Equilibrium CT Texture Analysis for the Evaluation of Hepatic Fibrosis: Preliminary Evaluation against Histopathology and Extracellular Volume Fraction

平衡CT纹理分析在肝纤维化评估中的应用:与组织病理学和细胞外容积分数的初步比较

阅读:1

Abstract

BACKGROUND: Evaluate equilibrium contrast-enhanced CT (EQ-CT) texture analysis (EQ-CTTA) against histologically-quantified fibrosis, serum-based enhanced liver fibrosis panel (ELF) and imaging-based extracellular volume fraction (ECV) in chronic hepatitis. METHODS: This study was a re-analysis of image data from a previous prospective study. Pre- and equilibrium-phase post-IV contrast CT datasets were collected from patients with chronic hepatitis with contemporaneous liver biopsy and serum ELF measurement between April 2011 and July 2013. Biopsy samples were analysed to derive collagen proportionate area (CPA). EQ-CTTA was performed with a filtration histogram technique using texture analysis software, with texture quantification using statistical and histogram-based metrics (mean, skewness, standard deviation, entropy, etc.). Association between pre-contrast and EQ-CTTA against CPA, ECV and ELF was evaluated using Spearman's rank correlation coefficient (r(s)). RESULTS: Complete datasets collected in 29 patients (16 male; 13 female), mean age (range): 49 (22-66 years). Liver ECV, CPA and ELF had a median (interquartile range) of 0.26 (0.24-0.29); 5.0 (3.0-13.7) and 9.71 (8.39-10.92). Difference in segment VII hepatic CTTA (medium texture scale) between EQ-CT and pre-contrast images was significantly and positively associated with ELF score (mean: r(s) = 0.69, p < 0.001; skewness: r(s) = 0.57, p = 0.007). Significant negative associations were observed between pre-contrast and EQ-CT whole hepatic CTTA (coarse texture scale) with CPA (pre-contrast, SD: r(s) = -0.66, p < 0.001) and ECV (EQ-CT, entropy: r(s) = -0.58, p = 0.006). CONCLUSIONS: Hepatic EQ-CTTA demonstrates significant association with validated markers of liver fibrosis, suggesting a role in non-invasive quantification of severity in diffuse fibrosis.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。