狼疮性肾炎
对比度(视觉)
医学
超声波
系统性红斑狼疮
超声造影
人工智能
计算机科学
放射科
模式识别(心理学)
内科学
疾病
作者
Shuping Wei,Yidan Zhang,Chunrui Liu,Baojie Wen,Jing Yao,Zhichao Xia,Xue Xu,Zhibin Jin
出处
期刊:Ndt Plus
[Oxford University Press]
日期:2025-10-08
摘要
Abstract Background Lupus nephritis (LN) is a severe complication of systemic lupus erythematosus (SLE), with renal biopsy as the diagnostic gold standard. However, biopsy is invasive. This study aims to evaluate the potential of contrast-enhanced ultrasound (CEUS) quantitative parameters as non-invasive predictors in differentiation proliferative LN from non-proliferative LN. Methods Fifty-eight biopsy-confirmed LN patients who underwent CEUS within three days before biopsy were included retrospectively. Patients were categorized into 38 cases of proliferative LN (Class III, IV, and III/IV + V) and 20 cases of non-proliferative LN (Class II and purely Class V). The clinical and laboratory data, conventional US parameters, and CEUS quantitative parameters derived from time-intensity curves (TICs) were analyzed. Multivariate logistic regression and receiver operating characteristic (ROC) curve analysis were performed to determine significant predictors and evaluate the diagnostic performance. Results Patients with proliferative LN exhibited significantly higher absolute Time to Peak(∆TTP), Half Descending Time (DT/2) and TIC-Area Under Curve (TIC-AUC) values than non-proliferative LN patients (P < 0.05). Logistic regression analysis identified TIC-AUC and anti-dsDNA antibody as independent predictors of proliferative LN. ROC analysis revealed that anti-dsDNA positive had an AUC of 0.745, with sensitivity of 87.5% and specificity of 61.5% for predicting proliferative LN. For TIC-AUC, a cutoff value of 8049.0 yielded an AUC of 0.810, sensitivity of 68.8%, and specificity of 84.6% for predicting proliferative LN. Conclusions CEUS quantitative parameters, particularly TIC-AUC, provide a non-invasive approach for identifying proliferative LN, and complement conventional laboratory markers. These findings demonstrate the potential of CEUS in improving LN diagnosis and facilitating clinical evaluation.
科研通智能强力驱动
Strongly Powered by AbleSci AI