Scholarlink

Objective <p>To develop a machine-learning-based model and construct a nomogram that integrates ClinCheck features and clinical risk factors for accurately predicting open gingival embrasures (OGE) between mandibular central incisors after clear aligner treatment (CAT).</p> Methods <p>A total of 297 patients (163 normal and 134 with OGE) who underwent Invisalign<sup>®</sup> treatment were enrolled. A ClinCheck model was developed based on predicted OGE-area in the final step from initial ClinCheck treatment plan. Twenty-three clinical features were extracted from electronic medical records and ClinCheck tooth movement metrics. Predictors were selected through Least Absolute Shrinkage and Selection Operator (LASSO) regression to establish a clinical model. Additionally, a nomogram incorporating ClinCheck features and clinical predictors was constructed via logistic regression and validated with bootstrap resampling. The performances of these models were evaluated through receiver operating characteristic (ROC) curves, area under curves (AUC), and decision curve analyses (DCA).</p> Results <p>Six clinical features, including age, gingival papilla angle, interproximal reduction, crown morphology and two types of tooth movement, were selected through LASSO regression. The combined model that consisted of OGE-area and clinical features demonstrated superior predictive capacity (AUC: 0.880; 95% <i>CI</i>: 0.840–0.916), outperforming both clinical model (AUC: 0.815; 95% <i>CI</i>: 0.767–0.861; <i>P</i> &lt; 0.001) and ClinCheck model (AUC: 0.860; 95% <i>CI</i>: 0.817–0.900; <i>P</i> &lt; 0.05). The corrected C-statistic of the combined nomogram was 0.888, and the calibration curve exhibited great performance with a mean absolute error of 0.012. In the DCA curve, the combined model showed higher net benefit than the clinical model when the threshold probability exceeded 0.13, and higher than the ClinCheck model between 0.13 and 0.63.</p> Conclusion <p>The integration of clinical features and ClinCheck in the machine-learning-based model demonstrated favorable predictive capabilities for OGE between lower central incisors. This comprehensive nomogram may contribute to precisely prediction and prevention of OGE in clinical practice.</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

A novel machine-learning-based model for prediction of open gingival embrasures between mandibular central incisors after clear aligners treatment: a retrospective cohort study

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年10月15日

查看全文
  • Guifeng Li,
  • Feng Guo,
  • Jun Chen,
  • Houxuan Li,
  • Lang Lei

摘要

Objective

To develop a machine-learning-based model and construct a nomogram that integrates ClinCheck features and clinical risk factors for accurately predicting open gingival embrasures (OGE) between mandibular central incisors after clear aligner treatment (CAT).

Methods

A total of 297 patients (163 normal and 134 with OGE) who underwent Invisalign® treatment were enrolled. A ClinCheck model was developed based on predicted OGE-area in the final step from initial ClinCheck treatment plan. Twenty-three clinical features were extracted from electronic medical records and ClinCheck tooth movement metrics. Predictors were selected through Least Absolute Shrinkage and Selection Operator (LASSO) regression to establish a clinical model. Additionally, a nomogram incorporating ClinCheck features and clinical predictors was constructed via logistic regression and validated with bootstrap resampling. The performances of these models were evaluated through receiver operating characteristic (ROC) curves, area under curves (AUC), and decision curve analyses (DCA).

Results

Six clinical features, including age, gingival papilla angle, interproximal reduction, crown morphology and two types of tooth movement, were selected through LASSO regression. The combined model that consisted of OGE-area and clinical features demonstrated superior predictive capacity (AUC: 0.880; 95% CI: 0.840–0.916), outperforming both clinical model (AUC: 0.815; 95% CI: 0.767–0.861; P < 0.001) and ClinCheck model (AUC: 0.860; 95% CI: 0.817–0.900; P < 0.05). The corrected C-statistic of the combined nomogram was 0.888, and the calibration curve exhibited great performance with a mean absolute error of 0.012. In the DCA curve, the combined model showed higher net benefit than the clinical model when the threshold probability exceeded 0.13, and higher than the ClinCheck model between 0.13 and 0.63.

Conclusion

The integration of clinical features and ClinCheck in the machine-learning-based model demonstrated favorable predictive capabilities for OGE between lower central incisors. This comprehensive nomogram may contribute to precisely prediction and prevention of OGE in clinical practice.