陳中明2006-07-262018-06-292006-07-262018-06-292003-07-31http://ntur.lib.ntu.edu.tw//handle/246246/22366為了增加診斷的正確性及降低人為因 素，在這兩年期計畫中，我們發展智慧型 電腦輔助診斷系統以協助良性與惡性乳房 腫瘤的區別診斷，進而降低組織切片的必 要性。在此計畫中，我們提出幾近不受參 數影響的影像特徵與高效率的分類系統。To augment the diagnostic accuracy with the effect of human factor minimized, in this two-year project, we propose to develop an intelligent CAD system to assist differential diagnosis of benign and malignant lesions. The ultimate goal is to provide a suggestive diagnosis for the benign and malignant lesions revealed in ultrasound images such that the number of biopsies may be minimized. At the end of this project, we propose a classification system of high efficiency with nearly setting-independent features.application/pdf117600 bytesapplication/pdfzh-TW國立臺灣大學醫學工程學研究所乳房超音波影像乳房腫瘤電 腦輔助診斷特徵擷取分類Breast Ultrasound ImagesBreast LesionCADFeature ExtractionClassificationreporthttp://ntur.lib.ntu.edu.tw/bitstream/246246/22366/1/912213E002068.pdf