電腦斷層心臟血管攝影是檢查冠狀動脈鈣化很重要的工具之一，它能夠提供準確的鈣化分數與血管阻塞的狀況評估，而良好的電腦斷層影像品質才能使醫師在進行診斷病灶時有更精確的判斷。本次研究利用田口動態分析法針對三個不同尺寸的擬人壓克力假體進行心臟電腦斷層掃瞄，評估各項參數設定之最佳化組合。使用田口方法的L18直交表進行18組不同參數設定的掃描，並且設定五個調控因子，分別是 (A) 管電壓峰值(kVp)；(B) 管電流與時間乘積(mAs)；(C) 螺距比(Pitch)；(D) 照野(FOV)；(E) 切片厚度(Slice thickness)；其中，A因子分配兩個水準，其餘因子分配三個水準。所使用到的設備為Philips brilliance 64-slice CT，並將自製的線群假體分別置入50、70、90公斤的壓克力假體中進行電腦斷層掃描用以模擬成人心臟實際情況。掃描後的影像經三位資深醫事放射師以隨機的順序進行評分，且重複評分三次以降低主觀認定所造成的誤差。結果得知影響影像品質的重要因子為Pitch，次要因子則是kVp。在50公斤、BMI 21.62假體的最佳化影像品質之參數組合為：管電壓140 kVp、管電流時間乘積700 mAs、螺距比0.16、照野250 mm、切片厚度0.90 mm；在70公斤、BMI 24.22假體的最佳化影像品質之參數組合為：管電壓140 kVp、管電流時間乘積700 mAs、螺距比0.20、照野220 mm、切片厚度0.8 mm；在90公斤、BMI 27.80假體之最佳化影像品質之參數組合為：管電壓140 kVp、管電流時間乘積800 mAs、螺距比0.18、照野220 mm、切片厚度0.9 mm；田口動態的最佳化影像品質之參數組合為：管電壓140 kVp、管電流時間乘積800 mAs、螺距比0.18、照野220 mm、切片厚度0.9 mm。利用不同尺寸的擬人假體搭配田口動態分析法成功地獲得不同體型的病人於冠狀動脈電腦斷層掃描的最佳化掃描參數組合。經t檢定評估後得知最佳化參數設定的空間解析度分別為2.62、2.55、2.66 mm，均優於原本臨床常規參數設定的空間解析度，此研究結果除了可以改善臨床影像品質提供醫師在診斷上更清晰的影像，更能節省檢查的時間，造福病患。

Computer Tomography Cardiovascular Photography is one of the most useful techniques for identifying coronary artery calcification. It can provide accurate assessment of calcification scores and condition of blood vessel obstruction. An optimal computer tomography image quality assists radiologists to obtain more accurate judgments in diagnosing lesions in reality. Taguchi dynamic analysis was used to evaluate the optimal combination of factor settings for computer tomography scan with three different acrylic phantoms in this study. Eighteen combinations of various factor settings were organized using Taguchi’s L18 orthogonal array. The five factors for CT scan included (A) peak voltage (kVp), (B) tube current time product (mAs), (C) spiral CT pitch, (D) Field of view (FOV) and (E) slice thickness, respectively. The CT model was Philips brilliance 64-slice CT. The customized line group phantom was placed in the thorax phantom to simulate an adult’s cardiac tissue. Accordingly, three images derived from each combination was graded and averaged by three well trained radiologists under double-blinded criteria to ascertain the reproducibility and accuracy to minimize the systematic uncertainty. The derived results were further reviewed through ANOVA to verify its importance in reality and the dominant factor was identified as Pitch. The optimal settings for 50, 70, and 90 kg phantom was 140 kVp, 700 mAs, 0.16 spiral CT pitch, 250 mm FOV, 0.90 mm slice thickness; 140 kVp, 700 mAs, 0.20 spiral CT pitch, 220 mm FOV, 0.80 mm slice thickness; and 140 kVp, 800 mAs, 0.18 spiral CT pitch, 220 mm FOV, and 0.90 mm slice thickness, respectively. Thus, the optimal setting for CT image for cardiac patients was accomplished in this study. Further, the corresponded spatial resolution for CT was 2.62, 2.55, and 2.66 mm, respectively for 50, 70, and 90 kg phantom according to a revised t-test algorithm. The image quality was greatly improved by Taguchi dynamic analysis than did the conventional setting in this study.

摘要 I
Abstract III
目錄 V
圖目錄 X
表目錄 XII
第一章 前言 1
1.1 研究背景 1
1.2 研究動機 3
1.3 研究目的 3
1.4 文章架構 4
第二章 背景說明與文獻分析 5
2.1冠狀動脈心臟病 5
2.1.1 定義 5
2.1.2 病理 6
2.1.3 臨床症狀 6
2.1.4 診斷檢查 6
2.1.5 治療方式 6
2.2相關文獻分析 7
2.2.1電腦斷層的影像品質 7
2.2.2 田口分析法 9
第三章 田口穩健設計法 12
3.1 田口方法概述 12
3.2 田口直交表 13
3.3 動態田口分析法 15
3.4 變異數分析(ANOVA) 16
第四章 材料與方法 18
4.1 實驗設備 18
4.1.1飛利浦64切電腦斷層掃描儀(Philips brilliance 64-slice CT) 18
4.1.2 壓克力線群假體(Polymethylmethacrylate line group phantom) 19
4.1.3 擬人壓克力假體 20
4.2 實驗設計 22
4.2.1 掃描因子的選用 22
4.2.1.1 管電壓峰值(kVp) 22
4.2.1.2 管電流與時間乘積(mAs) 22
4.2.1.3 螺距比(Pitch) 22
4.2.1.4 照野(Field of view, FOV) 23
4.2.1.5 切片厚度(Slice thickness) 23
4.2.2 田口直交表的選擇 23
4.3 實驗設計與規劃 25
4.3.1 執行電腦斷層掃描 25
4.3.2 影像評定標準 26
4.3.3 統計分析工具 27
4.3.4 影像品質量化 27
4.3.5 實驗步驟及流程 31
第五章 結果 32
5.1 電腦斷層影像的原始數據 32
5.1.1 體重50公斤擬人假體之原始數據 32
5.1.1.1 電腦斷層掃描的影像 34
5.1.1.2 專家的影像評分 35
5.1.1.3 因子反應分析 36
5.1.1.4 變異數分析(ANOVA) 39
5.1.2 體重70公斤擬人假體之原始數據 40
5.1.2.1 電腦斷層掃描的影像 42
5.1.2.2 專家的影像評分 43
5.1.2.3 因子反應分析 44
5.1.2.4 變異數分析(ANOVA) 47
5.1.3 體重90公斤擬人假體之原始數據 48
5.1.3.1 電腦斷層掃描的影像 50
5.1.3.2 專家的影像評分 51
5.1.3.3 因子反應分析 52
5.1.3.4 變異數分析(ANOVA) 55
5.2 動態田口分析結果 56
5.2.1 專家的影像評分結果 58
5.2.2 因子反應分析 59
第六章 討論 61
6.1因子反應分析 61
6.1.1體重50公斤擬人假體之因子反應分析 61
6.1.2體重70公斤擬人假體之因子反應分析 62
6.1.3體重90公斤擬人假體之因子反應分析 63
6.1.4 田口動態分析 63
6.2　假體在不同切面之空間解析度上的差異 65
6.2.1 體重50公斤擬人假體之空間解析度分析 65
6.2.2 體重70公斤擬人假體之空間解析度分析 68
6.2.3 體重90公斤擬人假體之空間解析度分析 70
6.2.4 田口動態分析法之空間解析度分析 72
6.3 動態最佳化的靈敏度分析 75
6.4 t檢定與幅高半寬的差異 77
第七章 結論 79
第八章 未來展望 81
參考文獻 82

[1] 衛生福利部統計處，歷年死因統計 https://dep.mohw.gov.tw/DOS/cp-3352-33579-113.html
[2] 張小平，冠心病危險因素與冠狀動脈病變程度的關係，中華醫學雜誌 1998；78(1)：49-51.
[3] 衛生福利部國民健康署-認識冠心病 https://www.hpa.gov.tw/Pages/Detail.aspx?nodeid=632&pid=1188
[4] Meir J. Stampfer. Estrogen replacement therapy and coronary heart disease: A quantitative assessment of the epidemiologic evidence. Journal of Preventive Medicine 1991；20(1)：47-63.
[5] Wasserman, J. Gender-Specific Effects of modifiable health risk factor on coronary heart disease and related expenditures. Journal of occupation Environment Medicine 2000； 42(11)：1060-69.
[6] Ming-Hui Gui. The comparison of coronary angiographic profiles between diabetic and nondiabetic patients with coronary artery disease in a Chinese population. Diabetes Research and Clinical Practice 2009；85(2)： 213-219
[7] Jaume Figueras. Extent of myocardial ischemia during coronary occlusion in single vessel disease. A comparison between patients with exercise-induced angina and patients with recurrent angina at rest. 2008；127(3)：433-435
[8] 黃曉鈴，淺談冠狀動脈疾病，新光吳火獅紀念醫院-家庭醫學醫科-新光醫訊第176期 http://www.skh.org.tw/Mnews/176/3-1.htm
[9] 林勝峰、廖清海、楊孟龍，談運動猝死之隱形殺手—冠狀動脈疾病，長榮運動休閒學刊2007年6月；第一期。63-69頁
[10] 張其任，冠狀動脈疾病及其介入治療，美國世界日報2015年6月27日 http://www.chang-gung.com/ae/news-1.aspx?rows=8&page=1&type=Photo&id=225&mid=119&bid=3
[11] 陳良光、呂坤木、蘇誠道，64 切與256 切電腦斷層血管攝影於冠狀動脈繞道術後評估之輻射劑量與影像品質研究，中華放射線技術學雜誌C J Radiologic Tech 2015; 39(3): 159-166
[12] 呂坤木、盧建利、陳良光，64 及256 切小兒心臟電腦斷層於不同造影模式之影像品質與輻射劑量的探討，中華放射線技術學雜誌 C J Radiologic Tech 2017; 41(3): 123-130
[13] 林玟吟，探討64切冠狀動脈電腦斷層攝影造影參數與影像品質之關聯性，碩士論文，義守大學資訊工程研究所，2015
[14] Simom D, Francesa P, Aloha M, et al. Image quality and radiation dose of a prospectively electrocardiography-triggered high-pitch data acquisition strategy for coronary CT angiography: The multicenter, randomized PROTECTION IV study. Journal of Medical Imaging and Health Informatics. 2015; 278-285
[15] 蕭貴仁、陳俊吉、潘榕光，利用田口方法最佳化錐形射東電腦斷層影像之品質特性，中華放射線技術學雜誌C J Radiologic Tech 20 14; 38(4 ): 2 13-219
[16] 陳欣妤、李季達、潘榕光，利用田口方法找出256 切CT 解析度最佳化設定，臺灣應用輻射與同位素雜誌 Taiwanese Journal of Applied Radiation and Isotopes. Mar 2017; Vol. 13, No. 1, P1433-1440
[17] Pan LF, Chen CC, Pan LK, et al. Optimization of the imaging quality of 64- slice CT acquisition protocol using Taguchi analysis: A phantom study. Bio-Medical Materials and Engineering 26 (2015) S1651–S1658
[18] Pan LF, Chu KH, Pan LK, et al. Optimizing left anterior oblique (LAO) caudal imaging in coronary angiography using the Taguchi method: A phantom study with clinical verification. Int J Cardiovasc Imaging DOI 10.1007/s10554-017-1129-8
[19] Kuo HN, Yi JE, Pan LK, et al. Taguchi dynamic analysis application to computer tomography number-mass density linear dependence optimization. Computer Assisted Surgery ISSN: (Print) 2469-9322
[20] Samrit K, Pan LF, Lin CH, et al. The Optimization of the Single Photon Emission Computed Tomography Image Quality via Taguchi Analysis: A Feasibility Study of a V-Shaped Phantom: Journal of Medical Imaging and Health Informatics Vol. 7, 143–148, 2017
[21] 李輝煌。田口方法：品質設計的原理與實務，四版。高立圖書有限公司，2000：249-309
[22] Taguchi G. Introduction to Quality Engineering. Asian Productivity Organization 1990.
[23] Barker TB. The concept of quality. In: Engineering quality by design , Marcel Dekker, Inc. 1990; Chapt 1.
[24] Ross PJ. Orthogonal array selection and utilization. In: Taguchi techniques for quality engineering , McGraw-Hill Inc. Second ed. 1996; Chapt 3.
[25] Lochner RH. Two-level experiments: Full factorial designs. In:Designing for quality-An introduction to the best of Taguchi and westernmethods of statistical experimental design. Kraus Organization Limited.1990; Chapt 3.
[26] M. N. Dhavlika. Combined Taguchi and dual response method for optimization of a centerless grinding operation. Journal of Materials Processing Technology 2003；132(1-3)：90-94.
[27] Tzeng Yih-fong. Dimensional quality optimisation of high-speed CNC milling process with dynamic quality characteristic. Robotics and Computer-Integrated Manufacturing 2005；21：506-517.
[28] Shrikant Kulkarni. Taguchi design of experiments for optimization of ionic conductivity in nanocrystalline Gadolinium doped Ceria. Ceramics International 2015；41(7)： 8973-8980.
[29] Alireza Zamani Aghaie. A general optimized geometry of angled ribs for enhancing the thermo-hydraulic behavior of a solar air heater channel -A Taguchi approach. Renewable Energy 2015；83： 47-54.
[30] C. Panagiotopoulou. Application of the Taguchi approach for the composition optimization of alkali activated fly ash binders. Construction and Building Materials 2015；91：17-22.
[31] Harun Tanyildizi. Application of Taguchi method for optimization of concrete strengthened with polymer after high temperature. Construction and Building Materials 2015；79：97-103.
[32] Hassan Mehboob. Optimal design of a functionally graded biodegradable composite bone plate by using the Taguchi method and finite element analysis. Composite Structures 2015； 119：166-173.
[33] Tan C. Nguyen. Using Taguchi and ANOVA methods to study the combined effects of drilling parameters on dynamic barite sag. Journal of Petroleum Science and Engineering 2014；121：126-133.
[34] I.H. Ibrahim. Applying Taguchi’s off-line quality control method and ANOVA on the maneuverability of the F-5E intake. Mathematical and Computer Modelling 2009；49(7-8)：1359-1371.
[35] 黃正仲，放射科學－放射師必知的物理、生物和防護(第九版)，力大圖書有限公司，2010，367-393
[36] 林昶仲，CT攝影技術學，合記圖書出版社，2015
[37] Pan LF, Erdene E, Pan LK, Optimization of the imaging quality of 64-slice CT acquisition protocol using Taguchi analysis: A phantom study. Bio-Medical Materials and Engineering 26 (2015) S1651–S1658
[38] 張文釧、廖學鍵、陳健龍，利用β阻斷劑提昇冠狀動脈電腦斷層影像品質之探討，澄清醫護管理雜誌，第六卷第二期，2010