隨著科技不斷的進步，在量測技術上，也逐漸朝著精準、快速、高穩定性和非侵入式的掃描來發展。光學同調斷層掃描(Optical Coherence Tomography, OCT)是一種無侵入式的低同調性光學成像技術干涉儀(Low Coherence Interferometry, LCI)，它已應用於許多生物醫學成像問題，透過將光打入樣本後反射出來的訊號做解調來獲得橫截面或三維圖像。
然而並不是每一次掃描所得到的樣本資訊都是清楚的，所以本論文我們透過使用深度學習(Deep Learning, DL)的Pix2Pix生成對抗網路(Generative Adversarial Network, GAN)結合OCT的技術來加強圖像品質。透過使用Pix2Pix GAN對大量有嚴重色散的原始圖像(Source image)和清晰的目標圖像(Target image)進行訓練與驗證，經過深度學習訓練之後的模型可以幫我們把所得到的嚴重色散的原始圖像生成出清晰的生成圖像(Generated image)，論文的後面還會去比較生成對抗網路是否會因為訓練圖片數據量的多寡和訓練的次數而影響轉換出的圖片好壞。

With the continuous advancement of technology, the measurement techniques are gradually developing towards accurate, fast, highly stable and non-invasive scans. Optical Coherence Tomography (OCT) is a non-invasive Low-Coherence Interferometer(LCI) that has been applied to many biomedical imaging problems to obtain cross-sectional or three-dimensional images by demodulating the signal reflected from the light hitting the sample.
However, not all of the information in the sample from each scan is clear. In this paper, we use Deep Learning (DL) for Pix2Pix Generative Adversarial Network (GAN) combined with OCT to enhance the image quality. By using Pix2Pix GAN to train and validate a large number of severely dispersive source images and clear target images, the model after deep learning can help us to generate a clear generated image from the obtained severely dispersive source images. Later in the paper, we will compare whether the generative adversarial network will affect the quality of the transformed images due to the amount of training data and the number of training epoch.

摘要 ...i
Abstract ...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 研究動機和目的 ...2
1.3 論文架構 ...2
第二章 光學同調斷層掃描技術背景與原理介紹 ...3
2.1 光學同調斷層掃描(Optical Coherence Tomography,OCT) ...3
2.2 時域光學同調斷層掃描(Time-domain OCT) ...5
2.3 頻域光學同調斷層掃描(Frequency-domain OCT) ...7
第三章 Pix2Pix GAN背景與原理介紹和文獻探討 ...11
3.1 Pix2Pix GAN介紹 ...11
3.2 生成器架構 ...12
3.3 鑑別器架構 ...14
3.4 卷積層結構 ...15
3.5 圖像的定量分析 ...15
3.6 文獻探討 ...17
第四章 實驗內容與結果 ...23
4.1 實驗儀器 ...23
4.2 實驗架構 ...25
4.2.1 使用生成對抗網路技術於頻域光學同調斷層掃描圖像加強系統 ...25
4.2.2 準備數據群組 ...26
4.3 實驗結果 ...27
4.3.1 對數據群組一進行訓練 ...27
4.3.2 對數據群組二進行訓練 ...33
第五章 結論 ...39
參考文獻 ...40
Extended Abstract ...42

[1] D. Huang, E. A. Swanson, C. P. Lin, J. S. Schuman, W. G. Stinson, W. Chang, M. Hee, T. Flotte, K. Gregory, C. A. Puliafito, and J. G. Fujimoto, “Optical coherence tomography,” Science, vol.254, pp. 1178-1181, 1991.
[2] P. A. Flournoy, R. W. McClure, and G. Wyntjes, “White-Light Interferometric Thickness Gauge,” Appl. Opt., vol.11, pp.1907-1915, 1972.
[3] A. F. Fercher, C. K. Hitzenberger,“Measurement of intraocular distances by backscattering spectral interferometry,” Opt. Commun, vol. 117, pp. 443-448, 1995.
[4] P. Isola, J. Y. Zhu, T. Zhou, and A. A. Efros, ‘‘Image-to-image translation with conditional adversarial networks,’’ IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 5967-5976, 2017.
[5] A. Mino, and G. Spanakis, “LoGAN: Generating Logos with a Generative Adversarial Neural Network Conditioned on color,” IEEE International Conference on Machine Learning and Applications (ICMLA), pp. 965-970, 2018.
[6] O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” MICCAI, vol. 9351, 2015.
[7] Q. Hao, K. Zhou, J. Yang, Y. Hu, Z. Chai, Y. Ma, G. Liu, Y. Zhao, S. Gao, and J. Liu, ‘‘High signal-to-noise ratio reconstruction of low bit-depth optical coherence tomography using deep learning,’’ J. Biomed. Opt., vol. 25, no. 12, 2020.
[8] Ugur Demir, Gozde B.Unal, “Patch-Based Image Inpainting with Generative Adversarial Networks,” ArXiv, vol. 1803, 2018.
[9] 維基百科，結構相似性，取自
https://zh.m.wikipedia.org/zh-tw/%E7%B5%90%E6%A7%8B%E7%9B%B8%E4%BC %BC%E6%80%A7
[10] J. G. Fujimoto, C. Pitris, S. A. Boppart, and M. E. Brezinski, "Optical Coherence Tomography: An Emerging Technology for Biomedical Imaging and Optical Biopsy," Neoplasia, vol. 2, no. 1, pp. 9-25, 2000.
[11] S. Asrani, L. Essaid, B. D. Alder, and C. S. Turla, “Artifacts in spectraldomain optical coherence tomography measurements in glaucoma,” JAMA Ophthalmol, vol. 132, no. 4, pp. 396-402, 2014.
[12] J. Lin, S. Zhong, Q. Zhang, and W. Chen, "Five-Frame Variable Phase-Shifting Method for Full-Range Spectral-Domain Optical Coherence Tomography," Applied Sciences, vol. 8, no. 9, 2018.
[13] Y. Zhang et al., "Neural network-based image reconstruction in swept-source optical coherence tomography using undersampled spectral data," Light: Science & Applications, vol. 10, no. 1, p. 155, 2021.
[14] S. Cao et al., "Super-resolution technology to simultaneously improve optical & digital resolution of optical coherence tomography via deep learning," in 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), pp. 1879-1882, 2020.
[15] C.M. Huang, E. Wijanto and H.C. Cheng , “Applying a Pix2Pix Generative Adversarial Network to a Fourier-Domain Optical Coherence Tomography System for Artifact Elimination,” IEEE Access, vol. 9, pp. 103311-103324, 2021
[16] W. Bao, Y. Shen, T. Chen, P. Li, and Z. Ding, “High-speed high-precision and ultralong-range complex spectral domain dimensional metrology,”Opt. Exp., vol. 23, no. 9, 2015.
[17] H.C. Cheng, J.F. Huang, and Y.H. Hsieh, “Numerical analysis of One-Shot and Full Range FD-OCT by Orthogonality of Polarized Light,” Opt. Commun., Vol. 282, pp. 3040-3045, 2009.
[18] X. Liu, M. Ke, X. Yao, J. Chua, L. Schmetterer, and B. Tan, "Stable complex conjugate artifact removal in OCT using circularly polarized light as reference," Optics Letters, vol. 45, no. 14, pp. 3977-3980, 2020.
[19]許明軒，2011, “使用正交極化光於全域式頻率域光學同調斷層掃描之研究與實驗驗證”，國立虎尾科技大學光電與材料科技研究所論文
[20] Y. Zhu and W. R. Gao, “Single-shot wavelength-independent phase-shifting method for full-field optical coherence tomography,” Appl. Opt, vol. 58, p. 806, 2019.
[21] T. Zhang, J. Cheng, H. Fu, Z. Gu, Y. Xiao, K. Zhou, S. Gao, R. Zheng, and J. Liu, “Noise adaptation generative adversarial network for medical image analysis,” IEEE Trans. Med. Imag, vol. 39, no. 4, pp. 1149-1159, Apr. 2020.
[22] Y.L. Yang, Z.H. Ding, K. Wang, L. Wu, and L. Wu, “Full-field optical coherence tomography by achromatic phase shifting with a rotating half-wave plate,” J. Opt.,vol 12, 2010
[23] S. K. Devalla et al., "A Deep Learning Approach to Denoise Optical Coherence Tomography Images of the Optic Nerve Head," Scientific Reports, vol. 9, no. 1, p. 14454, 2019.
[24] H.C. Cheng and C.T. Huang, “Measurement of Thickness and Refractive Index of Optical Samples Simultaneously Using Full-Range One-Shot Frequency-Domain Optical Coherence Tomography,” Fiber and Integrated Optics, Vol. 34, Iss. 3, pp.171-182, 2015.
[25] H. Cheong et al., "OCT-GAN: single step shadow and noise removal from optical coherence tomography images of the human optic nerve head," Biomedical Optics Express, vol. 12, no. 3, pp. 1482-1498, 2021.
[26] B. Buchroithner, A. Prylepa, P. J. Wagner, S. E. Schausberger, D. Stifter, and B. Heise, "Full-field optical coherence tomography in a balanced detection mode," Applied Optics, vol. 57, no. 29, pp. 8705-8710, 2018.