Department of Oral and Maxillofacial Surgery, Radboud University Nijmegen Medical Centre, P.O. Box 9101, 6500 HB, Nijmegen, the Netherlands; Department of Artificial Intelligence, Radboud University, Nijmegen, the Netherlands; Department of Oral and Maxillofacial Surgery, Universitätsklinikum Münster, Münster, Germany.
Promaton Co. Ltd., Amsterdam 1076 GR, the Netherlands.
J Dent. 2021 Dec;115:103864. doi: 10.1016/j.jdent.2021.103864. Epub 2021 Oct 29.
The aim of this study is to automatically detect, segment and label teeth, crowns, fillings, root canal fillings, implants and root remnants on panoramic radiographs (PR(s)).
As a reference, 2000 PR(s) were manually annotated and labeled. A deep-learning approach based on mask R-CNN with Resnet-50 in combination with a rule-based heuristic algorithm and a combinatorial search algorithm was trained and validated on 1800 PR(s). Subsquently, the trained algorithm was applied onto a test set consisting of 200 PR(s). F1 scores, as a measure of accuracy, were calculated to quantify the degree of similarity between the annotated ground-truth and the model predictions. The F1-score considers the harmonic mean of precison (positive predictive value) and recall (specificity).
The proposes method achieved F1 scores up to 0.993, 0.952 and 0.97 for detection, segmentation and labeling, respectivley.
The proposed method forms a promising foundation for the further development of automatic chart filing on PR(s).
Deep learning may assist clinicians in summarizing the radiological findings on panoramic radiographs. The impact of using such models in clinical practice should be explored.
本研究旨在自动检测、分割和标记全景影像（PR(s)）中的牙齿、牙冠、填充物、根管填充物、种植体和残根。
作为参考，对 2000 张 PR(s)进行了手动注释和标记。基于 Resnet-50 的掩模 R-CNN 的深度学习方法与基于规则的启发式算法和组合搜索算法相结合，在 1800 张 PR(s)上进行了训练和验证。随后，将经过训练的算法应用于由 200 张 PR(s)组成的测试集。使用 F1 分数（准确性的度量）来计算模型预测与注释的地面实况之间的相似程度。F1 分数考虑了精度（阳性预测值）和召回率（特异性）的调和平均值。
所提出的方法在检测、分割和标记方面分别达到了高达 0.993、0.952 和 0.97 的 F1 分数。
该方法为全景影像自动归档的进一步发展奠定了有前景的基础。
深度学习可能有助于临床医生总结全景影像的放射学发现。应探讨在临床实践中使用此类模型的影响。
