透過您的圖書館登入
IP:18.219.14.63
  • 學位論文

肝硬化之超音波影像分析

Ultrasonic image analysis of liver cirrhosis

指導教授 : 邵耀華

摘要


本研究採用肝臟組織的灰階影像分析與肝靜脈和門靜脈的能量都卜勒血流動態影像PDVI 分析探討臨床肝硬化的指標。在肝臟組織灰階影像分析中,我們利用灰階共生矩陣,計算肝臟影像的特徵參數指標,分析正常肝臟與肝硬化影像,部分參數指標在分類肝硬化特徵具統計學上之差異。 而另一方面,在正常控制組(N=10)肝靜脈的 PDVI會比肝硬化患者(N=16)之肝靜脈PDVI 較有週期性或規律性。由 PDVI 發展出之指標 PDVImax PDVImin,正常控制組的值是2.44 ± 0.85,有肝硬化者是 1.52 ± 0.26,統計上具有顯著差異( p = 0.007, Threshold = 1.8, sensitivity 93.75%, specificity 80% )。而門靜脈之PDVI比值在控制組與肝硬化組之間差別並不大,統計上沒有顯著差異。 正常肝臟組織具有彈性,肝靜脈血管組織內血液流通順暢,而肝臟纖維化變硬之後,將使得血管變得較不具彈性,血液流通不順,使得PDVI的脈動性明顯地發生變化,正常肝臟的 PDVI 之脈動量會比較大,兩者相比較,肝靜脈的脈動性有著顯著的差別,因此對未來分辨正常與肝硬化的差異,此一指標是具有意義的。

並列摘要


In this study, the gray level B-mode ultrasound image analysis of liver parenchyma and hepatic vein and portal vein Power Doppler Vascular Index (PDVI) analysis was used to differentiate liver cirrhosis. The characteristic parameters extracted from grey-level co-occurrence matrix was statistical significant in differentiating the normal liver and the cirrhotic liver. The PDVI waveforms measured from hepatic vein is more regular or periodic in the normal livers than in the cirrhotic livers. The ratio of maximum PDVI to minimum PDVI ( PDVImax : PDVImin ) for hepatic vein were 2.44 ± 0.85 in control normal, and 1.52 ± 0.26 in patients with liver cirrhosis, and is statistical significant. ( p = 0.007, with threshold of 1.8, sensitivity 93.75%, specificity 80%). Such a difference of PDVI ratio in the portal vein was not as significant as in the hepatic vein. The hepatic vein for normal control subjects can be considered as an elastic vessel. When the liver becomes fibrosis, the vessel would lose its elasticity. Therefore, the pulsation of PDVI of blood vessel decreases, and such the difference would be easily detectable in the hepatic vein than the portal vein. Generally, use the PDVI ratio as an clinical index for differentiating normal livers and cirrhotic livers would to be meaningful.

參考文獻


32.蔡漪琳 (2005);以肝臟超音波影像紋理特徵區別肝硬化之研究;國立臺灣大學醫學工程學研究所碩士論文
34.包舜華 (2004);超音波動態影像分析技術與臨床醫學應用;國立臺灣大學應用力學研究所博士論文
35.林家慶 (2005);以犬隻腎臟超音波探討血液灌流指數之物理意義;國立臺灣大學應用力學研究所碩士論文
2.Wu, C.M., Y.C. Chen, et al. (1992). "Texture features for classification of ultrasonic liver images." IEEE Trans. Med. Imaging 11(2):141-152.
3.Sun, Y.N., M.H. Horng, et al. (1996). "Ultrasonic image analysis for liver diagnosis: A noninvasive alternative to determine liver disease." IEEE Engineering in Medicine and Biology 15(6):93-101.

延伸閱讀