A pilot study of a deep learning approach to detect marginal bone loss around implants

Background Recently, there has been considerable innovation in artificial intelligence (AI) for healthcare. Convolutional neural networks (CNNs) show excellent object detection and classification performance. This study assessed the accuracy of an artificial intelligence (AI) application for the detection of marginal bone loss on periapical radiographs. Methods A Faster region-based convolutional neural network (R-CNN) was trained. Overall, 1670 periapical radiographic images were divided into training (n = 1370), validation (n = 150), and test (n = 150) datasets. The system was evaluated in terms of sensitivity, specificity, the mistake diagnostic rate, the omission diagnostic rate, and the positive predictive value. Kappa (κ) statistics were compared between the system and dental clinicians. Results Evaluation metrics of AI system is equal to resident dentist. The agreement between the AI system and expert is moderate to substantial (κ = 0.547 and 0.568 for bone loss sites and bone loss implants, respectively) for detecting marginal bone loss around dental implants. Conclusions This AI system based on Faster R-CNN analysis of periapical radiographs is a highly promising auxiliary diagnostic tool for peri-implant bone loss detection.

Differences in Demographic and Radiographic Characteristics between Patients with Visible and Invisible T1 Slopes on Lateral Cervical Radiographic Images

Introduction: The T1 slope is important for cervical surgical planning, and it may be invisible on radiographic images. The prevalence of T1 invisible cases and the differences in demographic and radiographic characteristics between patients whose T1 slopes are visible or invisible remains unexplored. Methods: This pilot study aimed to evaluate the differences in these characteristics between outpatients whose T1 slopes were visible or invisible on radiographic images. Patients (n = 60) who underwent cervical radiography, whose T1 slope was confirmed clearly, were divided into the visible (V) group and invisible (I) group. The following radiographic parameters were measured: (1) C2-7 sagittal vertical axis (SVA), (2) C2-7 angle in neutral, flexion, and extension positions. Results: Based on the T1 slope visibility, 46.7% of patients were included in group I. The I group had significantly larger C2-7 SVA than the V group for males (p < 0.05). The C2-7 SVA tended to be larger in the I group, without significant difference for females (p = 0.362). Discussion: The mean C2-7 angle in neutral and flexion positions was not significantly different between the V and I groups for either sex. The mean C2-7 angle in the extension position was greater in the V group. The T1 slope was invisible in males with high C2-7 SVA.

Development of a Computational Tool for the Estimation of Alveolar Bone Loss in Oral Radiographic Images

The present study evaluated a newly developed computational tool (CT) to assess the alveolar bone space and the alveolar crest angle and compares it to dentist assessment (GT). The novel tool consisted of a set of processes initiated with image enhancement, points localization, and angle and area calculations. In total, we analyzed 148 sites in 39 radiographic images, and among these, 42 sites were selected and divided into two groups of non-periodontitis and periodontitis. The alveolar space area (ASA) and alveolar crest angle (ACA) were estimated. The agreement between the computer software and the ground truth was analyzed using the Bland–Altman plot. The sensitivity and specificity of the computer tool were measured using the ROC curve. The Bland–Altman plot showed an agreement between the ground truth and the computational tool in all of the parameters assessed. The ROC curve showed 100% sensitivity and 100% specificity for 12.67 mm of the alveolar space area. The maximum percentage of sensitivity and specificity were 80.95% for 13.63 degrees of the alveolar crest angle. Computer tool assessment provides accurate disease severity and treatment monitoring for evaluating the alveolar space area (ASA) and the alveolar crest angle (ACA).

A Novel Approach towards Automatic Contour Identification of Jaw Cysts from Digital Panoramic Radiographs to improvise the Treatment planning

Panoramic dental x-ray, a two-dimensional dental x-ray that captures the entire mouth in a single image, is used for the initial screening of various dental anomalies. One such is Jaw bone cyst, which, if not identified earlier, may lead to complications which in turn may lead to disfigurement and loss of function. Hence processing of radiographic images plays a vital role in identifying and locating the cystic region and extracting related features to assist clinical experts in further analysis. Objective: To develop an application of active contour model, known as Geodesic Active Contour, to generate Panoramic Dental X-Ray, a single 2 D X-ray image of the entire mouth highlighting the dental specifications. Methods: The process involves the image conversion from the OPG image into grayscale, Contrast adjustment using intensity level slicing, edge smoothing, segmentation, and cyst segmentation by Morphological Geodesic Active Contour to obtain the results. Hence processing of radiographic images plays a vital role in identifying and locating the cystic region. It is crucial in extracting related features to assist clinical experts in further analysis. Conclusion: When efficient and accurate diagnostic methods exist, the treatment and cure become easy and concrete. Based on the morphological snake and level sets, it aims at identifying the boundary by minimizing the energy. Results: Using the structural similarity index, an accuracy of 97.6% is obtained. Advances in Knowledge: This process is advantageous as it is simpler, faster, and does not suffer from instability problems. Morphological methods improve their functional gradient descent by improving stability and speed. The hysteresis algorithm exhibits better edge detection performance, a significant reduction in computational time and scalability.

Intractable Trigeminal Neuralgia Secondary to Osteoma of the Clivus: A Case Report and Literature Review

Skull base osteomas (SBOs) are benign tumors that are frequently detected on radiographic images by coincidence. They are known for being slow-growing tumors and rarely symptomatic. The therapeutic approach for SBOs can differ substantially. Depending on the symptoms, size, and location of the tumor, this can range from serial observation to vigorous surgical extirpation. Clival osteoma is extremely rare. We report a case of clival osteoma, causing intractable trigeminal neuralgia due to the pressure effect on the trigeminal nerve at Meckel's cave. We also provide a review of pertinent literature. A 37-year-old woman presented with intractable trigeminal neuralgia. Cranial magnetic resonance imaging (MRI) demonstrated a large, lobulated, extra-axial lesion involving the right cerebellopontine angle and epicentering the clivus. Pathologically, the specimen was proven to be osteoma. The patient reported complete symptom resolution over a 4-year follow-up period. To the best of the authors' knowledge, this is the first clinical case of intractable trigeminal neuralgia due to clival osteoma.

Automated Prediction of Extraction Difficulty and Inferior Alveolar Nerve Injury for Mandibular Third Molar Using a Deep Neural Network

Extraction of mandibular third molars is a common procedure in oral and maxillofacial surgery. There are studies that simultaneously predict the extraction difficulty of mandibular third molar and the complications that may occur. Thus, we propose a method of automatically detecting mandibular third molars in the panoramic radiographic images and predicting the extraction difficulty and likelihood of inferior alveolar nerve (IAN) injury. Our dataset consists of 4903 panoramic radiographic images acquired from various dental hospitals. Seven dentists annotated detection and classification labels. The detection model determines the mandibular third molar in the panoramic radiographic image. The region of interest (ROI) includes the detected mandibular third molar, adjacent teeth, and IAN, which is cropped in the panoramic radiographic image. The classification models use ROI as input to predict the extraction difficulty and likelihood of IAN injury. The achieved detection performance was 99.0% mAP over the intersection of union (IOU) 0.5. In addition, we achieved an 83.5% accuracy for the prediction of extraction difficulty and an 81.1% accuracy for the prediction of the likelihood of IAN injury. We demonstrated that a deep learning method can support the diagnosis for extracting the mandibular third molar.

Diagnostic effect of artificial intelligence solution for referable thoracic abnormalities on chest radiography: a multicenter respiratory outpatient diagnostic cohort study

Objectives We aim ed to evaluate a commercial artificial intelligence (AI) solution on a multicenter cohort of chest radiographs and to compare physicians' ability to detect and localize referable thoracic abnormalities with and without AI assistance. Methods In this retrospective diagnostic cohort study, we investigated 6,006 consecutive patients who underwent both chest radiography and CT. We evaluated a commercially available AI solution intended to facilitate the detection of three chest abnormalities (nodule/masses, consolidation, and pneumothorax) against a reference standard to measure its diagnostic performance. Moreover, twelve physicians, including thoracic radiologists, board-certified radiologists, radiology residents, and pulmonologists, assessed a dataset of 230 randomly sampled chest radiographic images. The images were reviewed twice per physician, with and without AI, with a 4-week washout period. We measured the impact of AI assistance on observer's AUC, sensitivity, specificity, and the area under the alternative free-response ROC (AUAFROC). Results In the entire set (n = 6,006), the AI solution showed average sensitivity, specificity, and AUC of 0.885, 0.723, and 0.867, respectively. In the test dataset (n = 230), the average AUC and AUAFROC across observers significantly increased with AI assistance (from 0.861 to 0.886; p = 0.003 and from 0.797 to 0.822; p = 0.003, respectively). Conclusions The diagnostic performance of the AI solution was found to be acceptable for the images from respiratory outpatient clinics. The diagnostic performance of physicians marginally improved with the use of AI solutions. Further evaluation of AI assistance for chest radiographs using a prospective design is required to prove the efficacy of AI assistance. Key Points • AI assistance for chest radiographs marginally improved physicians’ performance in detecting and localizing referable thoracic abnormalities on chest radiographs. • The detection or localization of referable thoracic abnormalities by pulmonologists and radiology residents improved with the use of AI assistance.

Spontaneous pre-axial polydactyly in Swiss mice

ABSTRACT: Spontaneous polydactyly has been described in several species, but only one report about it in Swiss mice. The aim of the current study was to report the spontaneous occurrence of pre-axial polydactyly in Swiss mice. Clinical examination showed one extra toe laterally to the first digit, in the plantar region, alopecia in the back, altered face growth anatomy and changed perineal region anatomy. Pre-axial polydactyly in the tibial side, fused metatarsals and Y-shaped free phalanges were evidenced in the radiographic images. Pre-axial polydactyly observed in the plantar region differed from that in reports on albino Swiss mice with post-axial polydactyly (Po/Po+) phenotype featured by one extra toe in the ulnar side of one, or both, front limbs, which is the dominant feature. The observed findings highlight the importance of both clinical examinations and close attention by professionals involved in rodents’ breeding on physical changes resulting from different causes, including the genetic ones, since they reveal mutations and, sometimes, new biomodels.