• Korean Journal of Radiology
• /
• v.22 no.11
• /
• pp.1918-1928
• /
• 2021

Objective: With the recent development of various MRI-conditional cardiac implantable electronic devices (CIEDs), the accurate identification and characterization of CIEDs have become critical when performing MRI in patients with CIEDs. We aimed to develop and evaluate a deep learning-based algorithm (DLA) that performs the detection and characterization of parameters, including MRI safety, of CIEDs on chest radiograph (CR) in a single step and compare its performance with other related algorithms that were recently developed. Materials and Methods: We developed a DLA (X-ray CIED identification [XCID]) using 9912 CRs of 958 patients with 968 CIEDs comprising 26 model groups from 4 manufacturers obtained between 2014 and 2019 from one hospital. The performance of XCID was tested with an external dataset consisting of 2122 CRs obtained from a different hospital and compared with the performance of two other related algorithms recently reported, including PacemakerID (PID) and Pacemaker identification with neural networks (PPMnn). Results: The overall accuracies of XCID for the manufacturer classification, model group identification, and MRI safety characterization using the internal test dataset were 99.7% (992/995), 97.2% (967/995), and 98.9% (984/995), respectively. These were 95.8% (2033/2122), 85.4% (1813/2122), and 92.2% (1956/2122), respectively, with the external test dataset. In the comparative study, the accuracy for the manufacturer classification was 95.0% (152/160) for XCID and 91.3% for PPMnn (146/160), which was significantly higher than that for PID (80.0%,128/160; p < 0.001 for both). XCID demonstrated a higher accuracy (88.1%; 141/160) than PPMnn (80.0%; 128/160) in identifying model groups (p < 0.001). Conclusion: The remarkable and consistent performance of XCID suggests its applicability for detection, manufacturer and model identification, as well as MRI safety characterization of CIED on CRs. Further studies are warranted to guarantee the safe use of XCID in clinical practice.

• Journal of the Korean Society of Radiology
• /
• v.83 no.6
• /
• pp.1298-1311
• /
• 2022

Purpose To develop and validate a deep learning-based screening tool for the early diagnosis of scoliosis using chest radiographs with a semi-supervised generative adversarial network (GAN). Materials and Methods Using a semi-supervised learning framework with a GAN, a screening tool for diagnosing scoliosis was developed and validated through the chest PA radiographs of patients at two different tertiary hospitals. Our proposed method used training GAN with mild to severe scoliosis only in a semi-supervised manner, as an upstream task to learn scoliosis representations and a downstream task to perform simple classification for differentiating between normal and scoliosis states sensitively. Results The area under the receiver operating characteristic curve, negative predictive value (NPV), positive predictive value, sensitivity, and specificity were 0.856, 0.950, 0.579, 0.985, and 0.285, respectively. Conclusion Our deep learning-based artificial intelligence software in a semi-supervised manner achieved excellent performance in diagnosing scoliosis using the chest PA radiographs of young individuals; thus, it could be used as a screening tool with high NPV and sensitivity and reduce the burden on radiologists for diagnosing scoliosis through health screening chest radiographs.

• Journal of the Korean Society of Radiology
• /
• v.83 no.5
• /
• pp.1081-1089
• /
• 2022

Purpose To compare the computerized Greulich-Pyle based bone age with elbow bone age. Materials and Methods A total of 2126 patients (1525 girls; 601 boys) whose elbow bone age was within the evaluable range by the Sauvegrain method, and who simultaneously underwent hand radiography, were enrolled in the study. The 1st-bone age and VUNO score of the hand were evaluated using VUNOMed-BoneAge software. The correlation between the hand and elbow bone age was analyzed according to the child's gender and the probability of 1st-bone age. Results The correlation between VUNO score and elbow bone age (r = 0.898) was higher than the correlation between 1st-bone age and elbow bone age (r = 0.879). Moreover, the VUNO score showed a better correlation with the elbow bone age in patients with a 1st-bone age probability of less than 70%, or in girls. Elbow bone age was more advanced compared to hand bone age, and this difference increased until the middle of puberty and gradually decreased in the latter half. Conclusion The computerized Greulich-Pyle based hand bone age showed a significant correlation with the elbow bone age at puberty. However, since the elbow bone age tends to advance faster than the hand bone age, caution is required while judging the bone age during puberty.

• Journal of the Korean Society of Radiology
• /
• v.83 no.6
• /
• pp.1286-1297
• /
• 2022

Purpose To assess the usefulness of various metal artifact reduction (MAR) methods in patients with hip prostheses. Materials and Methods This retrospective study included 47 consecutive patients who underwent hip arthroplasty and dual-energy CT. Conventional polyenergetic image (CI), orthopedic-MAR (OMAR), and virtual monoenergetic image (VMI, 50-200 keV) were tested for MAR. Quantitative analysis was performed in seven regions around the prostheses. Qualitative assessments included evaluation of the degree of artifacts and the presence of secondary artifacts. Results The lowest amount of image noise was observed in the O-MAR, followed by the VMI. O-MAR also showed the lowest artifact index, followed by high-keV VMI in the range of 120-200 keV (soft tissue) or 200 keV (bone). O-MAR had the highest contrast-to-noise ratio (CNR) in regions with severe hypodense artifacts, while VMI had the highest CNR in other regions, including the periprosthetic bone. On assessment of the CI of pelvic soft tissues, VMI showed a higher structural similarity than O-MAR. Upon qualitative analysis, metal artifacts were significantly reduced in O-MAR, followed by that in VMI, while secondary artifacts were the most frequently found in the O-MAR (p < 0.001). Conclusion O-MAR is the best technique for severe MAR, but it can generate secondary artifacts. VMI at high keV can be advantageous for evaluating periprosthetic bone.

• Journal of the Korean Society of Radiology
• /
• v.81 no.3
• /
• pp.654-664
• /
• 2020

Purpose To determine the frequency of ossification of the transverse ligament of the atlas (OTLA) and to investigate the associated findings on cervical spine CT and plain radiography. Materials and Methods We reviewed 5201 CT scans of the cervical spine of 3975 consecutive patients over an 11-year period for the presence of OTLA and compared them with those of age- and sex-matched controls. The frequency and associated findings of OTLA were investigated and statistically correlated. Results The overall frequency of OTLA was 1.1% (45 of 3975 patients) and increased with age (p < 0.005). The frequency of OTLA in patients over 80 years was 12%. The space available for the spinal cord (SAC) was smaller in patients with OTLA (p < 0.005). Mineralization of the complex of the anterior atlantooccipital membrane and Barkow ligament, ossification of the ligamentum flavum, and kyphosis of the cervical spine positively correlated to the presence of OTLA (p < 0.005). Conclusion OTLA was associated with age, SAC narrowing, cervical kyphosis, and ossification of other cervical ligaments and may be associated with degenerative spondylosis, systemic hyperostotic status, or mechanical stress or instability.

• Journal of the Korean Society of Radiology
• /
• v.82 no.4
• /
• pp.994-999
• /
• 2021

Late recurrence over 10 years after surgery and endobronchial metastasis are some of the specific biological behaviors of renal cell carcinoma (RCC). The current report describes a case of solitary endobronchial metastasis at a subsegmental bronchus that developed 20 years after curative nephrectomy for RCC. A 71-year-old male was admitted to our hospital for pneumonia. Chest radiography showed multifocal ill-defined nodular opacities in the right lower lung zone, suggesting pneumonia. Subsequent chest CT confirmed pneumonic infiltration in the right lung. However, a 4.3-cm, well-defined, elongated mass with a branching pattern was also identified in the right lower lobe, and a right nephrectomy scar was detected on the covered upper abdomen. The patient had undergone right nephrectomy 20 years ago due to clear cell RCC. After right lower lobectomy, the postoperative pathological diagnosis was endobronchial metastatic clear cell RCC. Endobronchial metastasis should be considered in a patient with a history of RCC who presents with a suspected endobronchial tumor, even decades after curative surgery.

• Korean Journal of Radiology
• /
• v.24 no.3
• /
• pp.259-270
• /
• 2023

Objective: It is unknown whether artificial intelligence-based computer-aided detection (AI-CAD) can enhance the accuracy of chest radiograph (CR) interpretation in real-world clinical practice. We aimed to compare the accuracy of CR interpretation assisted by AI-CAD to that of conventional interpretation in patients who presented to the emergency department (ED) with acute respiratory symptoms using a pragmatic randomized controlled trial. Materials and Methods: Patients who underwent CRs for acute respiratory symptoms at the ED of a tertiary referral institution were randomly assigned to intervention group (with assistance from an AI-CAD for CR interpretation) or control group (without AI assistance). Using a commercial AI-CAD system (Lunit INSIGHT CXR, version 2.0.2.0; Lunit Inc.). Other clinical practices were consistent with standard procedures. Sensitivity and false-positive rates of CR interpretation by duty trainee radiologists for identifying acute thoracic diseases were the primary and secondary outcomes, respectively. The reference standards for acute thoracic disease were established based on a review of the patient's medical record at least 30 days after the ED visit. Results: We randomly assigned 3576 participants to either the intervention group (1761 participants; mean age ± standard deviation, 65 ± 17 years; 978 males; acute thoracic disease in 472 participants) or the control group (1815 participants; 64 ± 17 years; 988 males; acute thoracic disease in 491 participants). The sensitivity (67.2% [317/472] in the intervention group vs. 66.0% [324/491] in the control group; odds ratio, 1.02 [95% confidence interval, 0.70-1.49]; P = 0.917) and false-positive rate (19.3% [249/1289] vs. 18.5% [245/1324]; odds ratio, 1.00 [95% confidence interval, 0.79-1.26]; P = 0.985) of CR interpretation by duty radiologists were not associated with the use of AI-CAD. Conclusion: AI-CAD did not improve the sensitivity and false-positive rate of CR interpretation for diagnosing acute thoracic disease in patients with acute respiratory symptoms who presented to the ED.

• Korean Journal of Radiology
• /
• v.23 no.10
• /
• pp.1009-1018
• /
• 2022

Objective: This study aimed to investigate the feasibility of using artificial intelligence (AI) to identify normal chest radiography (CXR) from the worklist of radiologists in a health-screening environment. Materials and Methods: This retrospective simulation study was conducted using the CXRs of 5887 adults (mean age ± standard deviation, 55.4 ± 11.8 years; male, 4329) from three health screening centers in South Korea using a commercial AI (Lunit INSIGHT CXR3, version 3.5.8.8). Three board-certified thoracic radiologists reviewed CXR images for referable thoracic abnormalities and grouped the images into those with visible referable abnormalities (identified as abnormal by at least one reader) and those with clearly visible referable abnormalities (identified as abnormal by at least two readers). With AI-based simulated exclusion of normal CXR images, the percentages of normal images sorted and abnormal images erroneously removed were analyzed. Additionally, in a random subsample of 480 patients, the ability to identify visible referable abnormalities was compared among AI-unassisted reading (i.e., all images read by human readers without AI), AI-assisted reading (i.e., all images read by human readers with AI assistance as concurrent readers), and reading with AI triage (i.e., human reading of only those rendered abnormal by AI). Results: Of 5887 CXR images, 405 (6.9%) and 227 (3.9%) contained visible and clearly visible abnormalities, respectively. With AI-based triage, 42.9% (2354/5482) of normal CXR images were removed at the cost of erroneous removal of 3.5% (14/405) and 1.8% (4/227) of CXR images with visible and clearly visible abnormalities, respectively. In the diagnostic performance study, AI triage removed 41.6% (188/452) of normal images from the worklist without missing visible abnormalities and increased the specificity for some readers without decreasing sensitivity. Conclusion: This study suggests the feasibility of sorting and removing normal CXRs using AI with a tailored cut-off to increase efficiency and reduce the workload of radiologists.

• 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.

• Korean Journal of Radiology
• /
• v.2 no.3
• /
• pp.138-144
• /
• 2001

Objective: Bronchogenic carcinoma can mimic or be masked by pulmonary tuberculosis (TB), and the aim of this study was to describe the radiologic findings and clinical significance of bronchogenic carcinoma and pulmonary TB which coexist in the same lobe. Materials and Methods: The findings of 51 patients (48 males and three females, aged 48-79 years) in whom pulmonary TB and bronchogenic carcinoma coexisted in the same lobe were analyzed. The morphologic characteristics of a tumor, such as its diameter and margin, the presence of calcification or cavitation, and mediastinal lymphadenopathy, as seen at CT, were retrospectively assessed, and the clinical stage of the lung cancer was also determined. Using the serial chest radiographs available for 21 patients, the possible causes of delay in the diagnosis of lung cancer were analyzed. Results: Lung cancers with coexisting pulmonary TB were located predominantly in the upper lobes (82.4%). The mean diameter of the mass was 5.3 cm, and most tumors (n=42, 82.4%) had a lobulated border. Calcification within the tumor was seen in 20 patients (39.2%), and cavitation in five (9.8%). Forty-two (82.4%) had mediastinal lymphadenopathy, and more than half the tumors (60.8%) were at an advanced stage [IIIB (n=11) or IV (n=20)]. The average delay in diagnosing lung cancer was 11.7 (range, 1-24) months, and the causes of this were failure to observe new nodules masked by coexisting stable TB lesions (n=8), misinterpretation of new lesions as aggravation of TB (n=5), misinterpretation of lung cancer as tuberculoma at initial radiography (n=4), masking of the nodule by an active TB lesion (n=3), and subtleness of the lesion (n=1). Conclusion: Most cancers concurrent with TB are large, lobulated masses with mediastinal lymphadenopathy, indicating that the morphologic characteristics of lung cancer with coexisting pulmonary TB are similar to those of lung cancer without TB. The diagnosis of lung cancer is delayed mainly because of masking by a tuberculous lesion, and this suggests that in patients in whom a predominant or growing nodule is present and who show little improvement of symptoms despite antituberculous or other medical therapy, coexisting cancer should be suspected.