Purpose: This study explored the feasibility of utilizing a deep learning artificial intelligence (AI) model to detect ileocolic intussusception on grayscale ultrasound images. Methods: This retrospective observational study incorporated ultrasound images of children who underwent emergency ultrasonography for suspected ileocolic intussusception. After excluding video clips, Doppler images, and annotated images, 40,765 images from two tertiary hospitals were included (positive-to-negative ratio: hospital A, 2,775:35,373; hospital B, 140:2,477). Images from hospital A were split into a training set, a tuning set, and an internal test set (ITS) at a ratio of 7:1.5:1.5. Images from hospital B comprised an external test set (ETS). For each image indicating intussusception, two radiologists provided a bounding box as the ground-truth label. If intussusception was suspected in the input image, the model generated a bounding box with a confidence score (0-1) at the estimated lesion location. Average precision (AP) was used to evaluate overall model performance. The performance of practical thresholds for the modelgenerated confidence score, as determined from the ITS, was verified using the ETS. Results: The AP values for the ITS and ETS were 0.952 and 0.936, respectively. Two confidence thresholds, CTopt and CTprecision, were set at 0.557 and 0.790, respectively. For the ETS, the perimage precision and recall were 95.7% and 80.0% with CTopt, and 98.4% and 44.3% with CTprecision. For per-patient diagnosis, the sensitivity and specificity were 100.0% and 97.1% with CTopt, and 100.0% and 99.0% with CTprecision. The average number of false positives per patient was 0.04 with CTopt and 0.01 for CTprecision. Conclusion The feasibility of using an AI model to diagnose ileocolic intussusception on ultrasonography was demonstrated. However, further study involving bias-free data is warranted for robust clinical validation.

Purpose: This study explored the feasibility of utilizing a deep learning artificial intelligence (AI) model to detect ileocolic intussusception on grayscale ultrasound images. Methods: This retrospective observational study incorporated ultrasound images of children who underwent emergency ultrasonography for suspected ileocolic intussusception. After excluding video clips, Doppler images, and annotated images, 40,765 images from two tertiary hospitals were included (positive-to-negative ratio: hospital A, 2,775:35,373; hospital B, 140:2,477). Images from hospital A were split into a training set, a tuning set, and an internal test set (ITS) at a ratio of 7:1.5:1.5. Images from hospital B comprised an external test set (ETS). For each image indicating intussusception, two radiologists provided a bounding box as the ground-truth label. If intussusception was suspected in the input image, the model generated a bounding box with a confidence score (0-1) at the estimated lesion location. Average precision (AP) was used to evaluate overall model performance. The performance of practical thresholds for the modelgenerated confidence score, as determined from the ITS, was verified using the ETS. Results: The AP values for the ITS and ETS were 0.952 and 0.936, respectively. Two confidence thresholds, CTopt and CTprecision, were set at 0.557 and 0.790, respectively. For the ETS, the perimage precision and recall were 95.7% and 80.0% with CTopt, and 98.4% and 44.3% with CTprecision. For per-patient diagnosis, the sensitivity and specificity were 100.0% and 97.1% with CTopt, and 100.0% and 99.0% with CTprecision. The average number of false positives per patient was 0.04 with CTopt and 0.01 for CTprecision. Conclusion The feasibility of using an AI model to diagnose ileocolic intussusception on ultrasonography was demonstrated. However, further study involving bias-free data is warranted for robust clinical validation.

Purpose: This study explored the feasibility of utilizing a deep learning artificial intelligence (AI) model to detect ileocolic intussusception on grayscale ultrasound images. Methods: This retrospective observational study incorporated ultrasound images of children who underwent emergency ultrasonography for suspected ileocolic intussusception. After excluding video clips, Doppler images, and annotated images, 40,765 images from two tertiary hospitals were included (positive-to-negative ratio: hospital A, 2,775:35,373; hospital B, 140:2,477). Images from hospital A were split into a training set, a tuning set, and an internal test set (ITS) at a ratio of 7:1.5:1.5. Images from hospital B comprised an external test set (ETS). For each image indicating intussusception, two radiologists provided a bounding box as the ground-truth label. If intussusception was suspected in the input image, the model generated a bounding box with a confidence score (0-1) at the estimated lesion location. Average precision (AP) was used to evaluate overall model performance. The performance of practical thresholds for the modelgenerated confidence score, as determined from the ITS, was verified using the ETS. Results: The AP values for the ITS and ETS were 0.952 and 0.936, respectively. Two confidence thresholds, CTopt and CTprecision, were set at 0.557 and 0.790, respectively. For the ETS, the perimage precision and recall were 95.7% and 80.0% with CTopt, and 98.4% and 44.3% with CTprecision. For per-patient diagnosis, the sensitivity and specificity were 100.0% and 97.1% with CTopt, and 100.0% and 99.0% with CTprecision. The average number of false positives per patient was 0.04 with CTopt and 0.01 for CTprecision. Conclusion The feasibility of using an AI model to diagnose ileocolic intussusception on ultrasonography was demonstrated. However, further study involving bias-free data is warranted for robust clinical validation.

Purpose: This study explored the feasibility of utilizing a deep learning artificial intelligence (AI) model to detect ileocolic intussusception on grayscale ultrasound images. Methods: This retrospective observational study incorporated ultrasound images of children who underwent emergency ultrasonography for suspected ileocolic intussusception. After excluding video clips, Doppler images, and annotated images, 40,765 images from two tertiary hospitals were included (positive-to-negative ratio: hospital A, 2,775:35,373; hospital B, 140:2,477). Images from hospital A were split into a training set, a tuning set, and an internal test set (ITS) at a ratio of 7:1.5:1.5. Images from hospital B comprised an external test set (ETS). For each image indicating intussusception, two radiologists provided a bounding box as the ground-truth label. If intussusception was suspected in the input image, the model generated a bounding box with a confidence score (0-1) at the estimated lesion location. Average precision (AP) was used to evaluate overall model performance. The performance of practical thresholds for the modelgenerated confidence score, as determined from the ITS, was verified using the ETS. Results: The AP values for the ITS and ETS were 0.952 and 0.936, respectively. Two confidence thresholds, CTopt and CTprecision, were set at 0.557 and 0.790, respectively. For the ETS, the perimage precision and recall were 95.7% and 80.0% with CTopt, and 98.4% and 44.3% with CTprecision. For per-patient diagnosis, the sensitivity and specificity were 100.0% and 97.1% with CTopt, and 100.0% and 99.0% with CTprecision. The average number of false positives per patient was 0.04 with CTopt and 0.01 for CTprecision. Conclusion The feasibility of using an AI model to diagnose ileocolic intussusception on ultrasonography was demonstrated. However, further study involving bias-free data is warranted for robust clinical validation.

Purpose: This study explored the feasibility of utilizing a deep learning artificial intelligence (AI) model to detect ileocolic intussusception on grayscale ultrasound images. Methods: This retrospective observational study incorporated ultrasound images of children who underwent emergency ultrasonography for suspected ileocolic intussusception. After excluding video clips, Doppler images, and annotated images, 40,765 images from two tertiary hospitals were included (positive-to-negative ratio: hospital A, 2,775:35,373; hospital B, 140:2,477). Images from hospital A were split into a training set, a tuning set, and an internal test set (ITS) at a ratio of 7:1.5:1.5. Images from hospital B comprised an external test set (ETS). For each image indicating intussusception, two radiologists provided a bounding box as the ground-truth label. If intussusception was suspected in the input image, the model generated a bounding box with a confidence score (0-1) at the estimated lesion location. Average precision (AP) was used to evaluate overall model performance. The performance of practical thresholds for the modelgenerated confidence score, as determined from the ITS, was verified using the ETS. Results: The AP values for the ITS and ETS were 0.952 and 0.936, respectively. Two confidence thresholds, CTopt and CTprecision, were set at 0.557 and 0.790, respectively. For the ETS, the perimage precision and recall were 95.7% and 80.0% with CTopt, and 98.4% and 44.3% with CTprecision. For per-patient diagnosis, the sensitivity and specificity were 100.0% and 97.1% with CTopt, and 100.0% and 99.0% with CTprecision. The average number of false positives per patient was 0.04 with CTopt and 0.01 for CTprecision. Conclusion The feasibility of using an AI model to diagnose ileocolic intussusception on ultrasonography was demonstrated. However, further study involving bias-free data is warranted for robust clinical validation.

Objective: To evaluate the diagnostic performance of a deep learning algorithm for the automated detection of developmental dysplasia of the hip (DDH) on anteroposterior (AP) radiographs. Materials and Methods: Of 2601 hip AP radiographs, 5076 cropped unilateral hip joint images were used to construct a dataset that was further divided into training (80%), validation (10%), or test sets (10%). Three radiologists were asked to label the hip images as normal or DDH. To investigate the diagnostic performance of the deep learning algorithm, we calculated the receiver operating characteristics (ROC), precision-recall curve (PRC) plots, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) and compared them with the performance of radiologists with different levels of experience. Results: The area under the ROC plot generated by the deep learning algorithm and radiologists was 0.988 and 0.988–0.919, respectively. The area under the PRC plot generated by the deep learning algorithm and radiologists was 0.973 and 0.618– 0.958, respectively. The sensitivity, specificity, PPV, and NPV of the proposed deep learning algorithm were 98.0, 98.1, 84.5, and 99.8%, respectively. There was no significant difference in the diagnosis of DDH by the algorithm and the radiologist with experience in pediatric radiology (p = 0.180). However, the proposed model showed higher sensitivity, specificity, and PPV, compared to the radiologist without experience in pediatric radiology (p < 0.001). Conclusion The proposed deep learning algorithm provided an accurate diagnosis of DDH on hip radiographs, which was comparable to the diagnosis by an experienced radiologist.

Objective: To evaluate the diagnostic performance of a deep learning algorithm for the automated detection of developmental dysplasia of the hip (DDH) on anteroposterior (AP) radiographs. Materials and Methods: Of 2601 hip AP radiographs, 5076 cropped unilateral hip joint images were used to construct a dataset that was further divided into training (80%), validation (10%), or test sets (10%). Three radiologists were asked to label the hip images as normal or DDH. To investigate the diagnostic performance of the deep learning algorithm, we calculated the receiver operating characteristics (ROC), precision-recall curve (PRC) plots, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) and compared them with the performance of radiologists with different levels of experience. Results: The area under the ROC plot generated by the deep learning algorithm and radiologists was 0.988 and 0.988–0.919, respectively. The area under the PRC plot generated by the deep learning algorithm and radiologists was 0.973 and 0.618– 0.958, respectively. The sensitivity, specificity, PPV, and NPV of the proposed deep learning algorithm were 98.0, 98.1, 84.5, and 99.8%, respectively. There was no significant difference in the diagnosis of DDH by the algorithm and the radiologist with experience in pediatric radiology (p = 0.180). However, the proposed model showed higher sensitivity, specificity, and PPV, compared to the radiologist without experience in pediatric radiology (p < 0.001). Conclusion The proposed deep learning algorithm provided an accurate diagnosis of DDH on hip radiographs, which was comparable to the diagnosis by an experienced radiologist.

On page 370, Article Title has been incorrectly marked Etiologies and Predictors of ST-Segment Elevation Myocardial Infarction. The correct title is Etiologies and Predictors of False-Positive Diagnosis of ST-Segment Elevation Myocardial Infarction.

Endomyocardial biopsy (EMB) is one of the reliable methods for the diagnosis of various cardiac diseases. However, EMB can cause various complications. The purpose of this study is to evaluate the complication of transfemoral EMB with both fluoroscopic and two-dimensional (2-D) echocardiographic guidance. A total of 228 patients (148 men; 46.0+/-14.6 yr-old) who underwent EMB at Kyungpook National University Hospital from January 2002 to June 2012 were included. EMB was performed via the right femoral approach with the guidance of both echocardiography and fluoroscopy. Overall, EMB-related complications occurred in 21 patients (9.2%) including one case (0.4%) with cardiac tamponade requiring emergent pericardiocentesis, four cases (1.8%) with small pericardial effusion without pericardiocentesis, two cases (0.9%) with hemodynamically unstable ventricular tachycardia (VT), one case (0.4%) with nonsustained VT, one case (0.4%) with tricuspid regurgitation, twelve cases (5.3%) with right bundle branch block. There was no occurrence of either EMB-related death or cardiac surgery. Left ventricular ejection fraction was significantly lower (32.0+/-18.7% vs 42.0+/-19.1%, P=0.023) and left ventricular end-diastolic dimension was larger (60.0+/-10.0 mm vs 54.2+/-10.2 mm, P=0.013) in patients with EMB related complications than in those without. It is concluded that transfemoral EMB with fluoroscopic and 2-D echocardiographic guidance is a safe procedure with low complication rate.
Adult ; Biopsy/*adverse effects ; Cardiac Tamponade/etiology ; Echocardiography/*adverse effects ; Endocardium/*ultrasonography ; Female ; Fluoroscopy/*adverse effects ; Heart Diseases/*pathology ; Heart Ventricles/metabolism ; Humans ; Male ; Middle Aged ; Pericardial Effusion/etiology ; Tachycardia, Ventricular/etiology ; Ventricular Function

BACKGROUND AND OBJECTIVES: Rapid diagnosis of ST-segment elevation myocardial infarction (STEMI) is essential for the appropriate management of patients. We investigated the prevalence, etiologies and predictors of false-positive diagnosis of STEMI and subsequent inappropriate catheterization laboratory activation in patients with presumptive diagnosis of STEMI. SUBJECTS AND METHODS: Four hundred fifty-five consecutive patients (62+/-13 years, 345 males) with presumptive diagnosis of STEMI between August 2008 and November 2010 were included. RESULTS: A false-positive diagnosis of STEMI was made in 34 patients (7.5%) with no indication of coronary artery lesion. Common causes for the false-positive diagnosis were coronary spasm in 10 patients, left ventricular hypertrophy in 5 patients, myocarditis in 4 patients, early repolarization in 3 patients, and previous myocardial infarction and stress-induced cardiomyopathy in 2 patients each. In multivariate logistic regression analysis, symptom-to-door time >12 hours {odds ratio (OR) 4.995, 95% confidence interval (CI) 1.384-18.030, p=0.014}, presenting symptom other than chest pain (OR 7.709, 95% CI 1.255-39.922, p=0.027), absence of Q wave (OR 9.082, CI 2.631-31.351, p<0.001) and absence of reciprocal changes on electrocardiography (ECG) (OR 17.987, CI 5.295-61.106, p<0.001) were independent predictors of false-positive diagnosis of STEMI. CONCLUSION: In patients whom STEMI was planned for primary coronary intervention, the false-positive diagnosis of STEMI was not rare. Correct interpretation of ECGs and consideration of ST-segment elevation in conditions other than STEMI may reduce inappropriate catheterization laboratory activation.
Cardiomyopathies ; Catheterization ; Catheters ; Chest Pain ; Coronary Vessels ; Electrocardiography ; False Positive Reactions ; Humans ; Hypertrophy, Left Ventricular ; Logistic Models ; Myocardial Infarction ; Myocarditis ; Prevalence ; Spasm