Pediatric Radiology最新文献

Background: Repeated full-spine radiography for scoliosis follow-up in children results in increased radiation exposure, especially to anterior radiosensitive organs. Optimizing projection direction and beam filtration is essential for dose reduction.

Objective: To quantitatively evaluate the age-dependent effects of anteroposterior (AP) and posteroanterior (PA) projections, with and without a 0.1-mm copper filter, on organ and effective doses in pediatric full-spine radiography.

Materials and methods: Monte Carlo simulations were performed using the Particle and Heavy Ion Transport code System with 5-, 10-, and 15-year-old female hybrid phantoms. Full-spine radiography from the first cervical vertebra to both femoral heads was modeled under AP and PA conditions, with or without copper filtration. Organ doses were calculated, with active bone marrow and bone surface evaluated using the "International Commission on Radiological Protection Publication 116" dose response functions. Percentage depth dose analysis was performed to assess the effect of body thickness.

Results: PA projection markedly reduced doses to anterior radiosensitive organs, with maximum reductions of approximately 93% for the breast (AP/PA ratio 14) and over 80% for the thyroid. Copper filtration provided additional reductions of 15-19% in AP and 5-6% in PA. In contrast, dose increases were observed in posterior and deep-seated organs such as the kidneys and active bone marrow. Effective dose was reduced by about half with PA and further decreased with copper filtration.

Conclusion: PA projection and copper filtration are effective strategies for reducing radiation exposure to anterior radiosensitive organs and lowering effective dose in pediatric full-spine radiography. However, dose increases in deep-seated organs were also observed, highlighting the need for protocol optimization according to patient age and organ location.

Background: Synovial-joint abnormalities in children can be caused by different conditions, including autoimmune arthritis, infection, and neoplasm. An ultrasound-guided biopsy targeting the synovial membrane can aid in determining the etiology when the cause is unclear.

Objective: To determine the diagnostic performance, findings, and outcomes of ultrasound-guided joint biopsy in children.

Materials and method: This is a retrospective study on patients who underwent ultrasound-guided joint biopsy from May 2000 to December 2024. Patient demographics, clinical information, imaging, procedure details, pathology findings, adverse events, and clinical outcomes were collected and reviewed.

Results: Thirty-one patients (25 females) with a mean age of 10.2 years underwent 34 biopsies. Presenting symptoms were pain (33/34), mobility issues (33/34), and swelling (20/34). Effusion (19/28), joint capsule thickening (24/28), and contrast enhancement (20/28) were the most common MRI findings, while joint capsule thickening (29/29) and effusion (19/29) were the most frequent ultrasound findings. The most common joints biopsied were the hip (16/34), knee (9/34), and ankle (4/34). Core needle biopsy was performed in all cases. The mean number of passes was 4.5 (SD 1.8), obtaining a mean of 4.1 cores (SD 1.9). Biopsy was diagnostic in 20/34 (59% [CI 41-76%]) joints, and only one patient required surgical biopsy. Synovitis was the most common diagnosis (14/34), followed by pigmented villonodular synovitis (2/34). No major adverse events were observed.

Conclusion: Ultrasound-guided joint biopsy in children has moderate diagnostic performance; however, it can be clinically impactful, even when non-diagnostic, helping in joint disease management, potentially preventing surgery, with low adverse event incidence.

Background: Myelination is a key biomarker of healthy brain maturation, and its disruption can signal neurodevelopmental disorders.

Objective: The study aimed to enhance the accuracy and interpretability of brain age prediction in early infancy by incorporating the biological process of myelination as an attention mechanism into deep learning models.

Materials and methods: A fully automated deep learning framework, called myelination-attention-empowered model (MAENet), was developed through retrospective analysis of structural magnetic resonance imaging (sMRI) data from 603 participants who met the inclusion criteria, aged 0-2 years, collected in a local hospital between July 2017 and June 2024. The MAENet consisted of four modules: a multiscale information fusion channel (MSIF-channel) on the T2WI brain image, a myelination-empowered feature extraction channel (MEFE-channel) on an automated and standardized segmentation of the white matter image, a communication mechanism that enabled inter-channel information flow and enhanced the MSIF-channel's sensitivity to myelination-related features, and a myelination-attention mechanism that dynamically emphasized myelination-sensitive regions.

Results: The proposed MAENet model exhibited superior performance over multiple deep learning models, including ResNet-50, VGG, Inception, SFCN, Skewed, FiA-Net, and TSAN. The mean absolute error (MAE) between the predicted brain age and chronological age was significantly reduced by 18%-41% in the subgroup of 0-1-year-old infants, 25%-37% in the subgroup of 1-2-year-old infants, and 18%-40% in the whole group of 0-2-year-old infants in the experimental comparison (P < 0.05). The brain regions attended to by the MAENet model were visualized and consistent with the well-known developmental trajectories of white matter myelination in early infancy.

Conclusion: The MAENet model demonstrated a significant improvement in brain age prediction accuracy in 0-2-year-olds by effectively leveraging the developmental process of myelination.

Background: While single-energy hand computed tomography angiography (CTA) often yields suboptimal visualization of distal vessels, dual-energy computed tomography (CT) with low-keV virtual monoenergetic image (VMI) reconstruction enhances small-vessel conspicuity.

Objective: To evaluate the value of dual-energy CT VMIs in pediatric hand CTA.

Materials and methods: This retrospective study included 49 pediatric patients. Seven image series per patient were generated from dual-energy data: an M_0.5 image (50% 70 kVp+50% tin-filtered 150 kVp), a 70-kVp image, and five VMIs at 40-80 keV (10-keV increments). Objective metrics (attenuation, vessel noise, signal-to-noise ratio, contrast-to-noise ratio) and subjective scores were assessed for five vessels: the radial artery, the ulnar artery, the common palmar digital artery, and the proximal and distal parts of the proper palmar digital artery. Subjective image quality was independently evaluated by two radiologists using a 4-point Likert scale.

Results: The 40-keV VMIs provided the highest vascular attenuation across all vessels, albeit with the highest noise. Subjective scores for the radial, ulnar, and common palmar digital arteries showed no significant differences among the 40-keV, 50-keV, and 70-kVp series. However, for the small distal proper palmar digital arteries and total image quality, the 40-keV series was rated superior to the other series. No significant differences in image quality existed between the 70-kVp and 50-keV images.

Conclusion: For pediatric hand CTA, 40-keV VMIs provide optimal vascular conspicuity for small distal vessels, yielding the highest diagnostic confidence and total image quality score, and this benefit outweighs the associated increase in vessel noise.

Background: Extralobar pulmonary sequestration can undergo torsion within the pleural cavity, which represents a rare and the most severe complication in childhood. To date, no data have been published on the use of contrast-enhanced ultrasound (CEUS) in extralobar pulmonary sequestration with torsion.

Objective: The objective of this study was to retrospectively analyse the sonographic features of pediatric extralobar pulmonary sequestration with torsion on gray-scale ultrasound (US) and CEUS.

Materials and methods: A retrospective observational study was conducted in eight children with clinically and histologically confirmed extralobar pulmonary sequestration with torsion between January 2020 and September 2024. Gray-scale US findings were available for all eight cases, and CEUS features were obtained and reviewed in detail in four of these patients.

Results: All lesions were solitary, with a right-to-left ratio of 5:3. On gray-scale US, torsional extralobar pulmonary sequestration demonstrated a regular morphology and well-defined margins in all cases. Heterogeneous echotexture was observed in five cases, including cystic structures in two cases and linear branching structures in two cases. On CEUS, absence of enhancement in the early pulmonary arterial phase was identified in all four patients (100%). In the delayed bronchial arterial phase, stem-shaped enhancement confined to the base of the mass was observed in three patients (75%), including one case in which a feeding artery was visualised. Peripheral ring-shaped enhancement during the bronchial arterial phase was present in all four cases (100%). Other associated pulmonary findings included pleural effusion (8/8, 100%) and consolidation (4/8, 50%).

Conclusion: On gray-scale ultrasound, extralobar pulmonary sequestration with torsion typically appears as a well-defined mass with a regular shape in the lower thoracic cavity. On CEUS, stem-shaped enhancement at the base of the mass during the delayed bronchial arterial phase may represent a useful imaging feature for predicting extralobar pulmonary sequestration with torsion.

Data mining is the systematic process of extracting useful knowledge from large multimodal datasets and is increasingly enabled by artificial intelligence (AI) methods. Pediatric radiology is a natural field for data mining because multimodal data sources, including images, reports, metadata, and electronic health records, together capture rich information on anatomy, disease, treatment, and outcomes. In the current era, the boundaries between data mining and AI are increasingly blurred. AI assists in key steps of the mining workflow through automated labeling, information extraction, and representation learning, while data mining provides the high-quality curated datasets that underpin model performance, generalizability, and safety. This review, therefore, examines both domains together, emphasizing their interdependence in the pediatric context. We describe core concepts and workflows of data mining in pediatric radiology, including data collection, linkage, annotation, analysis, validation, and governance, and outline how modern AI tools such as deep learning, large language models, multimodal fusion, and federated learning support advanced pattern discovery across limited and heterogeneous pediatric datasets. We summarize current and emerging clinical applications across diagnosis, prognosis, radiation dose monitoring, operational analytics, reporting safety nets, and continual learning. We then discuss current challenges related to data quality and standardization, ethics, regulation, workflow integration, resource disparities, sustainability, and explainability. Finally, we highlight future perspectives, including synthetic data generation, foundation models, structured reporting, and pediatric-focused ethical frameworks that aim to enable safe, transparent, and equitable integration of AI-driven data mining to improve outcomes in children.