Department of Oral Diagnostics, Digital Health and Health Services Research, Charité - Universitätsmedizin Berlin, Germany.
Department of Oral Medicine and Radiology, King George's Medical University, Lucknow, India.
J Dent. 2020 Sep;100:103425. doi: 10.1016/j.jdent.2020.103425. Epub 2020 Jul 4.
We aimed to apply deep learning to detect caries lesions of different radiographic extension on bitewings, hypothesizing it to be significantly more accurate than individual dentists.
3686 bitewing radiographs were assessed by four experienced dentists. Caries lesions were marked in a pixelwise fashion. The union of all pixels was defined as reference test. The data was divided into a training (3293), validation (252) and test dataset (141). We applied a convolutional neural network (U-Net) and used the Intersection-over-Union as validation metric. The performance of the trained neural network on the test dataset was compared against that of seven independent using tooth-level accuracy metrics. Stratification according to lesion depth (enamel lesions E1/2, dentin lesions into middle or inner third D2/3) was applied.
The neural network showed an accuracy of 0.80; dentists' mean accuracy was significantly lower at 0.71 (min-max: 0.61-0.78, p < 0.05). The neural network was significantly more sensitive than dentists (0.75 versus 0.36 (0.19-0.65; p = 0.006), while its specificity was not significantly lower (0.83) than those of the dentists (0.91 (0.69-0.98; p > 0.05); p > 0.05). The neural network showed robust sensitivities at or above 0.70 for both initial and advanced lesions. Dentists largely showed low sensitivities for initial lesions (all except one dentist showed sensitivities below 0.25), while those for advanced ones were between 0.40 and 0.75.
To detect caries lesions on bitewing radiographs, a deep neural network was significantly more accurate than dentists.
Deep learning may assist dentists to detect especially initial caries lesions on bitewings. The impact of using such models on decision-making should be explored.
我们旨在应用深度学习技术来检测口内片上不同放射学延伸的龋损病变，假设其准确性明显高于单个牙医。
由四位经验丰富的牙医评估 3686 张口内片。以像素级方式标记龋损病变。所有像素的并集被定义为参考测试。数据分为训练集（3293 张）、验证集（252 张）和测试集（141 张）。我们应用了卷积神经网络（U-Net），并使用交并比作为验证指标。在测试集上，将训练好的神经网络的性能与七位独立牙医使用牙级准确性指标进行比较。根据病变深度（釉质病变 E1/2、牙本质病变进入中或内三分之一 D2/3）进行分层。
神经网络的准确率为 0.80；牙医的平均准确率明显较低，为 0.71（最小值-最大值：0.61-0.78，p<0.05）。神经网络的敏感性明显高于牙医（0.75 与 0.36（0.19-0.65；p=0.006），而特异性并不明显低于牙医（0.83 与 0.91（0.69-0.98；p>0.05）；p>0.05）。神经网络对初发和进展性病变的敏感性均在 0.70 以上，具有稳健性。牙医对初发病变的敏感性普遍较低（除一位牙医外，其余牙医的敏感性均低于 0.25），而对进展性病变的敏感性在 0.40 到 0.75 之间。
在口内片上检测龋损病变时，深度神经网络明显比牙医更准确。
深度学习技术可能有助于牙医在口内片上检测特别是初发龋损病变。应探索使用此类模型对决策的影响。
