Multi‐Institutional Development and Validation of a Nomogram Based on New Microvascular Ultrasound Techniques to Predict the Severity of Renal Fibrosis in Chronic Kidney Disease
Yao Zhang, Jingchen Liang, Xingyue Huang, Wenjun Zhang, Wei Xu, Xinde Gong, Wen Liu, Changhua Zhou, Yuexiang Peng, Qing ZhouObjectives
To develop and validate a nomogram integrating ultrasound microvascular imaging (super‐resolution ultrasound [SRUS], superb microvascular imaging [SMI]), resistive index (RI), and clinical features for the non‐invasive assessment of renal fibrosis severity in chronic kidney disease (CKD).
Materials and Methods
A total of 208 CKD patients (141 in the training cohort and 67 in the external validation cohort) were included in the study. All patients underwent renal ultrasound and biopsy. Based on the pathological findings, patients were classified into mild and moderate‐to‐severe fibrosis groups. Candidate predictors were selected using least absolute shrinkage and selection operator (LASSO) regression, followed by correlation and multicollinearity analysis. A multivariable logistic regression model was used to construct a nomogram. Model performance was evaluated by receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA).
Results
The nomogram, incorporating SRUS, SMI, RI, hypertension, blood urea nitrogen (BUN), and estimated glomerular filtration rate (eGFR), showed good predictive performance. The area under the curve (AUC) for the nomogram was 0.944 (95% confidence interval [CI]: 0.908–0.981) in the training cohort and 0.907 (95% CI: 0.835–0.980) in the external validation cohort. Calibration analyses showed acceptable agreement between predicted and observed probabilities, with calibration intercepts/slopes of 0.003/1.002 in the training cohort and −0.296/0.767 in the external validation cohort.
Conclusion
This nomogram may provide a non‐invasive approach for assessing renal fibrosis severity and supporting risk stratification in patients with CKD.