabilitation. According to whether a flap was elevated during surgery, patients were allocated to either a flap group or a flapless group. All implants were provisionally restored immediately after surgery using a fiber-reinforced resin splint combined with natural teeth or resin crowns. Cone-beam computed tomography (CBCT) scans obtained preoperatively, immediately after surgery, and at 4 months postoperatively were used to assess preoperative buccal bone plate thickness and postoperative changes in buccal bone thickness around the implants in both groups. <b>Results:</b> No statistically significant differences in buccal bone thickness were observed between the flap and flapless groups at any measurement level immediately after surgery or at 4 months postoperatively (all <i>P</i>0.05). At 4 months postoperatively, buccal bone thickness at most measurement levels in both groups was significantly reduced compared with the immediate postoperative values (all <i>P</i>0.05), except at 12 mm apical to the implant platform in the flapless group, where no significant bone resorption was detected (all <i>P</i>0.05). No significant differences were found between the two groups in buccal bone thickness changes at any measurement level at 4 months postoperatively (all <i>P</i>0.05). <b>Conclusions:</b> For immediate implant placement following dental trauma in the esthetic zone, when strict case selection and a FRC splint-supported immediate provisionalization protocol are applied, flap and flapless surgical approaches do not appear to differ significantly in their ability to maintain short-term buccal hard tissue stability.</p>","PeriodicalId":23965,"journal":{"name":"中华口腔医学杂志","volume":"61 7","pages":"1013-1019"},"PeriodicalIF":0.0,"publicationDate":"2026-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148377122","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"[A comparative study of four deep learning instance segmentation models for tooth segmentation and tooth numbering identification in intraoral occlusal images].","authors":"X Z Liu, X Zhou, L T Yu, Q Chen, L Wu","doi":"10.3760/cma.j.cn112144-20260225-00130","DOIUrl":"10.3760/cma.j.cn112144-20260225-00130","url":null,"abstract":"<p><p><b>Objective:</b> To compare the performance of four deep learning instance segmentation models in tooth contour segmentation and Fédération Dentaire Internationale (FDI) tooth numbering recognition on intraoral occlusal images, and to preliminarily evaluate the clinical acceptability of the best-performing model. <b>Methods:</b> A total of 700 adult participants, comprising 343 males and 357 females aged 35-87 years, including individuals with complete dentition and those with dentition defects, who attended the Department of Prosthodontics, Hospital of Stomatology, China Medical University, between March 2022 and October 2024 and underwent standardized intraoral occlusal photography were enrolled, yielding 1 400 images; after quality screening, 1 300 images from 650 participants were finally included. The dataset was divided by images into a training set (910 images), a validation set (195 images), and a test set (195 images) at a ratio of 7.0∶1.5∶1.5, and only the training set was augmented to 2 730 images. Twenty-eight tooth categories, excluding third molars, were annotated according to the FDI two-digit notation system. Under unified training conditions, four models: Mask R-CNN, RTMDet, YOLOv8-seg, and YOLOv11-seg were compared, and test-set performance was evaluated using segmentation mean average precision at an IoU threshold of 0.5 (seg_mAP@50), precision, recall, and F1-score, while five attending prosthodontists were recruited to compare tooth numbering accuracy with that of the best-performing model and to assess its practical applicability. <b>Results:</b> YOLOv11-seg outperformed the other models across all metrics, achieving a seg_mAP@50 of 98.2%, a precision of 96.3%, a recall of 96.8%, and an F1-score of 96.5%. The tooth numbering accuracy of this model was 97.5% (2 145/2 200), showing no statistically significant difference from the mean accuracy of the five attending prosthodontists (98.6%±1.3%) (<i>t</i>=1.89, <i>P</i>=0.130), and 98.6% of the model outputs were rated as clinically acceptable. <b>Conclusions:</b> YOLOv11-seg achieved highly accurate tooth segmentation and tooth numbering recognition on intraoral occlusal images, with performance comparable to that of attending prosthodontists, thus providing a technical basis for subsequent dentition-defect analysis based on tooth-level information.</p>","PeriodicalId":23965,"journal":{"name":"中华口腔医学杂志","volume":"61 7","pages":"972-978"},"PeriodicalIF":0.0,"publicationDate":"2026-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148377191","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"[Progress in basic research on hydrogel-based drug delivery systems for the treatment of peri-implantitis].","authors":"M K Liu, Z H Zhang, L Lang, H H Liu","doi":"10.3760/cma.j.cn112144-20251012-00399","DOIUrl":"10.3760/cma.j.cn112144-20251012-00399","url":null,"abstract":"<p><p>Peri-implantitis is a common complication following oral implantation that seriously jeopardizes long-term implant stability and prosthetic outcomes. Conventional therapeutic modalities, including non-surgical and surgical interventions, exhibit inherent limitations. Hydrogel-based drug delivery systems (hDDSs), as an emerging platform, are increasingly recognized for their therapeutic potential of favorable biocompatibility, sustained drug-release kinetics, and advantage of local administration in basic studies of peri-implantitis, thus offering broad translational prospects. This review outlines the hDDSs currently employed in peri-implantitis research and summarizes the latest advances. Moreover, we analyze and compare the strengths and limitations of various hydrogel drug-loading strategies, discuss the challenges hindering clinical translation, aiming to provide guidance and inspiration for future clinical management and research of peri-implantitis.</p>","PeriodicalId":23965,"journal":{"name":"中华口腔医学杂志","volume":"61 7","pages":"1066-1073"},"PeriodicalIF":0.0,"publicationDate":"2026-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148377231","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"[A retrieval-augmented generation-based clinical decision support system for the restoration of endodontically treated teeth].","authors":"J R Zhai, J W Guo, W Y Su, T M Sha, P Ding, Y Wang","doi":"10.3760/cma.j.cn112144-20260409-00225","DOIUrl":"10.3760/cma.j.cn112144-20260409-00225","url":null,"abstract":"&lt;p&gt;&lt;p&gt;&lt;b&gt;Objective:&lt;/b&gt; To develop and evaluate Clinical Decision Surpport System-augmented by retrieval-augmented generation (RAG)-Professional (CDSS-RAG-Pro), a RAG-based clinical decision support system, through multi-level comparison of multiple large language models (LLM), and to provide an objective decision-making aid for the clinical restoration of endodontically treated teeth. &lt;b&gt;Methods:&lt;/b&gt; A structured knowledge base was constructed by systematically retrieving and incorporating authoritative prosthodontic sources, including the Guideline for the Restoration of Tooth Defects, Dentition Defects and Edentulism (2022 Edition). A RAG module was developed accordingly and integrated into three LLM (DeepSeek-V3, Qwen 2.5-72B-Instruct, and GPT-3.5-turbo) to form the CDSS-RAG-Pro clinical decision support system for post-endodontic restoration. To evaluate the decision-making capability of this system, an expert panel was established. Thirty typical clinical cases were selected from the Department of Prosthodontics, Hospital of Stomatology, Sun Yat-sen University, from May 2024 to December 2025. The three LLM generated restorative plans under both RAG and non-RAG modes. Using a Likert scale, the expert panel scored the restorative plans across three dimensions: decision accuracy, logical coherence, and professional readability. Meanwhile, the expert panel drafted standard restorative reports. Two objective metrics, F1 score of the recall-oriented understudy for gisting evaluation based on longest common subsequence (ROUGE-L F1) and F1 score of the bidirectional encoder representations from Transformer-based semantic similarity score (BERTScore F1), were used to evaluate the semantic similarity between the generated restorative plans and the standard restorative reports. ROUGE-L F1 focused on assessing the model' s mastery of textual knowledge such as restorative guidelines, while BERTScore F1 focused on semantic consistency in reasoning patterns and decision-making logic. Generalized estimating equations were used to analyze differences in decision performance across LLM and between RAG and non-RAG modes. &lt;b&gt;Results:&lt;/b&gt; Compared with the non-RAG mode, all models showed improved performance across all evaluation metrics in the RAG mode, with statistically significant differences (all &lt;i&gt;P&lt;/i&gt;0.05). The highest BERTScore F1 in the RAG mode was 0.91 (DeepSeek-V3). Significant differences were observed among the three LLM across all evaluation metrics (all &lt;i&gt;P&lt;/i&gt;0.05). Significant interaction effects were found between model type and RAG mode for decision accuracy score (interaction χ&lt;sup&gt;2&lt;/sup&gt;=10.11, &lt;i&gt;P&lt;/i&gt;0.001) and ROUGE-L F1 (interaction χ&lt;sup&gt;2&lt;/sup&gt;=8.57, &lt;i&gt;P&lt;/i&gt;0.05). &lt;b&gt;Conclusions:&lt;/b&gt; The RAG-based CDSS-RAG-Pro improves the accuracy and interpretability of general-purpose LLM in decision-making for post-endodontic tooth restoration. This study demonstrates the unique clinical safety and scalability advantages of RAG as a lightwei","PeriodicalId":23965,"journal":{"name":"中华口腔医学杂志","volume":"61 7","pages":"986-995"},"PeriodicalIF":0.0,"publicationDate":"2026-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148376389","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}