Deep-learning systems for diagnosing cleft palate on panoramic radiographs in patients with cleft alveolus.

OBJECTIVES: The aim of the present study was to create effective deep learning-based models for diagnosing the presence or absence of cleft palate (CP) in patients with unilateral or bilateral cleft alveolus (CA) on panoramic radiographs. METHODS: The panoramic images of 491 patients who had unilateral or bilateral cleft alveolus were used to create two models. Model A, which detects the upper incisor area on panoramic radiographs and classifies the areas into the presence or absence of CP, was created using both object detection and classification functions of DetectNet. Using the same data for developing Model A, Model B, which directly classifies the presence or absence of CP on panoramic radiographs, was created using classification function of VGG-16. The performances of both models were evaluated with the same test data and compared with those of two radiologists. RESULTS: The recall, precision, and F-measure were all 1.00 in Model A. The area under the receiver operating characteristic curve (AUC) values were 0.95, 0.93, 0.70, and 0.63 for Model A, Model B, and the radiologists, respectively. The AUCs of the models were significantly higher than those of the radiologists. CONCLUSIONS: The deep learning-based models developed in the present study have potential for use in supporting observer interpretations of the presence of cleft palate on panoramic radiographs.

Performance of deep learning technology for evaluation of positioning quality in periapical radiography of the maxillary canine.

OBJECTIVES: The aim of the present study was to create and test an automatic system for assessing the technical quality of positioning in periapical radiography of the maxillary canines using deep learning classification and segmentation techniques. METHODS: We created and tested two deep learning systems using 500 periapical radiographs (250 each of good- and bad-quality images). We assigned 350, 70, and 80 images as the training, validation, and test datasets, respectively. The learning model of system 1 was created with only the classification process, whereas system 2 consisted of both the segmentation and classification models. In each model, 500 epochs of training were performed using AlexNet and U-net for classification and segmentation, respectively. The segmentation results were evaluated by the intersection over union method, with values of 0.6 or more considered as success. The classification results were compared between the two systems. RESULTS: The segmentation performance of system 2 was recall, precision, and F measure of 0.937, 0.961, and 0.949, respectively. System 2 showed better classification performance values than those obtained by system 1. The area under the receiver operating characteristic curve values differed significantly between system 1 (0.649) and system 2 (0.927). CONCLUSIONS: The deep learning systems we created appeared to have potential benefits in evaluation of the technical positioning quality of periapical radiographs through the use of segmentation and classification functions.

Detection and classification of unilateral cleft alveolus with and without cleft palate on panoramic radiographs using a deep learning system.

Although panoramic radiography has a role in the examination of patients with cleft alveolus (CA), its appearances is sometimes difficult to interpret. The aims of this study were to develop a computer-aided diagnosis system for diagnosing the CA status on panoramic radiographs using a deep learning object detection technique with and without normal data in the learning process, to verify its performance in comparison to human observers, and to clarify some characteristic appearances probably related to the performance. The panoramic radiographs of 383 CA patients with cleft palate (CA with CP) or without cleft palate (CA only) and 210 patients without CA (normal) were used to create two models on the DetectNet. The models 1 and 2 were developed based on the data without and with normal subjects, respectively, to detect the CAs and classify them into with or without CP. The model 2 reduced the false positive rate (1/30) compared to the model 1 (12/30). The overall accuracy of Model 2 was higher than Model 1 and human observers. The model created in this study appeared to have the potential to detect and classify CAs on panoramic radiographs, and might be useful to assist the human observers.

Effects of 1 year of training on the performance of ultrasonographic image interpretation: A preliminary evaluation using images of Sjögren syndrome patients.

PURPOSE: This study investigated the effects of 1 year of training on imaging diagnosis, using static ultrasonography (US) salivary gland images of Sjögren syndrome patients. MATERIALS AND METHODS: This study involved 3 inexperienced radiologists with different levels of experience, who received training 1 or 2 days a week under the supervision of experienced radiologists. The training program included collecting patient histories and performing physical and imaging examinations for various maxillofacial diseases. The 3 radiologists (observers A, B, and C) evaluated 400 static US images of salivary glands twice at a 1-year interval. To compare their performance, 2 experienced radiologists evaluated the same images. Diagnostic performance was compared between the 2 evaluations using the area under the receiver operating characteristic curve (AUC). RESULTS: Observer A, who was participating in the training program for the second year, exhibited no significant difference in AUC between the first and second evaluations, with results consistently comparable to those of experienced radiologists. After 1 year of training, observer B showed significantly higher AUCs than before training. The diagnostic performance of observer B reached the level of experienced radiologists for parotid gland assessment, but differed for submandibular gland assessment. For observer C, who did not complete the training, there was no significant difference in the AUC between the first and second evaluations, both of which showed significant differences from those of the experienced radiologists. CONCLUSION: These preliminary results suggest that the training program effectively helped inexperienced radiologists reach the level of experienced radiologists for US examinations.

Cone-beam computed tomography classification of the mandibular second molar root morphology and its relationship to panoramic radiographic appearance.

OBJECTIVES: This study aimed to clarify the relationship between the panoramic radiographic appearance and the longitudinal cone-beam computed tomography (CBCT) classification of root configurations of the mandibular second molar. METHODS: Panoramic radiographs of 1058 mandibular second molars were classified into five types according to the number and configuration of the roots. These molars were also examined with CBCT at four levels between the pulp chamber and the root apex, and axial images perpendicular to the root axis were categorized into three patterns: single (fused root with small grooves on both buccal and lingual sides or a round root with one canal); double (two separate roots with a trabecular appearance between them); and C-shaped (root with a deep groove opening only on the lingual or buccal side relative to the opposite side). Based on these patterns and their scan levels, the CBCT root morphology appearance in each tooth unit was classified into seven groups. Relationships were investigated between these seven CBCT groups and the five panoramic root types. RESULTS: In panoramic types 1 and 2 (with separate roots), 85% had roots with a double pattern (groups II and III) on the CBCT images. In panoramic types 3 and 4 (with fused roots), 85% had C-shaped CBCT patterns at the lower scan levels. CONCLUSIONS: When panoramic images show fused root types, CBCT examinations should be planned to clarify the root canal configuration.

Comparison of 3 deep learning neural networks for classifying the relationship between the mandibular third molar and the mandibular canal on panoramic radiographs.

OBJECTIVE: The aim of this study was to compare time and storage space requirements, diagnostic performance, and consistency among 3 image recognition convolutional neural networks (CNNs) in the evaluation of the relationships between the mandibular third molar and the mandibular canal on panoramic radiographs. STUDY DESIGN: Of 600 panoramic radiographs, 300 each were assigned to noncontact and contact groups based on the relationship between the mandibular third molar and the mandibular canal. The CNNs were trained twice by using cropped image patches with sizes of 70 × 70 pixels and 140 × 140 pixels. Time and storage space were measured for each system. Accuracy, sensitivity, specificity, and the area under the receiver operating characteristic curve (AUC) were determined. Intra-CNN and inter-CNN consistency values were calculated. RESULTS: Time and storage space requirements depended on the depth of CNN layers and number of learned parameters, respectively. The highest AUC values ranged from 0.88 to 0.93 in the CNNs created by 70 × 70 pixel patches, but there were no significant differences in diagnostic performance among any of the models with smaller patches. Intra-CNN and inter-CNN consistency values were good or very good for all CNNs. CONCLUSIONS: The size of the image patches should be carefully determined to ensure acquisition of high diagnostic performance and consistency.

Usefulness of a deep learning system for diagnosing Sjögren's syndrome using ultrasonography images.

OBJECTIVES: We evaluated the diagnostic performance of a deep learning system for the detection of Sjögren's syndrome (SjS) in ultrasonography (US) images, and compared it with the performance of inexperienced radiologists. METHODS: 100 patients with a confirmed diagnosis of SjS according to both the Japanese criteria and American-European Consensus Group criteria and 100 non-SjS patients that had a dry mouth and suspected SjS but were definitively diagnosed as non-SjS were enrolled in this study. All the patients underwent US scans of both the parotid glands (PG) and submandibular glands (SMG). The training group consisted of 80 SjS patients and 80 non-SjS patients, whereas the test group consisted of 20 SjS patients and 20 non-SjS patients for deep learning analysis. The performance of the deep learning system for diagnosing SjS from the US images was compared with the diagnoses made by three inexperienced radiologists. RESULTS: The accuracy, sensitivity and specificity of the deep learning system for the PG were 89.5, 90.0 and 89.0%, respectively, and those for the inexperienced radiologists were 76.7, 67.0 and 86.3%, respectively. The deep learning system results for the SMG were 84.0, 81.0 and 87.0%, respectively, and those for the inexperienced radiologists were 72.0, 78.0 and 66.0%, respectively. The AUC for the inexperienced radiologists was significantly different from that of the deep learning system. CONCLUSIONS: The deep learning system had a high diagnostic ability for SjS. This suggests that deep learning could be used for diagnostic support when interpreting US images.

Deep-learning classification using convolutional neural network for evaluation of maxillary sinusitis on panoramic radiography.

OBJECTIVES: To apply a deep-learning system for diagnosis of maxillary sinusitis on panoramic radiography, and to clarify its diagnostic performance. METHODS: Training data for 400 healthy and 400 inflamed maxillary sinuses were enhanced to 6000 samples in each category by data augmentation. Image patches were input into a deep-learning system, the learning process was repeated for 200 epochs, and a learning model was created. Newly-prepared testing image patches from 60 healthy and 60 inflamed sinuses were input into the learning model, and the diagnostic performance was calculated. Receiver-operating characteristic (ROC) curves were drawn, and the area under the curve (AUC) values were obtained. The results were compared with those of two experienced radiologists and two dental residents. RESULTS: The diagnostic performance of the deep-learning system for maxillary sinusitis on panoramic radiographs was high, with accuracy of 87.5%, sensitivity of 86.7%, specificity of 88.3%, and AUC of 0.875. These values showed no significant differences compared with those of the radiologists and were higher than those of the dental residents. CONCLUSIONS: The diagnostic performance of the deep-learning system for maxillary sinusitis on panoramic radiographs was sufficiently high. Results from the deep-learning system are expected to provide diagnostic support for inexperienced dentists.