A novel diagnostic model to grade the impairment of split renal function for patients with obstructive hydronephrosis based on enhanced CT imaging

To establish a new diagnostic model to predict the split renal function impairment in patients with obstructive hydronephrosis. A set of retrospective data was analyzed, including enhanced CT of 382 kidney data from 191 patients with obstructive hydronephrosis collected from 2017 to 2024. These kidneys were divided into renal dysfunction and non-renal dysfunction group. Risk factors for renal dysfunction were identified using logistic regression analysis, and a new diagnostic model was established and then the ROC curve was used to evaluate the effect of the diagnostic model to diagnose in diagnosing renal function impairment. In the training set, renal cortex volume, venous phase renal medulla CT values (VP-HuRM), and hydronephrosis were identified as independent risk factors for renal function impairment with odds ratios of 0.959 (0.946–0.972), 0.987 (0.977–0.996), and 4.625 (2.110–10.138), respectively. The diagnostic accuracy of the model (AUC) for detecting renal function impairment was 0.896 (0.858–0.934), with the Cut-off value of-0.124. The AUC for distinguishing between the non-renal dysfunction group and mild renal dysfunction group, and between severe renal dysfunction group and mild renal dysfunction group were: 0.852 (0.801–0.903), 0.848 (0.780–0.916), respectively. The AUCs in the validation set were: 0.928 (0.882–0.973), 0.885 (0.815–0.954), 0.886 (0.797–0.975), respectively. The diagnostic model based on CT with enhancement has certain clinical value in predicting the degree of split renal function impairment, aiding in the clinical accurate diagnosis of split glomerular filtration rate (sGFR) in patients with obstructive hydronephrosis and holding significant clinical value.