[Application of Deep Learning in Differential Diagnosis of Ameloblastoma and Odontogenic Keratocyst Based on Panoramic Radiographs].

Objective To evaluate the accuracy of different convolutional neural networks (CNN),representative deep learning models,in the differential diagnosis of ameloblastoma and odontogenic keratocyst,and subsequently compare the diagnosis results between models and oral radiologists. Methods A total of 1000 digital panoramic radiographs were retrospectively collected from the patients with ameloblastoma (500 radiographs) or odontogenic keratocyst (500 radiographs) in the Department of Oral and Maxillofacial Radiology,Peking University School of Stomatology.Eight CNN including ResNet (18,50,101),VGG (16,19),and EfficientNet (b1,b3,b5) were selected to distinguish ameloblastoma from odontogenic keratocyst.Transfer learning was employed to train 800 panoramic radiographs in the training set through 5-fold cross validation,and 200 panoramic radiographs in the test set were used for differential diagnosis.Chi square test was performed for comparing the performance among different CNN.Furthermore,7 oral radiologists (including 2 seniors and 5 juniors) made a diagnosis on the 200 panoramic radiographs in the test set,and the diagnosis results were compared between CNN and oral radiologists. Results The eight neural network models showed the diagnostic accuracy ranging from 82.50% to 87.50%,of which EfficientNet b1 had the highest accuracy of 87.50%.There was no significant difference in the diagnostic accuracy among the CNN models (P=0.998,P=0.905).The average diagnostic accuracy of oral radiologists was (70.30±5.48)%,and there was no statistical difference in the accuracy between senior and junior oral radiologists (P=0.883).The diagnostic accuracy of CNN models was higher than that of oral radiologists (P<0.001). Conclusion Deep learning CNN can realize accurate differential diagnosis between ameloblastoma and odontogenic keratocyst with panoramic radiographs,with higher diagnostic accuracy than oral radiologists.

Diagnostic efficacy of CBCT, MRI, and CBCT-MRI fused images in distinguishing articular disc calcification from loose body of temporomandibular joint.

OBJECTIVES: To evaluate the diagnostic efficacy of CBCT-MRI fused images for articular disc calcification of temporomandibular joint (TMJ). MATERIALS AND METHODS: Twenty patients (24 TMJs) whose image examinations showed dense bodies in the TMJ space were included in the study. The locations of dense bodies evaluated by the three experts were used as a reference standard. Three oral and maxillofacial radiology residents evaluated whether the dense bodies were disc calcification or not, with a five-point scale for four sets of images (CBCT alone, MRI alone, both CBCT and MRI observed at a time, and CBCT-MRI fused images) randomly and independently. Each set of images was observed at least 1 week apart. A second evaluation was performed after 4 weeks. Intraclass correlation coefficients were calculated to assess the intra- and inter-observer agreement. The areas under the ROC curves (AUCs) were compared between the four image sets using Z test. RESULTS: Ten cases were determined as articular disc calcifications, and fourteen cases were recognized as loose bodies in the TMJ spaces. The average AUC index for the CBCT-MRI fused images was 0.95 and significantly higher than the other sets (p < 0.01). The intra- and inter-observer agreement in the CBCT-MRI fused images (0.90-0.91, 0.93) was excellent and higher than those in the other images. CONCLUSIONS: CBCT-MRI fused images can significantly improve the observers' reliability and accuracy in determining articular disc calcification of the TMJ. CLINICAL RELEVANCE: The multimodality image fusion is feasible in detecting articular disc calcification of the TMJ which are hard to define by CBCT or MRI alone. It can be utilized especially for inexperienced residents to shorten the learning curve and improve diagnostic accuracy.

Age Estimation using Panoramic Radiographs by Transfer Learning.

OBJECTIVE: To assess the accuracy of transfer learning models for age estimation from panoramic photographs of permanent dentition of patients with an equal sex and age distribution and provide a new method of age estimation. METHODS: The panoramic photographs of 3000 patients with an equal sex and age distribution were divided into three groups: a training set (n = 2400), validation set (n = 300) and test set (n = 300). The ResNet, EffiecientNet, VggNet and DenseNet transfer learning models were trained with the training set. The models were subsequently tested using the data in the test set. The mean absolute errors were calculated and the different features extracted by the deep learning models in different age groups were visualixed. RESULTS: The mean absolute error (MAE) and root mean square error (RMSE) of the optimal transfer learning model EfficientNet-B5 in the test set were 2.83 and 4.59, respectively. The dentition, maxillary sinus, mandibular body and mandibular angle all played a role in age estimation. CONCLUSION: Transfer learning models can extract different features in different age groups and can be used for age estimation in panoramic radiographs.

Diagnostic efficacy of CBCT, MRI and CBCT-MRI fused images in determining anterior disc displacement and bone changes of temporomandibular joint.

OBJECTIVES: To evaluate the diagnostic efficacy of CBCT-MRI fused image for anterior disc displacement and bone changes of temporomandibular joint (TMJ), which are the main imaging manifestations of temporomandibular disorders (TMD). METHODS: Two hundred and thirty-one TMJs of 120 patients who were diagnosed with TMD were selected for the study. The anterior disc displacement, bone defect and bone hyperplasia evaluated by three experts were used as a reference standard. Three residents individually evaluated all the three sets of images, which were CBCT images, MRI images and CBCT-MRI fused images from individual CBCT and MRI images in a random order for the above-mentioned three imaging manifestations with a five-point scale. Each set of images was observed at least 1 week apart. A second evaluation was performed 4 weeks later. Intra- and interobserver agreements were assessed using the intraclass correlation coefficient (ICC). The areas under the ROC curves (AUCs) of the three image sets were compared with a Z test, and p < 0.05 was considered statistically significant. RESULTS: One hundred and forty-five cases were determined as anterior disc displacement, 84 cases as bone defect and 40 cases as bone hyperplasia. The intra- and interobserver agreements in the CBCT-MRI fused image set (0.76-0.91) were good to excellent, and the diagnostic accuracy for bone changes was significantly higher than that of MRI image set (p＜0.05). CONCLUSIONS: CBCT-MRI fused images can display the disc and surrounding bone structures simultaneously and significantly improve the observers' reliability and diagnostic accuracy, especially for inexperienced residents.