Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/92910
標題: Malignant Hypertensive Retinopathy Diagnosing System Based on Image
以影像為基礎惡性高血壓性視網膜病變診斷系統
作者: 郭安陞
An-Shang Kuo
關鍵字: Hypertension
Image segmentation
Hypertensive retinopathy
Retinal image
Keith-Wagener-Barker grade
高血壓
影像切割
高血壓視網膜
視網膜影像
Keith-Wagener-Barker grade
引用: References [1] 台灣高血壓學會, '高血壓衛教'. From http://www.ths.org.tw/ [2] 衛生福利部國民健康署, '2007年台灣地區高血壓、高血糖、高血脂之追蹤調查研究', 2007. From http://www.hpa.gov.tw/BHPNet/Web/Index/index.aspx [3] Eye Wiki, 'Hypertensive retinopathy' From http://eyewiki.aao.org/Hypertensive_retinopathy [4] Keith NM, Wagener HP, Barker NW, 'Some different typs of essential hypertension: their course and prognosis'. Am J Med Sci 197, pp. 332–343, 1939. [5] Louise O'Toole, MMedSci, FRCSI(Ophth), MRCOphth, FEBO 'Hypertensive Retinopathy' Optometry Today, 2011. From http://www.optometry.co.uk/ [6] Tien Y. Wong, M.D., Ph.D., and Paul Mitchell, M.D., Ph.D, 'Hypertensive Retinopathy', N Engl J Med 2004, pp. 2310-2317,2004. [7] N. Otsu, 'A Threshold Selection Method from Gray-level Histogram,' IEEE Transactions on System Man Cybernetics, vol. 9, no. 1, pp. 62-66, 1979. [8] Y. K. Chan and C. C. Chang, 'Image Matching Using Run-length Feature,' Pattern Recognition Letters, vol. 22, no.5, pp. 447-455, 2001. [9] Y. K. Chan and Yu-Ching Liu, 'Image Segmentation of Diabetic Retinopathy', pp.20-22,2012. [10] Sezgin M., and B. Sankur, 'Survey over Image Thresholding Techniques and Quantitative Performance Evaluation,' Journal of Electronic Imaging, Vol. 13, No. 1, pp. 146-165, 2004. [11] Gelasca E. D., J. Byun, B. Obara, and B. S. Manjunath, 'Evaluation and Benchmark for Biological Image Segmentation,' 15th IEEE International Conference on Image Processing (ICIP 2008), San Diego, CA, USA, pp. 1816-1819, 2008. [12] Haupt R. L., and S. E. Haupt, 'Practical Genetic Algorithms,' 2nd ed. Hoboken, N.J., John Wiley, 2004. [13] Carlone P., G. S. Palazzo, and R. Pasquino, 'Pultrusion Manufacturing Process Development: Cure Optimization by Hybrid Computational Methods,' Computers and Mathematics with Applications, Vol. 53, pp.1464–1471, 2007. [14] Wikipedia, The Free Encyclopedia, 'Gamma_correction,' From https://en.wikipedia.org/wiki/Gamma_correction [15] R. Gonzalez, R. Woods. ' Digital Image Processing', Prentice-Hall, 2002. [16] S. Eddins, 'The Watershed Transform: Strategies for Image Segmentation,' Newsletters―MATLAB News & Notes, 2002. From http://www.mathworks.com/company/newsletters/news_notes/win02/watershed.html [17] D. C. Huang, R. T. Chen, Y. K. Chan, and X. Jiang, 'An Automatic Indirect Immunofluorescence Based Cell Segmentation and Counting System,' National Digital Library of Theses and Dissertations in Taiwan, 2010. [18] Wikipedia, The Free Encyclopedia, 'Linear smoothers,' From https://en.wikipedia.org/wiki/Smoothing#Linear_smoothers [19] A. Hoover, V. Kouznetsova and M. Goldbaum, 'Locating Blood Vessels in Retinal Images by Piece-wise Threhsold Probing of a Matched Filter Response', IEEE Transactions on Medical Imaging , vol. 19 no. 3, pp. 203-210, 2000. From http://www.ces.clemson.edu/~ahoover/stare/ [20] A. Hoover and M. Goldbaum, 'Locating the optic nerve in a retinal image using the fuzzy convergence of the blood vessels', IEEE Transactions on Medical Imaging , vol. 22 no. 8, pp. 951-958, 2003. From http://www.ces.clemson.edu/~ahoover/stare/ [21] High-Resolution Fundus (HRF) Image Database. From https://www5.cs.fau.de/research/data/fundus-images/ [22] 衛生福利部統計處, '103年度死因統計', 2015. From http://www.mohw.gov.tw/cht/DOS/Statistic.aspx?f_list_no=312&fod_list_no=5488
摘要: More and more people suffer hypertension because of their diet, unhealthy lifestyle, eating habits or hereditary disease. It may cause retinopathy that high blood pressure can damage the vessels supplying blood to retina. When the cardiologist finds a hypertension patient who has the symptoms of hypertensive retinopathy, the cardiologist will help him make referrals to ophthalmology for a complete eye diagnosis. Early detection can lead to early treatment. However, early recognition of hypertensive retinopathy signs is not always easy to diagnose. Due to a lack of ophthalmic knowledge, the hypertensive retinopathy may be misdiagnosed by the cardiologist. This study hence provides a malignant hypertensive retinopathy diagnosis system (MHRD system) to promptly and accurately diagnose malignant hypertensive retinopathy based on retina image. The symptoms of malignant hypertensive retinopathy include constricted and tortuous arteries, retinal hemorrhage, hard exudates, cotton wool spots, retinal edema, and papilledema. These symptoms create some bright and dark areas, and lead to vascular distortion and uneven thickness in a gray-level retinal image. When given a retinal image, the MHRD system hence segments the bright areas, dark areas, and vessels from the retinal image. It then diagnoses the malignant hypertensive retinopathy based on the characteristics of the segmented bright areas, dark areas, and vessels. The experiment results show that the MHRD system can give over 85% diagnostic accuracy.
長久以來要判斷病患的病況都仰賴醫生親自肉眼觀察對各項病徵,但是光靠醫生主觀地對醫療影像肉眼辨識診斷,即便是經過長時間執業、訓練的醫生難免也會有疏忽,未必能百分之百準確的診斷。本篇研究的動機即是透過影像切割的技術準確地將病徵切割出來,再藉由這些病徵作為影像辨識的特徵來協助醫生診斷,如此一來不但可以省下病患在眼科及高血壓相關疾病專科醫生之間互相轉診的時間,亦能夠迅速幫助醫生診斷。 在台灣,由於生活文化的影響以及病患自我觀察診斷的技術和觀念不普遍,導致有許多高血壓的潛在病患和高風險群,藉由高血壓病變的視網膜影像之病徵辨識可以協助醫生和病患掌握高血壓的病情。高血壓病患之視網膜影像中隨著高血壓病情的加重會出現種種症狀,這些症狀依照Keith-Wagener-Barker Grade(KWB Grade)可以分為四期,例如一、二期的視網膜微血管彎曲硬化、血管壁變薄(視網膜影像中血管中呈現反光、變白),到第三期的棉絮斑(本文中視為白點)和第四期的出血(本文中視為黑點)。本篇論文提出針對高血壓視網膜病變之影像辨識方法,首先針對視網膜影像中的血管、棉絮斑和出血利用灰階值不同的特性進行切割,而後再透過平滑化計算血管彎曲程度、血管中心之亮度以及白點和黑點在視網膜影像中的分散程度作為特徵值來做影像辨識。 本實驗中所使用的視網膜影像來自開放式資料庫-HRF資料庫和STARE資料庫,這些高血壓病變之視網膜影像藉由本實驗中提出的切割和辨識方法,實驗結果顯示對於高血壓視網膜病變KWB各期的辨識準確度達到85%以上
URI: http://hdl.handle.net/11455/92910
其他識別: U0005-1008201522030400
文章公開時間: 10000-01-01
Appears in Collections:資訊管理學系

文件中的檔案:

取得全文請前往華藝線上圖書館



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.