Abstract
The severe stage of Diabetic Retinopathy (DR) is characterized by the growth of new blood vessels which is called Neovascularization (NV). The abnormally grown blood vessels on the disc are breakable in nature thus the patient is at high risk of sudden blindness. Therefore, the significance of early and accurate detection of Neovascularization on Disc (NVD) should not be neglected. This paper presents an automatic detection of the optic disc using a Controlled Differential Evolution (CDE) algorithm. Further, the Region of Interest (ROI) is created automatically by extending the extreme boundaries of the optic disc by 100 pixels to ensure the presence of NV around the optic disc also. From the ROI so created, blood vessels are segmented using multi-scale Gabor filtering and subsequently, both the morphological and textural features are extracted. Simultaneously, statistical features are directly extracted from the earlier created ROI. Finally, the fundus image is classified by a Support Vector Machine (SVM) using the extracted features from all three feature sets. From each individual image, 16 features are extracted and the feature dimension is reduced to 13 using a sequential backward feature (SBF) selection algorithm. The optimal features are obtained from a total of 205 fundus images, which consists of 99 NVD positive and 106 NVD negative images. This paper attains an average accuracy of 98.75%, the specificity of 100%, the sensitivity of 97.8%, and area under the curve (AUC) as 100% when tested over image selected randomly.
Funding source: Science and Engineering Research Board
Award Identifier / Grant number: EMR/2017/000885
Acknowledgments
The authors of this research article thank the Department of Science and Technology (DST), India for granting this work under the Extramural Research (EMR) funding scheme of Science Engineering and Research Board (SERB) under grant no – EMR/2017/000885.
Research funding: This study is funded by the Department of Science and Technology (DST), India for granting this work under the Extramural Research (EMR) funding scheme of Science Engineering and Research Board (SERB) under grant no – EMR/2017/000885.
Author contributions: All authors have accepted responsibility for the entire content of this manuscript and approved its submission.
Competing interests: All authors declare that they have no conflict of interest.
Informed consent: Informed consent was obtained from all individuals included in this study.
Ethical approval: This article does not contain any studies with human participants or animals performed by any of the authors.
References
1. Macular Hole - The American Society of Retina Specialists - The American Society of Retina Specialists. Available from: https://imagebank.asrs.org/discover-new/files/4/25?q=nvd.Search in Google Scholar
2. Goatman, KA, Fleming, AD, Philip, S, Williams, GJ, Olson, JA, Sharp, PF. Detection of new vessels on the optic disc using retinal photographs. IEEE Trans Med Imaging 2011;30:972–9. https://doi.org/10.1109/tmi.2010.2099236.Search in Google Scholar PubMed
3. Agurto, C, Yu, H, Murray, V, Marios, S P, Simon, B, Wendall, B, et al. Detection of neovascularization in the optic disc using an AM-FM representation, granulometry, and vessel segmentation. In: Proceedings annual international conference of the IEEE Engineering in Medicine and Biology Society; 2012, pp. 4946–9.10.1109/EMBC.2012.6347102Search in Google Scholar PubMed
4. Agurto, C, Yu, H, Murray, V, Marios, S P, Simon, B, Wendall, B, et al. A multiscale decomposition approach to detect abnormal vasculature in the optic disc. Comput Med Imaging Graph 2015;43:137–49. https://doi.org/10.1016/j.compmedimag.2015.01.001.Search in Google Scholar PubMed PubMed Central
5. Welikala, RA, Fraz, MM, Dehmeshki, J, Hoppe, A, Tah, V, Mann, S, et al. Genetic algorithm based feature selection combined with dual classification for the automated detection of proliferative diabetic retinopathy. Comput Med Imaging Graph 2015;43:64–77. https://doi.org/10.1016/j.compmedimag.2015.03.003.Search in Google Scholar PubMed
6. Gupta, G, Kulasekaran, S, Ram, K, Joshi, N, Sivaprakasam, M, Gandhi, R. Local characterization of neovascularization and identification of proliferative diabetic retinopathy in retinal fundus images. Comput Med Imaging Graph 2017;55:124–32. https://doi.org/10.1016/j.compmedimag.2016.08.005.Search in Google Scholar PubMed
7. Patton, N, Aslam, TM, MacGillivray, T, Jan, J D, Dhillon, B, Robert, H, et al. Retinal image analysis: concepts, applications and potential. Prog Retina Eye Res 2006;25:99–127. https://doi.org/10.1016/j.preteyeres.2005.07.001.Search in Google Scholar PubMed
8. Teng, T, Lefley, M, Claremont, D. Progress towards automated diabetic ocular screening: a review of image analysis and intelligent systems for diabetic retinopathy. Med Biol Eng Comput 2002;40:2–13. https://doi.org/10.1007/bf02347689.Search in Google Scholar PubMed
9. Eswari, S, Raja Rajeswari, S. Detection methods of new vessels in retinal fundus images - a survey. In: 2016 International conference on computing technologies and intelligent data engineering ICCTIDE; 2016. https://doi.org/10.1109/ICCTIDE.2016.7725355.Search in Google Scholar
10. Hassan, SSA, Bong, DBL, Premsenthil, M. Detect on of neovascularization in diabetic retinopathy. J Digit Imaging 2012;25:437–44. https://doi.org/10.1007/s10278-011-9418-6.Search in Google Scholar PubMed PubMed Central
11. Akram, MU, Tariq, A, Khan, SA. Detection of neovascularization for screening of proliferative diabetic retinopathy. Lect Notes Comput Sci (including Subser Lect Notes Artif Intell Lect Notes Bioinformatics) 2012;7325:372–9. https://doi.org/10.1007/978-3-642-31298-4_44.Search in Google Scholar
12. Usman Akram, M, Khalid, S, Tariq, A, Younus Javed, M. Detection of neovascularization in retinal images using multivariate m-Mediods based classifier. Comput Med Imaging Graph 2013;37:346–57. https://doi.org/10.1016/j.compmedimag.2013.06.008.Search in Google Scholar PubMed
13. Lee, J, Zee, BCY, Li, Q. Detection of neovascularization based on fractal and texture analysis with interaction effects in diabetic retinopathy. PLoS ONE 2013;8. https://doi.org/10.1371/journal.pone.0075699.Search in Google Scholar PubMed PubMed Central
14. Roychowdhury, S, Koozekanani, DD, Parhi, KK. Automated detection of neovascularization for proliferative diabetic retinopathy screening. Proc Annu Int Conf IEEE Eng Med Biol Soc EMBS 2016:1300–3.10.1109/EMBC.2016.7590945Search in Google Scholar PubMed
15. Roychowdhury, S, Koozekanani, DD, Parhi, KK. DREAM: diabetic retinopathy analysis using machine learning. IEEE J Biomed Health Inf 2014;18:1717–28. https://doi.org/10.1109/jbhi.2013.2294635.Search in Google Scholar PubMed
16. Welikala, RA, Dehmeshki, J, Hoppe, A, et al. Automated detection of proliferative diabetic retinopathy using a modified line operator and dual classification. Comput Methods Progr Biomed 2014;114:247–61. https://doi.org/10.1016/j.cmpb.2014.02.010.Search in Google Scholar PubMed
17. Kaggle Dataset. Available from: https://www.kaggle.com/c/diabetic-retinopathy-detection/data.Search in Google Scholar
18. DRIVE Dataset, Staal, JJ, Abramoff, MD, Niemeijer, M, Viergever, MA, van Ginneken, B. Ridge based vessel segmentation in color images of the retina. IEEE Trans Med Imag 2004;23:501–9. https://doi.org/10.1109/tmi.2004.825627 [online].Search in Google Scholar PubMed
19. STARE Dataset. Available from: https://www.ces.clemson.edu/∼ahoover/stare [online].Search in Google Scholar
20. Lay, B, Messidor Dataset. Available from: https://www.adcis.net/en/Download-Third-Party/MESSIDOR [online].Search in Google Scholar
21. Kauppi, T, Valentina, K, Joni, KT, Lasse, L, Irisis, S, Asta, R, et al. DIARETDB1 diabetic retinopathy database and evaluation protocol; 2007. Available from: https://www.it.lut.fi/project/imageret /DIARET.10.5244/C.21.15Search in Google Scholar
22. Odstrcilik, J, Kolar, R, Budai, A, Hornegger, J, Jan, J, Kubena, T, et al. Retinal vessel segmentation by improved matched filtering: evaluation on a new high-resolution fundus image database. Image Process 2013;7:373–83. https://doi.org/10.1049/iet-ipr.2012.0455.Search in Google Scholar
23. Storn, R, Price, K. Differential evolution – a simple and efficient heuristic for global optimization over continuous spaces. J Global Optim 1997;11:341–59. https://doi.org/10.1023/a:1008202821328.10.1023/A:1008202821328Search in Google Scholar
24. Kaelo, P, Ali, MM. A numerical study of some modified differential evolution algorithms. Eur J Oper Res 2006;169:1176–84. https://doi.org/10.1016/j.ejor.2004.08.047.Search in Google Scholar
25. Cuevas, E, Díaz, M, Manzanares, M, Zaldivar, D, Perez-Cisneros, M. An improved computer vision method for white blood cells detection. Comput Math Methods Med 2013. https://doi.org/10.1155/2013/137392.Search in Google Scholar PubMed PubMed Central
26. Omran, MGH, Engelbrecht, AP, Salman, A. Bare bones differential evolution. Eur J Oper Res 2009;196:128–39. https://dx.doi.org/10.12732/dsa.v27i4.1.10.1016/j.ejor.2008.02.035Search in Google Scholar
27. Greco, R, Vanzi, I. New few parameters differential evolution algorithm with application to structural identification. J Traffic Transp Eng (Engl Ed) 2019;6:1–14. https://doi.org/10.1016/j.jtte.2018.09.002.Search in Google Scholar
28. Lv, J, Yang, X. Approximate controllability of hilfer fractional neutral stochastic differential equations. 2018; 4: 691–713.10.12732/dsa.v27i4.1Search in Google Scholar
29. Boccignone, G, Napoletano, P, Caggiano, V, Ferraro, M. A multiresolution diffused expectation-maximization algorithm for medical image segmentation. Comput Biol Med 2007;37:83–96. https://doi.org/10.1016/j.compbiomed.2005.10.002.Search in Google Scholar PubMed
30. Boccignone, G, Ferraro, M, Napoletano, P. Diffused expectation maximisation for image segmentation. Electron Lett 2012;2:7–8. https://doi.org/10.1049/el:20045792.10.1049/el:20045792Search in Google Scholar
31. Gonzalez, RC, Woods, RE. Digital_image_processing_2ndEdb., Pearson Education. pp. 371–426. https://www.researchgate.net/publication/333856607_Digital_Image_Processing_Second_Edition.Search in Google Scholar
32. Tan, NM, Liu, J, Wong, DNK, Lim, JH, Li, H, Yu, W, et al. Automatic detection of left and right eye in retinal fundus images. In: 13th international conference on biomedical engineering IFMBE proceedings; 2009, pp. 610–1.10.1007/978-3-540-92841-6_150Search in Google Scholar
33. Yassur, Y, Pickle, LW, Fine, SL, Singerman, L, Orth, DH, Patz, A. Optic disc neovascularisation in diabetic retinopathy: I. A system for grading proliferation at the optic nerve head in patients with proliferative diabetic retinopathy. 1980:69–76. https://doi.org/10.1136/bjo.64.2.69.Search in Google Scholar PubMed PubMed Central
34. Yu, S, Xiao, D, Kanagasingam, Y. Machine learning based automatic neovascularization detection on optic disc region. IEEE J Biomed Health Inf 2018;22:886–94. https://doi.org/10.1109/jbhi.2017.2710201.Search in Google Scholar PubMed
35. Zhiqiang, L, Jianming, W, Junyu, Z, et al. The statistical meaning of kurtosis and its new application to identification of persons based on seismic signals. Sensors 2008;8:5106–19. https://doi.org/10.3390/s8085106.Search in Google Scholar PubMed PubMed Central
36. Maruyama, Y. Measures of multivariate skewness and kurtosis with principal components. Jpn J Appl Stat 2010;36:139–45. https://doi.org/10.1080/03610920701867587.Search in Google Scholar
37. Cristianini, N, Shawe-Taylor, J. An introduction to support vector machines and other kernel-based learning methods. Cambridge: Cambridge University Press; 2000. https://doi.org/10.1017/CBO9780511801389.Search in Google Scholar
© 2020 Walter de Gruyter GmbH, Berlin/Boston
