Centre for Translational Medicine, Semmelweis University, Tűzoltó utca 37-47 1072, Budapest, Hungary.
Centre for Translational Medicine, Semmelweis University, Tűzoltó utca 37-47 1072, Budapest, Hungary; Department of Biophysics and Radiation Biology, Semmelweis University, Tűzoltó utca 37-47, 1072, Budapest, Hungary.
J Dent. 2024 Dec;151:105388. doi: 10.1016/j.jdent.2024.105388. Epub 2024 Oct 11.
This systematic review and meta-analysis aimed to investigate the diagnostic accuracy of Artificial Intelligence (AI) for approximal carious lesions on bitewing radiographs.
This study included randomized controlled trials (RCTs) and non-randomized controlled trials (non-RCTs) reporting on the diagnostic accuracy of AI for approximal carious lesions on bitewing radiographs. The risk of bias was assessed using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool. A systematic search was conducted on November 4, 2023, in PubMed, Cochrane, and Embase databases and an updated search was performed on August 28, 2024. The primary outcomes assessed were sensitivity, specificity, and overall accuracy. Sensitivity and specificity were pooled using a bivariate model.
Of the 2,442 studies identified, 21 met the inclusion criteria. The pooled sensitivity and specificity of AI were 0.94 (confidence interval (CI): ± 0.78-0.99) and 0.91 (CI: ± 0.84-0.95), respectively. The positive predictive value (PPV) ranged from 0.15 to 0.87, indicating a moderate capacity for identifying true positives among decayed teeth. The negative predictive value (NPV) ranged from 0.79 to 1.00, demonstrating a high ability to exclude healthy teeth. The diagnostic odds ratio was high, indicating strong overall diagnostic performance.
AI models demonstrate clinically acceptable diagnostic accuracy for approximal caries on bitewing radiographs. Although AI can be valuable for preliminary screening, positive findings should be verified by dental experts to prevent unnecessary treatments and ensure timely diagnosis. AI models are highly reliable in excluding healthy approximal surfaces.
AI can assist dentists in detecting approximal caries on bitewing radiographs. However, expert supervision is required to prevent iatrogenic damage and ensure timely diagnosis.
本系统评价和荟萃分析旨在调查人工智能（AI）在牙合翼片上近中龋损的诊断准确性。
本研究纳入了报告 AI 在牙合翼片上近中龋损的诊断准确性的随机对照试验（RCT）和非随机对照试验（非 RCT）。使用诊断准确性研究的质量评估工具（QUADAS-2）评估偏倚风险。于 2023 年 11 月 4 日在 PubMed、Cochrane 和 Embase 数据库中进行系统检索，并于 2024 年 8 月 28 日进行了更新检索。主要评估指标为敏感度、特异度和总体准确性。使用双变量模型对敏感度和特异度进行汇总。
在 2442 项研究中，有 21 项符合纳入标准。AI 的汇总敏感度和特异度分别为 0.94（置信区间：±0.78-0.99）和 0.91（置信区间：±0.84-0.95）。阳性预测值（PPV）范围为 0.15 至 0.87，表明 AI 有中度能力识别龋坏牙齿中的真阳性。阴性预测值（NPV）范围为 0.79 至 1.00，表明 AI 有高度能力排除健康牙齿。诊断比值比很高，表明总体诊断性能良好。
AI 模型在牙合翼片上近中龋损的诊断中具有临床可接受的准确性。虽然 AI 可用于初步筛查，但阳性结果应由口腔专家进行验证，以防止不必要的治疗并确保及时诊断。AI 模型在排除健康近中面方面具有高度可靠性。
AI 可协助牙医在牙合翼片上检测近中龋损。然而，为了防止医源性损伤和确保及时诊断，需要专家监督。
