<div><h3>Statement of problem</h3><div>To ensure long-term stability and performance, removable partial dentures (RPDs) must be fitted precisely. Although 3-dimensional (3D) printing has been widely used, studies comparing various methods of manufacturing and designs for mandibular removable partial denture (RPD) frameworks are lacking.</div></div><div><h3>Purpose</h3><div>The aim of this in vitro study was to compare the trueness of RPD metal frameworks with 2 different major connector design types (lingual bar and lingual plate) fabricated with direct and indirect metal 3D printing with those fabricated with the conventional lost wax technique.</div></div><div><h3>Material and methods</h3><div>A Type IV stone cast of a Kennedy classification II modification 1 partially edentulous mandibular arch was prepared as the reference cast. A total of 30 definitive casts were fabricated from the reference cast and scanned into standard tessellation language (STL) files. Ten of these casts were used to fabricate cobalt chromium (Co-Cr) frameworks with the conventional lost wax technique (CLW group), 10 were used to fabricate frameworks by printing into a castable resin pattern followed by conventional casting (RPC group), and 10 were used to print metal frameworks directly using a selective laser melting printer (DSLM group). For each fabrication method, the group was divided into 2 design types: 5 casts for lingual plate frameworks and 5 for lingual bar frameworks (<em>n</em>=5). All metal frameworks were scanned and superimposed with the definitive casts with the Geomagic Control X software program. Gap discrepancies were measured as mean ±standard deviation (trueness), and the data were statistically analyzed with the 2-way ANOVA test (α=.05) to determine the interaction of the fabrication methods and design types on trueness. The Tukey HSD test was used to compare mean trueness among groups (α=.05).</div></div><div><h3>Results</h3><div>The CLW group demonstrated the highest overall gap discrepancies in the lingual plate frameworks, measuring 0.207 ±0.035 mm, whereas the DSLM group recorded the lowest value at 0.141 ±0.022 mm. No statistically significant difference was found between the DSLM and RPC groups (<em>P</em>>.05). The DSLM group exhibited the lowest mean gap for the lingual bar frameworks, 0.091 ±0.016 mm, with no significant difference between the RPC and CLW groups (<em>P</em>>.05). The 2-way ANOVA indicated that trueness was significantly affected by fabrication methods and design types. The color mapping of the lingual plate and bar in the DSLM frameworks shows minimal deviations relative to other groups.</div></div><div><h3>Conclusions</h3><div>The direct and indirect 3D printing of lingual plate RPD frameworks demonstrated better trueness compared with conventional casting methods. Direct 3D metal printing showed better fit and lower discrepancy for lingual bar designs. Both conventional and 3D printing methods demonstrated clini
{"title":"Comparing the trueness of 3D printing and conventional casting for removable partial denture metal framework fabrication in different mandibular major connectors designs: An in vitro study","authors":"Chalermkiet Sasithornvechakul DDS , Pisaisit Chaijareenont DDS, MS, PhD , Pattarika Angkasith DDS, MS","doi":"10.1016/j.prosdent.2025.10.035","DOIUrl":"10.1016/j.prosdent.2025.10.035","url":null,"abstract":"<div><h3>Statement of problem</h3><div>To ensure long-term stability and performance, removable partial dentures (RPDs) must be fitted precisely. Although 3-dimensional (3D) printing has been widely used, studies comparing various methods of manufacturing and designs for mandibular removable partial denture (RPD) frameworks are lacking.</div></div><div><h3>Purpose</h3><div>The aim of this in vitro study was to compare the trueness of RPD metal frameworks with 2 different major connector design types (lingual bar and lingual plate) fabricated with direct and indirect metal 3D printing with those fabricated with the conventional lost wax technique.</div></div><div><h3>Material and methods</h3><div>A Type IV stone cast of a Kennedy classification II modification 1 partially edentulous mandibular arch was prepared as the reference cast. A total of 30 definitive casts were fabricated from the reference cast and scanned into standard tessellation language (STL) files. Ten of these casts were used to fabricate cobalt chromium (Co-Cr) frameworks with the conventional lost wax technique (CLW group), 10 were used to fabricate frameworks by printing into a castable resin pattern followed by conventional casting (RPC group), and 10 were used to print metal frameworks directly using a selective laser melting printer (DSLM group). For each fabrication method, the group was divided into 2 design types: 5 casts for lingual plate frameworks and 5 for lingual bar frameworks (<em>n</em>=5). All metal frameworks were scanned and superimposed with the definitive casts with the Geomagic Control X software program. Gap discrepancies were measured as mean ±standard deviation (trueness), and the data were statistically analyzed with the 2-way ANOVA test (α=.05) to determine the interaction of the fabrication methods and design types on trueness. The Tukey HSD test was used to compare mean trueness among groups (α=.05).</div></div><div><h3>Results</h3><div>The CLW group demonstrated the highest overall gap discrepancies in the lingual plate frameworks, measuring 0.207 ±0.035 mm, whereas the DSLM group recorded the lowest value at 0.141 ±0.022 mm. No statistically significant difference was found between the DSLM and RPC groups (<em>P</em>>.05). The DSLM group exhibited the lowest mean gap for the lingual bar frameworks, 0.091 ±0.016 mm, with no significant difference between the RPC and CLW groups (<em>P</em>>.05). The 2-way ANOVA indicated that trueness was significantly affected by fabrication methods and design types. The color mapping of the lingual plate and bar in the DSLM frameworks shows minimal deviations relative to other groups.</div></div><div><h3>Conclusions</h3><div>The direct and indirect 3D printing of lingual plate RPD frameworks demonstrated better trueness compared with conventional casting methods. Direct 3D metal printing showed better fit and lower discrepancy for lingual bar designs. Both conventional and 3D printing methods demonstrated clini","PeriodicalId":16866,"journal":{"name":"Journal of Prosthetic Dentistry","volume":"135 3","pages":"Pages 586.e1-586.e7"},"PeriodicalIF":4.8,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145426863","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Response to Letter to the Editor regarding, “Evaluation of color matching accuracy using artificial intelligence applications and a spectrophotometer: A photometric analysis”","authors":"Nurşen Şahin DDS, Necati Kaleli DDS, PhD, Çağrı Ural DDS, PhD","doi":"10.1016/j.prosdent.2025.12.037","DOIUrl":"10.1016/j.prosdent.2025.12.037","url":null,"abstract":"","PeriodicalId":16866,"journal":{"name":"Journal of Prosthetic Dentistry","volume":"135 3","pages":"Pages 644.e1-644.e2"},"PeriodicalIF":4.8,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146096845","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2025-05-22DOI: 10.1016/j.prosdent.2025.04.038
Mohammadjavad Shirani DDS, MSc
Statement of problem
The accuracy of DeepSeek and the latest versions of ChatGPT and Gemini in responding to prosthodontics questions needs to be evaluated. Additionally, the extent to which the performance of these chatbots changes through user interactions remains unexplored.
Purpose
The purpose of this longitudinal repeated-measures experimental study was to compare the performance of ChatGPT (4o), DeepSeek (R1), and Gemini (2 Pro) in answering multiple-choice (MC) and short-answer (SA) fixed prosthodontics questions over 4 consecutive weeks after exposure to correct responses.
Material and methods
A total of 40 questions (20 MC and 20 SA) were developed based on the sixth edition of Contemporary Fixed Prosthodontics. Following a standardized protocol, these questions were posed to ChatGPT, DeepSeek, and Gemini on 4 consecutive Saturdays using 10 independent accounts per chatbot. After each session, correct answers were provided to the chatbots, and, before the next session, their memory and history were cleared. Responses were scored as correct (1) or incorrect (0) for MC questions and correct (2), partially correct (1), or incorrect (0) for SA questions. Weighted accuracy was calculated accordingly. The Kendall W coefficient was used to assess agreement among the 10 accounts per chatbot. The effects of chatbot type, time (week), and their interaction on performance were analyzed using generalized estimating equations (GEEs), followed by pairwise comparisons using the Mann-Whitney U test and Wilcoxon signed-rank test with Bonferroni adjustments for multiple comparisons (α=.05).
Results
All chatbots showed significant reproducibility, with Gemini exhibiting the highest repeatability for SA questions, followed by ChatGPT for MC questions. Accuracy ranged between 43% and 71%. ChatGPT and DeepSeek demonstrated significantly better performance in MC questions compared with Gemini (P<.017). However, in the third week, Gemini outperformed DeepSeek in SA questions (P=.007). Over time, Gemini showed continuous improvement in SA questions, whereas DeepSeek exhibited a performance surge in the fourth week. ChatGPT’s performance remained stable throughout the study period.
Conclusions
The overall accuracy of the studied chatbots in answering MC and SA prosthodontics questions was not satisfactory. Among them, ChatGPT was the most reliable for MC questions, while ChatGPT and Gemini performed best for SA questions. Gemini (for SA questions) and DeepSeek (for MC and SA questions) demonstrated improvement after exposure to correct responses.
{"title":"Comparing the performance of ChatGPT 4o, DeepSeek R1, and Gemini 2 Pro in answering fixed prosthodontics questions over time","authors":"Mohammadjavad Shirani DDS, MSc","doi":"10.1016/j.prosdent.2025.04.038","DOIUrl":"10.1016/j.prosdent.2025.04.038","url":null,"abstract":"<div><h3>Statement of problem</h3><div>The accuracy of DeepSeek and the latest versions of ChatGPT and Gemini in responding to prosthodontics questions needs to be evaluated. Additionally, the extent to which the performance of these chatbots changes through user interactions remains unexplored.</div></div><div><h3>Purpose</h3><div>The purpose of this longitudinal repeated-measures experimental study was to compare the performance of ChatGPT (4o), DeepSeek (R1), and Gemini (2 Pro) in answering multiple-choice (MC) and short-answer (SA) fixed prosthodontics questions over 4 consecutive weeks after exposure to correct responses.</div></div><div><h3>Material and methods</h3><div>A total of 40 questions (20 MC and 20 SA) were developed based on the sixth edition of Contemporary Fixed Prosthodontics. Following a standardized protocol, these questions were posed to ChatGPT, DeepSeek, and Gemini on 4 consecutive Saturdays using 10 independent accounts per chatbot. After each session, correct answers were provided to the chatbots, and, before the next session, their memory and history were cleared. Responses were scored as correct (1) or incorrect (0) for MC questions and correct (2), partially correct (1), or incorrect (0) for SA questions. Weighted accuracy was calculated accordingly. The Kendall W coefficient was used to assess agreement among the 10 accounts per chatbot. The effects of chatbot type, time (week), and their interaction on performance were analyzed using generalized estimating equations (GEEs), followed by pairwise comparisons using the Mann-Whitney U test and Wilcoxon signed-rank test with Bonferroni adjustments for multiple comparisons (α=.05).</div></div><div><h3>Results</h3><div>All chatbots showed significant reproducibility, with Gemini exhibiting the highest repeatability for SA questions, followed by ChatGPT for MC questions. Accuracy ranged between 43% and 71%. ChatGPT and DeepSeek demonstrated significantly better performance in MC questions compared with Gemini (<em>P</em><.017). However, in the third week, Gemini outperformed DeepSeek in SA questions (<em>P</em>=.007). Over time, Gemini showed continuous improvement in SA questions, whereas DeepSeek exhibited a performance surge in the fourth week. ChatGPT’s performance remained stable throughout the study period.</div></div><div><h3>Conclusions</h3><div>The overall accuracy of the studied chatbots in answering MC and SA prosthodontics questions was not satisfactory. Among them, ChatGPT was the most reliable for MC questions, while ChatGPT and Gemini performed best for SA questions. Gemini (for SA questions) and DeepSeek (for MC and SA questions) demonstrated improvement after exposure to correct responses.</div></div>","PeriodicalId":16866,"journal":{"name":"Journal of Prosthetic Dentistry","volume":"135 3","pages":"Pages 571-576"},"PeriodicalIF":4.8,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144132585","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2025-10-15DOI: 10.1016/j.prosdent.2025.10.005
Charles Goodacre DDS, MSD , Gary Goldstein DDS
<div><h3>Statement of problem</h3><div>For unknown reasons, the original term golden ratio (GR) was morphed in dentistry to golden proportion (GP), a term used in current dental literature and recognized by artificial intelligence. The emergence of computer-aided design and computer-aided manufacture (CAD-CAM) technology in prosthodontics poses the question, “Is the GP being used to design anterior milled crowns and digitally assisted dentures?” To perform these tasks, a digital library of tooth forms needs to be present in the software program.</div></div><div><h3>Purpose</h3><div>The purpose of this paper was to determine whether the GP is present in exocad tooth arrangements used in the CAD-CAM fabrication of anterior crowns and the most common tooth molds used with AvaDent digitally assisted dentures.</div></div><div><h3>Material and methods</h3><div>The default library of 17 tooth molds used with the exocad software program for milling anterior crowns was examined to determine whether any of the molds matched the GP of 0.62. Also, the 5 most common Ivoclar and Dentsply denture tooth molds selected for fabricating AvaDent digitally assisted complete dentures were examined to determine whether the GP was present in the tooth arrangements. In addition, a PubMed search of golden proportion was completed to determine whether it is being recommended for use in CAD-CAM dentistry. The search of golden proportion and dental esthetics was completed using the filters: Case Reports, Clinical Trial, Randomized Controlled Trial, and Systematic Review. Additional articles were culled from the reference lists in the articles.</div></div><div><h3>Results</h3><div>There were 2 early publications proposing the use of the golden proportion in dentistry, 6 studies that examined the relationship of the GP to natural tooth proportions, 7 systematic reviews, and 5 case reports. Measurements of the 17 exocad default tooth molds determined there were 4 molds with a proportional relationship between the width of the maxillary lateral incisor (MLI) relative to the maxillary central incisor (MCI) of 0.61, 0.63, 0.65, and 0.68 (MLI÷MCI), close to the GP of 0.62. All of the others had proportions between 0.73 and 0.87. None of the proportions of the maxillary canine (MCa) to the MLI were close to the GP and ranged from 0.83 to 0.99 (MCa÷MLI). When measurements were made of the 5 most common Ivoclar Blueline and Dentsply Portrait denture teeth molds, it was determined the MLI to MCI proportion did not match the GP and ranged from 0.71 to 0.82. Likewise, the proportion of the MCa to the MLI ranged from 0.72 to 0.80. No studies or systematic reviews supported a relationship between the GP and natural dentitions, yet there were 5 clinical reports published between 2004 and 2025 that recommended the clinical use of the GP, but none involved the use of CAD-CAM dentistry.</div></div><div><h3>Conclusions</h3><div>The golden proportion is not a valid guide for the design of anteri
{"title":"Should the golden ratio be used in CAD-CAM dentistry?","authors":"Charles Goodacre DDS, MSD , Gary Goldstein DDS","doi":"10.1016/j.prosdent.2025.10.005","DOIUrl":"10.1016/j.prosdent.2025.10.005","url":null,"abstract":"<div><h3>Statement of problem</h3><div>For unknown reasons, the original term golden ratio (GR) was morphed in dentistry to golden proportion (GP), a term used in current dental literature and recognized by artificial intelligence. The emergence of computer-aided design and computer-aided manufacture (CAD-CAM) technology in prosthodontics poses the question, “Is the GP being used to design anterior milled crowns and digitally assisted dentures?” To perform these tasks, a digital library of tooth forms needs to be present in the software program.</div></div><div><h3>Purpose</h3><div>The purpose of this paper was to determine whether the GP is present in exocad tooth arrangements used in the CAD-CAM fabrication of anterior crowns and the most common tooth molds used with AvaDent digitally assisted dentures.</div></div><div><h3>Material and methods</h3><div>The default library of 17 tooth molds used with the exocad software program for milling anterior crowns was examined to determine whether any of the molds matched the GP of 0.62. Also, the 5 most common Ivoclar and Dentsply denture tooth molds selected for fabricating AvaDent digitally assisted complete dentures were examined to determine whether the GP was present in the tooth arrangements. In addition, a PubMed search of golden proportion was completed to determine whether it is being recommended for use in CAD-CAM dentistry. The search of golden proportion and dental esthetics was completed using the filters: Case Reports, Clinical Trial, Randomized Controlled Trial, and Systematic Review. Additional articles were culled from the reference lists in the articles.</div></div><div><h3>Results</h3><div>There were 2 early publications proposing the use of the golden proportion in dentistry, 6 studies that examined the relationship of the GP to natural tooth proportions, 7 systematic reviews, and 5 case reports. Measurements of the 17 exocad default tooth molds determined there were 4 molds with a proportional relationship between the width of the maxillary lateral incisor (MLI) relative to the maxillary central incisor (MCI) of 0.61, 0.63, 0.65, and 0.68 (MLI÷MCI), close to the GP of 0.62. All of the others had proportions between 0.73 and 0.87. None of the proportions of the maxillary canine (MCa) to the MLI were close to the GP and ranged from 0.83 to 0.99 (MCa÷MLI). When measurements were made of the 5 most common Ivoclar Blueline and Dentsply Portrait denture teeth molds, it was determined the MLI to MCI proportion did not match the GP and ranged from 0.71 to 0.82. Likewise, the proportion of the MCa to the MLI ranged from 0.72 to 0.80. No studies or systematic reviews supported a relationship between the GP and natural dentitions, yet there were 5 clinical reports published between 2004 and 2025 that recommended the clinical use of the GP, but none involved the use of CAD-CAM dentistry.</div></div><div><h3>Conclusions</h3><div>The golden proportion is not a valid guide for the design of anteri","PeriodicalId":16866,"journal":{"name":"Journal of Prosthetic Dentistry","volume":"135 3","pages":"Pages 593.e1-593.e8"},"PeriodicalIF":4.8,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145308471","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2025-10-30DOI: 10.1016/j.prosdent.2025.10.008
Vidhi H. Sheth BDS, Naisargi P. Shah BDS, MDS, Romi Jain BDS, MDS, Nikhil Bhanushali BDS, MDS, Vishrut Bhatnagar BDS, MDS
{"title":"Comments regarding: Sheth et al. Development and validation of a risk-of-bias tool for assessing in vitro studies conducted in dentistry: The QUIN","authors":"Vidhi H. Sheth BDS, Naisargi P. Shah BDS, MDS, Romi Jain BDS, MDS, Nikhil Bhanushali BDS, MDS, Vishrut Bhatnagar BDS, MDS","doi":"10.1016/j.prosdent.2025.10.008","DOIUrl":"10.1016/j.prosdent.2025.10.008","url":null,"abstract":"","PeriodicalId":16866,"journal":{"name":"Journal of Prosthetic Dentistry","volume":"135 3","pages":"Pages 647-648"},"PeriodicalIF":4.8,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145422072","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2025-11-04DOI: 10.1016/j.prosdent.2025.10.031
Catina Prochnow DDS, MSciD, PhD , Duvan Cala Castillo DDS, MSciD , Rafaela Oliveira Pilecco DDS, MSciD, PhD , Amanda Maria de Oliveira Dal Piva DDS, MSciD, PhD , João Paulo Mendes Tribst DDS, MSciD, PhD , Luiz Felipe Valandro DDS, MSciD, PhD , Rafael Ratto de Moraes DDS, MSciD, PhD , Gabriel Kalil Rocha Pereira DDS, MSciD, PhD
Statement of problem
Residue from evaluation pastes may remain on the bonding surface of indirect restorations, requiring specific cleaning protocols. However, it is not clear how much the composition of different evaluation pastes affects the bond strength to resin cement.
Purpose
The purpose of this in vitro study was to determine the effectiveness of using 37% phosphoric acid to clean different evaluation pastes and its effect on the long-term bond strength of lithium disilicate ceramics to resin cement.
Material and methods
Lithium disilicate slices (IPS e.max CAD) with simulated computer-aided design and computer-aided manufacturing (CAD-CAM) topography were crystallized, etched (20 seconds - 5% hydrofluoric acid -HF), and received the application of different evaluation pastes (CTRL – no contamination; AC – Allcem; NX – NX3; ML – Multilink; RX – RelyX). All specimens were actively cleaned for 60 seconds with 37% phosphoric acid, except for the CTRL group. After applying silane, resin cement cylinders were produced and light-polymerized. Micro-shear bond strength testing was carried out on half of the cylinders (n=40) at baseline (24 hours) and after aging (220 days of water storage at 37 °C and 25 000 heat cycles between 5 °C and 55 °C). Failure modes, surface topography via scanning electron microscopy (SEM), and chemical composition through energy-dispersive X-ray spectroscopy (EDS) were assessed. Statistical analysis used 2-way analysis of variance (ANOVA) and Bonferroni post hoc tests (α=.05).
Results
At baseline, the AC and NX groups exhibited bond strength similar to the CTRL group. All groups experienced a considerable decline in bond strength after aging, but there was no difference between the AC group and the CTRL group. In both scenarios, the RX and ML groups displayed the weakest bonds. SEM analysis revealed similar surface topography, confirming the EDS results, where similar elemental composition was observed among the groups.
Conclusions
Cleaning lithium disilicate surfaces with 37% phosphoric acid after contamination with an evaluation paste is not universally effective for the tested pastes. Some pastes leave a residue that still compromises bond strength in the long term. Therefore, alternative cleaning methods should be considered.
{"title":"Influence of different evaluation pastes on the bond of resin cement to lithium disilicate ceramic: An in vitro study","authors":"Catina Prochnow DDS, MSciD, PhD , Duvan Cala Castillo DDS, MSciD , Rafaela Oliveira Pilecco DDS, MSciD, PhD , Amanda Maria de Oliveira Dal Piva DDS, MSciD, PhD , João Paulo Mendes Tribst DDS, MSciD, PhD , Luiz Felipe Valandro DDS, MSciD, PhD , Rafael Ratto de Moraes DDS, MSciD, PhD , Gabriel Kalil Rocha Pereira DDS, MSciD, PhD","doi":"10.1016/j.prosdent.2025.10.031","DOIUrl":"10.1016/j.prosdent.2025.10.031","url":null,"abstract":"<div><h3>Statement of problem</h3><div>Residue from evaluation pastes may remain on the bonding surface of indirect restorations, requiring specific cleaning protocols. However, it is not clear how much the composition of different evaluation pastes affects the bond strength to resin cement.</div></div><div><h3>Purpose</h3><div>The purpose of this in vitro study was to determine the effectiveness of using 37% phosphoric acid to clean different evaluation pastes and its effect on the long-term bond strength of lithium disilicate ceramics to resin cement.</div></div><div><h3>Material and methods</h3><div>Lithium disilicate slices (IPS e.max CAD) with simulated computer-aided design and computer-aided manufacturing (CAD-CAM) topography were crystallized, etched (20 seconds - 5% hydrofluoric acid -HF), and received the application of different evaluation pastes (CTRL – no contamination; AC – Allcem; NX – NX3; ML – Multilink; RX – RelyX). All specimens were actively cleaned for 60 seconds with 37% phosphoric acid, except for the CTRL group. After applying silane, resin cement cylinders were produced and light-polymerized. Micro-shear bond strength testing was carried out on half of the cylinders (n=40) at baseline (24 hours) and after aging (220 days of water storage at 37 °C and 25 000 heat cycles between 5 °C and 55 °C). Failure modes, surface topography via scanning electron microscopy (SEM), and chemical composition through energy-dispersive X-ray spectroscopy (EDS) were assessed. Statistical analysis used 2-way analysis of variance (ANOVA) and Bonferroni post hoc tests (α=.05).</div></div><div><h3>Results</h3><div>At baseline, the AC and NX groups exhibited bond strength similar to the CTRL group. All groups experienced a considerable decline in bond strength after aging, but there was no difference between the AC group and the CTRL group. In both scenarios, the RX and ML groups displayed the weakest bonds. SEM analysis revealed similar surface topography, confirming the EDS results, where similar elemental composition was observed among the groups.</div></div><div><h3>Conclusions</h3><div>Cleaning lithium disilicate surfaces with 37% phosphoric acid after contamination with an evaluation paste is not universally effective for the tested pastes. Some pastes leave a residue that still compromises bond strength in the long term. Therefore, alternative cleaning methods should be considered.</div></div>","PeriodicalId":16866,"journal":{"name":"Journal of Prosthetic Dentistry","volume":"135 3","pages":"Pages 580.e1-580.e7"},"PeriodicalIF":4.8,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145452300","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2025-11-12DOI: 10.1016/j.prosdent.2025.10.051
Sohil A. Kazim BDS, MDS , Charles J. Goodacre DDS, MSD , Joseph Y.K. Kan DDS, MS , Gary R. Goldstein DDS
<div><h3>Statement of problem</h3><div>Clinical dictation in dental practice is time-consuming and prone to quality variability. Although generative artificial intelligence (AI) systems such as Microsoft Copilot, OpenAI ChatGPT, Google Gemini, and OpenEvidence offer potential alternatives for automating and enhancing this process, their ability to produce clinically reliable and contextually appropriate dictations has not been evaluated.</div></div><div><h3>Purpose</h3><div>The purpose of this evaluation study was to assess the performance of Copilot, 2 versions of ChatGPT (v4 and 5), Gemini (v2.5), and OpenEvidence in generating accurate, complete, and clinically relevant dental dictations based on varying prompt structures.</div></div><div><h3>Material and methods</h3><div>Two different representative clinical scenarios were generated. One described the placement of a single mandibular implant with the dictation focused on the surgical steps, while the second described a prosthodontic procedure and the steps required to place a crown on a single maxillary implant. Each scenario was entered into Copilot, 2 different versions of ChatGPT, Gemini, and OpenEvidence using a structured prompt format (SP). All inputs were conducted in a clean session. All responses from Copilot, both included versions of ChatGPT, Gemini, and OpenEvidence software programs, were compared for both scenarios. Minimum specific prompts were also generated for the same clinical scenarios and tested to determine the least amount of clinician input required to produce acceptable dictations. Overall responses were compared with the more detailed structured prompts to assess differences in output quality.</div></div><div><h3>Results</h3><div>Across both surgical and restorative procedures, all 5 AI software programs (Copilot, ChatGPT-4, ChatGPT-5, Gemini, and OpenEvidence) produced clinically accurate dictations when using the SP format. For the surgical procedure, core steps—including preoperative assessment, anesthesia, incision design, osteotomy preparation, implant placement, closure, and postoperative instructions—were consistent. For the restorative procedure, all software programs documented atraumatic healing abutment removal, healthy peri-implant mucosa, correct crown fit, occlusion, shade matching, and standardized tightening (35 Ncm). No substantial differences in procedural accuracy were identified; discrepancies were primarily stylistic or related to the level of descriptive detail. With minimum specific prompts, the dictations were shorter and less descriptive but still captured the essential procedural steps for both clinical scenarios and were considered clinically acceptable in all included AI software programs.</div></div><div><h3>Conclusions</h3><div>Microsoft Copilot, both versions of OpenAI ChatGPT (GPT-4 and GPT-5), Google Gemini, and OpenEvidence can effectively assist clinicians in generating clinical dental procedure dictations, particularly when guided
{"title":"Can artificial intelligence (AI)-based software programs generate accurate clinical dictation?","authors":"Sohil A. Kazim BDS, MDS , Charles J. Goodacre DDS, MSD , Joseph Y.K. Kan DDS, MS , Gary R. Goldstein DDS","doi":"10.1016/j.prosdent.2025.10.051","DOIUrl":"10.1016/j.prosdent.2025.10.051","url":null,"abstract":"<div><h3>Statement of problem</h3><div>Clinical dictation in dental practice is time-consuming and prone to quality variability. Although generative artificial intelligence (AI) systems such as Microsoft Copilot, OpenAI ChatGPT, Google Gemini, and OpenEvidence offer potential alternatives for automating and enhancing this process, their ability to produce clinically reliable and contextually appropriate dictations has not been evaluated.</div></div><div><h3>Purpose</h3><div>The purpose of this evaluation study was to assess the performance of Copilot, 2 versions of ChatGPT (v4 and 5), Gemini (v2.5), and OpenEvidence in generating accurate, complete, and clinically relevant dental dictations based on varying prompt structures.</div></div><div><h3>Material and methods</h3><div>Two different representative clinical scenarios were generated. One described the placement of a single mandibular implant with the dictation focused on the surgical steps, while the second described a prosthodontic procedure and the steps required to place a crown on a single maxillary implant. Each scenario was entered into Copilot, 2 different versions of ChatGPT, Gemini, and OpenEvidence using a structured prompt format (SP). All inputs were conducted in a clean session. All responses from Copilot, both included versions of ChatGPT, Gemini, and OpenEvidence software programs, were compared for both scenarios. Minimum specific prompts were also generated for the same clinical scenarios and tested to determine the least amount of clinician input required to produce acceptable dictations. Overall responses were compared with the more detailed structured prompts to assess differences in output quality.</div></div><div><h3>Results</h3><div>Across both surgical and restorative procedures, all 5 AI software programs (Copilot, ChatGPT-4, ChatGPT-5, Gemini, and OpenEvidence) produced clinically accurate dictations when using the SP format. For the surgical procedure, core steps—including preoperative assessment, anesthesia, incision design, osteotomy preparation, implant placement, closure, and postoperative instructions—were consistent. For the restorative procedure, all software programs documented atraumatic healing abutment removal, healthy peri-implant mucosa, correct crown fit, occlusion, shade matching, and standardized tightening (35 Ncm). No substantial differences in procedural accuracy were identified; discrepancies were primarily stylistic or related to the level of descriptive detail. With minimum specific prompts, the dictations were shorter and less descriptive but still captured the essential procedural steps for both clinical scenarios and were considered clinically acceptable in all included AI software programs.</div></div><div><h3>Conclusions</h3><div>Microsoft Copilot, both versions of OpenAI ChatGPT (GPT-4 and GPT-5), Google Gemini, and OpenEvidence can effectively assist clinicians in generating clinical dental procedure dictations, particularly when guided ","PeriodicalId":16866,"journal":{"name":"Journal of Prosthetic Dentistry","volume":"135 3","pages":"Pages 596.e1-596.e7"},"PeriodicalIF":4.8,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145513086","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2025-05-21DOI: 10.1016/j.prosdent.2025.04.023
Fan Cheng DMD , Xiaohui Liu MDSc , Jianxiang Tao DMD, PhD
In this treatment, a polyetheretherketone (PEEK) resin-bonded prosthesis and a bonding guide fabricated using the digital approach were delivered to a patient with periodontitis and the loss of her left mandibular lateral incisor, simplifying the clinical procedures, improving the accuracy of the prosthesis, and reducing the risk of debonding compared with a conventional resin-bonded prosthesis.
{"title":"A polyetheretherketone resin-bonded prosthesis fabricated using the digital approach in a patient with periodontitis: A clinical report","authors":"Fan Cheng DMD , Xiaohui Liu MDSc , Jianxiang Tao DMD, PhD","doi":"10.1016/j.prosdent.2025.04.023","DOIUrl":"10.1016/j.prosdent.2025.04.023","url":null,"abstract":"<div><div><span>In this treatment, a polyetheretherketone (PEEK) resin-bonded prosthesis and a bonding guide fabricated using the digital approach were delivered to a patient with </span>periodontitis<span> and the loss of her left mandibular lateral incisor, simplifying the clinical procedures, improving the accuracy of the prosthesis, and reducing the risk of debonding compared with a conventional resin-bonded prosthesis.</span></div></div>","PeriodicalId":16866,"journal":{"name":"Journal of Prosthetic Dentistry","volume":"135 3","pages":"Pages 438-441"},"PeriodicalIF":4.8,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144127981","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2025-11-01DOI: 10.1016/j.prosdent.2025.10.034
Shuai Qi MSD , Haoxuan Shan ME , Yujie Fu PhD , Yufei Chen PhD , Qi Zhang PhD
Statement of problem
Current caries management has emphasized minimally invasive, biologically driven strategies that demand a higher level of precision in caries diagnosis. Artificial intelligence (AI)-driven tools for classifying caries on cone beam computed tomography (CBCT) scans may improve diagnostic accuracy and streamline clinical treatment planning. However, clinically oriented and interpretable AI solutions remain lacking.
Purpose
The purpose of this study was to develop and validate an interpretable AI framework, CariesAI-3D, for accurate and robust classification of dentin caries severity on CBCT images.
Material and methods
A high-quality CBCT dataset comprising 2148 CBCT images of single teeth was established, including sound teeth, moderate caries, deep caries, and extremely deep caries. The dataset was divided into a 5-fold cross-validation set (1826) for model training and validation and an independent test set (322) for final evaluation. CariesAI-3D was developed as a multitask learning network incorporating a spatial-attention feature fusion module (SA-FFM) for caries classification. Its performance was evaluated against 6 baseline models (ResNet-18, ResNet-34, ResNet-50, DenseNet-121, DenseNet-169, and MobileNet-V2) using cross-validation. An ablation study was conducted to evaluate the effectiveness of the SA-FFM. Caries classification performance was assessed using accuracy, precision, recall, F1-score, and the area under the receiver operating characteristic (ROC) curve (AUC). The mean absolute difference (MAD) between cross-validation and independent test sets was calculated to quantify model generalization. Statistical significance was assessed using a corrected resampled t test (α=.05).
Results
CariesAI-3D significantly outperformed the baseline models on the cross-validation set, achieving an accuracy of 0.886, precision of 0.882, recall of 0.873, and F1-score of 0.876. The ablation study confirmed that CariesAI-3D with SA-FFM demonstrated better accuracy than both the backbone model and the model with the element-wise feature addition. Furthermore, CariesAI-3D exhibited strong generalization on the independent test set, achieving class-wise AUC values between 0.947 and 0.998, with metric-wise MAD ranging from 0.011 to 0.033. Class activation mapping (CAM) demonstrated that the model’s predictions were highly correlated with caries and pulp regions.
Conclusions
By integrating multitask learning with an SA-FFM, CariesAI-3D achieved the accurate and interpretable classification of dentin caries severity on CBCT images, demonstrating significant advancements over conventional methods.
{"title":"A clinically oriented and interpretable AI framework for classifying dentin caries severity on CBCT images","authors":"Shuai Qi MSD , Haoxuan Shan ME , Yujie Fu PhD , Yufei Chen PhD , Qi Zhang PhD","doi":"10.1016/j.prosdent.2025.10.034","DOIUrl":"10.1016/j.prosdent.2025.10.034","url":null,"abstract":"<div><h3>Statement of problem</h3><div>Current caries management has emphasized minimally invasive, biologically driven strategies that demand a higher level of precision in caries diagnosis. Artificial intelligence (AI)-driven tools for classifying caries on cone beam computed tomography (CBCT) scans may improve diagnostic accuracy and streamline clinical treatment planning. However, clinically oriented and interpretable AI solutions remain lacking.</div></div><div><h3>Purpose</h3><div>The purpose of this study was to develop and validate an interpretable AI framework, CariesAI-3D, for accurate and robust classification of dentin caries severity on CBCT images.</div></div><div><h3>Material and methods</h3><div>A high-quality CBCT dataset comprising 2148 CBCT images of single teeth was established, including sound teeth, moderate caries, deep caries, and extremely deep caries. The dataset was divided into a 5-fold cross-validation set (1826) for model training and validation and an independent test set (322) for final evaluation. CariesAI-3D was developed as a multitask learning network incorporating a spatial-attention feature fusion module (SA-FFM) for caries classification. Its performance was evaluated against 6 baseline models (ResNet-18, ResNet-34, ResNet-50, DenseNet-121, DenseNet-169, and MobileNet-V2) using cross-validation. An ablation study was conducted to evaluate the effectiveness of the SA-FFM. Caries classification performance was assessed using accuracy, precision, recall, F1-score, and the area under the receiver operating characteristic (ROC) curve (AUC). The mean absolute difference (MAD) between cross-validation and independent test sets was calculated to quantify model generalization. Statistical significance was assessed using a corrected resampled <em>t</em> test (α=.05).</div></div><div><h3>Results</h3><div>CariesAI-3D significantly outperformed the baseline models on the cross-validation set, achieving an accuracy of 0.886, precision of 0.882, recall of 0.873, and F1-score of 0.876. The ablation study confirmed that CariesAI-3D with SA-FFM demonstrated better accuracy than both the backbone model and the model with the element-wise feature addition. Furthermore, CariesAI-3D exhibited strong generalization on the independent test set, achieving class-wise AUC values between 0.947 and 0.998, with metric-wise MAD ranging from 0.011 to 0.033. Class activation mapping (CAM) demonstrated that the model’s predictions were highly correlated with caries and pulp regions.</div></div><div><h3>Conclusions</h3><div>By integrating multitask learning with an SA-FFM, CariesAI-3D achieved the accurate and interpretable classification of dentin caries severity on CBCT images, demonstrating significant advancements over conventional methods.</div></div>","PeriodicalId":16866,"journal":{"name":"Journal of Prosthetic Dentistry","volume":"135 3","pages":"Pages 594.e1-594.e8"},"PeriodicalIF":4.8,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145431613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2025-11-12DOI: 10.1016/j.prosdent.2025.10.040
Catherine Arreaza DDS, MS , Robert R. Seghi DDS, MS , Scott R. Schricker PhD , William M. Johnston MS, PhD , Paola C. Saponaro DDS, MS, FACP
Statement of problem
Occlusal devices have been used for therapeutic and diagnostic purposes. Although their effectiveness in preventing tooth wear has been reported, research regarding the wear of the occlusal device when opposing different materials is lacking.
Purpose
The purpose of this in vitro study was to measure and compare the volumetric wear of conventionally processed, milled, and 3-dimensionally (3D) printed occlusal device materials when opposed by human enamel, lithium disilicate, and 4 mol% yttria-stabilized zirconia.
Material and methods
Milled polymethyl methacrylate (ProArt CAD Splint; Ivoclar AG), printed resin (Next Dent Ortho Rigid), and conventional heat-polymerized polymethyl methacrylate (Vitacrilic Clear; Fricke International, Inc) were tested. Fifteen Ø14×2-mm disk-shaped specimens were fabricated from each manufacturing process. Each material group was divided into 3 groups of 5 specimens and tested against each of the 3 antagonist materials, human enamel, zirconia (IPS e.max ZirCAD MT; Ivoclar AG), and lithium disilicate (IPS e.max CAD; Ivoclar AG) shaped into a spherical stylus. Each specimen was subjected to 50 000 mastication cycles using an oral wear simulator (Oregon Health Sciences University (OHSU) Wear Machine), with a frequency of 1.1 Hz and a maximum 50-N load with vertical load and horizontal movement. The surfaces of the substrate and styli were before and after wear testing using a laboratory scanner (E4 scanner; 3Shape A/S). The surfaces of each specimen before and after wear testing were analyzed using a software program (WearCompare; Leeds Digital Dentistry). The volume difference between the before and after scans were determined for both the substrate and the stylus and analyzed using a 2-way ANOVA. The statistical significance between groups was determined using the Tukey-Kramer HSD test (α=.025).
Results
The results of the 2-way ANOVA showed that the antagonist material (P<.01) significantly influenced the resulting volumetric wear of the occlusal device materials, whereas the substrate material (P=.03) had no significant effect. The interaction between substrate type and antagonist material (P=.42) also did not significantly affect volumetric wear. Only the material type of the opposing antagonist material (P<.001) influenced the volume loss, whereas the substrate type (P=.05) and interactions (P=.93) between these 2 factors were statistically similar.
Conclusions
Significant differences in wear were found among printed, milled, and heat-polymerized polymethyl methacrylate occlusal device materials. The wear of these materials was influenced by both the type of material used and the antagonist materials. The printed occlusal device material exhibited the least amount of wear.
{"title":"Comparison of wear behavior of occlusal device materials manufactured by different processes","authors":"Catherine Arreaza DDS, MS , Robert R. Seghi DDS, MS , Scott R. Schricker PhD , William M. Johnston MS, PhD , Paola C. Saponaro DDS, MS, FACP","doi":"10.1016/j.prosdent.2025.10.040","DOIUrl":"10.1016/j.prosdent.2025.10.040","url":null,"abstract":"<div><h3>Statement of problem</h3><div>Occlusal devices have been used for therapeutic and diagnostic purposes. Although their effectiveness in preventing tooth wear has been reported, research regarding the wear of the occlusal device when opposing different materials is lacking.</div></div><div><h3>Purpose</h3><div>The purpose of this in vitro study was to measure and compare the volumetric wear of conventionally processed, milled, and 3-dimensionally (3D) printed occlusal device materials when opposed by human enamel, lithium disilicate, and 4 mol% yttria-stabilized zirconia.</div></div><div><h3>Material and methods</h3><div>Milled polymethyl methacrylate (ProArt CAD Splint; Ivoclar AG), printed resin (Next Dent Ortho Rigid), and conventional heat-polymerized polymethyl methacrylate (Vitacrilic Clear; Fricke International, Inc) were tested. Fifteen Ø14×2-mm disk-shaped specimens were fabricated from each manufacturing process. Each material group was divided into 3 groups of 5 specimens and tested against each of the 3 antagonist materials, human enamel, zirconia (IPS e.max ZirCAD MT; Ivoclar AG), and lithium disilicate (IPS e.max CAD; Ivoclar AG) shaped into a spherical stylus. Each specimen was subjected to 50 000 mastication cycles using an oral wear simulator (Oregon Health Sciences University (OHSU) Wear Machine), with a frequency of 1.1 Hz and a maximum 50-N load with vertical load and horizontal movement. The surfaces of the substrate and styli were before and after wear testing using a laboratory scanner (E4 scanner; 3Shape A/S). The surfaces of each specimen before and after wear testing were analyzed using a software program (WearCompare; Leeds Digital Dentistry). The volume difference between the before and after scans were determined for both the substrate and the stylus and analyzed using a 2-way ANOVA. The statistical significance between groups was determined using the Tukey-Kramer HSD test (α=.025).</div></div><div><h3>Results</h3><div>The results of the 2-way ANOVA showed that the antagonist material (<em>P</em><.01) significantly influenced the resulting volumetric wear of the occlusal device materials, whereas the substrate material <em>(P=</em>.03) had no significant effect. The interaction between substrate type and antagonist material (<em>P</em>=.42) also did not significantly affect volumetric wear. Only the material type of the opposing antagonist material (<em>P</em><.001) influenced the volume loss, whereas the substrate type (<em>P</em>=.05) and interactions (<em>P</em>=.93) between these 2 factors were statistically similar.</div></div><div><h3>Conclusions</h3><div>Significant differences in wear were found among printed, milled, and heat-polymerized polymethyl methacrylate occlusal device materials. The wear of these materials was influenced by both the type of material used and the antagonist materials. The printed occlusal device material exhibited the least amount of wear.</div></div>","PeriodicalId":16866,"journal":{"name":"Journal of Prosthetic Dentistry","volume":"135 3","pages":"Pages 614.e1-614.e7"},"PeriodicalIF":4.8,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145513101","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：