基于多维特征参数的肝纤维化超声定量分析及建模

肝纤维化是肝硬化和肝癌的早期病理特征，精准评判其进展-逆转是现代治疗学的迫切要求。但目前的评价系统无法准确评估早期肝纤维化，且较难排除脂肪变性和炎症程度的影响。结合对肝脏形态学、血流动力学及硬度-弹性学的既往研究，我们提出：基于多维特征参数的完整生物学信息，在定量化和数学模型基础上，能精准评估肝纤维化。我们将利用组织声学定量、参数微血管成像及剪切波弹性成像，采集肝纤维化的多种组织学信息，参考病理学规律，提取并量化各参数；运用相关性分析和多元回归分析，建立并优化评价系统；以相关临床终点数据为对照指标，并验证其临床评价能力，进一步明确F1-4各期的定量标准；并实现F1-2期纤维化的评价，真正达到无创性早期、分期诊断；并阐明肝纤维化、炎症及脂肪变三者的相互关系。本项目不仅会为肝纤维化分析的方法学提供新思路；而且还将为现代治疗学提供以客观性和连续性量化数据为基本特点的新型肝纤维化诊断工具。

Liver fibrosis is the pathology characteristic of liver cirrhosis and hepatocellular carcinoma in early stage. Precise assessment on its progression and reversion is crying needs in modern treatment. However, the ultrasound evaluation system is not able to precisely evaluate liver fibrosis in early stage at present and to exclude the influence of steatosis and inflammation. Combining the results of previous studies in liver morphology, hemodynamics and hardness-elasticity, we point out that it is possible to assess liver fibrosis precisely on the basis of complete information on the multi-dimensional characteristics of the biological parameters and quantitative and mathematical models. We will use acoustic structure quantification (ASQ), micro flow image (MFI) and shear wave elastography (SWE) to collect various tissue information of liver fibrosis, consult pathology pattern, extract and quantify the parameters. Using correlation analysis and multiple regression analysis to establish and optimize the evaluation system. We will set the relevant clinical endpoint data as comparison indicators and verify the capacity of clinical evaluation to further clarify the quantitative criteria of stage 1-4, and noninvasively evaluate and diagnose stage 1 and stage 2 of fibrosis. We also will clarify the relationship between liver fibrosis, inflammation and fatty degeneration. This project will not only provide new view and perspective for the methodology of analyzing liver fibrosis, but also will provide modern therapeutics the modern ancillary diagnostic tool on basic features of objective and continuous quantitative data.