使用碎形維度、降維方法及分類樹進行肝硬化超音波影像的統計分析

Abstract

超音波照影是肝硬化的初期診斷與追蹤的重要工具，可以藉由比較同一病人肝與脾的超音波影像紋理來判定此病人是否患有肝硬化。本篇研究是使用電腦應用統計特徵比較肝跟脾的超音波影像，去幫助診斷肝硬化。我們計劃根據國立交通大學統計學研究所黃怡銘(1999)碩士論文「肝硬化超音波影像紋理的統計比較分析」的研究方法，再加上碎形維度的特徵和分類樹的分析，找出更好的統計分析方法。這些方法將應用在臺大醫院收集到的臨床資料驗證。

Ultrasound imaging is an important tool for early detection and regular check-ups. The diagnosis can be performed by the analysis of echo textures of liver and the accompanied spleens. This study aims to investigate the statistical features derived from the ultrasound images of livers and the accompanied spleens. Based on the projection methods proposed in the thesis of Hwang (1999), we will further incorporate the features of fractal dimension and classification trees to improve the statistical analysis methods. These methods are tested by the clinical data collected at National Taiwan Hospital.