目的:使用多类全景射线照片数据集评估三个版本的深度学习卷积神经网络在对象检测和分割方面的诊断性能。
方法:本研究随机选择了总共600个正像图,并由单个操作员使用图像注释工具(COCOAnnotatorv.11.0.1)进行手动注释,以建立地面实况。注释类包括牙齿,上颌骨,下颌骨,下牙槽神经,牙髓和种植体支撑的牙冠/桥体,牙髓治疗,树脂基修复体,金属修复体,和植入物。然后将数据集分为训练,验证,和测试子集,用于训练你只看一次(YOLO)神经网络的版本5、7和8。结果被存储,并通过计算精度(P)进行了后验性能分析,召回(R),F1得分,交汇处(IoU),和平均平均精度(mAP)在0.5和0.5-0.95阈值。还绘制了混淆矩阵和召回精度图。
结果:YOLOv5s显示物体检测结果有所改善,平均R=0.634,P=0.781,mAP0.5=0.631,mAP0.5-0.95=0.392。YOLOv7m取得了最好的目标检测结果,平均R=0.793,P=0.779,mAP0.5=0.740,mAP0.5-0.95=0,481。对于对象分割,YOLOv8m获得了最好的平均结果(R=0.589,P=0.755,mAP0.5=0.591,mAP0.5-0.95=0.272)。
结论:YOLOv7m更适合对象检测,而YOLOv8m在对象分割方面表现出卓越的性能。目标检测中最常见的错误与背景分类有关。相反,在对象分割中,在不同的牙科治疗类别中,TruePositions存在错误分类的趋势。
结论:基于全景射线照片的一般诊断和治疗决策可以使用新的基于人工智能的工具来增强。然而,这些神经网络的可靠性应该经过训练和验证,以确保它们的泛化性。
To evaluate the diagnostic performance of three versions of a deep-learning convolutional neural network in terms of object detection and segmentation using a multiclass panoramic radiograph dataset.
A total of 600 orthopantomographies were randomly selected for this study and manually annotated by a single operator using an image annotation tool (COCO Annotator v.11.0.1) to establish ground truth. The annotation classes included teeth, maxilla, mandible, inferior alveolar nerve, dento- and implant-supported crowns/pontics, endodontic treatment, resin-based restorations, metallic restorations, and implants. The dataset was then divided into training, validation, and testing subsets, which were used to train versions 5, 7, and 8 of You Only Look Once (YOLO) Neural Network. Results were stored, and a posterior performance analysis was carried out by calculating the precision (P), recall (R), F1 Score, Intersection over Union (IoU), and mean average precision (mAP) at 0.5 and 0.5-0.95 thresholds. The confusion matrix and recall precision graphs were also sketched.
YOLOv5s showed an improvement in object detection results with an average R = 0.634, P = 0.781, mAP0.5 = 0.631, and mAP0.5-0.95 = 0.392. YOLOv7m achieved the best object detection results with average R = 0.793, P = 0.779, mAP0.5 = 0.740, and mAP0.5-0.95 = 0,481. For object segmentation, YOLOv8m obtained the best average results (R = 0.589, P = 0.755, mAP0.5 = 0.591, and mAP0.5-0.95 = 0.272).
YOLOv7m was better suited for object detection, while YOLOv8m demonstrated superior performance in object segmentation. The most frequent error in object detection was related to background classification. Conversely, in object segmentation, there is a tendency to misclassify True Positives across different dental treatment categories.
General diagnostic and treatment decisions based on panoramic radiographs can be enhanced using new artificial intelligence-based tools. Nevertheless, the reliability of these neural networks should be subjected to training and validation to ensure their generalizability.