Share：
Share this content in WeChat
X
[Chinese] [PDF] 10 1
• Clinical Article •
The value of nomogram model based on IVIM-MRI radiomics for the noninvasive assessment of renal fibrosis in chronic kidney disease
ZHA Tingting  JIANG Zhenxing  LIU Guoqiang  CHEN Ying  CHEN Jie  XING Wei 

DOI:10.12015/issn.1674-8034.2025.12.016.


[Abstract] Objective To develop and validate a nomogram based on intravoxel incoherent motion magnetic resonance imaging (IVIM-MRI) radiomics and clinical indicators for assessing the severity of renal fibrosis (RF) in patients with chronic kidney disease (CKD).Materials and Methods This was a case-control study. The clinical and imaging data from 132 CKD patients confirmed by renal biopsy at the First People's Hospital of Changzhou between September 2016 and July 2022 were retrospectively analyzed. The patients were randomly divided into a training set (n = 92) and a test set (n = 40) in a 7∶3 ratio. Based on the T score of the Oxford MEST-C classification, patients were grouped into a mild fibrosis group (T0, fibrosis ≤ 25%) and a moderate-to-severe fibrosis group (T1-T2, fibrosis > 25%). Clinical indicators showing significant differences between groups were selected for subsequent clinical modeling. All patients underwent IVIM-MRI before biopsy, and radiomic features were extracted from true diffusion coefficient (D)、pseudo-diffusion coefficient (D*) and perfusion fraction (f) maps. Feature selection was performed using the Mann-Whitney U test, Pearson correlation analysis, and least absolute shrinkage and selection operator regression. Four radiomics models (Rad_D, Rad_D*, Rad_f, and Rad_D_D*_f) and a clinical model were constructed using logistic regression (LR), random forest, and multilayer perceptron algorithms. The optimal-performing radiomics and clinical models were then integrated to build a nomogram. Model performance was evaluated using receiver operating characteristic (ROC) curve analysis, the DeLong test, decision curve analysis (DCA), and calibration curves.Results A total of 9, 8, 11, and 12 features were selected for the construction of the Rad_D, Rad_D*, Rad_f, and Rad_D_D*_f models, respectively. Among the four radiomics models, Rad_D_D*_f demonstrated the best performance in distinguishing between mild and moderate-to-severe RF. Among the three algorithms comparison, both the Rad_D_D*_f radiomics model and the clinical model achieved the highest diagnostic performance using the LR algorithm. The nomogram, combining the best-performing radiomics and clinical models, further improved diagnostic performance, with area under the curve (AUC) of 0.942 (95% CI: 0.896 to 0.989) and 0.820 (95% CI: 0.687 to 0.954) in the training and test sets, respectively. The DeLong test showed that the nomogram significantly outperformed the clinical model (P < 0.05). DCA and calibration curves confirmed that the nomogram provided higher net clinical benefit and good model calibration.Conclusions The nomogram integrating IVIM-MRI radiomics and clinical indicators enables noninvasive identification of RF severity in CKD patients, demonstrating potential clinical applicability. This tool may provide imaging-based support for the precise management and dynamic assessment of CKD.
[Keywords] renal fibrosis；chronic kidney disease；nomogram；radiomics；magnetic resonance imaging；intravoxel incoherent motion

ZHA Tingting   JIANG Zhenxing   LIU Guoqiang   CHEN Ying   CHEN Jie   XING Wei*  

Department of Radiology, Third Affiliated Hospital of Soochow University, Jiangsu Province Artificial Intelligence for Medical Images Engineering Research Center, Changzhou 213003, China

Corresponding author: XING W, E-mail: suzhxingwei@suda.edu.cn

Conflicts of interest   None.

Received  2025-07-31
Accepted  2025-12-06
DOI: 10.12015/issn.1674-8034.2025.12.016
DOI:10.12015/issn.1674-8034.2025.12.016.

[1]
VIVANTE A. Genetics of chronic kidney disease[J]. N Engl J Med, 2024, 391(7): 627-639. DOI: 10.1056/NEJMra2308577.
[2]
HUANG R S, FU P, MA L. Kidney fibrosis: from mechanisms to therapeutic medicines[J/OL]. Signal Transduct Target Ther, 2023, 8(1): 129 [2025-07-30]. https://pubmed.ncbi.nlm.nih.gov/36932062/. DOI: 10.1038/s41392-023-01379-7.
[3]
HOUBEN A J H M, MARTENS R J H, STEHOUWER C D A. Assessing microvascular function in humans from a chronic disease perspective[J]. J Am Soc Nephrol, 2017, 28(12): 3461-3472. DOI: 10.1681/ASN.2017020157.
[4]
ZHOU H, SI Y, SUN J T, et al. Effectiveness of functional magnetic resonance imaging for early identification of chronic kidney disease: A systematic review and network meta-analysis[J/OL]. Eur J Radiol, 2023, 160: 110694 [2025-07-30]. https://pubmed.ncbi.nlm.nih.gov/36642011/. DOI: 10.1016/j.ejrad.2023.110694.
[5]
ZHOU S P, WANG Q, ZHAI X, et al. Value of diffusion weighted imaging of intra-voxel incoherent motion in differentiating diabetic nephropathy from non-diabetic nephropathy in patients with type 2 diabetes[J]. Chin J Internal Medicine, 2023(11): 1288-1294. DOI: 10.3760/cma.j.cn112138-20230520-00265.
[6]
ZHA T T, CHEN J, YU S N, et al. Value of intravoxel incoherent motion magnetic resonance imaging in evaluating renal fibrosis in rabbits with renal artery stenosis[J]. Chin J Magn Reson Imag, 2023, 14(11): 128-135. DOI: 10.12015/issn.1674-8034.2023.11.021.
[7]
YU Y M, WANG W, WEN J Q, et al. Detection of renal allograft fibrosis with MRI: arterial spin labeling outperforms reduced field-of-view IVIM[J]. Eur Radiol, 2021, 31(9): 6696-6707. DOI: 10.1007/s00330-021-07818-9.
[8]
LIANG P, YUAN G J, LI S C, et al. Non-invasive evaluation of the pathological and functional characteristics of chronic kidney disease by diffusion kurtosis imaging and intravoxel incoherent motion imaging: comparison with conventional DWI[J/OL]. Br J Radiol, 2023, 96(1141): 20220644 [2025-07-30]. https://pubmed.ncbi.nlm.nih.gov/36400040/. DOI: 10.1259/bjr.20220644.
[9]
LIU J J, WANG R, QIU J X, et al. Investigation of renal perfusion and pathological changes in patients with acute kidney disease and tubulointerstitial nephritis using intravoxel incoherent motion and arterial spin labelling MRI: a prospective, observational study protocol[J/OL]. BMJ Open, 2024, 14(3): e076488 [2025-07-30]. https://pubmed.ncbi.nlm.nih.gov/38531564/. DOI: 10.1136/bmjopen-2023-076488.
[10]
MAO W, DING Y Q, DING X Q, et al. Capability of arterial spin labeling and intravoxel incoherent motion diffusion-weighted imaging to detect early kidney injury in chronic kidney disease[J]. Eur Radiol, 2023, 33(5): 3286-3294. DOI: 10.1007/s00330-022-09331-z.
[11]
PRASAD P V, LI L P, HACK B, et al. Quantitative blood oxygenation level dependent magnetic resonance imaging for estimating intra-renal oxygen availability demonstrates kidneys are hypoxemic in human CKD[J]. Kidney Int Rep, 2023, 8(5): 1057-1067. DOI: 10.1016/j.ekir.2023.02.1092.
[12]
LI S L, WANG F, SUN D. The renal microcirculation in chronic kidney disease: novel diagnostic methods and therapeutic perspectives[J/OL]. Cell Biosci, 2021, 11(1): 90 [2025-07-30]. https://pubmed.ncbi.nlm.nih.gov/34001267/. DOI: 10.1186/s13578-021-00606-4.
[13]
LAMBIN P, RIOS-VELAZQUEZ E, LEIJENAAR R, et al. Radiomics: extracting more information from medical images using advanced feature analysis[J]. Eur J Cancer, 2012, 48(4): 441-446. DOI: 10.1016/j.ejca.2011.11.036.
[14]
QIN X C, LIU X L, XIAO W H, et al. Interpretable deep-learning model based on superb microvascular imaging for noninvasive diagnosis of interstitial fibrosis in chronic kidney disease[J]. Acad Radiol, 2025, 32(5): 2730-2738. DOI: 10.1016/j.acra.2024.11.067.
[15]
TANG Y, LIU X L, ZHOU W, et al. Interpretable machine learning model based on superb microvascular imaging for non-invasive determination of crescent status of IgAN[J]. J Inflamm Res, 2024, 17: 5943-5955. DOI: 10.2147/JIR.S476716.
[16]
DENG X G, LIANG K, XIE C, et al. The value of nomogram based on ultrasound and elastography for renal fibrosis in patients with chronic kidney disease[J]. Chin J Ultrasound Med, 2024, 40(9): 1017-1020. DOI: 10.3969/j.issn.1002-0101.2024.09.017.
[17]
REN Y, YANG F, LI W W, et al. End-to-end CT radiomics-based pipeline for predicting renal interstitial fibrosis grade in CKD patients[J]. Acad Radiol, 2025, 32(6): 3464-3474. DOI: 10.1016/j.acra.2024.12.050.
[18]
BIAŁEK P, DOBEK A, FALENTA K, et al. Usefulness of radiomics and kidney volume based on non-enhanced computed tomography in chronic kidney disease: initial report[J]. Kidney Blood Press Res, 2025, 50(1): 161-170. DOI: 10.1159/000543305.
[19]
WANG Y W, XU F, HAN Q Y, et al. AI-based automatic estimation of single-kidney glomerular filtration rate and split renal function using non-contrast CT[J/OL]. Insights Imaging, 2025, 16(1): 84 [2025-07-30]. https://pubmed.ncbi.nlm.nih.gov/40192862/. DOI: 10.1186/s13244-025-01959-x.
[20]
Kidney Disease: Improving Global Outcomes CKD Work Group. KDIGO 2024 clinical practice guideline for the evaluation and management of chronic kidney disease[J]. Kidney Int, 2024, 105(4S): S117-S314. DOI: 10.1016/j.kint.2023.10.018.
[21]
TRIMARCHI H, BARRATT J, CATTRAN D C, et al. Oxford classification of IgA nephropathy 2016: an update from the IgA nephropathy classification working group[J]. Kidney Int, 2017, 91(5): 1014-1021. DOI: 10.1016/j.kint.2017.02.003.
[22]
HU W T, DAI Y M, LIU F, et al. Assessing renal interstitial fibrosis using compartmental, non-compartmental, and model-free diffusion MRI approaches[J/OL]. Insights Imaging, 2024, 15(1): 156 [2025-07-30]. https://pubmed.ncbi.nlm.nih.gov/38900336/. DOI: 10.1186/s13244-024-01736-2.
[23]
WONG C, LIU T, ZHANG C Y, et al. Preoperative detection of lymphovascular invasion in rectal cancer using intravoxel incoherent motion imaging based on radiomics[J]. Med Phys, 2024, 51(1): 179-191. DOI: 10.1002/mp.16821.
[24]
MA L D, LIAO S T, ZHANG X L, et al. Application of intravoxel incoherent motion in the prediction of intra-tumoral tertiary lymphoid structures in hepatocellular carcinoma[J]. J Hepatocell Carcinoma, 2025, 12: 383-398. DOI: 10.2147/JHC.S508357.
[25]
ZHENG Y Y, ZHANG H, CHEN H J, et al. Combined morphology and radiomics of intravoxel incoherent movement as a predictive model for the pathologic complete response before neoadjuvant chemotherapy in patients with breast cancer[J/OL]. Front Oncol, 2025, 15: 1452128 [2025-07-30]. https://pubmed.ncbi.nlm.nih.gov/40007999/. DOI: 10.3389/fonc.2025.1452128.
[26]
CHEN Z M, WANG Y L, GUNDA S T, et al. Integrating shear wave elastography and estimated glomerular filtration rate to enhance diagnostic strategy for renal fibrosis assessment in chronic kidney disease[J]. Quant Imaging Med Surg, 2024, 14(2): 1766-1777. DOI: 10.21037/qims-23-962.
[27]
CHEN C J, PAI T W, HSU H H, et al. Prediction of chronic kidney disease stages by renal ultrasound imaging[J]. Enterp Inf Syst, 2020, 14(2): 178-195. DOI: 10.1080/17517575.2019.1597386.

PREV The value of machine learning model constructed based on parameters of DCE-MRI and clinical risk factors in predicting the unstable state of rectal cancer microsatellites
NEXT Prediction of pathological grading in prostate adenocarcinoma based on multiparametric MRI habitat imaging
  


LINK


Tel & Fax: +8610-67113815    E-mail: editor@cjmri.cn