Nomogram Model for Predicting Early Onset of Chronic Kidney Disease Using Color Doppler Region of Interest Technique

Abstract

Purpose: The risk factors of chronic kidney disease were analyzed by using the region of interest quantitative technology of color doppler combined with QLab software, and a Nomogram was established to conduct an individualized assessment of patients with chronic kidney disease.Methods: A total of 500 patients with chronic kidney disease diagnosed in our hospital from June 2019 to March 2021 were selected as the chronic kidney disease group, and 300 healthy patients during the same period were selected as the control group. Univariate analysis was performed on the test indexes(Fasting blood glucose, total cholesterol, triglyceride, urea nitrogen , creatinine, uric acid, albumin, red blood cell count, glomerular filtration rate, urinary protein) and the vascularity index, flow index, and vascularization flow index measured by the color doppler region of interest quantitative technique. The above meaningful indicators were included in the Logistics regression analysis to obtain the independent risk factors of early chronic kidney disease. The independent risk factors were imported into R software to draw a nomogram model for predicting early chronic kidney disease and evaluate the model.Results: Single factor analysis results suggest age, hypertension, diabetes, hyperlipidemia, disease of heart head blood-vessel, body mass index, vascularity index, flow index, and vascularization flow index, fasting blood sugar, triglyceride, total cholesterol, urea nitrogen, creatinine, uric acid, glomerular filtration rate differences statistically significant (P < 0.05), while gender, renal length to diameter, the thickness of the cortex, Resistance index and peak systolic velocity and albumin difference has no statistical significance (P > 0.05). Logistics regression analysis showed that hypertension, diabetes, flow index, and vascularization flow index, urea nitrogen and albumin were independent risk factors for the early occurrence of chronic kidney disease. The C index of this nomogram using independent risk factors is 0.896 (95%CI: 0.862-0.930), which indicates that the nomogram has good discriminant power. The receiver operating curve of the histograph was AUC(area under the curve) 0.884 (95%CI: 0.860-0.908), with the optimal threshold of 0.663, specificity of 88.7%, sensitivity of 78.0%, accuracy of 82.0%, and positive predictive value of 91.9%. The ROC(receiver operator characteristic curve) of urea nitrogen, albumin ,flow index, and vascularization flow index were evaluated. The results indicated that the best cutoff value of urea nitrogen was 5.9mmol/L, flow index was 14.67, vascularization flow index was 4.6, and albumin was 40.26g/L. Conclusion: In the prediction of chronic kidney disease I-II stage, the quantitative technique of color Doppler region of interest has certain diagnostic value. The model established in this study has good discriminative power and can be applied to clinical practice, giving certain indicative significance.