電機資訊學院: 資訊工程學研究所指導教授: 張瑞峰黃耀賢Huang, Yao-SianYao-SianHuang2017-03-032018-07-052017-03-032018-07-052015http://ntur.lib.ntu.edu.tw//handle/246246/275587乳癌已經成為女性高致死率的疾病。除了治療之外,未來癌症的復發機率以及存活率也是重要的資訊。隨著醫療技術的進步,與乳癌相關之生物標記已定義並且進行相關研究。不同的生物標記組合,其治療的方式、復發機率以及存活年限不盡相同,尤其是三陰性腫瘤的預後以及復發是最差以及最快的。因此,及早取得生物標記資訊,有助於治療的方式及療後評估。超音波乳房攝影是一種便宜、非侵入式且不受乳房緻密度影響的影像,近年來已廣泛使用於乳房腫瘤檢查,隨著技術的進步,超音波影像搭配都卜勒效應原理,將腫瘤周圍之血管資訊顯示於影像上,提供更多資訊作為診斷依據。因此,為了建立有效率的的診斷程序,許多電腦輔助診斷系統以超音波影像為基礎被開發出來,將影像特徵量化提供醫師方便以及立即協助。在這篇論文中,我們藉由電腦輔助診斷系統量化超音波影像特徵,包含腫瘤以及血管特性,再根據量化後特徵資訊,評估不同生物標記種類腫瘤之間血管以及腫瘤形狀、紋路的關係。Breast cancer has become the most fatal disease among women. Beside the treatment of breast cancer, the relapse and survival rate are also the important information. With the advance of medical technology, the biomarkers of breast cancer have been defined and investigated. Different subtypes of biomarkers response to the different treatment therapy, relapse, and survival rate, especially the triple-negative tumors that have the worst prognosis and highest relapse rate. Hence, it is helpful to determine the treatment planning and evaluate the prognosis if the biomarker information is obtained as early as possible. Ultrasound (US) is a cheaper, non-invasive, effective against to eliminate the influence of dense tissue, and has been widely used in breast examination. With the growth of US technique, the angiogenesis around tumor has been displayed on the US imaging with the Doppler effect and provided additional information for diagnosis. Therefore, for constructing the efficient diagnosing processes, various computer-aided diagnosis (CAD) systems have been developed based on US to quantify the imaging characteristics for supplying a convenient and immediate assistance. In this study, the morphology and texture features of angiogenesis and tumor are quantified by using the proposed CAD systems and then the correlations among distant subtypes of biomarkers are analyzed and evaluated.6036007 bytesapplication/pdf論文公開時間: 2020/7/22論文使用權限: 同意有償授權(權利金給回饋學校)乳癌超音波電腦輔助診斷都卜勒效應三陰性腫瘤Breast cancerultrasoundcomputer-aided diagnosisDoppler effecttriple-negative breast cancer[SDGs]SDG3乳房超音波之生物標記分析Analyzing Biomarker in Breast Ultrasoundthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/275587/1/ntu-104-D96922017-1.pdf