Scholarlink

Objectives <p>To evaluate the feasibility of intraoral scanning (IOS) for the digital design and fabrication of removable partial dentures (RPDs) across different Kennedy classes, and to compare IOS-based models with conventional final impressions using reverse engineering and artificial intelligence (AI)-assisted analyses. The null hypothesis was that no statistically significant differences exist in 3D deviations between IOS and conventional impression models among Kennedy classes. A secondary hypothesis was that AI-assisted analysis could serve as a reliable supplement to conventional reverse engineering.</p> Materials and methods <p>Fifty partially edentulous patients underwent two methods. 3D deviations between corresponding maxillary and mandibular models were quantified using reverse engineering software. Subgroup analyses were performed according to Kennedy classification, patient-related factors, and anatomical regions. A PointNet++-based deep learning model was used exclusively for automated segmentation and alignment. Data normality was assessed using Shapiro-Wilk test. One-way or two-way ANOVA with Tukey’s post hoc test was applied. Clinical follow-up was conducted at 3, 6, and 12 months.</p> Results <p>Smaller deviations were observed in Kennedy Class III and IV cases, whereas larger deviations occurred in Class I and II cases, particularly in soft tissue regions such as the palate, gingival margins, and retromolar pads. Age and sex showed minimal influence, while higher Body Mass Index (BMI) was associated with increased soft tissue deviation. AI-assisted analysis produced results consistent with conventional reverse engineering.</p> Conclusions <p>IOS demonstrates stable performance in Kennedy Class III and IV cases and acceptable outcomes in selected Class I and II cases with appropriate clinical adjustment.</p> Clinical Relevance <p>IOS may be applied for RPD fabrication in carefully selected cases; however, free-end extension RPDs still require cautious case selection and post-insertion adjustment.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Evaluating the feasibility of intraoral scanning technology for removable partial dentures: ai-assisted measurement analysis and clinical investigation

  • 期刊论文

  • 发表日期：

    2026年3月17日

查看全文
  • Xinyu Yang,
  • Yumin Wu,
  • Mengke Qian,
  • Ruoyun Wu,
  • Haifeng Xie,
  • Chen Chen

摘要

Objectives

To evaluate the feasibility of intraoral scanning (IOS) for the digital design and fabrication of removable partial dentures (RPDs) across different Kennedy classes, and to compare IOS-based models with conventional final impressions using reverse engineering and artificial intelligence (AI)-assisted analyses. The null hypothesis was that no statistically significant differences exist in 3D deviations between IOS and conventional impression models among Kennedy classes. A secondary hypothesis was that AI-assisted analysis could serve as a reliable supplement to conventional reverse engineering.

Materials and methods

Fifty partially edentulous patients underwent two methods. 3D deviations between corresponding maxillary and mandibular models were quantified using reverse engineering software. Subgroup analyses were performed according to Kennedy classification, patient-related factors, and anatomical regions. A PointNet++-based deep learning model was used exclusively for automated segmentation and alignment. Data normality was assessed using Shapiro-Wilk test. One-way or two-way ANOVA with Tukey’s post hoc test was applied. Clinical follow-up was conducted at 3, 6, and 12 months.

Results

Smaller deviations were observed in Kennedy Class III and IV cases, whereas larger deviations occurred in Class I and II cases, particularly in soft tissue regions such as the palate, gingival margins, and retromolar pads. Age and sex showed minimal influence, while higher Body Mass Index (BMI) was associated with increased soft tissue deviation. AI-assisted analysis produced results consistent with conventional reverse engineering.

Conclusions

IOS demonstrates stable performance in Kennedy Class III and IV cases and acceptable outcomes in selected Class I and II cases with appropriate clinical adjustment.

Clinical Relevance

IOS may be applied for RPD fabrication in carefully selected cases; however, free-end extension RPDs still require cautious case selection and post-insertion adjustment.