文献速递牙体牙髓病

术前风险评估无法通过机器学习预测根管充填长度:一项纵向研究

Preoperative risk assessment does not allow to predict root filling length using machine learning: A longitudinal study.
PubMed 文献 · J Dent · 2022-11-26

该研究纳入2016—2020年在一所大学诊所完成根管治疗且有完整记录的555颗患牙(343例患者),分析术前风险因素与根管充填长度(RFL)的关联,并采用逻辑回归及多种机器学习算法预测RFL。结果显示,本科生操作、根管路径不清、根管缩窄及再治疗与RFL不成功相关;无论采用何种机器学习方法,在独立测试集上预测RFL的能力均有限。作者结论认为RFL取决于操作者和多个风险因素,机器学习对根管治疗技术结局的预测性能不足。

要点

  • 本科生操作(OR 2.74)、根管路径不清(OR 11.04)、根管缩窄(OR 2.56)、再治疗(OR 3.13)与RFL不成功相关。
  • 亚组分析显示牙医比本科生更能成功降低风险。
  • 无论采用何种机器学习方法,在独立测试集上预测RFL的能力均有限。

核心结论

  • 达到RFL取决于操作者和多个风险因素,利用机器学习算法对根管治疗技术结局的预测性能不足。

研究设计与样本

  • 纳入2016—2020年在一所大学诊所接受根管治疗且有完整记录的患牙,共555颗完成根管治疗(343例患者,女/男32.1/67.9%)。
  • RFL成功定义为距根尖0—2mm,次优RFL为超过2mm或超出根尖。
  • 采用逻辑回归进行关联分析,并采用逻辑回归及随机森林、支持向量机、决策树、梯度提升机、极端梯度提升进行预测建模。

主要发现

  • 在包含全部数据的关联分析中,本科生操作(OR 2.74,95%CI [1.61, 4.75],p<0.001)、根管路径不清(OR 11.04,95%CI [2.87, 44.88],p<0.001)、根管缩窄(OR 2.56,95%CI [1.49, 4.46],p<0.01)、再治疗(OR 3.13,95%CI [1.6, 6.41],p<0.001)时RFL不成功的可能性更高。
  • 亚组分析显示,牙医比本科生更能成功降低风险。
  • 在独立测试集上,无论采用何种机器学习方法,预测RFL的能力均有限。

关键数值与推荐

  • RFL成功定义为距根尖0—2mm,次优RFL为超过2mm或超出根尖。
  • 本科生操作OR 2.74,95%CI [1.61, 4.75],p<0.001。
  • 根管路径不清OR 11.04,95%CI [2.87, 44.88],p<0.001。
  • 根管缩窄OR 2.56,95%CI [1.49, 4.46],p<0.01。
  • 再治疗OR 3.13,95%CI [1.6, 6.41],p<0.001。

临床怎么用

  • 术前风险评估是牙髓治疗计划的相关步骤。
  • 单一影像学风险因素与达到(或未达到)最佳RFL显著相关,并显示出比更复杂的风险评估表更高的预测价值。
  • 原文未据此给出推荐。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

OBJECTIVES: First we aimed to identify significant associations between preoperative risk factors and achieving optimal root filling length (RFL) during orthograde root canal treatments (RCT) and second to predict successful RFL using machine learning. METHODS: Teeth receiving RCT at one university clinic from 2016-2020 with complete documentation were included. Successful RFL was defined to be 0-2mm of the apex, suboptimal RFL >2mm or beyond the apex. Logistic regression (logR) was used for association analyses; logR and more advanced machine learning (random forest (RF), support vector machine (SVM), decision tree (DT), gradient boosting machine (GBM) and extreme gradient boosting (XGB)) were employed for predictive modeling. RESULTS: 555 completed RCT (343 patients, female/male 32.1/67.9%) were included. In our association analysis (involving the full dataset), unsuccessful RFL was more likely in undergraduate students (US): OR 2.74, 95% CI [1.61, 4.75], p < 0.001), teeth with indistinct canal paths (OR 11.04, [2.87, 44.88], p < 0.001), root canals reduced in size (OR 2.56, [1.49, 4.46], p < 0.01), retreatments (OR 3.13, [1.6, 6.41], p < 0.001). Subgroup analyses revealed that dentists were more successful in mitigating risks than undergraduate students. Prediction of RFL on a separate testset was limitedly possible regardless of the machine learning approach. CONCLUSIONS: Achieving RFL is depending on the operator and several risk factors. The predictive performance on the technical outcome of a root canal treatment utilizing ML algorithms was insufficient. CLINICAL SIGNIFICANCE: Preoperative risk assessment is a relevant step in endodontic treatment planning. Single radiographic risk factors were significantly associated with achieving (or not achieving) optimal RFL and showed higher predictive value than a more complex risk assessment form.

来源:J Dent,查看原文。版权归原作者所有。

引用

  1. Herbst SR, Herbst CS, Schwendicke F, et al. Preoperative risk assessment does not allow to predict root filling length using machine learning: A longitudinal study. J Dent. 2022. PMID: 36442583. doi:10.1016/j.jdent.2022.104378

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。