Abstract
Metabolic dysfunction-Associated Steatotic Liver Disease (MASLD) affects 30–50% of obese children, yet accurate non-invasive quantification remains challenging. While magnetic resonance imaging-proton density fat fraction (MRI-PDFF) represents the reference standard, its limited accessibility necessitates alternative approaches. Forty pediatric patients (age 12.16 ± 2.56 years) referred for MASLD were prospectively enrolled for same-day ultrasound radiofrequency (RF) data acquisition and MRI-PDFF examination. Two artificial intelligence (AI) approaches using multiple input combinations of RF data, ultrasound-guided attenuation parameters (UGAP), and clinical parameters were developed for non-invasive pediatric liver fat quantification: radiomics-based models and deep learning models. The best radiomics model (XGBoost) and the best deep learning model (Mod-MHDNet) achieved optimal performance with multimodal inputs (R2 = 0.81 and 0.76, respectively). Bland–Altman analysis demonstrated excellent agreement with MRI-PDFF, with a mean bias of < 0.4% points for both approaches. AI analysis of ultrasound RF data enables accurate and accessible quantification of pediatric liver fat, offering a practical alternative for MASLD evaluation.
Data availability
The datasets generated and analyzed during the current study include patient-level imaging data and clinical information, which are not publicly available due to institutional and ethical restrictions. De-identified data may be made available from the corresponding author (B.K.J.) upon reasonable request and with approval from the Korea University Ansan Hospital Institutional Review Board.
Code availability
The underlying code for this study is not publicly available but may be made available to qualified researchers on reasonable request from the corresponding author.
References
Estes, C., Razavi, H., Loomba, R., Younossi, Z. & Sanyal, A. J. Modeling the epidemic of nonalcoholic fatty liver disease demonstrates an exponential increase in burden of disease. Hepatology 67, 123–133 (2018).
Goldner, D. & Lavine, J. E. Nonalcoholic fatty liver disease in children: unique considerations and challenges. Gastroenterology 158, 1967–1983 e1961 (2020).
Mitsinikos, T., Mrowczynski-Hernandez, P. & Kohli, R. Pediatric nonalcoholic fatty liver disease. Pediatr. Clin. North. Am. 68, 1309–1320 (2021).
Schwimmer, J. B. et al. Prevalence of fatty liver in children and adolescents. Pediatrics 118, 1388–1393 (2006).
Schwimmer, J. B. et al. Magnetic resonance imaging and liver histology as biomarkers of hepatic steatosis in children with nonalcoholic fatty liver disease. Hepatology 61, 1887–1895 (2015).
Awai, H. I., Newton, K. P., Sirlin, C. B., Behling, C. & Schwimmer, J. B. Evidence and recommendations for imaging liver fat in children, based on systematic review. Clin. Gastroenterol. Hepatol. 12, 765–773 (2014).
Serai, S. D., Panganiban, J., Dhyani, M., Degnan, A. J. & Anupindi, S. A. Imaging modalities in pediatric NAFLD. Clin. Liver Disease. 17, 200–208 (2021).
Ferguson, A. E., Xanthakos, S. A. & Siegel, R. M. Challenges in screening for pediatric nonalcoholic fatty liver disease. Clin. Pediatr. (Phila). 57, 558–562 (2018).
Han, A. et al. Noninvasive diagnosis of nonalcoholic fatty liver disease and quantification of liver fat with radiofrequency ultrasound data using One-dimensional convolutional neural networks. Radiology 295, 342–350 (2020).
Idilman, I. S. et al. Hepatic steatosis: Quantification by proton density fat fraction with MR imaging versus liver biopsy. Radiology 267, 767–775 (2013).
Berardis, S. & Sokal, E. Pediatric non-alcoholic fatty liver disease: An increasing public health issue. Eur. J. Pediatr. 173, 131–139 (2014).
Younossi, Z. et al. Global burden of NAFLD and NASH: Trends, predictions, risk factors and prevention. Nat. Rev. Gastroenterol. Hepatol. 15, 11–20 (2018).
Azizi, N. et al. Evaluation of MRI proton density fat fraction in hepatic steatosis: A systematic review and meta-analysis. Eur. Radiol. 35, 1794–1807 (2025).
D’Hondt, A., Rubesova, E., Xie, H., Shamdasani, V. & Barth, R. A. Liver fat quantification by ultrasound in children: A prospective study. AJR Am. J. Roentgenol. 217, 996–1006 (2021).
Noureddin, M. et al. Utility of magnetic resonance imaging versus histology for quantifying changes in liver fat in nonalcoholic fatty liver disease trials. Hepatology 58, 1930–1940 (2013).
Dillman, J. R., Thapaliya, S., Tkach, J. A. & Trout, A. T. Quantification of hepatic steatosis by ultrasound: Prospective comparison with MRI proton density fat fraction as reference standard. AJR Am. J. Roentgenol. 219, 784–791 (2022).
Hernaez, R. et al. Diagnostic accuracy and reliability of ultrasonography for the detection of fatty liver: A meta-analysis. Hepatology 54, 1082–1090 (2011).
Dardanelli, E. P. et al. Ultrasound Attenuation imaging: A reproducible alternative for the noninvasive quantitative assessment of hepatic steatosis in children. Pediatr. Radiol. 53, 1618–1628 (2023).
Fetzer, D. T. et al. US quantification of liver fat: Past, present, and future. Radiographics 43, e220178 (2023).
Frankland, M. P. et al. Diagnostic performance of ultrasound hepatorenal index for the diagnosis of hepatic steatosis in children. Pediatr. Radiol. 52, 1306–1313 (2022).
Jeon, S. K. et al. Quantitative ultrasound radiofrequency data analysis for the assessment of hepatic steatosis using the controlled Attenuation parameter as a reference standard. Ultrasonography 40, 136–146 (2021).
Park, J., Lee, J. M., Lee, G., Jeon, S. K. & Joo, I. Quantitative evaluation of hepatic steatosis using advanced imaging techniques: Focusing on new quantitative ultrasound techniques. Korean J. Radiol. 23, 13–29 (2022).
Polti, G. et al. Quantitative ultrasound fatty liver evaluation in a pediatric population: Comparison with magnetic resonance imaging of liver proton density fat fraction. Pediatr. Radiol. 53, 2458–2465 (2023).
Serai, S. D., Dhyani, M., Srivastava, S. & Dillman, J. R. MR and ultrasound for liver fat assessment in children: Techniques and supporting evidence. J. Magn. Reson. Imaging. 62, 691–706 (2025).
Oelze, M. L. & Mamou, J. Review of quantitative ultrasound: Envelope statistics and backscatter coefficient imaging and contributions to diagnostic ultrasound. IEEE Trans. Ultrason. Ferroelectr. Freq. Control. 63, 336–351 (2016).
Tahmasebi, A. et al. Ultrasound-based machine learning approach for detection of nonalcoholic fatty liver disease. J. Ultrasound Med. 42, 1747–1756 (2023).
Xie, Z., Ji, N., Xu, L. & Ma, J. Ultrasound radiofrequency image improves the tissue segmentation performance of deep learning models. In 2024 IEEE Ultrasonics, Ferroelectrics, and Frequency Control Joint Symposium (UFFC-JS) 1–3 (2024).
Yin, C. et al. Artificial intelligence in imaging for liver disease diagnosis. Front. Med. (Lausanne). 12, 1591523 (2025).
Mansournia, M. A., Waters, R., Nazemipour, M., Bland, M. & Altman, D. G. Bland-Altman methods for comparing methods of measurement and response to criticisms. Glob Epidemiol. 3, 100045 (2021).
Xie, Z., Han, J., Ji, N., Xu, L. & Ma, J. RFImageNet framework for segmentation of ultrasound images with spectra-augmented radiofrequency signals. Ultrasonics 146, 107498 (2025).
Ferraioli, G. et al. Noninvasive assessment of liver steatosis in children: The clinical value of controlled Attenuation parameter. BMC Gastroenterol. 17, 61 (2017).
Vos, M. B. et al. NASPGHAN clinical practice guideline for the diagnosis and treatment of nonalcoholic fatty liver disease in children: Recommendations from the expert committee on NAFLD (ECON) and the North American society of pediatric gastroenterology, hepatology and nutrition (NASPGHAN). J. Pediatr. Gastroenterol. Nutr. 64, 319–334 (2017).
Hunter, A. K. & Lin, H. C. Review of clinical guidelines in the diagnosis of pediatric nonalcoholic fatty liver disease. Clin. Liver Disease. 18, 40–44 (2021).
Zhang, L. et al. An international multidisciplinary consensus on pediatric metabolic dysfunction-associated fatty liver disease. Med 5, 797–815e792 (2024).
Mouzaki, M. et al. Assessment of nonalcoholic fatty liver disease progression in children using magnetic resonance imaging. J. Pediatr. 201, 86–92 (2018).
Clemente, M. G., Mandato, C., Poeta, M. & Vajro, P. Pediatric non-alcoholic fatty liver disease: Recent solutions, unresolved issues, and future research directions. World J. Gastroenterol. 22, 8078–8093 (2016).
Lin, G. et al. Epidemiology and lifestyle survey of non-alcoholic fatty liver disease in school-age children and adolescents in Shenyang, Liaoning. BMC Pediatr. 22, 286 (2022).
Cao, W. et al. Application of deep learning in quantitative analysis of 2-Dimensional ultrasound imaging of nonalcoholic fatty liver disease. J. Ultrasound Med. 39, 51–59 (2020).
Zamanian, H., Mostaar, A., Azadeh, P. & Ahmadi, M. Implementation of combinational deep learning algorithm for Non-alcoholic fatty liver classification in ultrasound images. J. Biomed. Phys. Eng. 11, 73–84 (2021).
Davis, L. M. et al. Ultrasound innovations in abdominal radiology: Techniques and clinical applications in pediatric imaging. Abdom. Radiol. (NY). 50, 1744–1762 (2025).
Zhao, Y. et al. Reproducibility of ultrasound-guided Attenuation parameter (UGAP) to the noninvasive evaluation of hepatic steatosis. Sci. Rep. 12, 2876 (2022).
Mayerhoefer, M. E. et al. Introduction to radiomics. J. Nucl. Med. 61, 488–495 (2020).
Rizzo, S. et al. Radiomics: The facts and the challenges of image analysis. Eur. Radiol. Exp. 2, 36 (2018).
van Griethuysen, J. J. M. et al. Computational radiomics system to Decode the radiographic phenotype. Cancer Res. 77, e104–e107 (2017).
Novitasari, D. C. R., Lubab, A., Sawiji, A. & Asyhar, A. H. Application of feature extraction for breast cancer using one order Statistic, GLCM, GLRLM, and GLDM. Adv. Sci. Technol. Eng. Syst. J. 4, 115–120 (2019).
Hanchuan, P., Fuhui, L. & Ding, C. Feature selection based on mutual information criteria of max-dependency, max-relevance, and min-redundancy. IEEE Trans. Pattern Anal. Mach. Intell. 27, 1226–1238 (2005).
Tibshirani, R. Regression shrinkage and selection via the Lasso. J. Roy. Stat. Soc.: Ser. B (Methodol.). 58, 267–288 (2018).
Chen, T., Guestrin, C. & Xgboost A scalable tree boosting system. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining 785–794 (2016).
Awad, M. & Khanna, R. Support vector regression. In Efficient Learning Machines: Theories, concepts, and Applications for Engineers and System Designers 67–80 (Springer, 2015).
Su, X., Yan, X. & Tsai, C. L. Linear regression. Wiley Interdiscip. Rev.: Comput. Stat. 4, 275–294 (2012).
Kumar, D. EfficientNet-B4 based deep learning for automated paddy leaf disease classification. In 2024 International Conference on Integration of Emerging Technologies for the Digital World (ICIETDW) 1–6 (IEEE, 2024).
Tan, M., Le, Q. & Efficientnet Rethinking model scaling for convolutional neural networks. In International conference on machine learning 6105–6114 (PMLR, 2019).
Tang, P. et al. Deep learning with multi-scale Temporal hybrid structure for robust crop mapping. ISPRS J. Photogrammetry Remote Sens. 209, 117–132 (2024).
Wang, J. & Wang, J. MHDNet: A multi-scale hybrid deep learning model for person Re-Identification. Electronics 13, 1435 (2024).
Chai, T. & Draxler, R. R. Root mean square error (RMSE) or mean absolute error (MAE)?–Arguments against avoiding RMSE in the literature. Geosci. Model Dev. 7, 1247–1250 (2014).
Chicco, D., Warrens, M. J. & Jurman, G. The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation. Peerj Comput. Sci. 7, e623 (2021).
Rosner, B., Glynn, R. J. & Lee, M. L. The Wilcoxon signed rank test for paired comparisons of clustered data. Biometrics 62, 185–192 (2006).
Acknowledgements
This study was supported by Korea University Ansan Hospital Grant (K2409221, O2310661).
Funding
This study was funded by Korea University Ansan Hospital (grant numbers O2310661 and K2409221).
Author information
Authors and Affiliations
Contributions
G.C. and S.H. designed the study, performed data processing and analysis, conducted statistical analysis, and drafted and revised the manuscript. M.S., G.C., and S.H. collected data. G.C., B.K.J., and S.H. contributed to the literature review and manuscript editing. All authors reviewed and approved the final manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Choi, G., Ham, S., Je, BK. et al. AI-powered ultrasound radiofrequency analysis for non-invasive pediatric liver fat quantification. Sci Rep (2026). https://doi.org/10.1038/s41598-026-37862-8
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-026-37862-8
