Pediatric MRI frequently requires propofol anesthesia to obtain motion-free diagnostic images. Operational imaging factors may influence examination duration and anesthetic exposure, but these relationships remain incompletely characterized. To evaluate associations between operational protocol-related operational factors, MRI duration, and propofol exposure during pediatric MRI. This retrospective study included 2,149 pediatric MRI examinations performed under propofol anesthesia between January 2020 and May 2025. Operational imaging variables included scanner field strength (3 T vs 1.5 T), contrast-enhanced MRI, and protocol type (multi-region vs single-region). Multivariable log-linear regression models evaluated associations between operational imaging factors and MRI duration. Multivariable linear regression models assessed associations between MRI duration and total propofol exposure (mg/kg). Among 2,149 pediatric MRI examinations (median age 5 years [IQR 2–8]; 1,299/2,149 [60.4
The Greulich-Pyle atlas is generally used in clinical practice; however, it involves radiation exposure. Our aim was to clarify the correlation between ultrasonic ossification ratios (ORs) and radiographic bone age, and to develop a repeatable ultrasonic parameter for assessing bone age. To determine the correlation between ultrasonic ORs and radiographic bone age, and to test the hypothesis that a summed ultrasonic ossification parameter provides a highly correlated, reproducible, radiation‑free method for pediatric bone age assessment. Ultrasonic imaging was performed on seven ossification centers and epiphyses in 144 consecutive patients aged 0.1–18.0 years. The ORs were measured and compared with the radiological bone age. The correlation between ultrasonic ORs and bone age assessed by the Greulich-Pyle atlas was analyzed by the Pearson correlation coefficient. The intra-observer and inter-observer reliability were evaluated using the intraclass correlation coefficient (ICC). The sum of the ORs of the distal radius, distal ulna, and proximal radius was highly correlated with radiographic bone age (boys: r=0.98; girls: r=0.97). Adding the ORs of other bones did not increase the value of Pearson’s r, while summing the ORs of all assessed bones slightly reduced it (boys: r=0.97; girls: r=0.96). The reliability of the sum of the ORs of the radius, ulna, and femur and the sum of the ORs of the distal radius, distal ulna, and proximal radius was identical (intra-observer ICC=0.99 [95
Fetal magnetic resonance imaging (MRI)-derived lung volume measurements are used for prenatal risk stratification in conditions associated with pulmonary hypoplasia, but manual segmentation is time-intensive. To evaluate whether automated fetal MRI lung segmentation enables rapid and reliable volumetric assessment across diverse pulmonary hypoplasia phenotypes. In this retrospective study, fetal MRI examinations performed from 2016 to 2025 with expert-adjudicated lung segmentations obtained during clinical care were used for model development and testing. Automated segmentation pipelines were developed for total lung volume (TLV) and percent predicted lung volume (PPLV). Performance was assessed using Dice similarity coefficient (DSC), absolute percentage error (APE), Bland-Altman analysis, intraclass correlation coefficient (ICC), case-level risk-category agreement, and inference time. Lowest-performing TLV outputs were adjudicated by a fetal radiologist to simulate a human-in-the-loop workflow. The final dataset included 906 fetal MRI examinations with complete and recoverable lung segmentations, acquired at a mean gestational age of 27.1 weeks, range 16-38 weeks. Automated TLV estimation showed excellent agreement with manual volumetry (ICC, 0.955; mean DSC, 0.81±0.10; median APE 11.1
Dengue infection is an increasingly recognized cause of neurological complications in children, with a wide and evolving spectrum of central and peripheral nervous system involvement. This pictorial review systematically illustrates the neuroradiological manifestations of pediatric dengue using a pathophysiology- and imaging pattern-based approach, correlating characteristic magnetic resonance imaging (MRI) findings with the current literature. Neuroimaging manifestations are broadly categorized into direct viral neuroinvasive, immune-mediated, and reversible edema patterns. Direct viral involvement includes meningoencephalitis, deep gray matter involvement, and rhombencephalitis, with severe cases demonstrating the characteristic "double doughnut" sign of necrotizing encephalopathy-like encephalitis and the "Jack-o'-lantern" sign of pontine involvement. Immune-mediated white matter injury includes acute leukoencephalopathy with restricted diffusion and acute disseminated encephalomyelitis, which differ in their diffusion characteristics and temporal relationship to infection. Reversible edema syndromes include reversible splenial lesion syndrome and the less commonly reported posterior reversible encephalopathy syndrome, both associated with favorable clinical and radiological outcomes. Spinal and peripheral nervous system manifestations include myelitis and Guillain-Barré syndrome, with MRI playing a key role in differentiating these entities from other infectious and inflammatory myeloradiculopathies. Importantly, intralesional hemorrhagic changes on susceptibility-weighted or gradient recalled echo imaging, particularly within the deep gray nuclei and brainstem, represent a valuable imaging marker favoring dengue over several viral and inflammatory mimickers. Recognition of these characteristic pattern-based neuroradiological manifestations and their underlying pathophysiological correlates improves diagnostic confidence, facilitates differentiation from competing diagnoses, and guides timely management of pediatric dengue neurological disease in endemic regions.
Pediatric head computed tomography (CT) is widely used for urgent neuroimaging, but children’s higher radiosensitivity makes protocol optimization and quantitative validation essential. Many available phantoms lack patient-specific pediatric anatomy and realistic brain-mimicking attenuation, limiting their relevance for neuro-CT research and benchmarking. To develop and validate a CT-derived, patient-specific pediatric brain phantom that preserves anatomically faithful geometry and reproduces brain-mimicking soft-tissue attenuation with physics-validated energy dependence. A fully anonymized head CT of a 5-year-old child was segmented in 3-D Slicer, exported to standard tessellation language (STL), and partitioned into hemispheres. A polylactic acid (PLA) master was three-dimensional (D)-printed to fabricate a reinforced silicone mold. A homogeneous brain-mimicking soft-tissue surrogate was produced by modifying an epoxy system with acetone. The assembled phantom was scanned at 80, 100, 120, and 140 kilovolt peak (kVp). Region of interest (ROI)-based CT numbers were measured across both hemispheres. Computational attenuation verification was performed using Particle and Heavy Ion Transport code System (PHITS) Monte Carlo simulations with a mono-energetic transmission method over 15-150 kiloelectronvolt (keV) and compared with PhyX–Photon Shielding and Dosimetry (PhyX-PSD) and National Institute of Standards and Technology XCOM database (NIST/XCOM) reference. The fabricated phantom preserved the main bilateral morphology and retained 96.85
Pediatric spinal CSF leaks are uncommon and encompass a spectrum of leak types with distinct imaging findings, diagnostic pathways, and treatment strategies. Brain and spine MRI guide the evaluation of suspected CSF leaks and the selection of advanced imaging techniques for precise leak localization. Leak types include dural tears, CSF-venous fistulas, and CSF-vascular malformation fistulas, each with distinct imaging and treatment considerations. This pictorial review illustrates the spectrum of pediatric spinal CSF leaks and provides a practical framework for their diagnosis and management.
Bone age assessment using the Greulich and Pyle atlas is widely used in pediatrics but is subject to substantial reader variability. As artificial intelligence (AI) systems increasingly support clinical workflows, understanding the magnitude and structure of human variability is essential for contextualizing AI performance. To quantify human variability in pediatric bone age assessment across multiple institutions and to compare human variability with two automated bone age estimation methods. In this multi-reader, multi-case study, 1,285 left-hand and wrist radiographs from five US academic centers were independently interpreted by four radiologists per case from a centrally administered pool of 22 radiologists recruited across nine institutions. Bone age estimates were obtained using the Greulich and Pyle atlas. Variability was assessed at the image, rater, and institution levels using mixed-effects modeling. Inter-reader variability was estimated while accounting for patient age and sex. Performance of two automated methods (BoneXpert and a deep-learning algorithm) was evaluated in an interchangeability analysis using a three-rater consensus reference. Accuracy metrics included mean absolute error (MAE), root mean square error (RMSE), and rates of substantial deviation (>1.8 years). Within-image variability among radiologists was higher in younger patients (<12 years) than in older patients (standard deviation (SD) 8.7 months vs. 4.2 months, P<0.001). Between-institution variability decreased substantially after adjustment for patient age (8.9 months to 0.3 months). Inter-rater variability remained stable after adjustment (1.2 months). Variability was greater in male patients and younger age groups. Several radiologists demonstrated systematic biases related to age or sex. In comparison with individual human readers, BoneXpert showed lower error (MAE 4.8 months vs. 6.5 months) and variability, while the deep-learning algorithm performed similarly to human readers (MAE 6.3 months). Substantial deviations occurred in 0.7
Foreign body ingestion constitutes a common emergency in children. Radiography is essential for confirming presence, type, and location of ingested objects. International guidelines recommend broad anatomical coverage from the neck to the pelvis; however, the impact of radiographic exposure parameters and collimation on radiation dose remains underexplored. To evaluate the effects of different scan protocols on radiation dose in pediatric foreign body ingestion. Two-hundred-twenty-one radiographs from 183 pediatric patients (median age 2.8 years [interquartile range 1.5–5.5 years]; 112 males, 71 females) performed between September 2019 and April 2025 were retrospectively analyzed. Radiographs were categorized as neck–upper abdomen, extended collimation including pelvis, or abdomen only. Notably, the first two protocols used chest radiograph exposure parameters. Data included patient demographics, foreign body type and location, radiographic protocol, dose-area product, and estimated effective dose. A foreign body was detected in 55.7
Advances in neonatal care have improved survival of extremely premature infants (<28 weeks gestational age (GA)), especially with necrotizing enterocolitis (NEC). To evaluate if imaging features of NEC differ between extreme premature and premature infants. We retrospectively reviewed the abdominal radiograph and/or ultrasound obtained immediately prior to surgically proven diagnosis of NEC in our neonatal intensive care unit infants January 2020-January 2025. Two blinded pediatric radiologists reviewed imaging, and a third acted as a tie-breaker in case of discordant review. Patients were stratified by GA (<28 weeks vs. ≥28 weeks) and birth weight (<1,500 g vs. ≥1,500 g). Chi-square test was used to assess differences in imaging findings. Ninety patients met inclusion criteria (median age 21.5 days); 43.3
The pediatric skull base develops through a complex sequence of events from fetal development through adolescence and may result in numerous normal variants and embryonic remnants that can mimic pathology on imaging. This review highlights the embryologic basis and imaging appearance of common central skull base developmental variants encountered in children. Focusing on synchondroses, clefts, foramina, notochordal remnants, and pneumatization-related variants, we emphasize age-dependent findings and key imaging features that provide radiologists with a reference to support confident interpretation and the ability to distinguish benign developmental variants from true pathology in children.
Bone age (BA) is the gold standard for skeletal maturity assessment but is not routinely incorporated into pediatric growth evaluation workflow because it requires dedicated hand radiographs and specialist interpretation. To develop an estimation model for BA from chest radiographs using deep neural networks. We retrospectively analyzed children aged 3–15 years who underwent both chest and hand radiography at a tertiary center over 20 years. Patients with skeletal dysplasia or chest wall abnormalities were excluded. Reference BA was determined from hand radiographs by pediatric endocrinologists using the Tanner-Whitehouse 2 radius-ulna-short bones (TW2-RUS) method. Three pretrained deep neural networks were fine-tuned to estimate BA from chest radiographs using sex-non-considering models and sex-considering models. Model performance was evaluated using the intraclass correlation coefficient (ICC), root mean squared error (RMSE), and related metrics. Of 180 screened patients, 101 were included, yielding 237 chest radiographs. Estimated BA showed good concordance with the reference standard, with ICCs up to 0.81 (95
Molecular characteristics of retinoblastoma cannot be assessed before treatment because tumor biopsy is contraindicated. Photoreceptorness reflects photoreceptor-related gene expression and tumor differentiation. Non-invasive imaging biomarkers that capture this biology are therefore needed. To evaluate whether quantitative radiomics derived from pretreatment magnetic resonance imaging can predict loss of photoreceptorness in retinoblastoma and validate this approach in an independent cohort. In this retrospective multicenter study, patients with retinoblastoma who underwent primary enucleation and had both pretreatment T2-weighted magnetic resonance imaging and genome-wide messenger RNA expression data were included. Tumors in the highest and lowest photoreceptorness quartiles were analyzed. Whole-tumor segmentations were used to extract radiomic features with PyRadiomics. Multiple machine-learning pipelines were evaluated using repeated stratified cross-validation, and the best-performing model was tested in an independent cohort. Forty-five patients (median age, 18 months [range, 2–70], 18 female) were included: 29 in the training cohort and 16 in the independent testing cohort. The best-performing model used recursive feature elimination with a random forest classifier and achieved a mean cross-validated area under the receiver operating characteristic curve of 0.83 in the training cohort. In the independent testing cohort, the model achieved an area under the receiver operating characteristic curve of 0.81 (95
To evaluate the feasibility of ultrafast shear-wave elastography (SWE) for quantifying thrombus stiffness in pediatric deep vein thrombosis (DVT), and to explore changes over time. In this single-center prospective study, pediatric patients (<18 years) with ultrasound-confirmed DVT were prospectively enrolled. Bedside SWE ultrasound was performed using a research platform (Vantage-256, Verasonics, Kirkland, Washington, USA) with a 6.9-MHz linear transducer. Mean shear-wave velocity was computed using MATLAB software. To assess shear-wave velocity (SWV) changes over time (exploratory analysis), SWE was planned at three timepoints (0-5 days, 6 days-6 weeks, and 6-16 weeks from DVT diagnosis). Feasibility was defined as the ability to obtain SWV measures. Changes in SWV over time in patients with ≥2 successful SWE were estimated using generalized estimating equations. The model was adjusted for patient age at DVT and sex. In total, 35 patients (median age 1.4 years [range, 0.0-17.6], 63
