Oral Radiology 

About: Oral Radiology is an academic journal published by Springer Science+Business Media. The journal publishes majorly in the area(s): Medicine & Oral and maxillofacial surgery. It has an ISSN identifier of 0911-6028. Over the lifetime, 1004 publications have been published receiving 5288 citations. The journal is also known as: Oral radiology (Print).

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

Abstract: To apply a deep-learning system for diagnosis of maxillary sinusitis on panoramic radiography, and to clarify its diagnostic performance. 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. 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. 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.

Evaluation of an artificial intelligence system for detecting vertical root fracture on panoramic radiography

Abstract: The aim of this study was to evaluate the use of a convolutional neural network (CNN) system for detecting vertical root fracture (VRF) on panoramic radiography. Three hundred panoramic images containing a total of 330 VRF teeth with clearly visible fracture lines were selected from our hospital imaging database. Confirmation of VRF lines was performed by two radiologists and one endodontist. Eighty percent (240 images) of the 300 images were assigned to a training set and 20% (60 images) to a test set. A CNN-based deep learning model for the detection of VRFs was built using DetectNet with DIGITS version 5.0. To defend test data selection bias and increase reliability, fivefold cross-validation was performed. Diagnostic performance was evaluated using recall, precision, and F measure. Of the 330 VRFs, 267 were detected. Twenty teeth without fractures were falsely detected. Recall was 0.75, precision 0.93, and F measure 0.83. The CNN learning model has shown promise as a tool to detect VRFs on panoramic images and to function as a CAD tool.

Trends in maxillofacial cone-beam computed tomography usage

Abstract: Cone-beam computed tomography (CBCT) is making inroads into dental practice worldwide, both in terms of adding the third dimension to diagnosis, and also in terms of enabling image-guided treatment strategies. This article reports trends in the early referral pattern of patients to a CBCT facility in the United States. With institutional review board approval, a retrospective study was made of sequential CBCT radiographic reports made by a specialist oral and maxillofacial radiology service from May 2004 through January 2006 (n = 329). Demographic and referral data were extracted from the reports. Descriptive statistics identified referral patterns, trends, and indications for CBCT. Comparisons were made with the Rogers' Product Innovation Adoption curve. The mean age of referred patients was 45 ± 21 years, and there was a predominance of women (62%). Oral and maxillofacial surgeons (51%) and periodontology specialists (17%) made most patient referrals. The listed reasons for CBCT referrals were dental implant planning (40%), suspected surgical pathology (24%), and temporomandibular joint analysis (16%). Other uses included planning extraction of impacted teeth and orthodontic assessment. Over the period of the study, the numbers of pathology diagnosis cases remained relatively constant, while adoption of CBCT for dental implant planning followed closely the first three stages of the Rogers' Product Innovation Adoption curve. Alongside this increased CBCT adoption for dental implant planning, there was an associated increased demand for use of DICOM image sets for laser modeling and provision of surgical guides. Diagnosis will probably remain a constant source of referral for CBCT examination by oral and maxillofacial radiologists. Nevertheless, more specialized applications such as laser-guided model fabrication and image-guided surgery are expanding indications for CBCT referrals by dentists and also expanding the horizons of clinical dental practice.

Observation of the retromolar foramen and canal of the mandible: a CBCT and macroscopic study

Abstract: The retromolar foramen (RMF) is an anatomical structure on the alveolar surface of the retromolar area. This foramen runs consecutive to the retromolar canal (RMC), which diverges from the mandibular canal. It is important to confirm the RMF and canal locations prior to surgical procedures, such as extraction of an impacted molar and bone harvesting as a donor site for bone graft surgery. This aim of this study was to investigate the RMF in Japanese cadaver mandibles using cone-beam computed tomography (CBCT) images and anatomical observations. Ninety sides of 46 cadaver mandibles were investigated in this study. CBCT images around the retromolar region were acquired for all of the mandibles. The frequency and anteroposterior and buccolingual locations of the RMF were examined on these images. Subsequently, four sides of three mandibles were dissected to confirm the contents of the RMC/RMF. In 24 of 46 (52%) mandibles and 34 of 90 (37%) sides, at least one RMF was observed in the images. In 26 dentate mandibles, 12 (48%) mandibles and 14 (33%) sides presented at least one RMF. The average location of the RMF was 14.4 mm posterior from the distal edge of the second molar. The buccolingual location was 3.0 mm lingual from the mandibular canal. Observations made during the cadaver dissections confirmed that the vessels and nerves diverged from the mandibular canal. The findings suggest that the RMF is not a rare anatomical structure and that practitioners should take this foramen into account in all anesthetic and surgical procedures involving the retromolar area.

Quantitative analysis of metallic artifacts caused by dental metals: comparison of cone-beam and multi-detector row CT scanners

Abstract: To quantitatively compare the streak artifacts produced by dental metals in a cone-beam computed tomography (CBCT) device and a multi-detector row computed tomography (MDCT) scanner in relation to metal types and imaging parameters. Cubes of aluminum, titanium, cobalt–chromium alloy, and type IV gold alloy were scanned with CBCT and MDCT scanners at tube voltages of 80 and 100 peak kV (kVp), and currents of 100 and 170 mAs by MDCT, and 102 and 170 mAs by CBCT. Artifact areas were quantified using ImageJ software. Artifact areas for the same metals and imaging parameters were smaller with CBCT than with MDCT under most conditions. Type IV gold alloy caused the largest artifact areas, followed by cobalt–chromium alloy, titanium, and aluminum, respectively. Higher tube voltage was associated with smaller artifact areas under most conditions, whereas increasing tube current had no consistent effect on artifact area using either CT device. CBCT was associated with smaller artifact areas than MDCT for the same parameters. Type IV gold alloy produced the largest artifact areas among the tested metals, but metallic artifacts could be reduced by increasing the tube voltage.