臺灣博碩士論文加值系統

子宮頸癌是台灣發生率與致死率高之疾病，各年齡層的女性都須防範。目前以子宮頸抹片檢查為防範子宮頸癌最普遍的方式。抹片之判讀在診斷中扮演非常重要的角色。因此本研究主要是透過影像處理技術，分析細胞特徵來協助醫師在抹片上異常細胞的偵測。

本研究主要是依據細胞染色特徵與細胞核參數來判斷異常細胞。細胞染色特徵由分析不同型態之細胞其細胞核與細胞質色彩在RGB比例變化來獲得。而細胞核參數部份，本研究利用影像量化與區域成長來判斷細胞核輪廓，藉以得到細胞核其色彩、面積與紋理等參數，並藉由ROC(Receiver Operating Characteristic)分析篩選各參數閥值來判斷細胞之型態。研究過程中，從40張抹片影像手動擷取細胞分析其染色與細胞核參數，進而得到自動圈選與細胞型態判斷之型態，並架構出自動處理整張抹片影像之方法。

研究結果顯示，細胞位置可由細胞核其色彩中(G/B)比值小於1來判斷，且異常細胞其B值比例有變大情況發生。而細胞型態辨識結果：正常與異常細胞核在RGB與面積參數之差異明顯，但抹片上染劑殘留與血球容易被誤判為異常細胞核，因此也利用細胞染色特徵來減低誤判。在10張正常與30張異常抹片辨識結果評估,其Accuracy為0.975、Sensitivity為1、Specificity為0.9，顯示本研究之處理方法對正常與異常抹片有良好之辨別率。而異常型態細胞之辨識部份，在異常細胞中融合型HSIL細胞可藉由細胞核聚集數量判斷，而在細胞核色彩與面積差異不大之HSIL與LSIL細胞，只能依據紋理參數作為判斷，因此LSIL影像有1/4被誤判為HSIL。在10張LSIL與20張HSIL抹片影像之辨識結果，其Accuracy為0.933、Sensitivity為1、 Specificity為0.8，測試結果以異常嚴重之抹片有比較大之辨識。

整體而言，本研究對於異常細胞核有不錯之偵測效率並能達到自動化處理。本研究之介面除了自動將抹片上細胞之型態分類外，也一併把細胞核圈選位置與其參數顯現出來，提供使用者參考與收集資料。而本研究採用影像量化與區域成長方法，降低處理步驟，對於大範圍與多細胞之影像將有不錯之圈選效率。

Cervical cancer is a disease of highly incidence and death rate in Taiwan, and the largest age group for the women must take precautions. At present, Pap smear screening is the most effective way to prevent cervical cancer.

This study detects abnormal cells according to features of stained cells and parameters of each nucleus. Features of stained cells obtained via RGB proportions of the Nucleus and Cytoplasm of cells in each lesion. In parameters for each nucleus, we use image quantification and region growing to chose Nucleus contour, herewith we obtain color, area, and textures. Then, ROC analysis ( Receiver Operating Characteristic Analysis ) was used to obtain parameters threshold in order to choose cell type. In this study, we analyze color and nucleus parameters for cells chosen from 40 Pap smear images, then automatically detect and class cells of each lesion, at last accomplish an automatic process for Pap smear.

Preliminary result shows the G/B ratio of nucleus color is smaller than 1, and abnormal cell has the higher blue percentage value. In discriminating result, it is obvious to differentiate between normal and abnormal nucleus according to their RGB and area parameters. However stain or bloods on the Pap smear were easily considered abnormal nucleus, so using other stain characteristics of cell to reduce error. In test for 10 normal and 30 abnormal images, the accuracy, sensitivity and specificity were 0.975,1, and 0.9, respectively. It indicated our study's process has a good identification in distinguishing normal and abnormal pap smears. In discriminating cell's lesion degree, HSIL syncytium detection can stand on nucleus's aggregate number, and we use texture parameters to choose HSIL and LSIL while nucleus has the similar color and area. Quarter of LSIL images were considered HSIL accordingly. In test for 10 LSIL and 20 HSIL images, the accuracy, sensitivity and specificity were 0.933,1, and 0.8, respectively. It indicated our study's process has a good identification in high grade lesion pap smears.

In conclusion, this system can provide an automatic process serves and identify abnormal nucleus efficiently. Moreover, both positions and parameters of nucleus can provide users compare resources and compile data through the interface of system. Image quantification and region growing which can reduce process cost were used in this study, we expect this can process large image more efficiently.

目 錄

摘 要...............I
Abstract ........ III
目 錄............ V
圖 索 引........ VIII
表 索 引.......... X
第一章 緒論............................................................................................................................ 1
1-1 前言................................................................................................................................. 1
1-2 研究背景.......................................................................................................................... 1
1-3 文獻回顧......................................................................................................................... 3
1-4 研究目的......................................................................................................................... 4
1-5 論文架構......................................................................................................................... 5
第二章 理論基礎...................................................................................................................... 6
2-1 子宮頸疾病...................................................................................................................... 6
2-1-2 HPV 感染................................................................................................................. 7
2-1-3 子宮頸癌與癌前病變............................................................................................. 7
2-2 子宮頸抹片影像............................................................................................................. 8
2-2-1 子宮頸抹片採樣...................................................................................................... 9
2-2-2 柏式染色(Pap`s Stain)............................................................................................. 9
2-2-3 薄層抹片(Thin Prep Pap Smear) .......................................................................... 10
2-2-4 細胞診斷報告....................................................................................................... 10
2-2-5 抹片偽陰性........................................................................................................... 11
2-2-6 細胞病理特徵........................................................................................................ 12
2-2-6-2 細胞大小與細胞形狀.................................................................................... 12
表2-1 抹片上可見的正常細胞特徵[11]............................................................... 13
2-2-6-3 細胞核大小.................................................................................................... 13
表2-2 抹片上細胞不正常變化其細胞核的平均大小[11].................................... 13
2-2-6-4 核濃染............................................................................................................ 13
2-2-7 CIN/SIL 診斷標準................................................................................................. 14
表2-3 LSIL 與HSIL 之組織學分類[7].................................................................. 14
2-2-7-1 LSIL 之細胞特徵........................................................................................... 14
2-2-7-2 HSIL 之細胞特徵........................................................................................... 15
2-3 影像處理方法............................................................................................................... 15
2-3-1-1 RGB 模型........................................................................................................ 15
2-3-2 影像量化.................................................................................................................... 16
2-3-3 區域成長法................................................................................................................ 17
2-3-4 紋理分析.................................................................................................................... 17
2-4 統計與評估分析........................................................................................................... 18
表2-4 決策矩陣...................................................................................................... 19
第三章 研究架構及方法........................................................................................................ 21
3-1 研究材料與設備........................................................................................................... 21
3-1-1 研究材料............................................................................................................... 21
3-1-2 研究設備............................................................................................................... 21
3-2 研究流程........................................................................................................................ 22
3-3 實驗設計........................................................................................................................ 23
3-3-1 細胞染色分析....................................................................................................... 23
3-3-2 細胞核輪廓圈選................................................................................................... 25
3-3-2-1 影像量化設定................................................................................................ 25
3-3-2-2 區域成長........................................................................................................ 26
3-3-2-3 判斷輪廓........................................................................................................ 27
3-3-3 自動處理流程....................................................................................................... 28
3-3-3-1 自動圈選........................................................................................................ 28
3-3-3-2 病變程度分析................................................................................................ 28
第四章 結果與討論................................................................................................................ 29
4-1 細胞染色分析結果....................................................................................................... 29
4-1-1 細胞染色表現....................................................................................................... 29
表4-1 各類細胞其B>G 區域之統計.................................................................... 31
4-1-2 影像細胞群族分類............................................................................................... 32
4-2 細胞核圈選................................................................................................................... 33
4-2-1 以不同qn 之圈選比較......................................................................................... 33
4-2-2 各影像量化方式在細胞核圈選比較................................................................... 34
表4-2 不同量化結果之正確圈選統計.................................................................. 35
4-3 細胞核參數統計結果.................................................................................................... 35
4-3-1 面積與色彩參數分析........................................................................................... 35
表4-3 各組之參數的平均值與標準差.................................................................. 36
表4-4 正常與各類異常細胞核參數之P 值......................................................... 36
表4-5 正常與各類異常細胞核參數之ROC 曲線面積....................................... 37
表4-6 各敏感度之操作點(閥值)與正確率關係表............................................... 38
表4-7 正常與異常細胞核參數之ROC 曲線面積............................................... 39
表4-8 正常與異常細胞核參數閥值組合之分組.................................................. 39
表4-9 各組閥值在正常與異常細胞之參數過濾結果.......................................... 40
4-3-2 紋理參數分析....................................................................................................... 40
表4-10 LSIL 與HSIL 細胞核紋理參數之平均值與標準差統計........................ 42
表4-11 LSIL 與HSIL 細胞核紋理參數之P 值.................................................... 43
表4-12 LSIL 與HSIL 細胞核紋理參數之ROC 曲線面積.................................. 43
表4-13 LSIL 與HSIL 紋理參數關係表............................................................... 43
4-3-3 細胞核面積與染色濃度....................................................................................... 44
表4-14 正常與異常細胞核在面積與RGB 參數之ROC 曲線面積................... 45
表4-15 在敏感度=1 時之操作點與正常細胞核過濾結果.................................. 45
4-3-4 細胞核與B>G 區域之比例................................................................................. 45
表4-16 細胞核在B>G 區域中所佔面積.............................................................. 46
4-4 影像自動圈選結果....................................................................................................... 47
4-4-1 細胞核位置分析................................................................................................... 48
表4-17 不同病變程度細胞核與量化區域之統計................................................ 49
4-4-2 影像量化改善....................................................................................................... 49
4-4-3 自動圈選流程....................................................................................................... 50
4-4-4 不同解析度之抹片影像測試............................................................................... 51
4-5 抹片影像自動處理測試............................................................................................... 52
4-5-1 正常與異常抹片影像判讀................................................................................... 53
表4-18 正常與異常抹片辨識結果評估................................................................ 55
4-5-2 異常細胞抹片影像判讀測試............................................................................... 56
表4-19 異常抹片辨識結果評估............................................................................ 56
4-6 研究結果與限制........................................................................................................... 57
4-7 分析與測試之介面....................................................................................................... 57
第五章 結論與未來展望........................................................................................................ 60
5-1 結論............................................................................................................................... 60
5-2 未來展望....................................................................................................................... 61
參考文獻................................................................................................................................. 63
圖 索 引
圖2-1 子宮頸癌演化過程(三總病理部提供) ......................................................................... 8
圖2-2 正常子宮頸抹片影像.................................................................................................. 10
圖2-3 橫切線色彩分佈圖...................................................................................................... 16
圖2-4 影像灰階轉換.............................................................................................................. 17
圖2-5 區域成長...................................................................................................................... 17
圖2-6 ROC 曲線分佈圖.......................................................................................................... 20
圖3-1 實驗流程圖.................................................................................................................. 22
圖3-2 角化與不成熟細胞影像.............................................................................................. 24
圖3-3 染色分析方法.............................................................................................................. 24
圖3-4 細胞核輪廓圈選流程.................................................................................................. 25
圖3-5 影像量化比較.............................................................................................................. 26
圖3-6 區域成長判斷過程...................................................................................................... 27
圖3-7 輪廓挑選...................................................................................................................... 27
圖4-1 不同型態細胞之RGB 分佈觀察................................................................................ 30
圖4-2 正異常細胞之B>G 區域............................................................................................ 31
圖4-3 融合型異常細胞影像.................................................................................................. 32
圖4-4 不同QN 之量化結果.................................................................................................... 33
圖4-5 不同之圈選結果........................................................................................................... 34
圖4-6 ROC 曲線示意圖.......................................................................................................... 37
圖4-7 橫切線色彩分布圖比較結果...................................................................................... 41
圖4-8 量化影像上之曲線分佈比較...................................................................................... 41
圖4-9 LSIL 與HSIL 細胞核紋理參數之平均值與標準差分布.......................................... 42
圖4-10 紋理參數之ROC 曲線示意圖.................................................................................. 43
圖4-11 不同型態細胞核之RGB 與面積分佈圖.................................................................. 44
圖4-12 細胞核在B>G 區域所佔比例.................................................................................. 45
圖4-13 細胞聚集判斷............................................................................................................ 46
圖4-14 無法正確辨識細胞核數之HSIL 影像..................................................................... 47
圖4-15 正常抹片影像細胞核位置判斷結果........................................................................ 48
圖4-16 細胞核區域位置選取................................................................................................ 49
圖4-17 不同量化之處理範圍比較........................................................................................ 50
圖4-18 正常抹片影像自動圈選結果.................................................................................... 51
圖4-19 測試不同解析度之影像............................................................................................ 51
圖4-20 不同解析度自動圈選數量........................................................................................ 52
圖4-21 不同比例HSIL 影像抹片細胞核圈選誤差............................................................. 52
圖4-22 自動處理流程............................................................................................................ 53
圖4-23 正常抹片誤判原因.................................................................................................... 54
圖4-24 正常細胞核誤判影像................................................................................................ 54
圖4-25 血球聚集其圈選面積與其所處區域........................................................................ 55
圖4-26 異常判斷為正常之細胞核........................................................................................ 56
圖4-27 抹片影像選擇............................................................................................................ 58
圖4-28 處理抹片影像............................................................................................................ 58
圖4-29 測試細胞核圈選介面................................................................................................ 59
圖4-30 細胞核型態自動分類與檢示參數............................................................................ 59
表 索 引
表2-1 抹片上可見的正常細胞特徵...................................................................................... 13
表2-2 抹片上細胞不正常變化其細胞核的平均大小........................................................... 13
表2-3 LSIL 與HSIL 之組織學分類...................................................................................... 14
表2-4 決策矩陣...................................................................................................................... 19
表4-1 各類細胞其B>G 區域之統計.................................................................................... 31
表4-2 不同量化結果之正確圈選統計.................................................................................. 35
表4-3 各組之參數的平均值與標準差.................................................................................. 36
表4-4 正常與各類異常細胞核參數之P 值.......................................................................... 36
表4-5 正常與各類異常細胞核參數之ROC 曲線面積........................................................ 37
表4-6 各敏感度之操作點(閥值)與正確率關係表............................................................... 38
表4-7 正常與異常細胞核參數之ROC 曲線面積................................................................ 39
表4-8 正常與異常細胞核參數閥值組合之分組.................................................................. 39
表4-9 各組閥值在正常與異常細胞之參數過濾結果.......................................................... 40
表4-10 LSIL 與HSIL 細胞核紋理參數之平均值與標準差統計........................................ 42
表4-11 LSIL 與HSIL 細胞核紋理參數之P 值.................................................................... 43
表4-12 LSIL 與HSIL 細胞核紋理參數之ROC 曲線面積.................................................. 43
表4-13 LSIL 與HSIL 紋理參數關係表................................................................................ 43
表4-14 正常與異常細胞核在面積與RGB 參數之ROC 曲線面積.................................... 45
表4-15 在敏感度=1 時之操作點與正常細胞核過濾結果................................................... 45
表4-16 細胞核在B>G 區域中所佔面積.............................................................................. 46
表4-17 不同病變程度細胞核與量化區域之統計................................................................ 49
表4-18 正常與異常抹片辨識結果評估................................................................................ 55
表4-19 異常抹片辨識結果評估............................................................................................ 56

[1]http://www.doh.gov.tw/statistic/data/死因摘要/93年/93.htm

[2]Vladimir Dadeshidze, Lars J. Olsson, and Richard A.Domanik , ”Segmentation of nuclear images in automated cervical cancer screening,”,SPIE.Society for Optical Engineering, Vol.2622, pp.723-727,1995

[3]R. Narayanswamy, ”Morphological feature detection for cancer screening”,SPIE. Society for Optical Engineering, Vol.2424, pp.265-275,1995

[4]Jean-Philippe Thiran, Benoit Macq, ” Morphological Feature Extraction for the Classification of Digital Images of Cancerous Tissues”, IEEE Transactions on Biomedical Engineering, Vol.43,pp.1011-1020,1996.

[5]S.W.Chan, K.S.Leung, W.S.Wong, “An expert system for the detection of cervical cancer cells using knowledgebased image analyzer”, Artificial Intelligence in Medicine, Vol.8, pp.67-90,1996.

[6]“Trans-vaginal color Doppler for predicting pathological response to preoperative chemo-radiation in locally advanced cervical carcinoma: A Preliminary studt”, Ultrasound in Medicine and Biology, Vol.25, No.7, pp.1941-1945,1999.

[7]阮正雄、傅曉喜與徐光正，子宮頸抹片與陰道鏡圖譜，嘉洲出版社，民國91年1月，台北

[8]黃博琪，”子宮頸細胞自動化電腦輔助影像分析系統之設計研發”， 中興大學生物醫學研究所碩士論文，民國九十二年。

[9]陳淑嬌，”梯度向量流主動輪廓模型於子宮頸抹片細胞邊界偵測之應用”，中興大學電機工程學系碩士論文。

[10]馮齡儀，”電腦輔助子宮頸抹片異常細胞辨識之初期研究”，中原大學醫學工程研究所碩士論文，民國九十三年。

[11]吳香達，子宮頸癌，茂昌圖書公司，民國91年9月，台北。

[12]Richart RM. Cervical intraepithelial neoplasia: A review. InSimmers SC(ed): Pathology Annual, Vol 8. Norwalk, Connecticut. Appleton-Century-Crofts, 1973, p. 301.

[13]繆紹綱，數位影像處理 活用-Matlab，全華科技圖書股份有限公司，p8-2∼8-28，民國八十八年。

[14]林宸生，數位信號-影像與語音處理，全華出版社，台北，民國86年12月。

[15]Metz CE. Some practical issues of experimental design and data analysis in radiological ROC studies. Invest Radiol 1989;24:234-245.

[16]Beth Dawson-Saunders and Robert G. Trapp, Basic and Clinical Biostatistics, Prentice-Hall International Inc., p233,1994.