The mean Hounsfield unit range acquired from different slices produces superior predictive accuracy for pyonephrosis in obstructive uropathy

INVESTIGATIVE AND CLINICAL UROLOGY(2024)

引用 0|浏览0
暂无评分
摘要
Purpose: To determine the non-contrast computer tomography imaging features of pyonephrosis and evaluate the predictive value of Hounsfield units (HUs) in different hydronephrotic region slices. Materials and Methods: We retrospectively reviewed data from patients with hydronephrosis who had renal-ureteral calculi. All patients were categorized into pyonephrosis and simple hydronephrosis groups. Baseline characteristics, the mean HU values in the maximal hydronephrotic region (uHU) slice, and the range of uHU in different slices (Delta uHU) were compared between the two groups. Univariate and multivariate analyses were performed to identify risk factors for pyonephrosis. Results: Among the 181 patients enrolled in the current study, 71 patients (39.2%) were diagnosed with pyonephrosis. The mean dilated pelvis surface areas were comparable between patients with pyonephrosis and simple hydronephrosis (822.61 mm(2) vs. 877.23 mm(2), p=0.722). Collecting system debris (p=0.022), a higher uHU (p=0.038), and a higher Delta uHU (p<0.001) were identified as independent risk factors for pyonephrosis based on multivariate analysis. The Delta uHU sensitivity and specificity were 88.7% and 86.4%, respectively, at a cutoff value of 6.56 (p<0.001), whereas the sensitivity and specificity for detecting pyonephrosis at a uHU cutoff value of 7.96 was 50.7% and 70.9%, respectively (p=0.003). Conclusions: Non- contrast computer tomography was shown to accurately distinguish simple hydronephrosis from pyonephrosis in patients with obstructive uropathy. Evaluation of the Delta uHU in different slices may be more reliable than the uHU acquired from a single slice in predicting pyonephrosis.
更多
查看译文
关键词
Hounsfield units,Hydronephrosis,Multidetector computed tomography,Pyonephrosis,Urinary tract infection
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要