Measuring synovial thickness on knee MRI in pediatric patients with arthritis: is contrast necessary?

BACKGROUND: The use of contrast-enhanced imaging has long been standard for magnetic resonance imaging (MRI) assessments of synovitis in juvenile idiopathic arthritis (JIA). However, advancements in MRI technology have allowed for reliable identification of synovium without contrast. OBJECTIVE: To assess the equivalence of unenhanced MRI with contrast-enhanced MRI in evaluating synovial thickness. MATERIALS AND METHODS: This is an institutional review board approved, retrospective study performed in a tertiary children's hospital. Pediatric JIA patients under 21 years old were included who underwent knee MRI scans (1.5 T or 3 T) without and with contrast between January 2012 and January 2022. Two radiologists independently measured synovial thickness at 6 knee sites on contrast-enhanced and unenhanced sequences. Numerical measurements and ordinal scores based on juvenile idiopathic arthritis magnetic resonance imaging scoring (JAMRIS) system were recorded, and tests of equivalence were conducted, as well as between-reader and within-reader reliability by concordance correlation coefficient (CCC). All tests were considered significant at the 5% level. RESULTS: A total of 38 studies from 35 patients (25 females, median age 14 years; interquartile range 7 to 15.7) were included. Equivalence was demonstrated at each of the 6 sites for both continuous measurements (P-values < 0.05) and ordinal scores (P-values < 0.05) based on the average over readers. Within-reader reliability was moderate to high (CCC 0.50-0.89), except for the cruciate ligaments site. Averaged over the 6 sites, reliability between readers was low for unenhanced (CCC 0.47, with 95% CI: [0.41, 0.53]) and moderate for contrast-enhanced (CCC 0.64, with 95% CI: [0.59, 0.69]) sequences. CONCLUSION: Unenhanced knee MRI is equivalent to contrast-enhanced MRI in assessment of synovial thickness using conventional MRI sequences. Contrast material helped improve inter-reader reliability.

Lung and Pleural Findings of Children with Pulmonary Vein Stenosis with and without Aspiration: MDCT Evaluation.

Purpose: To retrospectively compare the lung and pleural findings in children with pulmonary vein stenosis (PVS) with and without aspiration on multidetector computed tomography (MDCT). Materials and Methods: All consecutive children (≤18 years old) with PVS who underwent thoracic MDCT studies from August 2004 to December 2021 were categorized into two groups: children with PVS with aspiration (Group 1) and children with PVS without aspiration (Group 2). Two independent pediatric radiologists retrospectively evaluated thoracic MDCT studies for the presence of lung and pleural abnormalities as follows: (1) in the lung (ground-glass opacity (GGO), consolidation, nodule, mass, cyst(s), interlobular septal thickening, and fibrosis) and (2) in the pleura (thickening, effusion, and pneumothorax). Interobserver agreement between the two reviewers was evaluated by the proportion of agreement and the Kappa statistic. Results: The final study population consisted of 64 pediatric patients (36 males (56.3%) and 43 females (43.7%); mean age, 1.7 years; range, 1 day−17 years). Among these 64 patients, 19 patients (29.7%) comprised Group 1 and the remaining 45 patients (70.3%) comprised Group 2. In Group 1 (children with PVS with aspiration), the detected lung and pleural MDCT abnormalities were: GGO (17/19; 89.5%), pleural thickening (17/19; 89.5%), consolidation (16/19; 84.5%), and septal thickening (16/19; 84.5%). The lung and pleural MDCT abnormalities observed in Group 2 (children with PVS without aspiration) were: GGO (37/45; 82.2%), pleural thickening (37/45; 82.2%), septal thickening (36/45; 80%), consolidation (3/45; 6.7%), pleural effusion (1/45; 2.2%), pneumothorax (1/45; 2.2%), and cyst(s) (1/45; 2.2%). Consolidation was significantly more common in pediatric patients with both PVS and aspiration (Group 1) (p < 0.001). There was high interobserver agreement between the two independent reviewers for detecting lung and pleural abnormalities on thoracic MDCT studies (Kappa = 0.98; CI = 0.958, 0.992). Conclusion: Aspiration is common in pediatric patients with PVS who undergo MDCT and was present in nearly 30% of all children with PVS during our study period. Consolidation is not a typical radiologic finding of PVS in children without clinical evidence of aspiration. When consolidation is present on thoracic MDCT studies in pediatric patients with PVS, the additional diagnosis of concomitant aspiration should be considered.

Advanced imaging of disease unrelated to the coronavirus disease 2019 (COVID-19) during the pandemic: effect of patient demographics in a pediatric emergency department.

BACKGROUND: Coronavirus disease 2019 (COVID-19) disproportionately affected children from underrepresented minorities and marginalized populations, but little is understood regarding the pandemic's effect on non-COVID-19-related illnesses. OBJECTIVE: To examine the effect of the COVID-19 pandemic and related stay-at-home orders on pediatric emergency department (ED) imaging of non-COVID-19-related diseases across patient demographic groups. MATERIALS AND METHODS: We retrospectively reviewed radiology reports from advanced imaging (US, CT, MRI and fluoroscopy) on children in the ED during the month of April for the years 2017, 2018, 2019 and 2020, excluding imaging for respiratory illness and trauma. We used imaging results and the electronic medical record to identify children with positive diagnoses on advanced imaging, and whether these children were admitted to the hospital. Demographic variables included age, gender, race/ethnicity and insurance type. We used multivariable Poisson regression models to report rate ratio (RR) and binomial logistic regression models to report odds ratio (OR) with 95% confidence interval (CI). RESULTS: We included 1,418 ED encounters for analysis. Compared to pre-2020, fewer children underwent ED imaging in April 2020 (RR 0.63, 95% CI 0.52, 0.76). The odds of positive imaging results increased (OR 2.18, 95% CI 1.59, 3.00) overall, and for all racial/ethnic groups except Hispanic patients (OR 0.83, 95% CI 0.34, 2.03). No differences occurred in admission rates for positive imaging results in 2020 compared to pre-2020. CONCLUSION: In April 2020 compared to pre-2020, there were decreased imaging and increased positivity rates for imaging for non-respiratory and non-trauma ED visits. COVID-19 stay-at-home advisories might have resulted in triaging for urgent health care by families or referring clinicians during this month of the pandemic.

Pleuropulmonary MDCT Findings: Comparison between Children with Pulmonary Vein Stenosis and Prematurity-Related Lung Disease.

Purpose: To retrospectively compare the pleuropulmonary MDCT findings in children with pulmonary vein stenosis (PVS) and prematurity-related lung disease (PLD). Materials and Methods: All consecutive infants and young children (≤18 years old) who underwent thoracic MDCT studies from July 2004 to November 2021 were categorized into two groups­children with PVS (Group 1) and children with PLD without PVS (Group 2). Two pediatric radiologists independently evaluated thoracic MDCT studies for the presence of pleuropulmonary abnormalities as follows­(1) in the lung (ground-glass opacity (GGO), triangular/linear plaque-like opacity (TLO), consolidation, nodule, mass, cyst(s), interlobular septal thickening, and fibrosis); (2) in the airway (bronchial wall thickening and bronchiectasis); and (3) in the pleura (thickening, effusion, and pneumothorax). Interobserver agreement between the two reviewers was evaluated with the Kappa statistic. Results: There were a total of 103 pediatric patients (60 males (58.3%) and 43 females (41.7%); mean age, 1.7 years; range, 2 days−7 years). Among these 103 patients, 49 patients (47.6%) comprised Group 1 and the remaining 54 patients (52.4%) comprised Group 2. In Group 1, the observed pleuropulmonary MDCT abnormalities were­pleural thickening (44/49; 90%), GGO (39/49; 80%), septal thickening (39/49; 80%), consolidation (4/49; 8%), and pleural effusion (1/49; 2%). The pleuropulmonary MDCT abnormalities seen in Group 2 were­GGO (45/54; 83%), TLO (43/54; 80%), bronchial wall thickening (33/54; 61%), bronchiectasis (30/54; 56%), cyst(s) (5/54; 9%), pleural thickening (2/54; 4%), and pleural effusion (2/54; 4%). Septal thickening and pleural thickening were significantly more common in pediatric patients with PVS (Group 1) (p < 0.001). TLO, bronchial wall thickening, and bronchiectasis were significantly more frequent in pediatric patients with PLD without PVS (Group 2) (p < 0.001). There was high interobserver kappa agreement between the two independent reviewers for detecting pleuropulmonary abnormalities on thoracic MDCT angiography studies (k = 0.99). Conclusion: Pleuropulmonary abnormalities seen on thoracic MDCT can be helpful for distinguishing PVS from PLD in children. Specifically, the presence of septal thickening and pleural thickening raises the possibility of PVS, whereas the presence of TLO, bronchial wall thickening and bronchiectasis suggests PLD in the pediatric population.

Sociodemographic and Appointment Factors Affecting Missed Opportunities to Provide Neonatal Ultrasound Imaging.

PURPOSE: The aim of this study was to assess disparities in outpatient imaging missed care opportunities (IMCOs) for neonatal ultrasound by sociodemographic and appointment factors at a large urban pediatric hospital. METHODS: A retrospective review was performed among patients aged 0 to 28 days receiving one or more outpatient appointments for head, hip, renal, or spine ultrasound at the main hospital or satellite sites from 2008 to 2018. An IMCO was defined as a missed ultrasound or cancellation <24 hours in advance. Population-average correlated logistic regression modeling estimated the odds of IMCOs for six sociodemographic (age, sex, race/ethnicity, language, insurance, and region of residence) and seven appointment (type of ultrasound, time, day, season, site, year, and distance to appointment) factors. The primary analysis included unknown values as a separate category, and the secondary analysis used multiple imputation to impute genuine categories from unknown variables. RESULTS: The data set comprised 5,474 patients totaling 6,803 ultrasound appointments. IMCOs accounted for 4.4% of appointments. IMCOs were more likely for Black (odds ratio [OR], 3.31; P < .001) and other-race neonates (OR, 2.66; P < .001) and for patients with public insurance (OR, 1.78; P = .002). IMCOs were more likely for appointments at the main hospital compared with satellites (P < .001), during work hours (P = .021), and on weekends (P < .001). Statistical significance for primary and secondary analyses was quantitatively similar and qualitatively identical. CONCLUSIONS: Marginalized racial groups and those with public insurance had a higher rate of IMCOs in neonatal ultrasound. This likely represents structural inequities faced by these communities, and more research is needed to identify interventions to address these inequities in care delivery for vulnerable neonatal populations.

Emergency pediatric radiology imaging trends for non-COVID-19-related illnesses through different stages of the pandemic.

PURPOSE: To evaluate how the COVID-19 pandemic affected the imaging utilization patterns for non-COVID-19-related illness in a pediatric emergency department (ED). METHODS: We retrospectively reviewed radiology reports for ultrasound, CT, MRI, and fluoroscopy studies performed at a pediatric ED in April from 2017 to 2021, excluding studies for respiratory symptoms and trauma. Radiology reports and medical records were reviewed to determine if patients had a positive radiology diagnosis, the type of diagnosis, and whether it required hospital admission. Results from during the pandemic were compared to predicted rates based on pre-pandemic years. RESULTS: A total of 2198 imaging studies were included. During the COVID-19 pandemic, fewer ED imaging studies were performed compared to predicted. The decrease was greater in April 2020 (RR = 0.56, p < 0.001) than in April 2021 (RR = 0.80, p = 0.038). The odds of positive diagnosis was higher during the pandemic than before, and higher in 2020 (OR 2.53, p < 0.001) than in 2021 (OR 1.38, p = 0.008). The expected numbers of positive diagnoses and hospital admittances remained within the predicted range during the pandemic (p = 0.505-0.873). CONCLUSIONS: Although imaging volumes decreased during the studied months of the pandemic, the number of positive findings was unchanged compared to prior years. No differences were demonstrated in the percentage of patients admitted to the hospital with positive imaging findings. This suggests that, at our institution, the pandemic did not lead to a substantial number of missed diagnoses or severely delay the diagnosis of non-COVID-related conditions. While still lower than expected, imaging volumes increased in April 2021 suggesting a return towards baseline patient behavior as the pandemic conditions improved.

Thoracic CTA in infants and young children: Image quality of dual-source CT (DSCT) with high-pitch spiral scan mode (turbo flash spiral mode) with or without general anesthesia with free-breathing technique.

PURPOSE: To determine whether diagnostic quality thoracic computed tomography angiography (CTA) studies can be obtained without general anesthesia (GA) in infants and young children using dual-source computed tomography (DSCT) with turbo flash spiral mode (TFSM) and free-breathing technique. MATERIALS AND METHODS: All consecutive infants and young children (≤ 6 years old) who underwent thoracic CTA studies from January 2018 to October 2020 for suspected congenital thoracic disorders were categorized into two groups: with GA (Group 1) and without GA (Group 2). All thoracic CTA studies were performed on a DSCT scanner using TFSM and free-breathing technique. Two pediatric thoracic radiologists independently evaluated motion artifact in three lung zones (upper, mid, and lower). Degree of motion artifact was graded 0-3 (0, none; 1, mild; 2, moderate; and 3, severe). Logistic models adjusted for age and gender were used to compare the degree of motion artifact between lung zones. Interobserver agreement between reviewers was evaluated with kappa statistics. RESULTS: There were a total of 73 pediatric patients (43 males (59%) and 30 females (41%); mean age, 1.4 years; range, 0-5.9 years). Among these 73 patients, 42 patients (58%) underwent thoracic CTA studies with GA (Group 1) and the remaining 31 patients (42%) underwent thoracic CTA studies without GA (Group 2). Overall, the degree of motion artifact was higher for Group 2 (without GA). However, only a very small minority (1/31, 3%) of Group 2 (without GA) thoracic CTA studies had severe motion artifact. There was no significant difference between the two groups with respect to the presence of severe motion artifact (odds ratio [OR] = 6, p = .222). When two groups were compared with respect to the presence of motion artifact for individual lung zones, motion artifact was significantly higher in the upper lung zone for Group 2 (without GA) (OR = 20, p = .043). Interobserver agreement for motion artifact was high, the average Kappa being 0.81 for Group 1 and 0.95 for Group 2. CONCLUSION: Although the degree of motion artifact was higher in the group without GA, only a small minority (3%) of thoracic CTA studies performed without GA had severe motion artifact, rendering the study nondiagnostic. Therefore, the results of this study support the use of thoracic CTA without GA using DSCT with TFSM and free-breathing in infants and young children. In addition, given that motion artifact was significantly higher in the upper lung zone without GA, increased stabilization in the upper chest and extremities should be considered.

Free-breathing radial stack-of-stars three-dimensional Dixon gradient echo sequence in abdominal magnetic resonance imaging in sedated pediatric patients.

BACKGROUND: There is a strong need for improvements in motion robust T1-weighted abdominal imaging sequences in children to enable high-quality, free-breathing imaging. OBJECTIVE: To compare imaging time and quality of a radial stack-of-stars, free-breathing T1-weighted gradient echo acquisition (volumetric interpolated breath-hold examination [VIBE]) three-dimensional (3-D) Dixon sequence in sedated pediatric patients undergoing abdominal magnetic resonance imaging (MRI) against conventional Cartesian T1-weighed sequences. MATERIALS AND METHODS: This study was approved by the institutional review board with informed consent obtained from all subjects. Study subjects included 31 pediatric patients (19 male, 12 female; median age: 5 years; interquartile range: 5 years) undergoing abdominal MRI at 3 tesla with a free-breathing T1-weighted radial stack-of-stars 3-D VIBE Dixon prototype sequence, StarVIBE Dixon (radial technique), between October 2018 and June 2019 with previous abdominal MR imaging using conventional Cartesian T1-weighed imaging (traditional technique). MRI component times were recorded as well as the total number of non-contrast T1-weighted sequences. Two radiologists independently rated images for quality using a scale from 1 to 5 according to the following metrics: overall image quality, hepatic edge sharpness, hepatic vessel clarity and respiratory motion robustness. Scores were compared between the groups. RESULTS: Mean T1-weighted imaging times for all subjects were 3.63 min for radial exams and 8.01 min for traditional exams (P<0.001), and total non-contrast imaging time was 32.7 min vs. 43.9 min (P=0.002). Adjusted mean total MRI time for all subjects was 60.2 min for radial exams and 65.7 min for traditional exams (P=0.387). The mean number of non-contrast T1-weighted sequences performed in radial MRI exams was 1.0 compared to 1.9 (range: 0-6) in traditional exams (P<0.001). StarVIBE Dixon outperformed Cartesian methods in all quality metrics. The mean overall image quality (scale 1-5) was 3.95 for radial exams and 3.31 for traditional exams (P<0.001). CONCLUSION: Radial stack-of-stars 3-D VIBE Dixon during free-breathing abdominal MRI in pediatric patients offers improved image quality compared to Cartesian T1-weighted imaging techniques with decreased T1-weighted and total non-contrast imaging time. This has important implications for children undergoing sedation for imaging.

Radiology Trainees' Perceptions of Speaking up Culture Related to Safety and Unprofessional Behavior in Their Work Environments.

OBJECTIVE. The purpose of this study was to compare radiology trainees' perceptions of the culture regarding speaking up about patient safety and unprofessional behavior in the clinical environment and to assess the likelihood that they will speak up in the presence of a medical hierarchy. MATERIALS AND METHODS. The study included radiology trainees from nine hospitals who attended a communication workshop. Trainees completed questionnaires assessing their perceptions of the support provided by their clinical environment regarding speaking up about patient safety and unprofessional behavior. We also queried their likelihood of speaking up within a team hierarchy about an error presented in a hypothetical clinical vignette. RESULTS. Of 61 participants, 58 (95%) completed questionnaires. Of these 58 participants, 84% felt encouraged by colleagues to speak up about safety concerns, and 57% felt encouraged to speak up about unprofessional behavior (p < .001). Moreover, 17% and 34% thought speaking up about safety concerns and unprofessional behavior, respectively, was difficult (p < .02). Trainees were less likely to agree that speaking up about unprofessional behavior (compared with speaking up about safety concerns) resulted in meaningful change (66% vs 95%; p < .001). In a vignette describing a sterile technique error, respondents were less likely to speak up to an attending radiologist (48%) versus a nurse, intern, or resident (79%, 84%, and 81%, respectively; p < .001). Significant predictors of the likelihood of trainees speaking up to an attending radiologist included perceived potential for patient harm as a result of the error (odds ratio [OR], 6.7; p < .001), perceptions of safety culture in the clinical environment (OR, 5.0; p = .03), and race or ethnicity (OR, 3.1; p = .03). CONCLUSION. Radiology trainees indicate gaps in workplace cultures regarding speaking up, particularly concerning unprofessional behavior and team hierarchy.

Practical Imaging Strategies for Intussusception in Children.

OBJECTIVE. Intussusception is the most common cause of intestinal obstruction in young children. Radiology has a key role in its diagnosis and treatment. This systematic review summarizes the currently available evidence for best practices in radiologic management of pediatric intussusception. CONCLUSION. High diagnostic accuracy and lack of ionizing radiation make ultrasound (US) the preferred imaging modality for diagnosing intussusception. For intussusception reduction, fluoroscopy-guided pneumatic enema and US-guided hydrostatic enema are equally dependable and safe techniques. The areas that warrant further research in this field include the efficacy and safety of the US-guided pneumatic enema, potential benefits of sedation and general anesthesia for the reduction procedure, and the optimal management of intussusceptions potentially involving pathologic lead points.

Skeletal maturation and long-bone growth patterns of patients with progeria: a retrospective study.

BACKGROUND: Hutchinson-Gilford progeria syndrome (termed progeria in this Article) is a rare sporadic genetic disorder. One early clinical manifestation of progeria is abnormal skeletal growth, yet this growth has not been fully characterised. We aimed to characterise the skeletal maturation and long-bone growth patterns of patients with the clinical phenotype of progeria. METHODS: For this retrospective study, we reviewed skeletal surveys of patients (aged <20 years) with progeria obtained over a 9·5-year period. Most surveys included radiographs of the hands and long bones (humeri, radii, ulnas, tibias, and fibulas). Bone ages of these patients were estimated by the standards of Greulich and Pyle. Following the established methods for studying long-bone growth, the study cohort was separated into two overlapping age groups: longitudinal bone length measurements were made between physes for the childhood group (aged 12 years or younger) and from the upper margins of the proximal to the lower margin of the distal ossified epiphyses for the adolescent group (aged 10 years or older). Bone age estimates and bone length measurements were plotted against the chronological age of patients and compared with reference standards. Statistical analyses were based on mixed models. FINDINGS: 85 patients with progeria and 250 skeletal surveys were included in our study. For both sexes, bone age estimates showed a more advanced skeletal maturation rate throughout all chronological ages than the normal rate of 1 (p<0·0001), with the rate of maturation being 1·09 (SE 0·02) for boys and 1·14 (0·02) for girls. Longitudinal long-bone lengths began to deviate from normal standards by age 1-2 years. Growth curves for these long bones plateaued at about half the normal eventual bone length, and the half-life (the time taken to grow to half the eventual bone length) was also about half the time compared with normal standards. INTERPRETATION: Our study established growth curves that might serve as reference standards for skeletal maturation and long-bone growth of patients with the clinical phenotype of progeria. FUNDING: The Progeria Research Foundation, the US National Heart, Lung and Blood Institute, the Dana-Farber Cancer Institute Stop&Shop Pediatric Brain Tumor Program, the US National Center for Research Resources, US National Institutes of Health.

Perceived Impact of COVID-19 on Pediatric Radiology Departments Around the World: WFPI COVID-19 Task Force Survey Results from 6 Continents.

PURPOSE: To investigate how COVID-19 has impacted pediatric radiology practice around the world at the present time. MATERIALS AND METHODS: This study was based on a survey conducted by the World Federation of Pediatric Imaging (WFPI) COVID-19 task force. Pediatric radiology representatives from countries around the world were identified and asked to complete an electronic survey consisting of 17 questions related to the impact of, concerns surrounding, and education related to COVID-19 on pediatric radiology. For each outcome variable, a pair of 2-way frequency tables was constructed: one against COVID-19 prevalence, and another against region. Tests of independence of the tables' row and column factors were performed based on Fisher's exact test and odds ratios (OR) were estimated via logistic regression models corresponding to these 2-way tables. RESULTS: The survey (response rate 87%, 72/83) comprised representatives from 71 countries and Hong Kong across 6 continents. 66/72 respondents (92%) indicated that COVID-19 has resulted in moderate (21/72, 29%), significant (36/72, 50%), or complete (9/72, 13%) change in radiology departments/practices in their countries. The two most frequent concerns over the next four months were personal/family health (54/72, 75%) and exposure (48/72, 67%). 79% (57/72) of respondents indicated some level of discomfort in identifying pediatric COVID-19 imaging manifestations. Changes in resident education were reported by 68/72 (94%) respondents, and 22/72 (31%) were concerned that the likelihood of current trainees pursuing a career in pediatric radiology will be impacted. CONCLUSION: COVID-19 has had a substantial negative impact on pediatric radiology practice around the world.

Subperiosteal new bone formation with the distal tibial classic metaphyseal lesion: prevalence on radiographic skeletal surveys.

BACKGROUND: The classic metaphyseal lesion (CML) is a strong indicator of infant abuse, and the distal tibia is one of the most common sites for this injury. OBJECTIVE: To determine the prevalence of subperiosteal new bone formation accompanying the distal tibial CMLs identified on infant skeletal surveys. MATERIALS AND METHODS: Skeletal surveys performed for suspected infant abuse (2005-2017) were reviewed. Inclusion criteria were 1) anteroposterior (AP) and lateral radiographs of a distal tibial CML from the initial survey, 2) AP radiograph from the 2-week follow-up survey, 3) additional fractures, 4) child protection team consults and 5) mandated report filing for suspected abuse. We identified 22 distal tibial CMLs from 16 infants. Radiographs of these lesions were shown on the picture archiving and communication system to two blinded pediatric radiologists. Readers indicated the presence/absence of subperiosteal new bone formation on individual and combinations of images. RESULTS: Inter-reader agreements were fair (kappa=0.47). The prevalence of subperiosteal new bone formation on initial AP radiograph was 34%. Significant increases in the prevalence were found with the addition of follow-up AP radiograph (57%; P<0.001), initial lateral radiograph (57%; P=0.002) and follow-up AP plus initial lateral radiographs (71%; P<0.001). Statistically significant increases in prevalence were also noted when the third view was added to the other two views (increase of 14%; P=0.024). CONCLUSION: Even when skeletal surveys include initial AP, lateral and follow-up AP radiographs of the tibia, nearly one-third of distal tibial CMLs will fail to demonstrate subperiosteal new bone formation.

Improving Low-Dose Pediatric Abdominal CT by Using Convolutional Neural Networks.

PURPOSE: To evaluate the efficacy of convolutional neural networks (CNNs) to improve the image quality of low-dose pediatric abdominal CT images. MATERIALS AND METHODS: Images from 11 pediatric abdominal CT examinations acquired between June and July 2018 were reconstructed with filtered back projection (FBP) and an iterative reconstruction (IR) algorithm. A residual CNN was trained using the FBP image as the input and the difference between FBP and IR as the target such that the network was able to predict the residual image and simulate the IR. CNN-based postprocessing was applied to 20 low-dose pediatric image datasets acquired between December 2016 and December 2017 on a scanner limited to reconstructing FBP images. The FBP and CNN images were evaluated based on objective image noise and subjective image review by two pediatric radiologists. For each of five features, readers rated images on a five-point Likert scale and also indicated their preferred series. Readers also indicated their "overall preference" for CNN versus FBP. Preference and Likert scores were analyzed for individual and combined readers. Interreader agreement was assessed. RESULTS: The CT number remained unchanged between FBP and CNN images. Image noise was reduced by 31% for CNN images (P < .001). CNN was preferred for overall image quality for individual and combined readers. For combined Likert scores, at least one of the two score types (Likert or binary preference) indicated a significant favoring of CNN over FBP for low contrast, image noise, artifacts, and high contrast, whereas the reverse was true for spatial resolution. CONCLUSION: FBP images can be improved in image space by a well-trained CNN, which may afford a reduction in dose or improvement in image quality on scanners limited to FBP reconstruction.© RSNA, 2019.

The distal tibial classic metaphyseal lesion: medial versus lateral cortical injury.

BACKGROUND: The distal tibia is a common location for the classic metaphyseal lesion (CML). Prior radiologic-pathologic studies have suggested a tendency for medial, as opposed to lateral, cortical injury with the CML, but there has been no formal study of the geographic distribution of this strong indicator of abuse. OBJECTIVE: This study compares medial versus lateral cortical involvement of distal tibial CMLs in a clinical cohort of infants with suspected abuse. MATERIALS AND METHODS: Reports of 1,020 skeletal surveys performed for suspected abuse (July 2005-June 2016) were reviewed. Twenty-six distal tibial CMLs (14 unilateral, 6 bilateral) with anteroposterior (AP) and lateral projections on the initial skeletal survey and at least an AP view on the follow-up survey were identified in 20 infants. Two blinded pediatric radiologists determined if the medial and/or lateral margins of the distal tibial metaphysis were involved by the CML. RESULTS: Average interreader absolute agreement and kappa scores were 0.69-0.90 and 0.45-0.72, respectively. Average intrareader absolute agreement and kappa scores were 0.65-0.88 and 0.44-0.57, respectively. Analyses showed that the distal tibial CML almost always involved the medial cortical margin (reader 1=89%, reader 2=88%, pooled=89%) and the fracture infrequently involved the lateral cortical margin (reader 1=12%, reader 2=38%, pooled=26%). The percentage point difference between fracture involvement in medial and lateral margins was statistically significant from zero (P<0.001). CONCLUSION: The distal tibial CML is most often encountered medially; lateral involvement is uncommon. This observation should help guide the radiologic diagnosis and could have implications for understanding the biomechanics of this distinctive injury.

Simultaneous multi-slice accelerated turbo spin echo of the knee in pediatric patients.

PURPOSE: To compare knee MRI performed with the integrated parallel acquisition technique (PAT) and simultaneous multislice (SMS) turbo spin echo (TSE) T2-weighted (T2w) sequences with conventional TSE sequences in pediatric patients. MATERIALS AND METHODS: This was a retrospective IRB-approved study. Seventy-four subjects (26 male, 48 female, mean age 15.3 years, range 8-20) underwent 3-T MRI of the knee with a T2w TSE pulse sequence prototype with four-fold PAT and SMS acceleration as well as the standard PAT-only accelerated sequences. Images were anonymized and two study folders were created: one examination with only T2w PAT2 images (conventional examination) and one examination with only T2w SMS2/PAT2 sequences (SMS examination). Two readers rated examinations for 15 specific imaging findings and 5 quality metrics. Interreader agreement was measured. Signal to noise (SNR) and contrast to noise (CNR) were measured for SMS and conventional T2w sequences. RESULTS: Consensus review demonstrated diagnostic quality performance of SMS examinations with respect to all 15 structures. Average area under the curve (AROC) was 0.95 and 0.97 for readers 1 and 2, respectively. The conventional sequence was favored over SMS for four out of five quality metrics (p < 0.001). SNR and CNR were higher for the conventional sequences compared to SMS. CONCLUSION: SMS accelerated T2w TSE sequences offer a faster alternative for knee imaging in pediatric patients without compromise in diagnostic performance despite diminished SNR. The four-fold acceleration of SMS is beneficial to pediatric patients who often have difficulty staying still for long MRI examinations.

The special case of the 2 × 2 table: asymptotic unconditional McNemar test can be used to estimate sample size even for analysis based on GEE.

OBJECTIVES: Aligning the method used to estimate sample size with the planned analytic method ensures the sample size needed to achieve the planned power. When using generalized estimating equations (GEE) to analyze a paired binary primary outcome with no covariates, many use an exact McNemar test to calculate sample size. We reviewed the approaches to sample size estimation for paired binary data and compared the sample size estimates on the same numerical examples. STUDY DESIGN AND SETTING: We used the hypothesized sample proportions for the 2 × 2 table to calculate the correlation between the marginal proportions to estimate sample size based on GEE. We solved the inside proportions based on the correlation and the marginal proportions to estimate sample size based on exact McNemar, asymptotic unconditional McNemar, and asymptotic conditional McNemar. RESULTS: The asymptotic unconditional McNemar test is a good approximation of GEE method by Pan. The exact McNemar is too conservative and yields unnecessarily large sample size estimates than all other methods. CONCLUSION: In the special case of a 2 × 2 table, even when a GEE approach to binary logistic regression is the planned analytic method, the asymptotic unconditional McNemar test can be used to estimate sample size. We do not recommend using an exact McNemar test.

Blink patterns and lid-contact times in dry-eye and normal subjects.

PURPOSE: To classify blinks in dry eye and normal subjects into six subtypes, and to define the blink rate and duration within each type of blink, as well as the total lid-contact time/minute. MATERIALS AND METHODS: This was a single-centered, prospective, double-blind study of eleven dry-eye and ten normal subjects. Predefined subjects watched a video while blinks were recorded for 10 minutes. Partial blinks were classified by percentage closure of maximal palpebral fissure opening: 25%, 50%, 75%. Complete blinks were characterized as full (>0 seconds), extended (>0.1 seconds), or superextended (>0.5 seconds). The mean duration of each type of blink was determined and standardized per minute as total lid-contact time. RESULTS: Total blinks observed were 4,990 (1,414 normal, 3,756 dry eye): 1,809 (50.59%) partial and 1,767 (49.41%) complete blinks among dry-eye subjects versus 741 (52.90%) partial and 673 (47.60%) complete blinks among normal subjects. Only superextended blinks of ≥0.5-second duration were significantly more frequent in dry-eye subjects than normals (2.3% versus 0.2%, respectively; P=0.023). Total contact time was seven times higher in dry-eye subjects than normals (0.565 versus 0.080 seconds, respectively; P<0.001). Isolating only extended blinks (>0.1 second), the average contact time (seconds) was four times longer in dry-eye versus normal subjects (2.459 in dry eye, 0.575 in normals; P=0.003). Isolating only superextended blinks (>0.5 seconds), average contact time was also significantly different (7.134 in dry eye, 1.589 in normals; P<0.001). The contact rate for all full closures was 6.4 times longer in dry-eye (0.045 versus 0.007, P<0.001) than normal subjects. CONCLUSION: Dry-eye subjects spent 4.5% of a minute with their eyes closed, while normal subjects spent 0.7% of a minute with their eyes closed. Contact time might play a role in the visual function decay associated with increased blink rates.

Automated grading system for evaluation of ocular redness associated with dry eye.

BACKGROUND: We have observed that dry eye redness is characterized by a prominence of fine horizontal conjunctival vessels in the exposed ocular surface of the interpalpebral fissure, and have incorporated this feature into the grading of redness in clinical studies of dry eye. AIM: To develop an automated method of grading dry eye-associated ocular redness in order to expand on the clinical grading system currently used. METHODS: Ninety nine images from 26 dry eye subjects were evaluated by five graders using a 0-4 (in 0.5 increments) dry eye redness (Ora Calibra™ Dry Eye Redness Scale [OCDER]) scale. For the automated method, the Opencv computer vision library was used to develop software for calculating redness and horizontal conjunctival vessels (noted as "horizontality"). From original photograph, the region of interest (ROI) was selected manually using the open source ImageJ software. Total average redness intensity (Com-Red) was calculated as a single channel 8-bit image as R - 0.83G - 0.17B, where R, G and B were the respective intensities of the red, green and blue channels. The location of vessels was detected by normalizing the blue channel and selecting pixels with an intensity of less than 97% of the mean. The horizontal component (Com-Hor) was calculated by the first order Sobel derivative in the vertical direction and the score was calculated as the average blue channel image intensity of this vertical derivative. Pearson correlation coefficients, accuracy and concordance correlation coefficients (CCC) were calculated after regression and standardized regression of the dataset. RESULTS: The agreement (both Pearson's and CCC) among investigators using the OCDER scale was 0.67, while the agreement of investigator to computer was 0.76. A multiple regression using both redness and horizontality improved the agreement CCC from 0.66 and 0.69 to 0.76, demonstrating the contribution of vessel geometry to the overall grade. Computer analysis of a given image has 100% repeatability and zero variability from session to session. CONCLUSION: This objective means of grading ocular redness in a unified fashion has potential significance as a new clinical endpoint. In comparisons between computer and investigator, computer grading proved to be more reliable than another investigator using the OCDER scale. The best fitting model based on the present sample, and usable for future studies, was [Formula: see text] is the predicted investigator grade, and [Formula: see text] and [Formula: see text] are logarithmic transformations of the computer calculated parameters COM-Hor and COM-Red. Considering the superior repeatability, computer automated grading might be preferable to investigator grading in multicentered dry eye studies in which the subtle differences in redness incurred by treatment have been historically difficult to define.