BACKGROUND

Mobile clinics offer crucial healthcare services, including X-ray examinations, to underserved communities. Minimizing image repeats in this setting is vital due to radiation exposure, patient inconvenience, and cost implications.

This study investigated the prevalence and causes of image repeat in conventional radiography performed within mobile clinics in the Philippines.

A retrospective review analyzed data from five mobile clinics located in two highly urbanized cities in the Philippines from July to December 2023). Radiology staff assessed image quality, with suboptimal images requiring retakes. Reasons for rejection were categorized.

Out of 871 radiographs taken, 118 (13.55%) were repeated. Vertebrae and pelvic girdle images had the highest repeat rates (33.33%). Positioning errors were the most common cause (44.07%), followed by underexposure and overexposure.

This study identified a concerning repeat rate (13.55%) for mobile X-rays, primarily due to improper patient positioning, particularly for specific body parts. Targeted training programs and stricter protocols for mobile clinic staff are needed. Radiography education should also emphasize these skills, potentially through collaboration with mobile clinic operators to ensure graduates are prepared for the unique challenges of this environment.

OBJECTIVES: This study investigate the clinical and imaging features of Ewing sarcoma (ES) of the jaw. METHODS: Eight cases of pathologically diagnosed ES of the jaw from January 2010 to June 2022 were included in the study. Clinical and radiological features were retrospectively analyzed. RESULTS: Among the eight cases, the mean age at onset was 29.4 years, and the male to female ratio was 7∶1. The predilecting site was the posterior part of mandible, accounting for 75% of the cases. The lesions often exhibited early numbness of the lower lip and lymphadenopathy. The main radiographic manifestation of mandibular lesions was ill-defined radiolucency, mixed with fibrous or brush-like tumor matrix, and soft tissue mass. The maxillary ES lesions mainly presented as lytic bone destruction accompanied by adjacent soft tissue mass. Periosteal ossification was rarely seen. CONCLUSIONS The clinical and imaging characteristics of ES in the jaw are helpful for its diagnosis.
Male ; Humans ; Female ; Sarcoma, Ewing/pathology* ; Retrospective Studies ; Radiography ; Mandible/pathology* ; Lip ; Bone Neoplasms

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.
Humans ; Ameloblastoma/diagnostic imaging* ; Deep Learning ; Diagnosis, Differential ; Radiography, Panoramic ; Retrospective Studies ; Odontogenic Cysts/diagnostic imaging* ; Odontogenic Tumors

During interventional procedures,subjects are exposed to direct and scattered X-rays.Establishing diagnostic reference levels is an ideal way to optimize the radiation dose and reduce radiation hazard.In recent years,diagnostic reference levels in interventional radiology have been established in different countries.However,because of the too many indicators for characterizing the radiation dose,the indicators used to establish diagnostic reference levels vary in different countries.The research achievements in this field remain to be reviewed.We carried out a retrospective analysis of the definition,establishment method,application,and main factors influencing the dose difference of the diagnostic reference level,aiming to provide a basis for establishing the diagnostic reference level for interventional procedures in China.
Humans ; Diagnostic Reference Levels ; Radiology, Interventional/methods* ; Radiation Dosage ; Retrospective Studies ; Radiography

OBJECTIVE: To validate the effectiveness of a novel comprehensive classification for intertrochanteric fracture (ITF). METHODS: The study included 616 patients with ITF, including 279 males (45.29%) and 337 females (54.71%); the age ranged from 23 to 100 years, with an average of 72.5 years. Two orthopaedic residents (observers Ⅰ and Ⅱ) and two senior orthopaedic surgeons (observers Ⅲ and Ⅳ) were selected to classify the CT imaging data of 616 patients in a random order by using the AO/Orthopaedic Trauma Association (AO/OTA) classification of 1996/2007 edition, the AO/OTA classification of 2018 edition, and the novel comprehensive classification method at an interval of 1 month. Kappa consistency test was used to evaluate the intra-observer and inter-observer consistency of the three ITF classification systems. RESULTS: The inter-observer consistency of the three classification systems evaluated by 4 observers twice showed that the 3 classification systems had strong inter-observer consistency. Among them, the κ value of the novel comprehensive classification was higher than that of the AO/OTA classification of 1996/2007 edition and 2018 edition, and the experience of observers had a certain impact on the classification results, and the inter-observer consistency of orthopaedic residents was slightly better than that of senior orthopaedic surgeons. The intra-observer consistency of two evaluations of three classification systems by 4 observers showed that the consistency of the novel comprehensive classification was better for the other 3 observers, except that the consistency of observer Ⅳ in the AO/OTA classification of 2018 version was slightly higher than that of the novel comprehensive classification. The results showed that the novel comprehensive classification has higher repeatability, and the intra-observer consistency of senior orthopaedic surgeons was better than that of orthopaedic residents. CONCLUSION The novel comprehensive classification system has good intra- and inter-observer consistency, and has high validity in the classification of CT images of ITF patients; the experience of observers has a certain impact on the results of the three classification systems, and those with more experiences have higher intra-observer consistency.
Male ; Female ; Humans ; Young Adult ; Adult ; Middle Aged ; Aged ; Aged, 80 and over ; Observer Variation ; Reproducibility of Results ; Hip Fractures/surgery* ; Tomography, X-Ray Computed/methods* ; Radiography

OBJECTIVE: To observe clinical effect of percutaneous minimally invasive osteotomy with 8-shaped bandage and hallux valgus splint fixation in treating moderate hallux valgus. METHODS: Totally 23 patients with moderate hallux valgus were treated with percutaneous minimally invasive osteotomy with 8-shaped bandage and hallux valgus splint fixation from August 2019 to January 2021, and 1 patient was loss to follow-up, and finally 22 patients(30 feet) were included, 4 males (6 feet) and 18 females(24 feet), aged from 27 to 66 years old with an average of(50.59±11.95) years old. Hallux valgus angle (HVA), intermetatarsal angle (IMA), metatarsal span (the distance between the first and the fifth metatarsal bones), changed of soft tissue width, American Orthopaedic Foot and Ankle Society(AOFAS) score, and Visual Analogue Scale (VAS) were collected and compared before operation and 6 months after operation. RESULTS: Twenty-two patients were followed up from 5.7 to 6.4 months with an average of (6.13±0.85) months. The first metatarsal osteotomy of patients were obtained bone union, and deformity of the toes was corrected. Complications such as avascular necrosis of metatarsal head and transfer metatarsalgia were not occurred. Postoperative HVA, IMA, metatarsal span, soft tissue width, VAS, AOFAS score at 6 months were significantly improved compared with pre-operation (P<0.01). According to AOFAS score at 6 months after operation, 10 feet were excellent, 18 good and 2 poor. Two feet with poor were excellent after prolonged 8-shaped bandage and hallux valgus splint fixation time. CONCLUSION Percutaneous minimally invasive osteotomy with 8-shaped bandage and hallux valgus splint fixation for the treatment of moderate hallux valgus could better correct deformity of hallux valgus, relieve foot symptoms, good recovery of postoperative function, and has a significant clinical efficacy.
Male ; Female ; Humans ; Adult ; Middle Aged ; Aged ; Hallux Valgus/diagnostic imaging* ; Splints ; Radiography ; Bunion ; Treatment Outcome ; Metatarsal Bones/surgery* ; Osteotomy ; Bandages

OBJECTIVE: To explore the effect of a modified three-point bending fracture device for establishing a rabbit model of closed tibial fracture. METHODS: The model of closed tibial fracture was established in 40 6-month-old male New Zealand white rabbits with a body weight of 2.5 to 3.0 kg, and the model was verified at 6 weeks after operation. Five rabbits underwent pre modeling without temporary external fixation before modeling, and then were fractured with a modified three-point bending fracture device;35 rabbits underwent formal modeling. Before modeling, needles were inserted, and splints were fixed externally, and then the fracture was performed with a modified three-point bending fracture device. The fracture model and healing process were evaluated by imaging and histopathology at 2 hours, 4 weeks, and 6 weeks after operation. RESULTS: Two hours after modeling, the prefabricated module showed oblique fracture in varying degrees and the broken end shifted significantly;Except for 1 comminuted fracture, 2 curved butterfly fractures and 2 without obvious fracture line, the rest were simple transverse and oblique fractures without obvious displacement in formal modeling group. According to the judgment criteria, the success rate of the model was 85.71%. Four weeks after modeling, the fixed needle and splint of the experimental rabbits were in good position, the fracture alignment was good, the fracture line was blurred, many continuous callus growths could be seen around the fracture end, and the callus density was high. Six weeks after modeling, many thick new bone trabeculae at the fracture, marginal osteoblasts attached, and a small number of macrophages were seen under the microscope. The intramembrane osteogenesis area was in the preparation bone stage, the medullary cavity at the fracture had been partially reopened, the callus was in the absorption plastic stage, and many osteoclasts were visible. The X-ray showed that the fracture line almost disappeared, part of the medullary cavity had been opened, the external callus was reduced around, the callus was in the plastic stage, and the bone cortex was continuous. It suggests that the fracture model showed secondary healing. CONCLUSION The improved three-point bending fracture device can establish a stable rabbit model of closed tibial fracture, and the operation is simple, which meets the requirements of closed fracture model in basic research related to fracture healing.
Rabbits ; Male ; Animals ; Bony Callus ; Fracture Healing ; Tibial Fractures/surgery* ; Osteogenesis ; Radiography

OBJECTIVE: In order to solve the technical problems, clinical researchers face the process of medical imaging analysis such as data labeling, feature extraction and algorithm selection, a medical imaging oriented multi-disease research platform based on radiomics and machine learning technology was designed and constructed. METHODS: Five aspects including data acquisition, data management, data analysis, modeling and data management were considered. This platform provides comprehensive functions such as data retrieve and data annotation, image feature extraction and dimension reduction, machine learning model running, results validation, visual analysis and automatic generation of analysis reports, thus an integrated solution for the whole process of radiomics analysis has been generated. RESULTS: Clinical researchers can use this platform for the whole process of radiomics and machine learning analysis for medical images, and quickly produce research results. CONCLUSIONS This platform greatly shortens the time for medical image analysis research, decreasing the work difficulty of clinical researchers, as well as significantly promoting their working efficiency.
Machine Learning ; Diagnostic Imaging ; Algorithms ; Radiography

Objective: To develop a multi-classification orthodontic image recognition system using the SqueezeNet deep learning model for automatic classification of orthodontic image data. Methods: A total of 35 000 clinical orthodontic images were collected in the Department of Orthodontics, Capital Medical University School of Stomatology, from October to November 2020 and June to July 2021. The images were from 490 orthodontic patients with a male-to-female ratio of 49∶51 and the age range of 4 to 45 years. After data cleaning based on inclusion and exclusion criteria, the final image dataset included 17 453 face images (frontal, smiling, 90° right, 90° left, 45° right, and 45° left), 8 026 intraoral images [frontal occlusion, right occlusion, left occlusion, upper occlusal view (original and flipped), lower occlusal view (original and flipped) and coverage of occlusal relationship], 4 115 X-ray images [lateral skull X-ray from the left side, lateral skull X-ray from the right side, frontal skull X-ray, cone-beam CT (CBCT), and wrist bone X-ray] and 684 other non-orthodontic images. A labeling team composed of orthodontic doctoral students, associate professors, and professors used image labeling tools to classify the orthodontic images into 20 categories, including 6 face image categories, 8 intraoral image categories, 5 X-ray image categories, and other images. The data for each label were randomly divided into training, validation, and testing sets in an 8∶1∶1 ratio using the random function in the Python programming language. The improved SqueezeNet deep learning model was used for training, and 13 000 natural images from the ImageNet open-source dataset were used as additional non-orthodontic images for algorithm optimization of anomaly data processing. A multi-classification orthodontic image recognition system based on deep learning models was constructed. The accuracy of the orthodontic image classification was evaluated using precision, recall, F1 score, and confusion matrix based on the prediction results of the test set. The reliability of the model's image classification judgment logic was verified using the gradient-weighted class activation mapping (Grad-CAM) method to generate heat maps. Results: After data cleaning and labeling, a total of 30 278 orthodontic images were included in the dataset. The test set classification results showed that the precision, recall, and F1 scores of most classification labels were 100%, with only 5 misclassified images out of 3 047, resulting in a system accuracy of 99.84%(3 042/3 047). The precision of anomaly data processing was 100% (10 500/10 500). The heat map showed that the judgment basis of the SqueezeNet deep learning model in the image classification process was basically consistent with that of humans. Conclusions: This study developed a multi-classification orthodontic image recognition system for automatic classification of 20 types of orthodontic images based on the improved SqueezeNet deep learning model. The system exhibitted good accuracy in orthodontic image classification.
Humans ; Male ; Female ; Child, Preschool ; Child ; Adolescent ; Young Adult ; Adult ; Middle Aged ; Deep Learning ; Reproducibility of Results ; Radiography ; Algorithms ; Cone-Beam Computed Tomography

Artificial intelligence, represented by deep learning, has received increasing attention in the field of oral and maxillofacial medical imaging, which has been widely studied in image analysis and image quality improvement. This narrative review provides an insight into the following applications of deep learning in oral and maxillofacial imaging: detection, recognition and segmentation of teeth and other anatomical structures, detection and diagnosis of oral and maxillofacial diseases, and forensic personal identification. In addition, the limitations of the studies and the directions for future development are summarized.
Artificial Intelligence ; Deep Learning ; Diagnostic Imaging ; Radiography ; Image Processing, Computer-Assisted