臺灣博碩士論文加值系統

腦下垂體是身體中激素的主要控制中心，它位於顱骨底之一個名為碟鞍的空隙，在眾多腦腫瘤中是腦下垂體瘤是屬於生長較為緩慢的，腦下垂體的腺瘤症狀可被分類如下: 激素的生產過量、腦下垂體機能不足、視野缺乏、非特異性特徵，本文所要探討的肢端肥大症主要是由於垂體分泌生長素的細胞出現增生所導致，這種症狀會造成腦下垂體分泌過多的生長素，在青春期之前，因為骨骺尚未閉合，因此會引起巨人症；而在青春期之後，則會導致肢端肥大症的發生。
由於樣本數量眾多，醫護人員在計算細胞核顆數時，常常需要耗費大量的時間以及人力來判斷結果好壞，所以本篇論文希望藉由影像處理的技術提升醫護人員計算上的效率，首先，我們先從R、G、B三種灰階影像找出較適合的一種，再針對這張影像執行去背景、去雜訊等步驟，接著將單顆及多顆結合的細胞核分開來，針對多顆結合的細胞核利用Opening演算法及watershed演算法進行切割的動作，最後計算影像中細胞核的個數。
本篇論文所使用的腦下垂體瘤影像皆來自台中榮民總醫院精神科，這些腦下垂體瘤影像藉由本實驗中提出的計數方法，在Precision、Recall及F-measure三種評估方法皆能達到80%以上的結果。

The pituitary gland is the hormone primary control center in the body. It is located at the sella-turcica which is in the bottoms crevice of the saddle skull. In the multitudinous brain tumors, it grows slowly. The Adenoma of Pituitary’s symptom may classify as follows: The hormone production excessive、The pituitary body function insufficiency、The field of vision lack、The non-specificity characteristic. Acromegaly is the symptom, which this study wants to discuss, is caused by the gland pituitary cell that secretes growth hormone appears hyperplasia. This kind of symptoms can cause the pituitary gland to secrete growth hormone excessively. If the symptom occurred before the puberty, it would cause gigantism because the epiphyseal wasn’t closure. If the symptom occurred after the puberty, it would cause Acromegaly.
Because the quantity of the samples is multitudinous, the health care will often require a lot of time and manpower to judge good and bad results when they calculate the number of the cells. This study hopes to use the technology of image processing to enhance the efficiency of health care in calculating. First, we start with the R, G, B three kinds of grayscale images to find a more suitable one. Then we use this image to execute some steps like remove the non-cells part and remove noise. Next, we separate the single cell and multiple cells, and we focus on the multiple cells. We use the Opening algorithm and the watershed algorithm to segment the multiple cells. Finally, we calculate the number of the cells in the image.
The pituitary gland images which this study used come from Taichung Veterans General Hospital Psychiatry. The pituitary gland images used our method can get more than 80% of the results in the Precision、Recall and F-measure.

摘要..................................i
Abstract..............................ii
List of Tables........................v
List of Figures.......................vi
Chapter 1 Introduction................1
Chapter 2 Related Work................6
2.1. Otsu’s Thresholding...........6
2.2. Region Labeling...............8
2.3. Mathematical morphology......12
2.4. Watershed....................14
2.5. Genetic Algorithms...........15
2.6. Evaluation System............16
Chapter3 Adenoma of Pituitary Cell Counting Method (APCC)
.............................18
3.1. The Segmentation of Cells and Non-Cells
.............................18
3.2. The Segmentation of Multiple Cells
.............................23
3.3. Genetic-Based Diagnosing Parameter Detector (GBSPD) .............................30
Chapter4 Experimental Results and Discussion
.............................34
4.1. The Segmented Result of APCC
.............................34
4.2. Experimental Result of APCC
.............................37
Chapter5 Conclusion..................41
Reference............................43

[1]A.F. Daly, M. Rixhon, C. Adam, et al. "High prevalence of Liege, pituitary adenomas: a cross-sectional study in the province of Belgium".J Clin Endocrinol Metab 2006; 91: 4769-75.
[2]A.H. Fischer, et al. "Hematoxylin and eosin staining of tissue and cell sections." Cold Spring Harbor Protocols 2008.5 (2008): pdb-prot4986.
[3]C. Li, C. Xu, C. Gui, and M. D. Fox, "Level Set Evolution Without Re-initialization: A New Variational Formulation", CVPR 2005.
[4]H.J.Schneider, C.Sievers, B.Saller, et al." High prevalence of biochemical acromegaly in primary care patients with elevated IGF-1 levels". Clin Endocrinol (Oxf). 2008; 69:432-5.
[5]I.M.Holdaway, C.Rajasoorya. "Epidemiology of acromegaly." Pituitary 1999; 2: 29-41.
[6]J.Serra"Image Analysis and Mathematical Morphology", ISBN 0-12-637240-3 (1982)
[7]L.Najman and M.Schmitt."Geodesic Saliency of Watershed contours and hierarchical segmentation".In IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 18, Num. 12 (1996), pages 1163–1173.
[8]N.Otsu,"A Threshold Selection Method from Gray-level Histogram," IEEE Transactions on System Man Cybernetics, vol. 9, no. 1, pp. 62-66, 1979.
[9]R.L.Haupt and S.E.Haupt,"Practical Genetic Algorithms," 2nd ed. Hoboken, N.J., John Wiley, 2004.
[10]S.H. Chiu, et al. "肢端肥大症." 內科學誌 22.1 (2011): 9-18.
[11]Wikipedia,"腦下垂體腫瘤" From https://zh.wikipedia.org/wiki/%E8%85%A6%E4%B8%8B%E5%9E%82%E9%AB%94
[12]Wikipedia,"肢端肥大症" From https://zh.wikipedia.org/wiki/%E8%82%A2%E7%AB%AF%E8%82%A5%E5%A4%A7%E7%97%87
[13]Wikipedia,"H&E Stain" From
https://en.wikipedia.org/wiki/H%26E_stain
[14]Wikipedia,"Mathematical morphology" From https://en.wikipedia.org/wiki/Mathematical_morphology
[15]Wikipedia,"Precision and recall" From https://en.wikipedia.org/wiki/Precision_and_recall
[16]Wikipedia,"Watershed" From https://en.wikipedia.org/wiki/Watershed_(image_processing)