Scholarlink

A Commentary on <p><b>Abbott LP, Saikia A, Anthonappa RP</b>.</p> <p>Artificial intelligence platforms in dental caries detection: a systematic review and meta-analysis. J Evid Based Dent Pract. 2025; <a href="https://doi.org/10.1016/j.jebdp.2024.102077">https://doi.org/10.1016/j.jebdp.2024.102077</a>.</p> Data sources <p>The search strategy for this review aimed to identify published articles that have used either clinical or X-ray images for AI model development. The search was carried out in eight electronic databases, including Scopus, Web of Science, MEDLINE, Educational Resources Information Centre, Institute of Electrical and Electronics, Science Direct, Directory of Open Access Journals, and JSTOR. Studies published in English from January 2000 to March were selected.</p> Study selection <p>Based on predefined criteria, 2538 articles were retrieved from the search; after deduplication and exclusion of articles that did not meet the inclusion/exclusion criteria, 45 articles were included in the review. Of these, 33 studies had used dental radiographs and 12 had used clinical images.</p> Data extraction and synthesis <p>All the included articles were assessed for quality using QUADAS-2 and the CLAIM checklist. The results from all included studies were narratively summarized, reporting on various parameters, ranges, and mean accuracy achieved, as well as details of the annotation tool, AI platform, etc. In addition, a meta-analysis was conducted that included seven studies.</p> Results <p>The mean accuracy was 78.2% (95% CI: 72–84.4%) for clinical image studies and 81.5% (95% CI: 72.7–90.3%) for studies that included X-ray images. Based on a meta-analysis, the overall sensitivity and specificity were 76% (95% CI: 65–85%) and 91% (95% CI: 86–95%), respectively. An HSROC curve was also generated, indicating an AUC of 92% (95% CI: 89–94%).</p> Conclusions <p>AI models exhibited high sensitivity and specificity for caries detection.</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

AI for caries detection: how close are we to clinical use?

  • 期刊论文

  • 发表日期：

    2026年1月28日

查看全文
  • Neeraj Gugnani,
  • Shalini Gugnani

摘要

A Commentary on

Abbott LP, Saikia A, Anthonappa RP.

Artificial intelligence platforms in dental caries detection: a systematic review and meta-analysis. J Evid Based Dent Pract. 2025; https://doi.org/10.1016/j.jebdp.2024.102077.

Data sources

The search strategy for this review aimed to identify published articles that have used either clinical or X-ray images for AI model development. The search was carried out in eight electronic databases, including Scopus, Web of Science, MEDLINE, Educational Resources Information Centre, Institute of Electrical and Electronics, Science Direct, Directory of Open Access Journals, and JSTOR. Studies published in English from January 2000 to March were selected.

Study selection

Based on predefined criteria, 2538 articles were retrieved from the search; after deduplication and exclusion of articles that did not meet the inclusion/exclusion criteria, 45 articles were included in the review. Of these, 33 studies had used dental radiographs and 12 had used clinical images.

Data extraction and synthesis

All the included articles were assessed for quality using QUADAS-2 and the CLAIM checklist. The results from all included studies were narratively summarized, reporting on various parameters, ranges, and mean accuracy achieved, as well as details of the annotation tool, AI platform, etc. In addition, a meta-analysis was conducted that included seven studies.

Results

The mean accuracy was 78.2% (95% CI: 72–84.4%) for clinical image studies and 81.5% (95% CI: 72.7–90.3%) for studies that included X-ray images. Based on a meta-analysis, the overall sensitivity and specificity were 76% (95% CI: 65–85%) and 91% (95% CI: 86–95%), respectively. An HSROC curve was also generated, indicating an AUC of 92% (95% CI: 89–94%).

Conclusions

AI models exhibited high sensitivity and specificity for caries detection.