Sichuan Key Laboratory of Medical Imaging, and Department of Radiology, Affiliated Hospital of North Sichuan Medical College, Nanchong, Sichuan, China; Department of Radiology, Peoples' Hospital of Deyang, Deyang, Sichuan, China.
Department of Radiology, Mianyang Central Hospital, Mianyang, Sichuan, China.
Acad Radiol. 2019 Jun;26(6):e90-e97. doi: 10.1016/j.acra.2018.06.018. Epub 2018 Jul 31.
As an extension of the conventional diffusion weighted imaging, diffusion kurtosis imaging (DKI) is based on the non-Gaussian diffusion model that can inherently account for restricted water diffusion within the complex microstructure of most tissues. This study aimed to investigate association of liver DKI derived parameter with stage of liver fibrosis.
Fifty-six healthy New Zealand white rabbits were enrolled into this study, among which 48 rabbits were randomly given carbon tetrachloride to model liver fibrosis, and 8 rabbits treated with normal saline served as control subjects. All rabbits underwent liver DKI followed by biopsy to stage fibrosis (stages F0-F4) on 6th, 8th, 10th, and 12th weekends after initiation of modeling fibrosis. Mean kurtosis (MK), fractional anisotropy (FA), and mean diffusion (MD) were derived from DKI data. Statistical analysis was to evaluate association of DKI derived parameter with stage of fibrosis.
FA (r = 0.512) and MK (r = 0.567) increased, and MD (r = -0.574) decreased with increasing stage of fibrosis from F0 to F4 (all p values < 0.05). Significant differences were found in all parameters between F0 and F3 or F4, F1 and F4, F0 and F1-4, and F0-1 and F2-4 (all p values < 0.05). FA and MD could distinguish between F0 from F2, MD, and MK could distinguish F1 from F3, F0-2 from F3-4, and F1-2 from F3-4, and MK and FA could distinguish F2 from F4, and F0-3 from F4 (all p values < 0.05). According to receiver operating characteristic analysis, MK could best predict stage ≥F1, ≥F2, ≥F3, and F4, and discriminate F1-2 from F3-4 with areas under receiver operating characteristic curve of 0.766-0.930.
DKI derived parameters can help stage fibrosis.
作为传统弥散加权成像的扩展，弥散峰度成像（DKI）基于非高斯扩散模型，可以内在地描述大多数组织复杂微结构中受限的水分子扩散。本研究旨在探讨肝脏 DKI 衍生参数与肝纤维化分期的关系。
本研究纳入 56 只新西兰大白兔，其中 48 只随机给予四氯化碳建立肝纤维化模型，8 只给予生理盐水作为对照。所有兔子均行肝脏 DKI 检查，然后在建模后第 6、8、10 和 12 周末行肝活检，根据纤维化分期（F0-F4 期）进行分期。从 DKI 数据中提取平均峰度（MK）、各向异性分数（FA）和平均弥散度（MD）。统计分析用于评估 DKI 衍生参数与纤维化分期的关系。
FA（r=0.512）和 MK（r=0.567）随纤维化分期（从 F0 至 F4）增加而升高，MD（r=-0.574）随纤维化分期增加而降低（均 P 值<0.05）。在 F0 至 F3 或 F4、F1 至 F4、F0 至 F1-4 和 F0-1 至 F2-4 之间，各参数均有显著差异（均 P 值<0.05）。FA 和 MD 可将 F0 与 F2 区分开，MD 和 MK 可将 F1 与 F3 区分开，F0-2 与 F3-4 区分开，F1-2 与 F3-4 区分开，MK 和 FA 可将 F2 与 F4 区分开，F0-3 与 F4 区分开（均 P 值<0.05）。根据受试者工作特征分析，MK 对预测纤维化分期≥F1、≥F2、≥F3 和 F4 及 F1-2 与 F3-4 具有最佳效能，其受试者工作特征曲线下面积为 0.766-0.930。
DKI 衍生参数有助于肝纤维化分期。
Zotero 插件