• Korean Journal of Radiology
• /
• v.23 no.3
• /
• pp.343-354
• /
• 2022

Objective: To develop and evaluate a deep learning-based artificial intelligence (AI) model for detecting skull fractures on plain radiographs in children. Materials and Methods: This retrospective multi-center study consisted of a development dataset acquired from two hospitals (n = 149 and 264) and an external test set (n = 95) from a third hospital. Datasets included children with head trauma who underwent both skull radiography and cranial computed tomography (CT). The development dataset was split into training, tuning, and internal test sets in a ratio of 7:1:2. The reference standard for skull fracture was cranial CT. Two radiology residents, a pediatric radiologist, and two emergency physicians participated in a two-session observer study on an external test set with and without AI assistance. We obtained the area under the receiver operating characteristic curve (AUROC), sensitivity, and specificity along with their 95% confidence intervals (CIs). Results: The AI model showed an AUROC of 0.922 (95% CI, 0.842-0.969) in the internal test set and 0.870 (95% CI, 0.785-0.930) in the external test set. The model had a sensitivity of 81.1% (95% CI, 64.8%-92.0%) and specificity of 91.3% (95% CI, 79.2%-97.6%) for the internal test set and 78.9% (95% CI, 54.4%-93.9%) and 88.2% (95% CI, 78.7%-94.4%), respectively, for the external test set. With the model's assistance, significant AUROC improvement was observed in radiology residents (pooled results) and emergency physicians (pooled results) with the difference from reading without AI assistance of 0.094 (95% CI, 0.020-0.168; p = 0.012) and 0.069 (95% CI, 0.002-0.136; p = 0.043), respectively, but not in the pediatric radiologist with the difference of 0.008 (95% CI, -0.074-0.090; p = 0.850). Conclusion: A deep learning-based AI model improved the performance of inexperienced radiologists and emergency physicians in diagnosing pediatric skull fractures on plain radiographs.

• Proceedings of the Korean Society of Veterinary Clinics Conference
• /
• 2008.10a
• /
• pp.169-169
• /
• 2008

• Journal of Korean Neurosurgical Society
• /
• v.30 no.1
• /
• pp.54-59
• /
• 2001

Objectives : With the advancement of a social life, the pediatric head injuries(PHI) occur greater than ever. Since the PHI differs from adult head injury with regards to mechanism of trauma, prognosis, and mortality, it is important to identify the characteristics of the PHI for its proper treatments and prognosis. Methods : For this study, a series of 365 PHI patients under 15 years of age who were admitted to our hospital, were evaluated from January 1991 to December 1996. The clinical variable studied were age, sex, Glasgow coma score(GCS), causes of trauma, diagnosis, symptoms, associated injuries and Glasgow outcome score (GOS). The characteristics of PHI were evaluated according to presentations of skull fractures, intracranial hemorrhages, associated injuries, GCS at admission and GOS. Results : Mean age of the studied patients was 6.51 years of age. The majority of PHI patients were under the 7 years of age(66.7%). The ratio of male to female was 2.2:1. Seasonally, PHI occurred more frequently during March to August(61.6%). The main causes of the injuries were accidental falls and traffic accidents(47.1% and 46.3%). One hundred ninety seven(54%) patients suffered from skull fractures and 110(30.1%) patients were developed intracranial hemorrhages and acute epidural hematomas(17.8%) which were the most common intracranial hemorrhages. There was statistical significance between skull fractures and intracranial hemorrhage (p=0.032) and between GCS and GOS(p=0.001). However, there was no statistical significance between skull fractures and intracranial hemorrhage(epidural hematomas, subdural hematomas, and intracerebral, intraventricular and subarachnoid hemorrhage)(p=0.061, 0.251 and 0.880). Also there were no significance of prognosis between under the seven and over the 8 years of age(p=0.349). Conclusions : The core management for PHI is prevention from its occurrences. However, when unexpected accident occurs, early diagnosis and treatment for PHI by through examination for associated injuries and other damages even if there is no skull fracture are essential in managing patient's outcome.

• Journal of The Korean Dental Society of Anesthesiology
• /
• v.14 no.4
• /
• pp.233-236
• /
• 2014

In order to reduce jaw fracture accompanied by basal skull or nasal fracture, submental intubation could be generally performed. Albeit submental intubation has been widely accepted, it could develop complications such as nerve injury, glandular duct injury, and orocutaneous fistula. Here, we suggest oral intubation for overcoming complications and providing more stable surgical environment in emergency case. Under oral intubation maintaining in retromolar triangle and buccal corridor space, intermaxillary fixation was successfully underwent in 38-years-old female patient with Le Fort I fracture accompanied by pneumocephalus.

• Journal of the Korean Society of Radiology
• /
• v.11 no.5
• /
• pp.387-392
• /
• 2017

When an infant has visited a hospital due to skull fracture, the rupture of a blood vessel, or skin wounds on the head resulted from an incident, accident, traffic accident, or disease, he/she becomes to undergo anterior/posterior and lateral skull imaging, which is a head test at the department of radiology. In the head test, if the adult skull imaging grid is applied to the imaging, the secondary radiation will be removed to enhance the contrast of the image. However, among the radiation exposure conditions, the tube voltage should be enhanced by 8~10 kVp leading to an increase in the patient exposure. The present study was conducted under assumption that if the same images can be obtained from infant skull imaging without using the skull imaging grid, the exposure dose will be reduced and the artifacts due to grid cut off can be prevented. The researcher measured the radiation dosage using a radiation meter and conducted the subjective evaluation (ROC, receiver operating characteristic) among medical image evaluation methods. Based on the results, when the images were taken without using the grid, the exposure dose was reduced by 0.019 mGy in the anterior/posterior imaging and by 0.02 mGy in the lateral imaging and the image evaluation score was higher by 4 points. In conclusion, if the images of the skulls of infants that visited the hospital are taken with out using the grid, the exposure dose can be reduced, the image artifacts due to grid cut off can be prevented, and the lifespan of the X-ray tube will be extended.

• Journal of Korean Neurosurgical Society
• /
• v.48 no.5
• /
• pp.452-454
• /
• 2010

Traumatic atlantoaxial rotatory fixation (AARF) with accompanying odontoid and C2 articular facet fracture is a very rare injury, and only one such case has been reported in the medical literature. We present here a case of a traumatic AARF associated with an odontoid and comminuted C2 articular facet fracture, and this was treated with skull traction and halo-vest immobilization for 3 months. After removal of the halo-vest immobilization, his neck pain was improved and his neck motion was preserved without any neurologic deficits although mild torticolis was still observed in closer inspection.

• Journal of Korean Neurosurgical Society
• /
• v.46 no.5
• /
• pp.451-458
• /
• 2009

Objective : The purpose of this study was to analyze risk factors that are associated with intracranial lesion, and to propose criteria for classification of mild head injury (MHI), and appropriate treatment guidelines. Methods : The study was based on 898 patients who were admitted to our hospital with Glasgow Coma Scale (GCS) score of 13 to 15 between 2003 and 2007. The patients' initial computerized tomography (CT) findings were reviewed and clinical findings that were associated with intracranial lesions were analyzed. Results : GCS score, loss of consciousness (LOC), age and skull fracture were identified as independent risk factors for intracranial lesions. Based on the data ana lysed in this study, MHI patients were divided into four subgroups : very low risk MHI patients are those with a GCS score of 15 and without a history of LOC or headache; low risk MHI patients have a GCS score of 15 and with LOC and/or headache; medium risk MHI patients are those with a GCS score of 15 and with a skull fracture, neurological deficits or with one or more of the risk factors; high risk MHI patients are those with a GCS score of 15 with abnormal CT findings and GCS score of 14 and 13. Conclusion : A more detailed classification of MHI based on brain CT scan findings and clinical risk factors can potentially improve patient diagnosis. In light of our findings, high risk MHI patients should be admitted and treated in same manner as those with moderate head injury.

• Maxillofacial Plastic and Reconstructive Surgery
• /
• v.27 no.5
• /
• pp.488-491
• /
• 2005

A 3-year-old female, 5kg, Shih-tzu developed an acute onset of depression, disorientation, hypersalivation, nystagmus after falling down 2 meter height place. In plain skull radiography, there was fracture line in the frontal and parietal bones and next day magnetic resonance imaging examination was performed. Magnetic resonance imaging of the brain was performed with 3.0 Tesla unit. Under general anesthesia, the dog was placed in prone with its head positioned in a birdcage coil. Transverse, sagittal and coronal fast spin echo images of the brain were obtained with the following pulse sequences: T1 weighted images (TR = 560 ms and TE = 18.6 ms) and T2 weighted images (TR = 3500 ms and TE = 80 ms). Magnetic resonance imaging showed epidural hematoma in the left frontal area resulting in compression of the adjacent brain parenchyma. Left lateral ventricle was compressed secondarily and the longitudinal fissure shifted to the right, representing mass effect. The lesion was iso-to slightly hyperintense on T1 weighted image and iso-slightly hypointense signal on T2 weighted image. At necropsy, there was a skull fracture and epidural hematoma in the left frontal area. Magnetic resonance imaging of epidural hematoma is reviewed.

• Korean Journal of Veterinary Research
• /
• v.42 no.1
• /
• pp.115-121
• /
• 2002

The present study was conducted to investigate the pattern of fracture of 50 Thoroughbred foals in Jeju from January 1997 to August 2001. A total 50 Thoroughbred foals were investigated to figure out the relationship between breeding condition and fracture. The fracture was diagnosed by physical and radiological examinations after lameness test. Most sites of fracture were limb, skull and vertebrae. Age analyzed that the most popular is the 1-year-old foals. Most places of the occurrence of the fracture were pasture, paddock, track and stable. Main cause of the fracture were play, training and foal's dam. These results suggest that there were the 1-year-old foals have a lot of fracture during play at pasture in winter.

• Journal of Korean Neurosurgical Society
• /
• v.66 no.1
• /
• pp.53-62
• /
• 2023

Objective : Deep learning is a machine learning approach based on artificial neural network training, and object detection algorithm using deep learning is used as the most powerful tool in image analysis. We analyzed and evaluated the diagnostic performance of a deep learning algorithm to identify skull fractures in plain radiographic images and investigated its clinical applicability. Methods : A total of 2026 plain radiographic images of the skull (fracture, 991; normal, 1035) were obtained from 741 patients. The RetinaNet architecture was used as a deep learning model. Precision, recall, and average precision were measured to evaluate the deep learning algorithm's diagnostic performance. Results : In ResNet-152, the average precision for intersection over union (IOU) 0.1, 0.3, and 0.5, were 0.7240, 0.6698, and 0.3687, respectively. When the intersection over union (IOU) and confidence threshold were 0.1, the precision was 0.7292, and the recall was 0.7650. When the IOU threshold was 0.1, and the confidence threshold was 0.6, the true and false rates were 82.9% and 17.1%, respectively. There were significant differences in the true/false and false-positive/false-negative ratios between the anterior-posterior, towne, and both lateral views (p=0.032 and p=0.003). Objects detected in false positives had vascular grooves and suture lines. In false negatives, the detection performance of the diastatic fractures, fractures crossing the suture line, and fractures around the vascular grooves and orbit was poor. Conclusion : The object detection algorithm applied with deep learning is expected to be a valuable tool in diagnosing skull fractures.