臺灣博碩士論文加值系統

阿茲海默症是世界上造成老年人口失智的主要病情之一。目前臨床上透過一系列問診和測驗的方式進行診斷，甚至需利用侵入式理學檢查來進一步確認，造成就醫程序的繁瑣與負擔。人工智慧方法在電腦輔助診斷上的成功，使得醫學影像在檢測疾病的應用層面得以有所突破。因此希望藉由不同的人工智慧模型利用非侵入且快速的醫療影像讓阿茲海默症的診斷不再是一項繁瑣而複雜的流程。
在這項研究中，我們使用三種機器學習與三種深度學習的遷移學習模型，各別接上二元分類器分別對視網膜的前三層和前三層混合數據集進行統計分析，並以視網膜厚度圖像分類阿茲海默症患者。最終，統計分析表明與對照組相比患者的視網膜第二、三層的厚度在特定區塊上有顯著差異；在只考慮單層數據集的三種機器學習模型分類效果是不理想的；最終在深度學習模型搭配上第二與第三層的混合數據集在分類阿茲海默症的正確率達97%，實驗結果期望成為臨床醫學分類阿茲海默症的一項輔助診斷。

Alzheimer's disease is one of the leading causes of dementia in the elderly population in the world. At present, dementia is diagnosed through a series of consultations and tests, and even invasive physical examinations are required to confirm the diagnosis. These steps make medical procedures cumbersome and burdensome. The success of Artificial Intelligence (AI) methods in computer aided diagnosis has made a breakthrough in the application of medical imaging for early-symptom detection for disease. Therefore, it is hoped that the diagnosis of Alzheimer's disease will no longer be a tedious process by using non-invasive and rapid medical imaging through different AI models.
In this study, we adopted three machine learning models and three deep transfer learning models, respectively connected with binary classifiers for the first three layers of the retina and mixed layers of the first three layers of the retina for statistical analysis and classification of Alzheimer's disease patients. Finally, statistical analysis showed significant differences in specific sectors of the second and third layers of the patients compared to the control group; three machine learning models can illustrate the unsatisfactory effect of considering only single layer datasets; the deep learning model paired with the second and third layers datasets has an accuracy rate of 97% in classifying Alzheimer's disease. The experimental results are expected to become an auxiliary diagnosis for clinical medical classification of Alzheimer's disease.

摘要 i
Abstract ii
目錄 iii
圖目錄 vi
表目錄 ix
第一章 序論 1
1.1前言 1
1.2文獻探討 1
1.3研究動機 2
1.4論文架構 3
第二章 資料預處理 4
2.1 光學同調斷層掃描術成像原理 4
2.2 視網膜厚度圖與預處理 5
2.3資料擴增 6
2.4網格數字自動辨識 8
2.5 模型評估 10
第三章 基礎理論 12
3.1 類神經網路 12
3.1.1 人工神經網路 12
3.1.2神經元模型 12
3.1.3 多層感知器 13
3.2 機器學習 14
3.2.1支援向量機 14
3.2.2 隨機森林 15
3.2.3 羅吉斯回歸 15
3.3 深度學習 16
3.3.1 卷積層 16
3.3.2 池化層 17
3.3.3 全連接層 18
3.4 遷移學習 18
3.4.1 Visual Geometry Group架構 19
3.4.2 Deep Residual Network架構 20
3.4.3 Inception架構 21
第四章 實驗結果與討論 22
4.1 實驗平台介紹 22
4.2測試資料 23
4.2.1 資料收集流程 23
4.2.2 資料擴增與數據集驗證 24
4.3統計分析結果與討論 27
4.4機器學習分類結果與討論 31
4.4.1 單層數據集 31
4.4.2 混合層數據集 33
4.5深度學習分類結果與討論 37
第五章 結論 43
參考文獻 45
附錄 49

[1]K. Blennow, M. J. de Leon, and H. Zetterberg, “Alzheimer's disease,” The Lancet, vol. 368, no. 9533, pp. 387-403, July 2006.
[2]M. Prince, E. Albanese, M. Guerchet, and M. Prina, World Alzheimer report 2014: dementia and risk reduction: an analysis of protective and modifiable risk factors. Alzheimer’s Disease International (ADI), 2014.
[3]M. N. Sabbagh, L.-F. Lue, D. Fayard, and J. Shi, “Increasing precision of clinical diagnosis of Alzheimer's disease using a combined algorithm incorporating clinical and novel biomarker data,” Neurology and Therapy, vol. 6, no. 1, pp. 83-95, July 2017.
[4]Jack, C. R., Jr., Bennett, D. A., Blennow, K., Carrillo, M. C., Dunn, B., Haeberlein, S. B., Holtzman, D. M., Jagust, W., Jessen, F., Karlawish, J., Liu, E., Molinuevo, J. L., Montine, T., Phelps, C., Rankin, K. P., Rowe, C. C., Scheltens, P., Siemers, E., Snyder, H. M., & Sperling, R., “NIA‐AA research framework: toward a biological definition of Alzheimer's disease,” Alzheimer's & Dementia, vol. 14, no. 4, pp. 535-562, April 2018.
[5]S. Y. Ong, C. Y Cheung, X. Li, E. L Lamoureux, M. K. Ikram, J. Ding, C. Y. Cheng, B. A. Haaland, S. M. Saw, N. Venketasubramanian, C. P. L. Chen, T. Y.Wong, “Visual impairment, age-related eye diseases, and cognitive function: the Singapore Malay Eye study,” Archives of Ophthalmology, vol. 130, no. 7, pp. 895-900, July 2012.
[6]C. Y. Cheung, M. K. Ikram, C. Chen, and T. Y. Wong, “Imaging retina to study dementia and stroke,” Progress in Retinal and Eye Research, vol. 57, pp. 89-107, March 2017.
[7]J. C. Blanks, S. Y. Schmidt, Y. Torigoe, K. V. Porrello, D. R. Hinton, and R. H. Blanks, “Retinal pathology in Alzheimer's disease. II. Regional neuron loss and glial changes in GCL,” Neurobiology of Aging, vol. 17, no. 3, pp. 385-395, May 1996.
[8]J. C. Blanks, Y. Torigoe, D. R. Hinton, and R. H. Blanks, “Retinal pathology in Alzheimer's disease. I. Ganglion cell loss in foveal/parafoveal retina,” Neurobiology of Aging, vol. 17, no. 3, pp. 377-384, May 1996.
[9]J. S. Schuman, T. P. Kloizman, H. Pakter, N. Wang, V. Guedes, L. Huang, L. Pieroth, W. Scott, M. R. Hee, J. G. Fujimoto, H. Ishikawa, R. A. Bilonick, L. Kagemann, G. Wollstein, “Optical coherence tomography and histologic measurements of nerve fiber layer thickness in normal and glaucomatous monkey eyes,” Investigative Ophthalmology & Visual Science, vol. 48, no. 8, pp. 3645-3654, August 2007.
[10]L. Guo, J. Duggan, and M. Cordeiro, “Alzheimer's disease and retinal neurodegeneration,” Current Alzheimer Research, vol. 7, no. 1, pp. 3-14, February 2010.
[11]Y. Lu, Z. Li, X. Zhang, B. Ming, J. Jia, R. Wang, D. Ma, “Retinal nerve fiber layer structure abnormalities in early Alzheimer's disease: evidence in optical coherence tomography,” Neuroscience Letters, vol. 480, no. 1, pp. 69-72, August 2010.
[12]D. A. Valenti, “Neuroimaging of retinal nerve fiber layer in AD using optical coherence tomography,” Neurology, vol. 69, no. 10, pp. 1060-1060, September 2007.
[13]Y. Ito, M. Sasaki, H. Takahashi, S. Nozaki, S. Matsuguma, K. Motomura, R. Ui, R. Shikimoto, R. Kawasaki, K. Yuki, N. Sawada, M. Mimura, K. Tsubota, S. Tsugane “Quantitative assessment of the retina using OCT and associations with cognitive function,” Ophthalmology, vol. 127, no. 1, pp. 107-118, January 2020.
[14]E. H.-T. Loh, Y.-T. Ong, N. Venketasubramanian, S. Hilal, N. Thet, T. Y. Wong, C.P. L. Chen and C. Y.-L. Cheung, “Repeatability and reproducibility of retinal neuronal and axonal measures on spectral-domain optical coherence tomography in patients with cognitive impairment,” Frontiers in Neurology, vol. 8, p. 359, August 2017.
[15]U. Schmidt-Erfurth, A. Sadeghipour, B. S. Gerendas, S. M. Waldstein, and H. Bogunović, “Artificial intelligence in retina,” Progress in Retinal and Eye Research, vol. 67, pp. 1-29, November 2018.
[16]D. S. Kermany, M. Goldbaum, W. Cai, C. C.S. Valentim, H. Liang, S. L. Baxter, A. McKeown, G. Yang, X. Wu, F. Yan, J. Dong, M. K. Prasadha, J. Pei, M. Y.L. Ting, J. Zhu, C. Li, S. Hewett, J. Dong, I. Ziyar, A. Shi, R. Zhang, L. Zheng, R. Hou, W. Shi, X. Fu, Y. Duan, V. A.N. Huu, C. Wen, E. D. Zhang, C. L. Zhang, O. Li, X. Wang, M. A. Singer, X. Sun, J. Xu, A. Tafreshi, M. A. Lewis, H. Xia, K. Zhang, “Identifying medical diagnoses and treatable diseases by image-based deep learning,” Cell, vol. 172, no. 5, pp. 1122-1131, February 2018.
[17]D.-K. Hwang, C.-C. Hsu, K.-J. Chang, D. Chao, C.-H. Sun, Y.-C. Jheng, A. A. Yarmishyn, J.-C. Wu, C.-Y. Tsai, M.-L. Wang, C.-H. Peng, K.-H. Chien, C.-L. Kao, T.-C. Lin, L.-C. Woung, S.-J. Chen, S.-H. Chiou, “Artificial intelligence-based decision-making for age-related macular degeneration,” Theranostics, vol. 9, no. 1, p. 232, January 2019.
[18]T. Schlegl, S. M Waldstein, H. Bogunovic, F. Endstraßer, A. Sadeghipour, A.-M. Philip, D. Podkowinski, B. S. Gerendas, G. Langs, U. Schmidt-Erfurth, “Fully automated detection and quantification of macular fluid in OCT using deep learning,” Ophthalmology, vol. 125, no. 4, pp. 549-558, April 2018.
[19]A. A. Jammal, A. C. Thompson, E. B. Mariottoni, S. I. Berchuck, C. N. Urata, T. Estrela, S. M. Wakil, V. P. Costa, F. A. Medeiros, “Human versus machine: comparing a deep learning algorithm to human gradings for detecting glaucoma on fundus photographs,” American journal of ophthalmology, vol. 211, pp. 123-131, March 2020.
[20]D. Le, M. Alam, Cham K. Yao, J. I. Lim, Y.-T. Hsieh, R. V. P. Chan, D. Toslak, X. Yao, “Transfer learning for automated OCTA detection of diabetic retinopathy,” Translational Vision Science & Technology, vol. 9, no. 2, pp. 35-35, July 2020.
[21]K. M. Lee, E. J. Lee, T.-W. Kim, and H. Kim, “Comparison of the abilities of SD-OCT and SS-OCT in evaluating the thickness of the macular inner retinal layer for glaucoma diagnosis,” PLoS ONE, vol. 11, no. 1, p. e0147964, January 2016.
[22]E. Marziani, S. Pomati, P. Ramolfo, M. Cigada, A. Giani, C. Mariani, G. Staurenghi, “Evaluation of retinal nerve fiber layer and ganglion cell layer thickness in Alzheimer's disease using spectral-domain optical coherence tomography,” Investigative Ophthalmology & Visual Science, vol. 54, no. 9, pp. 5953-5958, September 2013.
[23]江鴻志、張正宜, “OCT技術概述與生醫應用.” Retrieved from https://www.materialsnet.com.tw/DocView.aspx?id=27579/, 2022
[24]K. Gurney, An Introduction to Neural Networks. CRC Press Taylor & Francis Group, 1997.
[25]W.S. McCulloch, W. A. Pitts, “Logical calculus of the ideas immanent in nervous activity,” Bulletin of Mathematical Biophysics, pp.115-133, December 1943.
[26]M.-C. Popescu, V.E. Balas, L. P. Popescu, and N. Mastorakis, “Multilayer perceptron and neural networks,” WSEAS Transactions on Circuits and Systems, pp.579-588, July 2009.
[27]M.A. Hearst, S.T. Dumais, E. Osuna, J. Platt, B. Scholkopf, “Support vector machines,” IEEE Intelligent Systems and their Applications, vol. 13, no. 4, pp. 18-28, July 1998.
[28]L. Breiman, “Random forests,” Machine Learning 45.1: 5-32, pp. 5-32, October 2001.
[29]D. R. Cox, “The regression analysis of binary sequences.” Journal of the Royal Statistical Society: Series B (Methodological), 20(2), 215–232, pp.215-232, July 1958.
[30]S. Skansi, Introduction to Deep Learning from Logical Calculus to Artificial Intelligence. Springer Cham, 2018
[31]C. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens, Z. Wojna, “Rethinking the Inception Architecture for Computer Vision,” in Proceedings of the Computer Vision and Pattern Recognition Conference (CVPR), Las Vegas, NV, USA: IEEE, 2016.
[32]K. Simonyan and A. Zisserman, “Very deep convolutional networks for large-scale image recognition,” in Proceedings of the Computer Vision and Pattern Recognition Conference (CVPR), Columbus, OH, USA: IEEE, 2014.
[33]K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the Computer Vision and Pattern Recognition Conference (CVPR), Las Vegas, NV, USA: IEEE, 2016.
[34]L. Gao, Y. Liu, X. Li, Q. Bai, and P. Liu, “Abnormal retinal nerve fiber layer thickness and macula lutea in patients with mild cognitive impairment and Alzheimer's disease,” Archives of Gerontology and Geriatrics, vol. 60, no. 1, pp. 162-167, January 2015.
[35]Z. Shi, X. Ca, J. Hu, L. Jiang, X. Mei, H. Zheng, Y. Chen, M. Wang, J. Cao, W. Li, T. Li, C. Li, Y. Shen, “Retinal nerve fiber layer thickness is associated with hippocampus and lingual gyrus volumes in nondemented older adults,” Progress in Neuro-Psychopharmacology and Biological Psychiatry, vol. 99, November 2019.
[36]V. T. Chan, Z. Sun, S. Tang, L. J. Chen, A. Wong, C. C. Tham, T. Y. Wong, C. Chen, M. K. Ikram, H. E. Whitson, E. M. Lad, V. C. T. Mok, C. Y. Cheung, “Spectral-domain OCT measurements in Alzheimer’s disease: a systematic review and meta-analysis,” Ophthalmology, vol. 126, no. 4, pp. 497-510, April 2019.
[37]C. Y. Cheung, V. Mok, P. J. Foster, E. Trucco, C. Chen, and T. Y. Wong, “Retinal imaging in Alzheimer’s disease,” Journal of Neurology, Neurosurgery & Psychiatry, pp. 983-994, September 2021.
[38]L.P. Cunha, L.C. Lopes, L.V. Costa-Cunha, C.F. Costa, L.A. Pires, A.L. Almeida, M.L. Monteiro, “Macular thickness measurements with frequency domain-OCT for quantification of retinal neural loss and its correlation with cognitive impairment in Alzheimerʼs disease,” PLoS One, vol. 11, no. 4, p. e0153830, April 2016.
[39]E. Garcia-Martin, M. P. Bambo, M. L. Marques, M. Satue, S. Otin, J. M. Larrosa, V. Polo, L. E. Pablo, “Ganglion cell layer measurements correlate with disease severity in patients with Alzheimer's disease,” Acta Ophthalmologica, vol. 94, no. 6, pp. e454-e459, February 2016.
[40]E. Salobrar-García, R. de Hoz, A.I. Ramírez, I. López-Cuenca, P. Rojas, R. Vazirani, C. Amarante, R. Yubero, P. Gil, M.D. Pinazo-Durán, J.J. Salazar, J.M. Ramírez, “Changes in visual function and retinal structure in the progression of Alzheimer's disease,” PLoS One, vol. 14, no. 8, p. e0220535, August 2019.
[41]S.M. Golzan, K. Goozee, D. Georgevsky, A. Avolio, P. Chatterjee, K. Shen, V. Gupta, R. Chung, G. Savage, C.F. Orr, R.N. Martins, S.L. Graham, “Retinal vascular and structural changes are associated with amyloid burden in the elderly: ophthalmic biomarkers of preclinical Alzheimer’s disease,” Alzheimer's Research & Therapy, vol. 9, no. 1, pp. 1-9, March 2017.
[42]I. López-Cuenca, R. de Hoz, E. Salobrar-García, L. Elvira-Hurtado, P. Rojas, J.A. Fernández-Albarral, A. Barabash, J.J. Salazar, A.I. Ramírez, J.M. Ramírez, “Macular thickness decrease in asymptomatic subjects at high genetic risk of developing Alzheimer’s disease: An OCT study,” Journal of Clinical Medicine, vol. 9, no. 6, pp. 1728, June 2020.
[43]C.Y. Cheung, Y.T. Ong, S. Hilal, M.K. Ikram, S. Low, Y.L. Ong, N. Venketasubramanian, P. Yap, D. Seow, C.L. Chen, T.Y. Wong, “Retinal ganglion cell analysis using high-definition optical coherence tomography in patients with mild cognitive impairment and Alzheimer's disease,” Journal of Alzheimer's Disease, vol. 45, no. 1, pp. 45-56, January 2015.