Çelebi Elif, Akkaya Nurullah, Ünsal Gürkan
Department of Oral and Maxillofacial Radiology, School of Dental Medicine, Bahçeşehir University, Gayrettepe, Barbaros Boulevard No:153, Beşiktaş, 34357, Istanbul, Turkey.
Department of Computer Engineering, Near East University, Nicosia, Cyprus.
Oral Radiol. 2025 Sep 17. doi: 10.1007/s11282-025-00862-x.
This study aimed to develop and evaluate a deep convolutional neural network (CNN) model for the automatic segmentation of foreign bodies and ghost images in panoramic radiographs (PRs), which can complicate diagnostic interpretation.
A dataset of 11,226 PRs from four devices was annotated by two radiologists using the Computer Vision Annotation Tool. A U-Net-based CNN model was trained and evaluated using Intersection over Union (IoU), Dice coefficient, accuracy, precision, recall, and F1 score.
For foreign body segmentation, the model achieved validation Dice and IoU scores of 0.9439 and 0.9043, and test scores of 0.9657 and 0.9371. For ghost image segmentation, validation Dice and IoU were 0.8234 and 0.7388, with test scores of 0.8749 and 0.8145. Overall test accuracy exceeded 0.999.
The AI model showed high accuracy in segmenting foreign bodies and ghost images in PRs, indicating its potential to assist radiologists. Further clinical validation is recommended.
本研究旨在开发并评估一种深度卷积神经网络（CNN）模型，用于自动分割全景X线片（PR）中的异物和伪影图像，这些图像会使诊断解读变得复杂。
使用计算机视觉标注工具，由两名放射科医生对来自四台设备的11226张PR数据集进行标注。使用交并比（IoU）、Dice系数、准确率、精确率、召回率和F1分数对基于U-Net的CNN模型进行训练和评估。
对于异物分割，该模型的验证Dice和IoU分数分别为0.9439和0.9043，测试分数分别为0.9657和0.9371。对于伪影图像分割，验证Dice和IoU分别为0.8234和0.7388，测试分数分别为0.8749和0.8145。总体测试准确率超过0.999。
该人工智能模型在分割PR中的异物和伪影图像方面显示出高准确率，表明其有协助放射科医生的潜力。建议进行进一步的临床验证。
