Postgraduate Program in Dentistry and Health, Federal University of Bahia, Salvador, Brazil.
Computer Science Department, Federal Institute of Education, Science and Technology of Bahia, Senhor do Bonfim, Bahia, Brazil.
Dentomaxillofac Radiol. 2021 Oct 1;50(7):20210002. doi: 10.1259/dmfr.20210002. Epub 2021 Apr 29.
To analyse the automatic classification performance of a convolutional neural network (CNN), Google Inception v3, using tomographic images of odontogenic keratocysts (OKCs) and ameloblastomas (AMs).
For construction of the database, we selected axial multidetector CT images from patients with confirmed AM ( = 22) and OKC ( = 18) based on a conclusive histopathological report. The images ( = 350) were segmented manually and data augmentation algorithms were applied, totalling 2500 images. The k-fold × five cross-validation method ( = 2) was used to estimate the accuracy of the CNN model.
The accuracy and standard deviation (%) of cross-validation for the five iterations performed were 90.16 ± 0.95, 91.37 ± 0.57, 91.62 ± 0.19, 92.48 ± 0.16 and 91.21 ± 0.87, respectively. A higher error rate was observed for the classification of AM images.
This study demonstrated a high classification accuracy of Google Inception v3 for tomographic images of OKCs and AMs. However, AMs images presented the higher error rate.
分析卷积神经网络（CNN）——Google Inception v3 在牙源性角化囊肿（OKC）和造釉细胞瘤（AM）的断层图像中的自动分类性能。
为了构建数据库，我们根据明确的组织病理学报告，选择了经证实患有 AM（n=22）和 OKC（n=18）的患者的轴向多排 CT 图像。这些图像（n=350）由手动分割，并应用了数据扩充算法，总共得到了 2500 张图像。我们使用 k 折×五交叉验证法（n=2）来估计 CNN 模型的准确性。
五次迭代的交叉验证准确率和标准差（%）分别为 90.16±0.95、91.37±0.57、91.62±0.19、92.48±0.16 和 91.21±0.87。AM 图像的分类错误率较高。
本研究表明，Google Inception v3 对 OKC 和 AM 的断层图像具有较高的分类准确性。然而，AM 图像的错误率较高。
