Retinal Görüntülerde Eksuda Lezyonlarının Tespiti Üzerine Bir Çalışma

Göz yapısını etkileyen bir hastalık olarak bilinen şeker hastalığı gözün çalışmasını etkiler ve görme kaybına neden olur. Diyabetik retinopati hastalarının şeker seviyesindeki aşırı artışa bağlı olarak bu hastaların retinalarında çeşitli lezyonlar oluşur. Bu lezyonların otomatik tespiti bu hastalığın erken teşhisi için önemli bir unsurdur. Bu çalışmada, diyabetik retinopati hastalarının retinalarındaki eksuda lezyonlarını otomatik olarak tespit eden bir metot önerilmektedir. Bu metot SURF anahtar nokta algoritması ile özellik çıkartımı ve sonrasında Destek Vektör Makineleri, Çok Katmanlı Algılayıcı ve Rasgele Orman algoritmalarıyla lezyonların tespitini içermektedir. Önerilen yöntemin performansı DIARETDB0 ve DIARETDB1 veri tabanları üzerinde gözlemlenmiştir. Her iki veri tabanı içinde sırasıyla %95,8 ve %92,4 doğrulukla Rasgele Orman algoritması en başarılı sonuçları vermiştir. Bu algoritma ile tespit edilen eksuda lezyonları, ilgili veri tabanlarının kesin referans bilgilerine dayanılarak sert ve yumuşak eksuda olarak etiketlenmiştir.

A Study On The Detection Of Exudate Lesions In Retinal Fundus Images

Diabetes Mellitus, known also as diabetes affects the structure of the eye, impairs the working of it, and then causes loss of vision. Based on the excessive increase of glucose level in diabetic retinopathy patients, various lesions form in their retinas. Automatic detection of these lesions is relatively important for early detection of the disease. In this study, a method is suggested, which automatically detects the exudates in the retinas of diabetic retinopathy patients. This method consists of feature extraction using SURF keypoint algorithm, and detection of lesions using Support Vector Machine, Multi-layer Perceptron and Random Forest algorithms. The performance of proposed method was evaluated on DIARETDB0 and DIARETDB1 databases. For both databases, RF algorithm, which provides 95,8% and 92,4% accuracies respectively, was the most successful algorithm. Detected exudate lesions using this algorithm were labeled as hard and soft exudates based on ground truth information of considered databases.

Keywords:

diabetic retinopathy, exudate, hard exudate soft exudate, jaccard index, machine learning,

PDF

___

[1] S. J. McPhee, M. A. Papadakis, Current medical diagnosis & treatment, McGraw-Hill Medical, New York, 2010.
[2] D. S. Fong et al., “Diabetic Retinopathy”, Diabetes Care, 26(11), 99-102, 2003.
[3] K. I. Rother, “Diabetes Treatment — Bridging the Divide”, N. Engl. J. Med., 356(15), 1499–1501, 2007.
[4] K. G. M. M. Alberti, P. Z. Zimmet, “Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus. Provisional report of a WHO Consultation”, Diabet. Med., 15(7), 539–553, 1998.
[5] D. Daneman, “Type 1 diabetes”, Lancet, 367(9513), 847–858, 2006.
[6] M. A. Atkinson, G. S. Eisenbarth, A. W. Michels, “Type 1 diabetes”, Lancet, 383(9911), 69–82, 2014.
[7] P. M. Dodson, “Diabetic retinopathy: treatment and prevention”, Diabetes Vasc. Dis. Res., 4(3), 9–11, 2007.
[8] B. van G. J.J. Staal, M.D. Abramoff, M. Niemeijer, M.A. Viergever, “Ridge based vessel segmentation in color images of the retina”, IEEE transactions on medical imaging, 23(4), 501-509, 2004.
[9] J.M. Tarr, K. Kaul, K. Wolanska, E.M. Kohner, R. Chibber, "Retinopathy in Diabetes", Diabetes. Advances in Experimental Medicine and Biology, vol 771, Editor: Ahmad S.I., Springer, New York, NY, 88-106, 2013
[10] R. A. DeFronzo, A. Ralph et al., “Type 2 diabetes mellitus”, Nature Reviews Disease Primes, 1, 15019, 2015.
[11] S. İnan, “Diabetik Retinopati ve Etiyopatogenezi", Kocatepe Tıp Dergisi, 15(2), 207-217, 2014.
[12] S. Tripathi, K. K. Singh, B. K. Singh, A. Mehrotra, “Automatic detection of exudates in retinal fundus images using differential morphological profile”, International. Journal of Engineering Technology, 5(3), 2024–2029, 2013.
[13] H. Yazid, H. Arof, H. M. Isa, “Exudates segmentation using inverse surface adaptive thresholding”, Measurement, 45(6), 1599–1608, 2012.
[14] C. JayaKumari, R. Maruthi, “Detection of hard exudates in color fundus images of the human retina”, Procedia Engineering, 30, 297–302, 2012.
[15] T. Kauppi et al., “The diaretdb1 diabetic retinopathy database and evaluation protocol”, BMVC, 1, 1–10, 2007.
[16] A. Elbalaoui, M. Fakir, “Exudates detection in fundus images using mean-shift segmentation and adaptive thresholding”, Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 7(2), 145–153, 2019.
[17] A. M. N. Allam, A. A. H. Youssif, A. Z. Ghalwash, A. M, “Segmentation of Exudates via Color-based K-means Clustering and Statistical-based Thresholding”, Journal of Computer Science, 13(10), 524–536, 2017.
[18] C. I. Sánchez, M. García, A. Mayo, M. I. López, R. Hornero, “Retinal image analysis based on mixture models to detect hard exudates”, Medical Image Analysis, 13(4), 650–658, 2009.
[19] C. Eswaran, M. D. Saleh, J. Abdullah, “Projection based algorithm for detecting exudates in color fundus images”, 19th International Conference on Digital Signal Processing, Hong Kong, China, 459–463, 20-23 August, 2014.
[20] A. S. A. Alharthi, V. Emamian, “An Automated mechanism for early screening and diagnosis of diabetic retinopathy in human retinal images”, British Journal of Applied Science & Technology, 12(1), 1–15, 2016.
[21] S. Rajan, T. Das, R. Krishnakumar, “An analytical method for the detection of exudates in retinal images using invertible orientation scores”, in Proceedings of the World Congress on Engineering, vol. 1, London, UK, 29 June- 1 July, 2016.
[22] M. M. Fraz, W. Jahangir, S. Zahid, M. M. Hamayun, S. A. Barman, “Multiscale segmentation of exudates in retinal images using contextual cues and ensemble classification”, Biomedical Signal Processing and Control, 35, 50–62, 2017.
[23] B. Harangi, A. Hajdu, “Automatic exudate detection by fusing multiple active contours and regionwise classification”, Computers in Biology and Medicine, 54, 156–171, 2014.
[24] J. Kaur, D. Mittal, “A generalized method for the segmentation of exudates from pathological retinal fundus images”, Biocybernetics and Biomedical Engineering, 38(1), 27–53, 2018.
[25] Q. Liu et al., “A location-to-segmentation strategy for automatic exudate segmentation in colour retinal fundus images”, Computerized Medical Imaging and Graphics, 55, 78–86, 2017.
[26] A. Colomer, V. Naranjo, T. Janvier, J. M. Mossi, “Evaluation of fractal dimension effectiveness for damage detection in retinal background”, Journal of Computational and Applied Mathematics, 337, 341–353, 2018.
[27] Ö. Demir, B. Doğan, E. Ç. Bayezit, K. Yıldız, “Retina Fundus Floresan Anjiyografi Görüntülerinde Drüsen Alanlarının Otomatik Tespiti ve Büyüklüklerinin Hesaplanması”, Marmara Fen Bilimleri Dergisi, 30(2), 126–132, 2018.
[28] T. Kauppi et al., DIARETDB0: Evaluation Database and Methodology for Diabetic Retinopathy Algorithms, Machine Vision and Pattern Recognition Research Group, Lappeenranta University of Technology,Finland, 2006.
[29] A. Kumar, A. K. Gaur, M. Srivastava, “A Segment based Technique for Detecting Exudate from Retinal Fundus Image”, Procedia Technology, 6, 1–9, 2012.
[30] H. F. Jaafar, A. K. Nandi, W. Al-Nuaimy, “Automated detection of exudates in retinal images using a split-and-merge algorithm” in 18th European signal processing conference, Aalborg, Denmark, 1622–1626, 23-27 August, 2010.
[31] A. Değirmenci, İ. Çankaya, R. Demirci, "Gradyan Anahtarlamalı Gauss Görüntü Filtresi", Düzce Üniversitesi Bilim ve Teknoloji Dergisi, 6(1), 196-215, 2018.
[32] E. Tanyıldızı, S. Okur, “Retina Görüntülerindeki Kan Damarlarının Belirlenmesi”, Fırat Üniversitesi Mühendislik Bilimleri Dergisi, 28(2), 15-22, 2016.
[33] Y. V. Vizilter, Y. P. Pyt’ev, A. I. Chulichkov, L. M. Mestetskiy, “Morphological Image Analysis for Computer Vision Applications”, Computer Vision in Control Systems-1, Intelligent Systems Reference Library, vol 73, Editor: Favorskaya, M., Jain, L., Springer, Cham, 2015.
[34] J. Serra, “Introduction to mathematical morphology”, Computer Vision, Graphics and Image Processing, 35(3), 283– 305, 1986.
[35] X. Zhang, Mathematical Morphological Processing For Retinal Image Analysis, PhD Thesis, Oklahoma State University, 2005.
[36] E. D. Pisano et al., “Contrast limited adaptive histogram equalization image processing to improve the detection of simulated spiculations in dense mammograms”, Journal of Digital Imaging, 11(4), 193–200, 1998.
[37] M. Idrissa, M. Acheroy, “Texture classification using Gabor filters”, Pattern Recognition Letters, 23(9), 1095–1102, 2002.
[38] K. R. A. Biran, P. Sobhe Bidari, “Automatic Method for Exudates and Hemorrhages Detection from Fundus Retinal Images”, International Journal of Computer and Information Engineering, 10(9), 1599-1602, 2016.
[39] H. Bay, A. Ess, T. Tuytelaars, L. Van Gool, “Speeded-up robust features (SURF)”, Computer Vision and Image Understanding, 110(3), 346–359, 2008.
[40] E. Ogasawara, L. C. Martinez, D. de Oliveira, G. Zimbrao, G. L. Pap, M. Mattoso, “Adaptive Normalization: A novel data normalization approach for non-stationary time series”, in International Joint Conference on Neural Networks (IJCNN), Barcelona, Spain, 1–8, 18-23 July, 2010.
[41] M. Rahman, M. R. Hassan, R. Buyya, “Jaccard Index based availability prediction in enterprise grids”, Procedia Computer Science, 1(1), 2707–2716, 2010.