請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/81900完整後設資料紀錄
| DC 欄位 | 值 | 語言 |
|---|---|---|
| dc.contributor.advisor | 陳中明(Chung-Ming Chen) | |
| dc.contributor.author | Wei-Chieh Huang | en |
| dc.contributor.author | 黃瑋傑 | zh_TW |
| dc.date.accessioned | 2022-11-25T03:06:10Z | - |
| dc.date.available | 2026-10-22 | |
| dc.date.copyright | 2021-11-02 | |
| dc.date.issued | 2021 | |
| dc.date.submitted | 2021-10-22 | |
| dc.identifier.citation | '108年死因統計結果分析 - 衛生福利部。民國109年6月16日。檢自https://www.mohw.gov.tw/mp-1.html.' R. L. Siegel, K. D. Miller, and A. Jemal, 'Cancer statistics, 2019,' CA Cancer J Clin, vol. 69, no. 1, pp. 7-34, Jan 2019. N. J. T. E. P. J. P. Camarlinghi, 'Automatic detection of lung nodules in computed tomography images: training and validation of algorithms using public research databases,' vol. 128, no. 9, pp. 1-21, 2013. H. Li, P. Boiselle, J. Shepard, B. Trotman-Dickenson, and T. J. A. A. j. o. r. McLoud, 'Diagnostic accuracy and safety of CT-guided percutaneous needle aspiration biopsy of the lung: comparison of small and large pulmonary nodules,' vol. 167, no. 1, pp. 105-109, 1996. J. C. Nesbitt, J. B. Putnam Jr, G. L. Walsh, J. A. Roth, and C. F. J. T. A. o. t. s. Mountain, 'Survival in early-stage non-small cell lung cancer,' vol. 60, no. 2, pp. 466-472, 1995. J.-W. Gao et al., 'Pulmonary ground-glass opacity: computed tomography features, histopathology and molecular pathology,' vol. 6, no. 1, p. 68, 2017. N. K. Altorki, D. F. Yankelevitz, M. F. Vazquez, A. Kramer, and C. I. Henschke, 'Bronchioloalveolar carcinoma in small pulmonary nodules: clinical relevance,' in Seminars in thoracic and cardiovascular surgery, 2005, vol. 17, no. 2, pp. 123-127: Elsevier. D. Wu et al., 'Stratified learning of local anatomical context for lung nodules in CT images,' in 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2010, pp. 2791-2798: IEEE. F. Li, S. Sone, H. Abe, H. MacMahon, and K. J. R. Doi, 'Malignant versus benign nodules at CT screening for lung cancer: comparison of thin-section CT findings,' vol. 233, no. 3, pp. 793-798, 2004. N. L. S. T. R. T. J. N. E. J. o. Medicine, 'Reduced lung-cancer mortality with low-dose computed tomographic screening,' vol. 365, no. 5, pp. 395-409, 2011. W. Jiang et al., 'Clinical outcome of subcentimeter non‐small cell lung cancer after surgical resection: Single institution experience of 105 patients,' vol. 110, no. 3, pp. 233-238, 2014. K. J. C. m. i. Doi and graphics, 'Computer-aided diagnosis in medical imaging: historical review, current status and future potential,' vol. 31, no. 4-5, pp. 198-211, 2007. H. Greenspan, B. Van Ginneken, and R. M. J. I. t. o. m. i. Summers, 'Guest editorial deep learning in medical imaging: Overview and future promise of an exciting new technique,' vol. 35, no. 5, pp. 1153-1159, 2016. A. Krizhevsky, I. Sutskever, and G. E. J. A. i. n. i. p. s. Hinton, 'Imagenet classification with deep convolutional neural networks,' vol. 25, pp. 1097-1105, 2012. K. Simonyan and A. J. a. p. a. Zisserman, 'Very deep convolutional networks for large-scale image recognition,' 2014. K. He, X. Zhang, S. Ren, and J. Sun, 'Deep residual learning for image recognition,' in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 770-778. J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei, 'Imagenet: A large-scale hierarchical image database,' in 2009 IEEE conference on computer vision and pattern recognition, 2009, pp. 248-255: Ieee. S. G. Armato III et al., 'The lung image database consortium (LIDC) and image database resource initiative (IDRI): a completed reference database of lung nodules on CT scans,' vol. 38, no. 2, pp. 915-931, 2011. Z. Hussain, F. Gimenez, D. Yi, and D. Rubin, 'Differential data augmentation techniques for medical imaging classification tasks,' in AMIA annual symposium proceedings, 2017, vol. 2017, p. 979: American Medical Informatics Association. I. Goodfellow et al., 'Generative adversarial nets,' vol. 27, 2014. J. J. C. P. f. S. C. N. C. N. N. f. V. R. Gauthier, Winter semester, 'Conditional generative adversarial nets for convolutional face generation,' vol. 2014, no. 5, p. 2, 2014. Y. Jiang and J. J. A. S. Li, 'Generative adversarial network for image super-resolution combining texture loss,' vol. 10, no. 5, p. 1729, 2020. M. J. Chuquicusma, S. Hussein, J. Burt, and U. Bagci, 'How to fool radiologists with generative adversarial networks? A visual turing test for lung cancer diagnosis,' in 2018 IEEE 15th international symposium on biomedical imaging (ISBI 2018), 2018, pp. 240-244: IEEE. Y. Onishi et al., 'Automated pulmonary nodule classification in computed tomography images using a deep convolutional neural network trained by generative adversarial networks,' vol. 2019, 2019. Z. Xu et al., 'Correlation via synthesis: end-to-end nodule image generation and radiogenomic map learning based on generative adversarial network,' 2019. Y. Mirsky, T. Mahler, I. Shelef, and Y. Elovici, 'CT-GAN: Malicious tampering of 3D medical imagery using deep learning,' in 28th {USENIX} Security Symposium ({USENIX} Security 19), 2019, pp. 461-478. C. Han et al., 'Synthesizing diverse lung nodules wherever massively: 3D multi-conditional GAN-based CT image augmentation for object detection,' in 2019 International Conference on 3D Vision (3DV), 2019, pp. 729-737: IEEE. Y. Onishi et al., 'Investigation of pulmonary nodule classification using multi-scale residual network enhanced with 3DGAN-synthesized volumes,' vol. 13, no. 2, pp. 160-169, 2020. W. Shen et al., 'Multi-crop convolutional neural networks for lung nodule malignancy suspiciousness classification,' vol. 61, pp. 663-673, 2017. A. O. de Carvalho Filho, A. C. Silva, A. C. de Paiva, R. A. Nunes, and M. J. P. R. Gattass, 'Classification of patterns of benignity and malignancy based on CT using topology-based phylogenetic diversity index and convolutional neural network,' vol. 81, pp. 200-212, 2018. 陳稜鎔, '電腦斷層掃描肺腫瘤良惡性判別之深度學習影像特徵擷取電腦斷層掃描肺腫瘤良惡性判別之深度學習影像特徵擷取,' 國立台灣大學工學院暨醫學院醫學工程研究所, 2018. K. Liu, G. J. I. J. o. I. S. Kang, and Technology, 'Multiview convolutional neural networks for lung nodule classification,' vol. 27, no. 1, pp. 12-22, 2017. F. Han et al., 'Texture feature analysis for computer-aided diagnosis on pulmonary nodules,' vol. 28, no. 1, pp. 99-115, 2015. K. Jarrett, K. Kavukcuoglu, M. A. Ranzato, and Y. LeCun, 'What is the best multi-stage architecture for object recognition?,' in 2009 IEEE 12th international conference on computer vision, 2009, pp. 2146-2153: IEEE. J. Yu, Z. Lin, J. Yang, X. Shen, X. Lu, and T. S. Huang, 'Free-form image inpainting with gated convolution,' in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2019, pp. 4471-4480. W. Shi et al., 'Real-time single image and video super-resolution using an efficient sub-pixel convolutional neural network,' in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 1874-1883. A. Alekseev and A. Bobe, 'GaborNet: Gabor filters with learnable parameters in deep convolutional neural network,' in 2019 International Conference on Engineering and Telecommunication (EnT), 2019, pp. 1-4: IEEE. T. Abhishree, J. Latha, K. Manikantan, and S. J. P. C. S. Ramachandran, 'Face recognition using Gabor filter based feature extraction with anisotropic diffusion as a pre-processing technique,' vol. 45, pp. 312-321, 2015. L. Chen, G. Lu, and D. Zhang, 'Effects of different Gabor filters parameters on image retrieval by texture,' in 10th International Multimedia Modelling Conference, 2004. Proceedings., 2004, pp. 273-278: IEEE. T. R. Lim and A. T. Guntoro, 'Car recognition using Gabor filter feature extraction,' in Asia-Pacific Conference on Circuits and Systems, 2002, vol. 2, pp. 451-455: IEEE. P. Isola, J.-Y. Zhu, T. Zhou, and A. A. Efros, 'Image-to-image translation with conditional adversarial networks,' in Proceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 1125-1134. T. Miyato, T. Kataoka, M. Koyama, and Y. J. a. p. a. Yoshida, 'Spectral normalization for generative adversarial networks,' 2018. 李雅菁, '一致性結構單元點群飄移:放射線治療肺部電腦斷層掃瞄之多時間點影像對位演算法,' 國立台灣大學工學院暨醫學院醫學工程研究所, 2017. 林均達, '肺部斷層掃描之肺葉分割研究,' 國立台灣大學工學院暨醫學院醫學工程研究所, 2013. R. Adams, L. J. I. T. o. p. a. Bischof, and m. intelligence, 'Seeded region growing,' vol. 16, no. 6, pp. 641-647, 1994. A. F. Frangi, W. J. Niessen, K. L. Vincken, and M. A. Viergever, 'Multiscale vessel enhancement filtering,' in International conference on medical image computing and computer-assisted intervention, 1998, pp. 130-137: Springer. | |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/81900 | - |
| dc.description.abstract | 根據衛福部於民國108年之統計,癌症已連續38年排行台灣國人年度十大死因之首,其中又以肺癌為排行第一之癌症死因。由於近年來電腦斷層掃描技術日益進步,以電腦斷層掃描進行肺癌檢測成為臨床重要之診斷工具,有助於早期發現肺癌病灶並給予治療,進而有效提升肺癌患者之存活率。在肺癌診斷過程中關鍵在於偵測之肺結節良惡性判斷,肺結節之良惡性以將影響臨床診斷及治療手段,因此如何準確判定肺結節之良惡性為主要議題。 臨床上透過放射科醫師針對肺結節進行偵測、標註與良惡性判斷以擬定治療方針,然而根據不同經驗之放射科醫師主觀意識以及臨床上大量判斷之需求,導致結節診斷上面臨到一定的挑戰性。透過電腦輔助診斷(CAD)之判斷可提供具參考之客觀意見輔助醫師於臨床做出診斷及治療決策。目前CAD之建立可分為基於機器學習以及深度學習,主要以分類之特徵來源作為區分。深度學習透過在訓練過程中從訓練樣本中提取特徵並選取分類之依據,有效避免機器學習中有可能遇到之影像前處理與特徵提取中存在之人為誤差。然而深度學習在訓練過程中需要大量之訓練樣本以建立分類模型,在現有公開結節樣本數據庫所具有之樣本數量仍相對不足。 近年來,生成對抗網路模型(GAN)影像生成之能力受到許多研究團隊之重視,GAN建立兩組深度學習網路Generator以及Discriminator在訓練中進行二人非合作零和賽局(two-player non-cooperative zero-sum game)並目標達到納許平衡(Nash equilibrium),透過取得真實樣本之資料分布並在該分布內生成新的樣本以達到生成更佳之影像結果。然而其大多數應用於二維自然影像之生成,於三維醫學影像之生成面臨到不少挑戰,例如肺結節之紋理特徵對於結節良惡性分類有重要的影響力,透過GAN生成之肺結節樣本必須在特徵上有足夠相似之特徵以模擬真實結節樣本。 在生成對抗網路(GAN)技術之發展以及結節樣本數量對於深度學習相對不足之困境下,本研究開發一基於Gabor filter提取結節影像於空間頻域的紋理特徵並作為loss function計算其紋理相似性之Gabor-loss GAN模型,透過以下兩特點:(1)無限制結節樣本邊緣形狀、(2)考慮其紋理資訊之特徵以保持樣本真實性,進行擬真肺結節樣本之生成以克服肺結節樣本數量不足之困境。 在Gabor-loss GAN中,本研究首先將結節標註區域進行分割以保留結節形狀之多樣性;再者將結節標註區域轉置於無結節之肺實質背景中增加結節與背景之多樣性;第三透過使用Gabor filter對結節於空間頻域之紋理特徵進行提取並用於loss function之計算,在生成結節樣本上要求其於空間頻域紋理特徵具有模擬真實結節樣本之能力並同時保有其生成特徵之多樣性。在Gabor-loss GAN中根據訓練樣本之不同型態,可生成屬於良性結節、惡性結節或是具有GGO型態之結節樣本。此外,將基於Gabor-loss以及validation data計算之validation loss作為最佳生成模型之選擇指標。生成之結節樣本用於良惡性分類網路之數據增量,在加入了GAN模型生成之結節樣本作為訓練樣本進行訓練,其分類AUC皆可達0.93以上,其中又以使用了最大數量之數據增量達到最高之AUC為0.950 ± 0.019,其在Accuracy、Sensitivity以及Specificity也分別達到0.883 ± 0.016、0.90 ± 0、0.867 ± 0.034,在相同分類網路架構中成功透過GAN模型生成之結節樣本達到數據增量之目標。 為克服肺結節樣本數量不足之困境,本研究提出Gabor-loss GAN演算法藉由Gabor filter提取結節於空間頻域之紋理特徵進行loss function計算其紋理相似性並根據validation data計算之Gabor-loss作為生成模型選擇指標。透過Gabor-loss GAN生成之結節樣本成功增加結節與肺實質背景之多樣性以及在模擬真實結節具有之紋理特徵下增加結節紋理特徵之多樣性,並應用於結節良惡性分類之數據增量。 | zh_TW |
| dc.description.provenance | Made available in DSpace on 2022-11-25T03:06:10Z (GMT). No. of bitstreams: 1 U0001-2210202101204100.pdf: 5579931 bytes, checksum: c273b03c9962bd12f38884760320c2ca (MD5) Previous issue date: 2021 | en |
| dc.description.tableofcontents | 摘要 I Abstract III 目錄 V 圖目錄 VII 表目錄 IX 第一章 緒論 1 1.1 研究背景 1 1.2 研究動機與目的 5 第二章 文獻回顧 7 2.1 Data augmentation 7 2.1.1 Tradition augmentation 7 2.1.2 樣本與背景多樣性增量演算法 8 2.2 生成對抗網路(Generative Adversarial Nets) 8 2.3 Nodule classification 10 第三章 研究材料及方法 12 3.1 研究材料 12 3.2 研究方法 14 3.2.1 生成對抗網路模型 14 3.2.2 Classification CNN 38 3.3 Performance metrics 40 第四章 研究結果與討論 43 4.1 生成對抗網路模型訓練與結節樣本生成 43 4.1.1 良性結節樣本模型訓練與生成 44 4.1.2 惡性結節樣本模型訓練與生成 49 4.1.3 GGO結節樣本模型訓練與生成 54 4.2 結節良惡性分類網路 59 第五章 結論 64 Reference 66 | |
| dc.language.iso | zh-TW | |
| dc.subject | 卷積神經網路 | zh_TW |
| dc.subject | 數據增量 | zh_TW |
| dc.subject | 肺結節良惡性分類 | zh_TW |
| dc.subject | 深度學習 | zh_TW |
| dc.subject | 生成對抗網路 | zh_TW |
| dc.subject | Convolutional neural network | en |
| dc.subject | Generative adversarial nets | en |
| dc.subject | Data augmentation | en |
| dc.subject | Classification of benign and malignant pulmonary nodule | en |
| dc.subject | Deep learning | en |
| dc.subject | Gabor filter | en |
| dc.title | 以生成對抗網路生成電腦斷層掃描三維肺結節樣本:基於Gabor函數之紋理相似性量度與模型選擇指標 | zh_TW |
| dc.title | Generating 3D Pulmonary Nodule on Computed Tomography based on Generative Adversarial Nets: Gabor-based Texture Similarity Measurement and Model Selection Index | en |
| dc.date.schoolyear | 109-2 | |
| dc.description.degree | 碩士 | |
| dc.contributor.oralexamcommittee | 李佳燕(Hsin-Tsai Liu),林孟暐(Chih-Yang Tseng) | |
| dc.subject.keyword | 生成對抗網路,數據增量,肺結節良惡性分類,深度學習,卷積神經網路, | zh_TW |
| dc.subject.keyword | Generative adversarial nets,Data augmentation,Classification of benign and malignant pulmonary nodule,Deep learning,Gabor filter,Convolutional neural network, | en |
| dc.relation.page | 69 | |
| dc.identifier.doi | 10.6342/NTU202104006 | |
| dc.rights.note | 同意授權(全球公開) | |
| dc.date.accepted | 2021-10-25 | |
| dc.contributor.author-college | 工學院 | zh_TW |
| dc.contributor.author-dept | 醫學工程學研究所 | zh_TW |
| dc.date.embargo-lift | 2026-10-22 | - |
| 顯示於系所單位: | 醫學工程學研究所 | |
文件中的檔案:
| 檔案 | 大小 | 格式 | |
|---|---|---|---|
| U0001-2210202101204100.pdf 此日期後於網路公開 2026-10-22 | 5.45 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。
