Pediatric Radiology最新文献

Pediatric H3K27-altered glioma most frequently arises from midline cerebral parenchyma, whereas intracranial extra-axial locations are exceedingly rare. We report the case of an 8-year-old boy who presented with chronic-onset left-sided hearing impairment and peripheral facial paralysis. Radiological imaging identified an isolated cerebellopontine angle lesion distinct from surrounding structures. A vestibular schwannoma was diagnosed based on preoperative examinations. Surgical resection revealed an unusual facial nerve-infiltrating mass with distinct vascular features. Histopathological assessment yielded an unexpected diagnosis of H3K27-altered diffuse midline glioma, contrasting with initial preoperative impressions. This exceptional case illustrates diagnostic pitfalls in pediatric skull base pathologies and expands the recognized spectrum of diffuse midline gliomas.

Foundation models (FMs) are large deep learning models pre-trained on vast heterogeneous datasets through self-supervised learning that are adaptable to diverse downstream tasks with minimal fine-tuning. In pediatric radiology, where data scarcity, rare pathologies, and anatomical variability present significant hurdles, FMs offer a robust mechanism for broad feature learning. Complementary to traditional machine learning pipelines, FMs serve as flexible backbones that can be adapted to specific clinical needs through established techniques such as transfer learning and parameter-efficient fine-tuning. These models can facilitate multiple tasks, such as pathology detection and classification, lesion segmentation, report generation, and visual question answering, with the potential to improve diagnostic accuracy, workflow efficiency, and decision support in pediatric care. However, FMs' implementation in pediatric imaging faces key challenges, including pediatric unique disease spectra and anatomical variability, limited datasets, ethical and privacy issues, and the absence of pediatric-specific validation. Broader limitations include hallucinations, lack of model explainability, resource disparities, and risks of radiologists' deskilling. Future perspectives to overcome these barriers are represented by techniques such as federated and continual learning, and synthetic data generation. Our review introduces the principles and architectures of FMs, presents the current and emerging applications in pediatric radiology, and offers an overview of the challenges and future directions for FMs' safe, equitable, and effective integration into clinical practice. FMs are a promising frontier for transforming pediatric imaging and advancing child-centered healthcare.

Infection-triggered encephalopathy syndrome is a group of acute encephalopathies that develop in temporal association with febrile illnesses. Unlike infectious or autoimmune encephalitis, direct viral invasion or antibody-mediated pathology is usually absent. Instead, excitotoxicity and cytokine storm are considered key mechanisms. The core diagnostic criteria include a preceding febrile illness, presence of neurological symptoms (encephalopathy, seizures), and, most importantly, syndrome-specific magnetic resonance imaging findings. Therefore, neuroimaging is indispensable for confirming infection-triggered encephalopathy syndromes and differentiating them from infectious encephalitis, metabolic disorders, and other acute encephalopathies. Characteristic radiological features-such as the bright tree appearance in acute encephalopathy with biphasic seizures and late reduced diffusion, bithalamic involvement in acute necrotizing encephalopathy, and reversible splenial lesions in mild encephalopathy with reversible splenial lesions-are essential for diagnostic classification and prognostication. Pediatric radiologists must be familiar with these patterns to ensure early recognition and appropriate diagnosis. This review aimed to present the epidemiology, clinical characteristics, and neuroimaging features of infection-triggered encephalopathy syndromes, emphasizing the critical role of magnetic resonance imaging findings in diagnostic criteria.

Background: Contrast enema is a key imaging tool in the evaluation of children with suspected Hirschsprung disease, although the diagnostic accuracy of individual radiologic signs remains debated.

Objective: This study aimed to evaluate the diagnostic performance of specific radiological signs and clinical symptoms in Hirschsprung disease and to develop a simple predictive model based on logistic regression.

Materials and methods: We retrospectively reviewed 247 contrast enemas from 237 children evaluated for suspected Hirschsprung disease, including 59 studies from 49 histologically confirmed cases. Radiologic and clinical variables were analysed to assess the diagnostic accuracy of specific signs using standard performance metrics and receiver operating characteristic (ROC) curve analysis. A multivariable logistic regression model was developed and optimised to identify the minimal combination of predictors yielding the best diagnostic performance for Hirschsprung disease.

Results: Children with Hirschsprung disease presented with earlier symptom onset and earlier contrast enema evaluation compared with non-Hirschsprung disease patients (both P<0.001). Among radiologic findings, an abnormal rectosigmoid index and a visible transition zone were the strongest predictors of Hirschsprung disease. Clinical features such as abdominal distension, vomiting, and need for rectal irrigations showed significant associations. The final multivariable model demonstrated excellent diagnostic performance (sensitivity 0.83, specificity 0.87, area under the ROC curve (AUC-ROC)=0.85) and was transformed into a simplified clinical score (0-9 points) for practical application.

Conclusion: Specific radiological signs, particularly rectosigmoid index and transition zone, retain significant diagnostic value in Hirschsprung disease. When combined with clinical symptoms, they allow development of a simple predictive model that may support clinical decision-making.

This brief review focuses on the imaging of the craniocervical junction with particular attention to linear and angular craniometry. It attempts to answer the question of why there are so many measurements and how they are clinically useful. Congenital anomalies of the craniocervical junction will be discussed with an overview of craniocervical junction embryology and development. Entities affecting the craniocervical junction which will be discussed include platybasia, basilar invagination, atlanto-occipital assimilation, and assorted congenital anomalies.

Background: With the increasing use of magnetic resonance imaging (MRI) in children, radiologists frequently encounter incidental findings that may mimic pathology. One such underrecognized finding is T2-weighted hyperintensity in the superior extraconal orbital fat, which is occasionally mistaken for an infiltrative or neoplastic process.

Objective: Our objective was to characterize the imaging appearance, prevalence, and clinical associations of superior extraconal orbital fat T2 hyperintensity in pediatric MRI.

Materials and methods: We conducted a retrospective study of 143 pediatric patients (mean age 7.2±5.1 years) who underwent brain MRI with an orbit-specific protocol between 2015 and 2022. Patients were grouped based on the presence or absence of bilateral papilledema and whether imaging was performed under general anesthesia. Clinical data were extracted from the electronic medical records. Three neuroradiologists reviewed images for the presence of a hyperintense signal along the superior extraconal orbital fat. Interobserver agreement was calculated using Fleiss' kappa. Univariate and multivariable logistic regression analyses were performed to assess associations with age, anesthesia, gender, and magnet strength.

Results: Symmetric T2-hyperintense bands along the superior extraconal orbital fat were observed in 45.5% of patients. The finding was more common in younger children (4.4±3.9 years vs. 9.6±4.8 years; P<0.001). Multivariate analysis showed a significant negative correlation with age (P<0.001) and a positive correlation with papilledema (P=0.012), but no independent association with gender, anesthesia, or magnet strength. The hyperintensity was non-enhancing or only subtly enhancing. Clinical follow-up demonstrated no subsequent orbital or infiltrative pathology in the majority of patients, and most patients without documented follow-up underwent ambulatory MRI for evaluation of strabismus, which showed no evidence of infiltrative or other orbital disease.

Conclusion: Superior extraconal orbital fat T2 hyperintensity is a relatively common, likely non-pathologic MRI finding in pediatric patients, particularly in younger children, and the apparent association with anesthesia likely reflects age-related confounding. Awareness of this benign appearance may help avoid diagnostic confusion and prevent unnecessary workup or intervention.

Background: It is important to obtain an accurate volumetric characterization of fetal posterior fossa growth on magnetic resonance imaging (MRI).

Objective: This study analyzes the normal growth, sexual dimorphism, and lateral asymmetry of the fetal posterior fossa during the late second and third trimesters.

Materials and methods: A total of 100 fetuses (23-40 weeks gestational age (GA)) with normal brain development were retrospectively included in this study. The bilateral posterior fossa, cerebellar hemisphere, and brainstem (including the midbrain, pons, and medulla oblongata) were manually segmented on in vivo fetal MRIs. Normal growth, sexual dimorphism, and lateral asymmetry were analyzed after the fossa volumes were obtained.

Results: The total, left, and right posterior fossa volumes all linearly increased with GA. The volume of the right posterior fossa was significantly greater than that of the left and increased slightly faster. The cerebellar volume exponentially increased with increasing GA. The right cerebellar volume was significantly greater than the left cerebellar volume. The bilateral cerebellar volume had a similar growth rate, which accelerated and increased more quickly than did that of the posterior fossa volume after 35 weeks GA. The brainstem volume linearly increased with GA. The pontine volume increased the fastest, followed by the midbrain volume, and the medullary volume increased the slowest. The relationship between the volume of the cerebrospinal fluid in the posterior fossa and GA was described adequately by a second-order polynomial curve, which increased before 30 weeks GA but gradually decreased after 35 weeks GA. No sexual dimorphism was detected in any of the measurements.

Conclusions: The development of the fetal posterior fossa follows a specific spatiotemporal course, and volumetric measurements reveal structure-specific and GA-related changes with different rates and lateral asymmetries. These results are valuable for assessing normal fetal posterior fossa development in utero.

Palpable soft-tissue masses in the pediatric head and neck are common and encompass a broad differential that includes congenital, infectious, inflammatory, and neoplastic processes. Owing to its ability to characterize lesion morphology, vascularity, and internal architecture without ionizing radiation, ultrasonography (US) remains the primary imaging modality for initial evaluation. This review focuses on pediatric head and neck masses, emphasizing practical US-based assessment and differential diagnosis. Entities discussed include common congenital lesions such as dermoid cysts, thyroglossal duct cysts, and branchial cleft anomalies, salivary gland abnormalities (including sialadenitis, sialolithiasis, ranulas, and salivary tumors), thymic lesions (such as ectopic thymus and thymic cysts), and thyroid abnormalities (including ectopic thyroid tissue, thyroid nodules, and thyroid malignancy).

Background: Fontan-associated liver disease (FALD) is associated with morbidity and mortality in patients with palliated single ventricle congenital heart disease.

Objective: To develop machine learning models using radiomic features from T1-weighted, T2-weighted, and diffusion-weighted MRI with pertinent clinical variables to predict Fontan failure and correlates of FALD severity in patients who underwent the Fontan operation.

Materials and methods: In this retrospective study of abdominal MRI examinations and clinical record data from 131 Fontan palliation patients (age range 9.1 - 53.3 years old), radiomic features from the liver and spleen were extracted using axial T1-weighted, T2-weighted fat-suppressed, and diffusion-weighted sequences. Patients were categorized by a composite clinical outcome (i.e., Fontan failure) and by correlates of FALD severity, including liver shear stiffness and portal hypertension. Support vector machine (SVM) and multivariable logistic regression models were used to perform two-class classification using radiomic features and/or clinical data. All models were trained and evaluated using five-fold cross-validation (CV).

Results: The best radiomic-only model utilized T2-weighted imaging of both organs with logistic regression to predict the presence of portal hypertension, achieving an AUROC of 0.85±0.01. Clinical-only models showed inferior diagnostic accuracy with the highest AUROC of 0.70±0.08. Combining radiomic and clinical features also did not enhance performance compared to radiomic-only models, with the highest AUROC of 0.77±0.05. Ensemble modeling, which incorporated radiomics from all three MRI sequences, yielded AUROCs ranging from 0.33 to 0.72.

Conclusion: Models incorporating radiomic features from abdominal MRI in Fontan circulation patients demonstrate moderate diagnostic performance for predicting Fontan failure as well as correlates of FALD severity. These models outperformed models containing only clinical electronic health record data and did not improve with ensembled radiomic and clinical data.