透過您的圖書館登入
IP:3.21.162.87
  • 學位論文

頸動脈中層膜厚度與血管直徑變化對動脈硬化前期影響之研究

A Study on the Influence of Carotid Intima-Media Thickness and Differences in the Common Carotid Artery Diameter on Preclinical Atherosclerosis

指導教授 : 張俊郎
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


台灣地區歷年來十大死因以惡性腫瘤居於首位,而其次為心臟疾病以及腦血管疾病,在過去對於動脈硬化之研究侷限於頸動脈中層膜厚度(carotid artery intima–media thickness, CCA-IMT)對動脈粥狀硬化現象的影響程度,對於頸動脈血管直徑(diameter of the common carotid artery, CC-AD)是否會增加罹患心腦血管疾病之探討較少。由於動脈前期具有可逆性反應,因此建構一套輔助診斷系統,協助醫師的臨床診斷及預防愈顯其重要性。 本研究主要在探討血管內膜以及血管直徑相關變異之因子對動脈硬化前期影響程度,針對未發病之健康人的血管頸動脈血管直徑以及頸動脈中層膜厚度進行分析,透過醫院資料庫藉由資料探勘技術中的類神經網路以及決策樹進行客觀的資料分析,對動脈粥狀硬化有效臨床診斷指標CCA-IMT及CC-AD變化加以評估。 研究顯示類神經網路於CCA-IMT分類預測準確率為82.19%;CC-AD分類預測準確率為82.36%;在類神經網路結合決策樹演算法之分類預測準確率為82.51%;而決策樹方面分類準確率為97.28%;在類神經方面的分類結果顯示影響程度最大的主要因子是年齡,其他影響因子包含收縮壓、抽菸習慣、喝酒習慣、運動習慣以及低密度膽固醇等共同部分:而在決策樹方面分類結果顯示,第一層分類節點為腰臀比以及運動習慣,其次第二層為膽固醇、身高、喝酒習慣為主;本研究成果可提供醫師進行醫療服務時,作為全方位的早期評估與判斷的參考,同時可擬定早期防範的對策,降低發生的可能性,對於醫師之臨床診斷將有實質之助益。

並列摘要


Abstract Malignant neoplasm (cancer) tops the ten leading causes of death in Taiwan, followed by heart diseases and cerebral-vascular diseases. In the past, limited to and confined in the studies of the impact of carotid artery intima-medial thickness on atherosclerosis. Few studies were conducted focusing on whether the common carotid artery diameter (CCAD) would increase chances to develop cerebral-vascular diseases. It is crucial to develop a set of diagnostic system that can help doctors in clinical diagnosis and prevention. In this study, we mainly discussed the degree of influence of the inner layers of the blood vessels and the relative various factors associated with the blood vessel diameters. We conducted analysis on the CCAD and the CIMT of the healthy people, used the artificial neural network and decision tree of the data mining technology to perform analysis objectively on the database of the case hospital, and evaluated the differences of the effective clinical diagnostic indicators of atherosclerosis, CCA-IMT and CC-AD. The study results indicate that accuracy of artificial neural network in CCA-IMT classification prediction is 82.19%, while that of CC-AD is at 82.36%. Regarding the decision tree model, the classification accuracy is 97.28%, while the classification results using artificial neural networks show that the major factor with the major influence is age. Simultaneously, it can help thorough planning for early prevention strategies to reduce possibility of developing atherosclerosis, thus bringing practical advantages to clinical diagnosis.

參考文獻


1. 中央健保局電子報第121期http://www.nhi.gov.tw/epaper2/ItemList.asp?PapersID=147&ItemTypeID=7, 2009/04/30。
2. 內政部統計處,http://www.moi.gov.tw/stat/news_content.aspx?sn=3754,2009/04/30。
3. 行政院衛生署http://www.doh.gov.tw/CHT2006/index_populace.aspx ,2009/05/18。
4. 丁一賢、陳牧言 (2005),資料探勘,滄海書局,台北。
5. 江嘉琪 (2006),頸動脈疾病,急重創聯合網。http://www.sicu.org.tw/index_detail.php?querypage=icuprotocol_detail&qid=9&hunite=#,2009/04/28。

延伸閱讀