臺灣博碩士論文加值系統

由於腎功能指數較低者，使用顯影劑後會有機率導致急性腎功能衰退，而僅靠未顯影電腦斷層影像很難診斷所有疾病，且健康檢查電腦斷層項目中也以未施打顯影劑者居多，盼能不增加受檢者身體負擔前提下，提供一預測肝癌之參考資訊。
本研究為回溯性研究設計，從2016年12月至2017年6月間經放射科醫師完成報告之電腦斷層肝三相檢查案例中收集101例。將未發現明顯異常及良性病灶者歸為對照組，共56位；而經放射科醫師完成疑似肝細胞癌報告者為實驗組，共45位。接著利用多種子區域成長法擷取影像全肝區域，並計算區域內影像特徵數值統計分析，建立合理預測肝癌發生之模型。
描述性統計結果，實驗組中影像特徵平均值為138.84HU低於對照組之141.74HU，而實驗組之平均年齡64.22歲、AFP為531.43ng/ml、GOT為59.36U/L及GPT為53.04U/L，皆高於對照組之平均年齡57.05歲、AFP為9ng/ml、GOT為32.43U/L及GPT為33.19U/L，且P值皆小於0.05，是為肝癌顯著特徵。交叉列表比對發現本研究男性罹病之機率為女性1.88倍。在單變數羅吉斯迴歸中肝癌顯著特徵為影像特徵平均值、年齡、AFP、GOT及GPT(P<0.05)。最後利用多變數羅吉斯迴歸建立預測模型，ROC曲線驗證結果，其AUC可達0.899，當切點計算為0.526時擁有真陽性率80.5%及明確度88.6%，具有相當高之準確性，且Kappa係數也有0.657之一致性。
未顯影電腦斷層影像評估肝癌顯著特徵為影像特徵平均值、年齡、AFP、GOT及GPT，當預測模型判定高機率為肝癌且影像特徵平均值偏低、年齡偏高以及AFP、GOT及GPT高於正常值時，需進一步安排檢查以達早期發現早期治療之目的。未來希望能針對肝臟紋理作分析，提升圈選肝臟時之準確性。

Due to the low index of renal function who after use contrast media probability lead to acute kidney failure, and only by precontrast CT difficult to diagnose all diseases. And in physical examination CT items to the most are without contrast media. This study hopes to provide a reference for predicting HCC without increasing the physical burden.
This study is a retrospective study design, a total of 101 cases of CT triphase examinations reported by radiologists are expected to be collected from December 2016 to June 2017.Among them, those who did not find significant abnormalities and benign lesions were classified as control group, a total of 56 cases; and the radiologist completed the report of suspected HCC as experimental group, a total of 45 cases. Then we will using multi-seed region growing to obtain the features of precontrast CT liver images, and the statistical analysis of the intra-area image characteristics was calculated to establish a model for predicting the occurrence of HCC.
For descriptive statistics, the mean value in the experimental group was 138.84 HU, which was lower than the 141.74 HU of the control group, while the average age of the experimental group was 64.22 years, the AFP,GOT and GPT was 531.43 ng/ml,59.36 U/L and 53.04 U. /L, were higher than the average age of the control group 57.05 years old, AFP.GOT and GPT was 9ng/ml,32.43U/L and 33.19U/L, P<0.05 is a significant feature of HCC. Cross-table comparisons found that the chance of male rickets in this study was 1.88 times that of women. In the univariate logistic regression, the significant features of HCC were the mean value, age, AFP,GOT and GPT (P<0.05). Finally, using the multivariate logistic regression to establish the prediction model, and then use the ROC curve to verify ,and the AUC can reach 0.899. When the cut-off point is 0.526, it has a true positive rate of 80.5% and a clearness of 88.6%. It has a very high accuracy, and the Kappa also has a consistency of 0.657.
The significant features of HCC are the mean value, age, AFP, GOT, and GPT. The predictive model shows high probability of HCC, as lower mean value, older, higher AFP, GOT and GPT. Then, further inspections are needed to exam the situation of body. In the future, we hope to analyze the liver texture to improve the capture of the whole liver area.

摘要 i
英文摘要 ii
誌謝 iii
目錄 iv
表目錄 vi
圖目錄 vii
第一章 緒論 1
1.1 前言 1
1.2 研究動機與目的 11
第二章 文獻探討 12
2.1 電腦斷層數位影像定量分析 12
2.2 羅吉斯迴歸應用於醫學影像 15
第三章 研究方法與步驟 17
3.1 研究架構 17
3.2 研究對象與收集樣本 18
3.3 特徵定義 20
3.3.1 影像特徵定義 20
3.3.2 生理特徵 22
3.4統計分析及驗證效能 23
第四章 結果分析 26
4.1描述性統計分析 26
4.2交叉表結果分析 29
4.3羅吉斯迴歸模型 30
4.3.1單變數羅吉斯迴歸 30
4.3.2多變數羅吉斯迴歸 31
4.4驗證模型效能 32
第五章 結論與討論 35
5.1 結論 35
5.2 討論及未來研究方向 36
參考文獻 38
附錄 附錄-1

[1] Wen CP, Lin J, Yang YC, Tsai MK, Tsao CK, Etzel C, Huang M, Hsu CY, Ye Y, Mishra L, Hawk E, Wu X. Hepatocellular carcinoma risk prediction model for the general population: the predictive power of transaminases. J Natl Cancer Inst. 2012 17;104(20):1599-611
[2] Ma KW, Chok KSH, Chan ACY, Tam HSC, Dai WC, Cheung TT, Fung JYY, Lo CM. A new formula for estimation of standard liver volume using computed tomography-measured body thickness. Liver Transpl. 2017;23(9):1113-22
[3] Kokudo T, Hasegawa K, Uldry E, Matsuyama Y, Kaneko J, Akamatsu N, Aoki T, Sakamoto Y, Demartines N, Sugawara Y, Kokudo N, Halkic N. A new formula for calculating standard liver volume for living donor liver transplantation without using body weight. J Hepatol. 2015;63(4):848-54
[4] Vauthey JN, Abdalla EK, Doherty DA, Gertsch P, Fenstermacher MJ, Loyer EM, Lerut J, Materne R, Wang X, Encarnacion A, Herron D, Mathey C, Ferrari G, Charnsangavej C, Do KA, Denys A. Body surface area and body weight predict total liver volume in Western adults. Liver Transpl. 2002;8(3):233-40
[5] Combs CA, Jaekle RK, Rosenn B, Pope M, Miodovnik M, Siddiqi TA. Sonographic estimation of fetal weight based on a model of fetal volume. Obstet Gynecol. 1993;82(3):365-70
[6] Chan KS, Jiao F, Mikulski MA, Gerke A, Guo J, Newell JD Jr, Hoffman EA, Thompson B, Lee CH, Fuortes LJ. Novel Logistic Regression Model of Chest CT Attenuation Coefficient Distributions for the Automated Detection of Abnormal (Emphysema or ILD) Versus Normal Lung. Acad Radiol. 2016;23(3):304-14
[7] Silva M, Nemec SF, Dufresne V, Occhipinti M, Heidinger BH, Chamberlain R, Bankier AA. Normal spectrum of pulmonary parametric response map to differentiate lung collapsibility: distribution of densitometric classifications in healthy adult volunteers. Eur Radiol. 2016;26(9):3063-70
[8] Wang J, Li F, Li Q. Automated segmentation of lungs with severe interstitial lung disease in CT. Med Phys. 2009;36(10):4592-9
[9] Barbosa EM Jr, Song G, Tustison N, Kreider M, Gee JC, Gefter WB, Torigian DA. Computational analysis of thoracic multidetector row HRCT for segmentation and quantification of small airway air trapping and emphysema in obstructive pulmonary disease. Acad Radiol. 2011;18(10):1258-69
[10] Prosch H, Schaefer-Prokop CM, Eisenhuber E, Kienzl D, Herold CJ. CT protocols in nterstitial lung diseases--a survey among members of the European Society of Thoracic Imaging and a review of the literature. Eur Radiol. 2013;23(6):1553-63
[11] Zhao Y, Wang X, Wang Y, Zhu Z. Logistic regression analysis and a risk prediction model of pneumothorax after CT-guided needle biopsy. J Thorac Dis. 2017;9(11):4750-7
[12] Lee HY, Lee IJ. Assessment of Independent Risk Factors of Developing Pneumothorax During Percutaneous Core Needle Lung Biopsy: Focus on Lesion Depth. Iran J Radiol. 2016;13(4)
[13] Kakizawa H, Toyota N, Hieda M, Hirai N, Tachikake T, Matsuura N, Oda M, Ito K. Risk factors for severity of pneumothorax after CT-guided percutaneous lung biopsy using the single-needle method. Hiroshima J Med Sci. 2010;59(3):43-50
[14] Heerink WJ, de Bock GH, de Jonge GJ, Groen HJ, Vliegenthart R, Oudkerk M. Complication rates of CT-guided transthoracic lung biopsy: meta-analysis. Eur Radiol. 2017;27(1):138-48
[15] Furlan A, Marin D, Vanzulli A, Patera GP, Ronzoni A, Midiri M, Bazzocchi M, Lagalla R, Brancatelli G. Hepatocellular carcinoma in cirrhotic patients at multidetector CT: hepatic venous phase versus delayed phase for the detection of tumour washout. Br J Radiol. 2011;84(1001):403-12
[16] El-Serag HB, Rudolph KL. Hepatocellular carcinoma: epidemiology and molecular carcinogenesis. Gastroenterology. 2007;132(7):2557-76
[17] Liao TC, Liu CA, Chiu NC, Yeh YC, Chiou YY. Lymphoepithelioma-like cholangiocarcinoma: a mimic of hepatocellular carcinoma on imaging features.World J Gastroenterol. 2015 7;21(13):4089-95
[18] Bialecki ES, Di Bisceglie AM. Diagnosis of hepatocellular carcinoma. HPB (Oxford). 2005;7(1):26-34
[19] Yeh SH, Chen PJ. Gender disparity of hepatocellular carcinoma: the roles of sex hormones. Oncology. 2010;78 Suppl 1:172-9
[20] Hefaiedh R, Ennaifer R, Romdhane H, Ben Nejma H, Arfa N, Belhadj N, Gharbi L,Khalfallah T. Gender difference in patients with hepatocellular carcinoma. Tunis Med. 2013;91(8-9):505-8
[21] Maldonado G, Greenland S. Simulation study of confounder-selection strategies. American journal of epidemiology. 1993;138(11):923-36
[22] Suzuki K, Kohlbrenner R, Epstein ML, Obajuluwa AM, Xu J, Hori M. Computer-aided measurement of liver volumes in CT by means of geodesic active contour segmentation coupled with level-set algorithms. Med Phys. 2010;37(5):2159-66
[23] Massoptier L, Casciaro S. A new fully automatic and robust algorithm for fast segmentation of liver tissue and tumors from CT scans. Eur Radiol. 2008;18(8):1658-65
[24] Moghbel M, Mashohor S, Mahmud R, Saripan MI. Automatic liver tumor segmentation on computed tomography for patient treatment planning and monitoring. EXCLI J. 2016;15:406-23
[25] Kumar S.S., An automatic computer-aided diagnosis system for liver tumorson computed tomography images, Computers and Electrical Engineering, 2013;39,1516–26