Objectives To evaluate the diagnostic performance of a deep convolutional neural network (DCNN)-based computer-assisted diagnosis (CAD) system in the detection of osteoporosis on panoramic radiographs, through a comparison with diagnoses made by oral and maxillofacial radiologists.  Methods: Oral and maxillofacial radiologists with >10 years of experience reviewed the panoramic radiographs of 1268 females {mean [± standard deviation (SD)] age: 52.5 ± 22.3 years} and made a diagnosis of osteoporosis when cortical erosion of the mandibular inferior cortex was observed. Among the females, 635 had no osteoporosis [mean (± SD) age: 32.8 ± SD 12.1 years] and 633 had osteoporosis (72.2 ± 8.5 years). All panoramic radiographs were analysed using three CAD systems, single-column DCNN (SC-DCNN), single-column with data augmentation DCNN (SC-DCNN Augment) and multicolumn DCNN (MC-DCNN). Among the radiographs, 200 panoramic radiographs [mean (± SD) patient age: 63.9 ± 10.7 years] were used for testing the performance of the DCNN in detecting osteoporosis in this study. The diagnostic performance of the DCNN-based CAD system was assessed by receiver operating characteristic (ROC) analysis.  Results: The area under the curve (AUC) values obtained using SC-DCNN, SC-DCNN (Augment) and MC-DCNN were 0.9763, 0.9991 and 0.9987, respectively.  Conclusions: The DCNN-based CAD system showed high agreement with experienced oral and maxillofacial radiologists in detecting osteoporosis. A DCNN-based CAD system could provide information to dentists for the early detection of osteoporosis, and asymptomatic patients with osteoporosis can then be referred to the appropriate medical professionals.

No abstract

PurposeThe aim of this study was to investigate the incidence of bisphosphonate-related osteonecrosis of the jaw (BRONJ) after tooth extraction in patients with osteoporosis on oral bisphosphonates in Korea and to evaluate local factors affecting the development of BRONJ.Materials and MethodsThe clinical records of 320 patients who underwent dental extraction while receiving oral bisphosphonates were reviewed. All patients had a healing period of more than 6 months following the extractions. Each patient's clinical record was used to assess the incidence of BRONJ; if BRONJ occurred, a further radiographic investigation was carried out to obtain a more definitive diagnosis. Various local factors including age, gender, extraction site, drug type, duration of administration, and C-terminal telopeptide (CTx) level were retrieved from the patients' clinical records for evaluating their effect on the incidence of BRONJ.ResultsAmong the 320 osteoporotic patients who underwent tooth extraction, 11 developed BRONJ, reflecting an incidence rate of 3.44%. Out of the local factors that may affect the incidence of BRONJ, gender, drug type, and CTx level showed no statistically significant effects, while statistically significant associations were found for age, extraction site, and duration of administration. The incidence of BRONJ increased with age, was greater in the mandible than the maxilla, and was associated with a duration of administration of more than 3 years.ConclusionTooth extraction in patients on oral bisphosphonates requires careful consideration of their age, the extraction site, and the duration of administration, and close postoperative follow-up should be carried out to facilitate effective early management.

Dental panoramic radiography (DPR) is a method commonly used in dentistry for patient diagnosis. This study presents a new technique that combines a regional convolutional neural network (RCNN), Single Shot Multibox Detector, and heuristic methods to detect and number the teeth and implants with only fixtures in a DPR image. This technology is highly significant in providing statistical information and personal identification based on DPR and separating the images of individual teeth, which serve as basic data for various DPR-based AI algorithms. As a result, the mAP(@IOU = 0.5) of the tooth, implant fixture, and crown detection using the RCNN algorithm were obtained at rates of 96.7%, 45.1%, and 60.9%, respectively. Further, the sensitivity, specificity, and accuracy of the tooth numbering algorithm using a convolutional neural network and heuristics were 84.2%, 75.5%, and 84.5%, respectively. Techniques to analyze DPR images, including implants and bridges, were developed, enabling the possibility of applying AI to orthodontic or implant DPR images of patients.

PurposeThis study was performed to assess the compatibility of cone beam computed tomography (CBCT) synthesized cephalograms with conventional cephalograms, and to find a method for obtaining normative values for three-dimensional (3D) assessments.Materials and MethodsThe sample group consisted of 10 adults with normal occlusion and well-balanced faces. They were imaged using conventional and CBCT cephalograms. The CBCT cephalograms were synthesized from the CBCT data using OnDemand 3D software. Twenty-one angular and 12 linear measurements from each imaging modality were compared and analyzed using paired-t test.ResultsThe linear measurements between the two imaging modalities were not statistically different (p>0.05) except for the U1 to facial plane distance. The angular measurements between the two imaging modalities were not statistically different (p>0.05) with the exception of the gonial angle, ANB difference, and facial convexity.ConclusionTwo-dimensional cephalometric norms could be readily used for 3D quantitative assessment, if corrected for lateral cephalogram distortion.