RPS 1912 - Hot Topic: AI in paediatric radiology

Author Block: K. Saravanan¹, F. Abubacker Sulaiman², R. Praveenkumar², J. Lydia², D. Velan²; ¹Melmaruvathur/IN, ²Chennai/IN
Purpose: To develop and validate an artificial intelligence (AI)–based radiomics model integrating diffusion-weighted imaging (DWI) and T2-weighted MRI features of the placenta for early prediction of preeclampsia (PE) and fetal growth restriction (FGR). The objective is to identify microstructural and textural biomarkers preceding clinical manifestation, enabling proactive obstetric management and improved perinatal outcomes.
Methods or Background: This prospective study included 80 pregnant women (20–32 weeks gestation) who underwent placental MRI on a 3-Tesla scanner. DWI (b-values 0, 800 s/mm²) and high-resolution T2 sequences were analyzed. Placental volumes were segmented semi-automatically, and radiomic features (first-order, texture, and shape) were extracted. Machine learning models—including Random Forest and Gradient Boosting—were trained to classify risk for PE/FGR, with cross-validation and feature selection via LASSO regression.

Clinical parameters (blood pressure, uterine artery Doppler indices) were integrated into a multimodal model for comparison.
Results or Findings: The AI-radiomics model demonstrated an AUC of 0.93 for predicting PE and 0.91 for FGR, outperforming conventional Doppler parameters (AUC 0.76). Key discriminative features included gray-level co-occurrence entropy and ADC histogram kurtosis, reflecting placental microstructural heterogeneity. Combined clinical-radiomics integration improved specificity and early detection (mean 5 weeks before clinical onset).
Conclusion: AI-enhanced placental radiomics from DWI and T2 MRI enables noninvasive early prediction of preeclampsia and FGR with high accuracy. This approach holds promise for precision obstetric imaging, facilitating timely intervention and reducing perinatal morbidity.
Limitations: Single-center design, limited sample size, and variability in placental segmentation may restrict generalizability. Lack of histopathological correlation and absence of external validation cohorts warrant further multicentric studies for clinical translation.
Funding for this study: Not Applicable
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: Z. Bingjie, Y. X. Li, J. Qu, C. Xu, X. Chen, Y. Wu; Zheng Zhou/CN
Purpose: To evaluate the diagnostic performance of deep learning-accelerated whole-body MRI without contrast for treatment monitoring in pediatric lymphoma patients compared with 18F-FDG PET/CT.
Methods or Background: In this multicenter prospective study conducted across five institutions, 176 children (aged 0-14 years; 60% male) with histopathologically confirmed Hodgkin lymphoma (HL, 65%) or non-Hodgkin lymphoma (NHL, 35%) underwent deep learning-accelerated whole-body MRI without contrast and 18F-FDG PET/CT at baseline, after induction chemotherapy, and at end of therapy. MRI scans, performed on 3T scanners, utilized a proprietary deep learning reconstruction algorithm, with its performance evaluated for image quality and scan time reduction compared to standard MRI protocols. Apparent diffusion coefficient (ADC) from MRI was correlated with PET/CT standardized uptake value (SUV). Statistical analysis included Gwet’s AC for agreement and Pearson’s correlation for quantitative metrics.
Results or Findings: Deep learning-accelerated MRI without contrast achieved 95% sensitivity and 92% specificity for lesion detection compared to PET/CT (Gwet’s AC = 0.94 [0.89, 0.97]). Therapy response assessment showed 94% concordance (Gwet’s AC = 0.94). The deep learning algorithm reduced scan time by 80% without compromising image quality (signal-to-noise ratio equivalent to standard MRI). ADC values strongly correlated with SUV (r² = 0.94) for treatment response. Subgroup analysis showed excellent agreement at end of therapy (Gwet’s AC = 0.97), with higher concordance for HL (Gwet’s AC = 0.94) than NHL (Gwet’s AC = 0.66) after induction chemotherapy.
Conclusion: Deep learning-accelerated whole-body MRI without contrast offers a fast, radiation-free, contrast-free alternative to PET/CT for pediatric lymphoma treatment monitoring, with comparable diagnostic accuracy and significantly reduced scan times.
Limitations: The lower concordance in therapy response assessment for non-Hodgkin lymphoma compared to Hodgkin lymphoma after induction chemotherapy, potentially affecting diagnostic reliability in certain subtypes.
Funding for this study: This study was funded by the Henan Province Medical Science and Technology Research Program (Grand No.20233526)
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved by the Institutional Review Board of Henan Cancer Hospital (approval number: 2023110712). Written informed consent was obtained from all participants prior to enrollment.

Author Block: B. Wichtlhuber¹, E. Frodl¹, M. Sayed¹, T. Persigehl², M. Neitzel¹, J. Dietz¹, M. Eicke³, D. M. Renz³, A. M. Bucher¹; ¹Frankfurt/DE, ²Koeln/DE, ³Hannover/DE
Purpose: Paediatric lymph node segmentation faces critical data scarcity challenges. We hypothesised that incorporating adult CT data into deep learning training would overcome this limitation while maintaining clinical relevance for paediatric oncology applications, particularly for detecting pathologically enlarged nodes requiring follow-up.
Methods or Background: We analysed 418 head/neck CTs from University Hospital Frankfurt: 146 paediatric (<18 years) and 272 adult cases. A 3D full-resolution nnU-Net underwent five-fold cross-validation training (250 epochs) using three strategies: paediatric-only, adult-only, and combined training. Approximately 20% of cases were reserved for independent testing. Evaluation metrics included the Dice coefficient, Intersection over Union (IoU). Clinical relevance was assessed through volumetric coverage analysis, with particular focus on nodes >10mm short-axis diameter—the threshold for pathological enlargement requiring clinical action. Values are presented as median and standard deviation.
Results or Findings: Cross-age training outperformed paediatric-only approaches. The combined model achieved Dice=0.714 ±0.151 and IoU=0.575 ±0.176 on paediatric test data versus Dice=0.695 ±0.155 and IoU=0.553 ±0.179 for paediatric-only training, showing clear improvement. Interestingly, this benefit was unidirectional—paediatric patients gained from mixed training, whereas adult test performance remained unchanged (combined model Dice=0.643 ±0.125 vs. adult-only Dice=0.647 ±0.113). Clinical utility analysis showed strong performance: 88.97% of enlarged nodes (>10mm) achieved ≥10% volumetric coverage, ensuring reliable oncological detection. Moreover, 69.08% of all lymph nodes reached ≥50% coverage, supporting accurate volumetric measurements. These metrics translate into improved workflow efficiency, reducing missed findings while preserving precision for follow-up.
Conclusion: Cross-age training addresses the key challenge of paediatric data scarcity in medical imaging AI, improving technical metrics and enabling clinically meaningful detection of pathological lymphadenopathy. This approach supports robust AI use in paediatric radiology with limited annotated data, potentially accelerating adoption in oncology workflows.
Limitations: A limitation of the study is its single-center validation.
Funding for this study: Funding was provided by the German Federal Ministry of Education and Research through the RACOON project (reference number 01KX2021).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethics approval was obtained by University Medicine Frankfurt (Reference 2023-1459)

Author Block: A. F. Ibil, M. Faraşat, M. Özkol; Izmir/TR
Purpose: Superb Microvascular Imaging(SMI) is an advanced Doppler technique that allows detailed visualization of small, slow-flow vessels without contrast agents. This study aimed to assess the impact of antibiotic therapy on lymph node morphology and hemodynamics in lymphadenitis using SMI.
Methods or Background: Patients admitted to our hospital’s pediatric clinic were evaluated. SMI, Advanced Dynamic Flow(ADF) and Power Doppler(PD) US were applied to measure the Vascular Index(VI) within defined regions of interest(ROI) for comparison and diagnostic enhancement, both before and after antibiotics in lymphadenitis. For comparisons of quantitative variables that did not show normal distribution between two groups, the Wilcoxon test was used. The McNemar test and Pearson’s chi-square test were applied for the comparison of qualitative data. Statistical analyses were performed using SPSS Software version26.0.
Results or Findings: A total of 17 patients and 65 lymph nodes were evaluated in the study population. Wilcoxon analysis showed significant post-antibiotic vascular index reductions in PD cSMI, and mSMI, while ADF demonstrated a significant increase(all p<0.05). Wilcoxon analysis showed significantly reduced post-antibiotic vascular scores in SMI and PD(p<0.05), while ADF changes were not statistically significant (p>0.05), despite negative Z-scores indicating lower post-treatment values.
Conclusion: Our study demonstrated that combining PD, SMI, ADF, and gray-scale US findings may be diagnostically valuable for assessing the efficacy of antibiotic therapy and evaluating lymphadenitis. Recognizing typical features of lymphadenitis may also aid in early differentiation from malignancy and reduce the need for histopathology. As studies investigating ultrasound changes in pediatric lymphadenitis following antibiotic treatment are scarce, our work may serve as a reference.
Limitations: Limitations included patient compliance in young children, exclusion of irregular antibiotic use or follow-up, observer dependency in qualitative assessments, and the small number of pediatric patients receiving planned antibiotic therapy.
Funding for this study: This study received no financial support.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethical approval was obtained from the Clinical Research Ethics Committee of Manisa Celal Bayar University (decision no: 20.478.486/2322).

Author Block: L. Ibrahim, A. Ilivitzki, M. Freiman; Haifa/IL
Purpose: To investigate whether radiomic features from post–first-cycle PET and CT scans, alone or combined with clinical variables, improve the prediction of event-free survival (EFS) in pediatric high-risk Hodgkin lymphoma (HL, stage IIIB/IVB), compared with a baseline model incorporating clinical variables and radiological PET assessment.
Methods or Background: Outcome prediction in pediatric HL traditionally relies on risk-adapted protocols combining clinical factors with interim FDG PET/CT, mainly focused on the isotopic PET component. Prior evidence suggests that CT provides additional prognostic information, underscoring the importance of anatomical features.
In this study, post–first-cycle PET/CT scans from 137 patients enrolled in the Children’s Oncology Group AHOD0831 trial (NCT01026220) were analyzed. Up to five lesions per patient were segmented; PET images were SUV-standardized, and radiomic features from PET and CT were extracted and averaged using tumor volume weighting. Clinical variables included demographics, histology, stage, bulky disease, and radiological PET response. Machine learning models were trained with stratified 5-fold cross-validation, and performance metrics were averaged across folds
Results or Findings: EFS events occurred in 27 patients (19.7%). CT radiomics outperformed the baseline clinical + PET model (accuracy: 0.78 ± 0.10; balanced accuracy: 0.71 ± 0.09 vs. 0.70 ± 0.11; 0.66 ± 0.04) and exceeded PET radiomics (accuracy: 0.77 ± 0.07; balanced accuracy: 0.66 ± 0.07). The strongest predictive performance was achieved when CT radiomics were combined with clinical variables.
Conclusion: Radiomics from CT improved prognostic prediction beyond the standard clinical + PET model and outperformed PET radiomics, reinforcing the anatomical contribution to risk stratification. CT radiomics represent a promising complementary tool for early prognostic assessment in pediatric high-risk HL.
Limitations: This study was based on internal cross-validation within a single trial cohort (AHOD0831), without external or temporal validation, which may limit generalizability.
Funding for this study: None
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: N. XU; Shenzhen/CN
Purpose: This study aims to develop and validate a deep convolutional neural network algorithm, named HipSonoNeuNet model (HSNN), using multicenter hip ultrasound data.
Methods or Background: This multicenter cross-sectional study combined data from 22 Chinese hospitals, enrolling 3082 participants. A total of 7286 hip ultrasound images (1429 dynamic, 5857 static) were collected and were divided into three datasets. The study was conducted in three phases. Phase I trained the models using 2431 participants. Phase II compared diagnostic performance between radiologists of varied experience and the model across 500 participants. Phase III prospectively validated the model's generalizability with 151 participants .
Results or Findings: In Phase I, the HSNN yielded AUC of 0.99 (95%CI: 0.99-1.00), sensitivity of 1.00 (95% CI: 0.99-1.00), specificity of 0.91 (95% CI: 0.88-1.00), F1 score of 0.90 (95% CI: 0.87-1.00) on internal test dataset. In Phase II, the HSNN achieved an accuracy of 0.94 (95% CI: 0.88-1.00), AUC of 0.99 (95%CI: 0.99-1.00), sensitivity of 1.00 (95% CI: 0.99-1.00), specificity of 0.94 (95% CI: 0.87-1.00), F1 score of 0.58 (95% CI: 0.50-0.66), and strong agreement with expert (κ = 0.77). AI assistance improved all 7 junior radiologists' diagnostic performance (accuracy from 0.90 to 0.93, AUC from 0.80 to 0.95, sensitivity from 0.69 to 0.97) and reduced examination time with enhanced interobserver agreement. In Phase III, the model maintained robust performance (accuracy = 0.92, AUC = 0.99, sensitivity = 1.00, κ with experts = 0.76).
Conclusion: The HSNN demonstrates accurate, robust, and generalizable performance in DDH detection. It might potentially enhance diagnostic capabilities for radiologists, particularly in hospitals with varying levels of expertise.
Limitations: 1.DDH image imbalance may reduce model prediction stability.
2.China-only US images limit model performance across regions/ethnicities, affected by culture, genetics and healthcare resources.
Funding for this study: 1.Guangdong High-level Hospital Construction Fund(SZGSP012).
2.Shenzhen Clinical Research Center(20220819113341005)“Shenzhen Clinical Research Center for Child Health and Disease(szcrc2024_005)”
3.Guangdong Medical Research Funded Project (A2024019)
4.Shenzhen Science and Technology Innovation Commission General Program for Basic Research(JCYJ20220530160000001)
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study protocol was approved by the Ethics Committee of Shenzhen Children's Hospital (Approval No. 202308602)

Author Block: I. Bianco, D. Ippolito, C. Maino, C. R. G. L. O. M. Talei Franzesi, P. N. Franco, D. G. Gandola, R. Corso; Monza/IT
Purpose: To compare image quality and radiation dose between deep learning reconstruction (DLR) and hybrid iterative reconstruction (HIR) algorithms in unenhanced chest CT scans of pediatric patients.
Methods or Background: We retrospectively reviewed 142 unenhanced, single-phase chest CT scans performed for routine diagnostic purpose. 71 pediatric patients were examined using a 128-slice MDCT scanner (100kV) with a DLR algorithm (Precise Image), while the control group of 71 patients underwent scanning with a 256-slice MDCT scanner (100kV) using HIR algorithm (iDose4). Subjective image quality was assessed using a 5-point Likert scale, and objective quality was evaluated by measuring Hounsfield Units (HU) and Standard Deviations (SD) in lung parenchyma, tracheal lumen, air, aorta and muscle. Radiation dose metrics (CTDIvol and DLP) were recorded for both groups, and the estimated effective dose (EED) was calculated using age-specific k-factors.
Results or Findings: A total of 142 pediatric patients were included (median age:10 years).Inter-reader agreement for Likert scale image quality assessment was moderate (κ=0.432) with significantly higher subjective image quality scores for DLR group compared to HIR group (p=0.03).Quantitatively, air and muscle HU differed significantly between the two scanners (air, p=0.036; muscle, p<0.001). Image noise, assessed by SD was significantly lower in DLR group across the lung parenchyma (p=0.007), tracheal lumen (p=0.002) and muscle (p<0.001). Radiation dose metrics were significantly reduced in DLR group compared to control group (mean DLP: 49.6 vs 188.9 mGy·cm;CTDIvol: 1.6 vs 5.6 mGy;EED:0.8 vs 3.2 mSv;all p<0.001).Age-based sub-analysis using quartiles (0–5,6–12,>13 years) confirmed a consistent threefold reduction in DLP and CTDIvol, and a fourfold reduction in EED across all age groups (all p<0.001).
Conclusion: DLR in unenhanced chest CT allows significant radiation dose reduction in pediatric population while delivering superior image quality compared to HIR algorithms.
Limitations: Monocentric study
Funding for this study: None
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: