Performance Analysis of Hybrid SOM and AdaBoost Classifiers for Diagnosis of Hypertensive Retinopathy
محورهای موضوعی : Image Processing
Wiharto Wiharto
1
(Universitas Sebelas Maret,Surakarta, Indonesia)
Esti Suryani
2
(Universitas Sebelas Maret, Surakarta, Indonesia)
Murdoko Susilo
3
(Universitas Sebelas Maret, Surakarta, Indonesia)
کلید واژه: Hypertensive Retinopathy, Self-organizing Maps, Segmentation, Adaboost, Classification, Information Gain.,
چکیده مقاله :
The diagnosis of hypertensive retinopathy (CAD-RH) can be made by observing the tortuosity of the retinal vessels. Tortuosity is a feature that is able to show the characteristics of normal or abnormal blood vessels. This study aims to analyze the performance of the CAD-RH system based on feature extraction tortuosity of retinal blood vessels. This study uses a segmentation method based on clustering self-organizing maps (SOM) combined with feature extraction, feature selection, and the ensemble Adaptive Boosting (AdaBoost) classification algorithm. Feature extraction was performed using fractal analysis with the box-counting method, lacunarity with the gliding box method, and invariant moment. Feature selection is done by using the information gain method, to rank all the features that are produced, furthermore, it is selected by referring to the gain value. The best system performance is generated in the number of clusters 2 with fractal dimension, lacunarity with box size 22-29, and invariant moment M1 and M3. Performance in these conditions is able to provide 84% sensitivity, 88% specificity, 7.0 likelihood ratio positive (LR+), and 86% area under the curve (AUC). This model is also better than a number of ensemble algorithms, such as bagging and random forest. Referring to these results, it can be concluded that the use of this model can be an alternative to CAD-RH, where the resulting performance is in a good category.
The diagnosis of hypertensive retinopathy (CAD-RH) can be made by observing the tortuosity of the retinal vessels. Tortuosity is a feature that is able to show the characteristics of normal or abnormal blood vessels. This study aims to analyze the performance of the CAD-RH system based on feature extraction tortuosity of retinal blood vessels. This study uses a segmentation method based on clustering self-organizing maps (SOM) combined with feature extraction, feature selection, and the ensemble Adaptive Boosting (AdaBoost) classification algorithm. Feature extraction was performed using fractal analysis with the box-counting method, lacunarity with the gliding box method, and invariant moment. Feature selection is done by using the information gain method, to rank all the features that are produced, furthermore, it is selected by referring to the gain value. The best system performance is generated in the number of clusters 2 with fractal dimension, lacunarity with box size 22-29, and invariant moment M1 and M3. Performance in these conditions is able to provide 84% sensitivity, 88% specificity, 7.0 likelihood ratio positive (LR+), and 86% area under the curve (AUC). This model is also better than a number of ensemble algorithms, such as bagging and random forest. Referring to these results, it can be concluded that the use of this model can be an alternative to CAD-RH, where the resulting performance is in a good category.
[1] W. Wiharto and E. Suryani, “The Review of Computer Aided Diagnostic Hypertensive Retinopathy Based on The Retinal Image Processing,” in The 2nd Sriwijaya international Conference on Science, Engineering, and Technology [SICEST], Palembang, Indonesia, 2018, vol. 690, pp. 1–9. doi: 10.1088/1757-899X/620/1/012099.
[2] F. Garcia-Lamont, J. Cervantes, A. López, and L. Rodriguez, “Segmentation of images by color features: A survey,” Neurocomputing, vol. 292, pp. 1–27, May 2018, doi: 10.1016/j.neucom.2018.01.091.
[3] W. Wiharto and E. Suryani, “The Analysis Effect of Cluster Numbers On Fuzzy C-Means Algorithm for Blood Vessel Segmentation of Retinal Fundus Image,” in IEEE The 2nd International Conference on Information and Communications Technology, Yogyakarta, Indonesia, 2019, pp. 1–5. doi: 10.1109/ICOIACT46704.2019.8938583.
[4] W. Wiharto and E. Suryani, “The Segmentation Analysis of Retinal Image Based on K-means Algorithm for Computer-Aided Diagnosis of Hypertensive Retinopathy,” Indonesian Journal of Electrical Engineering and Informatics (IJEEI), vol. 8, no. 2, pp. 419–426, 2020, doi: 10.11591/ijeei.v8i2.1287.
[5] W. Wiharto, E. Suryani, and M. Susilo, “The Hybrid Method of SOM Artificial Neural Network and Median Thresholding for Segmentation of Blood Vessels in the Retina Image Fundus,” International Journal of Fuzzy Logic and Intelligent Systems, vol. 19, no. 4, pp. 323–331, 2019.
[6] F. Shafiei and S. Fekri-Ershad, “Detection of Lung Cancer Tumor in CT Scan Images Using Novel Combination of Super Pixel and Active Contour Algorithms,” Traitement du Signal, vol. 37, no. 6, pp. 1029–1035, Dec. 2020, doi: 10.18280/ts.370615.
[7] C. Lupascu and D. Tegolo, “Automatic unsupervised segmentation of retinal vessels using self-organizing maps and k-means clustering,” in Computational Intelligence Methods for …, Berlin, Heidelberg, 2011, vol. 6685, pp. 263–274. doi: 10.1007/978-3-642-21946-7_21.
[8] C. Budayan, I. Dikmen, and M. T. Birgonul, “Comparing the performance of traditional cluster analysis, self-organizing maps and fuzzy C-means method for strategic grouping,” Expert Systems with Applications, vol. 36, no. 9, pp. 11772–11781, 2009, doi: 10.1016/j.eswa.2009.04.022.
[9] S. Arumugadevi and V. Seenivasagam, “Comparison of Clustering Methods for Segmenting Color Images,” Indian Journal of Science and Technology, vol. 8, no. 7, pp. 670–677, 2015, doi: 10.17485/ijst/2015/v8i7/62862.
[10] O. A. Abbas, “Comparisons Between Data Clustering Algorithms,” The International Arab Journal of Information Technology, vol. 5, no. 3, pp. 320–325, 2008.
[11] K. K. Jassar, “Comparative Study and Performance Analysis of Clustering Algorithms,” IJCA Proceedings on International Conference on ICT for Healthcare, vol. ICTHC 2015, no. 1, pp. 1–6, 2016.
[12] P. Zhu et al., “The relationship of retinal vessel diameters and fractal dimensions with blood pressure and cardiovascular risk factors,” PLoS ONE, vol. 9, no. 9, pp. 1–10, 2014, doi: 10.1371/journal.pone.0106551.
[13] N. Popovic, M. Radunovic, J. Badnjar, and T. Popovic, “Fractal dimension and lacunarity analysis of retinal microvascular morphology in hypertension and diabetes,” Microvascular Research, vol. 118, no. 2018, pp. 36–43, 2018, doi: 10.1016/j.mvr.2018.02.006.
[14] H. A. Crystal et al., “Association of the Fractal Dimension of Retinal Arteries and Veins with Quantitative Brain MRI Measures in HIV-Infected and Uninfected Women,” PLoS ONE, vol. 11, no. 5, pp. 1–11, 2016, doi: 10.1371/journal.pone.0154858.
[15] E. V. L. Costa and R. A. Nogueira, “Fractal, multifractal and lacunarity analysis applied in retinal regionsof diabetic patients with and without non-proliferative diabetic retinopathy,” Fractal Geometry and Nonlinear Anal in Med and Biol, vol. 1, no. 3, pp. 112–119, 2016, doi: 10.15761/FGNAMB.1000118.
[16] W. Wiharto, E. Suryani, and M. Yahya Kipti, “Assessment of Early Hypertensive Retinopathy using Fractal Analysis of Retinal Fundus Image,” TELKOMNIKA (Telecommunication Computing Electronics and Control), vol. 16, no. 1, pp. 445-454, 2018, doi: 10.12928/telkomnika.v16i1.6188.
[17] M. F. Syahputra, I. Aulia, R. F. Rahmat, and others, “Hypertensive retinopathy identification from retinal fundus image using probabilistic neural network,” in 2017 International Conference on Advanced Informatics, Concepts, Theory, and Applications (ICAICTA), 2017, pp. 1–6.
[18] K. Narasimhan, V. C. Neha, and K. Vijayarekha, “Hypertensive retinopathy diagnosis from fundus images by estimation of AVR,” Procedia Engineering, vol. 38, no. 2012, pp. 980–993, 2012, doi: 10.1016/j.proeng.2012.06.124.
[19] N. Hutson, A. Karan, J. A. Adkinson, P. Sidiropoulos, I. Vlachos, and L. Iasemidis, “Classification of Ocular Disorders Based on Fractal and Invariant Moment Analysis of Retinal Fundus Images,” in 2016 32nd Southern Biomedical Engineering Conference (SBEC), Shreveport, LA, USA, Mar. 2016, pp. 57–58. doi: 10.1109/SBEC.2016.21.
[20] U. G. Abbasi and U. M. Akram, “Classification of blood vessels as arteries and veins for diagnosis of hypertensive retinopathy,” in 2014 10th International Computer Engineering Conference: Today Information Society What’s Next?, ICENCO 2014, Giza, Egypt, 2014, pp. 5–9. doi: 10.1109/ICENCO.2014.7050423.
[21] B. K. Triwijoyo, W. Budiharto, and E. Abdurachman, “The Classification of Hypertensive Retinopathy using Convolutional Neural Network,” Procedia Computer Science, vol. 116, pp. 166–173, 2017, doi: 10.1016/j.procs.2017.10.066.
[22] X. Miao and J. S. Heaton, “A comparison of random forest and Adaboost tree in ecosystem classification in east Mojave Desert,” in 2010 18th International Conference on Geoinformatics, Beijing, China, Jun. 2010, pp. 1–6. doi: 10.1109/GEOINFORMATICS.2010.5567504.
[23] C. A. Lupascu, D. Tegolo, and E. Trucco, “FABC: Retinal Vessel Segmentation Using AdaBoost,” IEEE Trans. Inform. Technol. Biomed., vol. 14, no. 5, pp. 1267–1274, Sep. 2010, doi: 10.1109/TITB.2010.2052282.
[24] N. Dey, A. B. Roy, M. Pal, and A. Das, “FCM Based Blood Vessel Segmentation Method for Retinal Images,” International Journal of Computer Science and Network (IJCSN), vol. 1, no. 3, pp. 1–5, 2012.
[25] G. B. Kande, T. S. Savithri, and P. Subbaiah, “Segmentation of vessels in fundus images using spatially weighted fuzzy c-means clustering algorithm,” International journal of computer science and network security, vol. 7, no. 12, pp. 102–109, 2007.
[26] R. A. Aras, T. Lestari, H. Adi Nugroho, and I. Ardiyanto, “Segmentation of retinal blood vessels for detection of diabetic retinopathy: A review,” Communications in Science and Technology, vol. 1, no. 2016, pp. 33–41, 2016, doi: 10.21924/cst.1.1.2016.13.
[27] Ş. Talu, C. Vlǎduţiu, L. A. Popescu, C. A. Lupaşcu, Ş. C. Vesa, and S. D. Ţǎlu, “Fractal and lacunarity analysis of human retinal vessel arborisation in normal and amblyopic eyes,” Human and Veterinary Medicine, vol. 5, no. 2, pp. 45–51, 2013.
[28] T. Acharya and K. R. Ray, Image Processing: Principles and Applications. USA: John Willey & Sons, 2005.
[29] B. B. Mandelbrot, The fractal geometry of nature, vol. 173. WH freeman New York, 1983.
[30] A. R. Backes and O. M. Bruno, “A new approach to estimate fractal dimension of texture images,” in In: Elmoataz A., Lezoray O., Nouboud F., Mammass D. (eds) Image and Signal Processing. ICISP 2008. Lecture Notes in Computer Science, Berlin, Heidelberg, 2008, vol. 5099, pp. 136–143. doi: 10.1007/978-3-540-69905-7_16.
[31] C. Allain and M. Cloitre, “Characterizing the lacunarity of random and deterministic fractal sets,” Physical Review A, vol. 44, no. 6, pp. 3552–3558, 1991, doi: 10.1103/PhysRevA.44.3552.
[32] M.-K. Hu, “Visual pattern recognition by moment invariants,” IRE transactions on information theory, vol. 8, no. 2, pp. 179–187, 1962.
[33] Z. Wu et al., “Application of image retrieval based on convolutional neural networks and Hu invariant moment algorithm in computer telecommunications,” Computer Communications, vol. 150, pp. 729–738, Jan. 2020, doi: 10.1016/j.comcom.2019.11.053.
[34] S. Lefkovits and L. Lefkovits, “Gabor Feature Selection Based on Information Gain,” Procedia Engineering, vol. 181, no. 2017, pp. 892–898, 2017, doi: 10.1016/j.proeng.2017.02.482.
[35] Y. Freund and R. E. Schapire, “Experiments with a New Boosting Algorithm,” in Proceedings of the Thirteenth International Conference on International Conference on Machine Learning, The University of Virginia, 1996, pp. 148–156.
[36] G. Eibl and K. Pfeiffer, “Multiclass Boosting forWeak Classifiers,” Journal of Machine Learning Research, vol. 6, no. 7, pp. 189–210, 2005.
[37] Y. Freund and R. E. Schapire, “A Decision-Theoretic Generalization of On-Line Learning and an Application to Boosting,” Journal of Computer and System Sciences, vol. 55, no. 1, pp. 119–139, Aug. 1997, doi: 10.1006/jcss.1997.1504.
[38] E. Ramentol, Y. Caballero, R. Bello, and F. Herrera, “SMOTE-RSB *: A hybrid preprocessing approach based on oversampling and undersampling for high imbalanced data-sets using SMOTE and rough sets theory,” Knowledge and Information Systems, vol. 33, no. 2, pp. 245–265, 2012, doi: 10.1007/s10115-011-0465-6.
[39] S. McGee, “Simplifying likelihood ratios,” J Gen Intern Med, vol. 17, no. 8, pp. 647–650, Aug. 2002, doi: 10.1046/j.1525-1497.2002.10750.x.
[40] F. Gorunescu, Data Mining: Concepts, Models and Techniques. Berlin, Heidelberg: Springer, 2011.
[41] M. Stella and S. Kumar, “Prediction and Comparison using AdaBoost and ML Algorithms with Autistic Children Dataset,” IJERT, vol. 9, no. 7, pp. 133–136, Jul. 2020, doi: 10.17577/IJERTV9IS070091.
[42] P. H. Prastyo, I. G. Paramartha, M. S. M. Pakpahan, and I. Ardiyanto, “Predicting Breast Cancer: A Comparative Analysis of Machine Learning Algorithms,” PROC. INTERNAT. CONF. SCI. ENGIN., vol. 3, no. 2020, pp. 455–459, 2020.
