تشخیص اتوماتیک بیمارهای شبکیه چشم با استفاده از مدل های ریاضیاتی پردازش تصویر، مبتنی بر یادگیری دیکشنری چند لایه

نوع مقاله: مقاله پژوهشی

نویسندگان

1 گروه فناوری اطلاعات، دانشکده کامپیوتر، دانشگاه قم، قم ، ایران

2 دانشکده برق و کامپیوتر، دانشگاه صنعتی قم

چکیده

هدف از این مطالعه بهبود عملکرد طبقه بندی روش های نوین، با استفاده از مدلی چند لایه به منظور کمک به تشخیص بیماری‌های شبکیه ی چشم است. این مدل از الگوریتم K-SVD پیشرفته، برای یادگیری ماتریس دیکشنری و الگو های پایه استفاده می کند تا بتواند با الگوپذیری از معماری چند لایه، ویژگی های بهتری را در تصاویر OCT شبکیه بیاموزد. همچنین در این معماری، علاوه بر استفاده از برچسب های کلاس داده های آموزشی، اطلاعات برچسب نیز در هر ستون پایه در ماتریس دیکشنری ترکیب می شود تا در کدگذاری تنک در طی فرآیند یادگیری دیکشنری بیشترین تبعیض اعمال شود که این منجر به موفقیت مراحل کدگزاری تنک و جمع بندی، در پیدا کردن نمایش موثر تری از داده به منظور طبقه بندی می گردد. برای اعتبار سنجی الگوریتم، از مجموعه داده های داک استفاده شده است. نتایج تجربی نشان می دهد که الگوریتم پیشنهادی این مقاله توانسته است با پیشی گرفتن از بسیاری از مدل های جدید یادگیری دیکشنری و نمایش تنک، بسیار خوب عمل نماید و با دقت خوبی منجر به طبقه بندی صحیح ٪95.85 برای تصاویر نرمال و صددرصد برای تصاویر بیمار (DME و AMD) شود.

چکیده تصویری

تشخیص اتوماتیک بیمارهای شبکیه چشم با استفاده از مدل های ریاضیاتی پردازش تصویر، مبتنی بر یادگیری دیکشنری چند لایه

کلیدواژه‌ها


عنوان مقاله [English]

Automatic recognition of retinal diseases using mathematical models of image processing, based on multilayer-dictionary learning

نویسندگان [English]

  • Azadeh Montazeri 1
  • Mahboubeh Shamsi 2
1 Department of Information Technology, Faculty of Computer, University of Qom, ًQom, Iran
2 Faculty of Electrical and Computer Engineering, Qom University of Technology, Qom, Iran
چکیده [English]

The purpose of this study is to improve the classification of new methods using a multi-layered model to address retinal diseases diagnosis. This paper presents a multi-layer dictionary learning method for classification tasks. Our multi-layer framework uses a label consistent in K-SVD algorithm to learn a discriminative dictionary for sparse coding in order to learn better features in retinal optical coherence tomography images. In addition to using class labels of training data, we also associate label information with each dictionary item (columns of the dictionary matrix) to enforce discrimination in sparse codes during dictionary learning process. In fact, it relies on a succession of sparse coding and pooling steps in order to find an effective representation of data for classification. Moreover, we apply Duke dataset for validating our algorithm. Our classifier leads to a correct classification rate of 95.85% and 100.00% for normal and abnormal (DME and AMD). Experimental results demonstrate that our algorithm outperforms compared to many recent proposed supervised dictionary learning and sparse representation techniques.

کلیدواژه‌ها [English]

  • Multi-layer Dictionary Learning
  • Sparse Representation
  • K-SVD Algorithm
  • Optical Coherence Tomography
  • Classification

[1] Schirrmacher F, Köhler T, Husvogt L, (2017). Quantile sparse image prior for spatio-temporal denoising of retinal OCT data.Medical Image Computing and Computer-Assisted Intervention,  12(4), 83-91.

[2] Sarbjeet Kaur, V., & Banga, K. (2014). Review paper of character recognition using image processing. International Journal of Computer Science & Engineering Technology, 5(1), 254-268.

[3] Fan Meng et al, (2017). A sparse dictionary learning-based adaptive patch in painting method for thick Clouds removal from high-spatial resolution remote sensing imagery. Sensors, 17(9), 123-135.

[4] Zheng H., Zhu, J., Yang, Z., & Jin, Z. (2017). Effective micro-expression recognition using relaxed K-SVD algorithm. International Journal of Machine Learning, 8(6), 2043–2049.

[5] Chan Wai Tim, S., Rombaut, M., & Pellerin, D, (2015). Rejection-based classification for action recognition using a spatio-temporal dictionary. Advanced Concepts for Intelligent Vision Systems15(8), 522-533.

[6] Zhang, Z. et al. (2016). Discriminative-dictionary-learning-based multilevel point-cluster features for ALS point cloud classification. IEEE Transactions on Geoscience and Remote Sensing, 54(12), 334-347.

[7] Zhang, Z. et al. (2016). Sparse codes auto-extractor for classification: A joint embedding and dictionary learning framework for representation. IEEE Signal Processing Society, 64(14), 553-561.

[8] de Moura J., Novo, J., & Rouco, J. (2017). Automatic Detection of Blood Vessels in Retinal OCT Images. Biomedical Applications Based on Natural and Artificial Computing, 10(4), 3-10.

[9] Oguz, I. et al. (2016). Optimal retinal cyst segmentation from OCT images. Medical Imaging, 15(9), 245-256.

[10] Sun, Y., Lim S., Sun, Z. (2017). Fully automated macular pathology detection in retina optical coherence tomography images using sparse coding and dictionary learning. Journal of biomedical optics, 22(1), 214-225

[11] Rickman, C. B. et al. (2013). Dry age-related macular degeneration: mechanisms, therapeutic targets, and imagingdry AMD mechanisms, targets, and imaging.Invest. Ophthalmol. Visual Sci, 54(14), 68– 80.

[12] Antonym B.J et al. (2013). A combined machine-learning and graph-based framework for the segmentation of retinal surfaces in SD-OCT volumes. Biomed. Opt. Express, 4(12), 2712–2728.

[13] Carass, A. et al. (2014). Multiple-object geometric deformable model for segmentation of macular OCT. Biomed. Opt. Express, 5(4), 1062–1074.

[14] Chium S. J. et al. (2012). Validated automatic segmentation of AMD pathology including drusen and geographic atrophy in SD-OCT images. Invest. Ophthalmol. Visual Sci, 53(1), 53–61.

[15] Sun, Y. et al. (2016). 3D automatic segmentation method for retinal optical coherence tomography volume data using boundary surface enhancement. J. Innovative Opt. Health Sci, 9(2), 165-177.

[16] Vermeer, K. et al. (2011), Automated segmentation by pixel classification of retinal layers in ophthalmic OCT images. Biomed. Opt. Express, 2(6), 1743–1756.

[17] Yang, Q. et al. (2011), Automated segmentation of outer retinal layers in macular OCT images of patients with retinitis pigmentosa, Biomed. Opt. Express 2(9), 2493–2503.

[18] Lang, A. et al. (2013), Retinal layer segmentation of macular OCT images using boundary classification. Biomed. Opt. Express, 4(7), 1133–1152.

[19] Garcia-Allende, P. B et al. (2011). Morphological analysis of optical coherence tomography images for automated classification of gastrointestinal tissues. Biomed. Opt. Express, 2(10), 2821–2836.

[20] Chan, S. et al, (2016), Multi-layer Dictionary Learning for Image Classification. Advanced Concepts for Intelligent Vision Systems, 4(15), 132-143.

[21] Jiang, Z., Lin, Z., & Davis, L., (2011). Learning a discriminative dictionary for sparse coding via label consistent K-SVD. IEEE computer society conference on computer vision and pattern recognition, 6(12), 1697-1704.

[22] Pande, P. et al. (2014), Automated classification of optical coherence tomography images for the diagnosis of oral malignancy in the hamster cheek pouch. J. Biomed. Opt, 19(8), 356-368.

[23] Alsaih, K., et al, (2017). Machine learning techniques for diabetic macular edema (DME) classification on SD-OCT images. Bio Medical Engineering, 16(1), 447-455.

[24] Liu, Y. et al. (2011). Automated macular pathology diagnosis in retinal OCT images using multi-scale spatial pyramid and local binary patterns in texture and shape encoding. Med. Image Anal, 15(5), 748–759.

[25] Zheng, Y., Hijazi, M.H.A, & Coenen, F. (2012). Automated ‘disease/no disease’ grading of age-related macular degeneration by an image mining approach.Invest. Ophthalmol. Visual Sci, 53(13), 8310–8318.

[26] Hijazi, M.H.A, Coenen, F., & Zheng, Y. (2012).Data mining techniques for the screening of age-related macular degenerationg. Knowled e-Based Syst. 29(1), 83–92.

[27] Kafieh, R., & Rabbani, H. (2013). Optical coherence tomography noise reduction over learned dictionaries with introduction of complex wavelet for noise reduction. Proceedings,Wavelets and Sparsity, 8858(26), 238-247.

[28] Albarrak, A. et al. (2012). Volumetric image mining based on decomposition and graph analysis: an application to retinal optical coherence tomography. Computational Intelligence and Informatics, 1344(64), 245-256.

[29] Wang, Q. et al, (2017). Synthesis K-SVD based analysis dictionary learning for pattern classification. Multimedia Tools and Applications,12(4), pp1-24.

[30] Srinivasan, P.P et al, (2014). Fully automated detection of diabetic macular edema and dry age-related macular degeneration from optical coherence tomography images. Biomed. Opt. Express, 5(10), 3568–3577.

[31] Esmaeili, M., Dehnavi, A. M., & Rabbani, H. (2017). Speckle noise reduction in optical coherence tomography using two-dimensional curvelet-based dictionary learning. Journal of medical, 17(2), 86-91.

[32] Yankelevsky, Y. et al, (2016). Structure-aware classification using supervised dictionary learning. Computer Vision and Pattern Recognition, 10(5), 679-692.

[33] Sun, Y., Li, S.H, Sun, Z.H. (2017). Fully automated macular pathology detection in retina optical coherence tomography images using sparse coding and dictionary learning. J. Biomed. Opt., 22(1), 160-171.

[34] Hassan, T. et al, (2015). Review of OCT and fundus images for detection of Macular Edema.  Imaging Systems and Techniques,155(45), 234-245.

[35] Shalev-Shwartz, S., & Shamir, O. (2017). Failures of gradient-based deep learning. International Journal of Machine Learning, 17(3), 456-466.

[36] Javidi, M.,  Pourreza, H.R., & Harati, A. (2017).  Vessel segmentation and microaneurysm detection using discriminative­­­ dictionary learning and sparse representation. Computer methods and programs in biomedicine, 139(1), 93-108.

 [37] Geimer, T., et al, (2017). A Kernel Ridge regression model for respiratory motion estimation in radiotherapy. Bildverarbeitung für die Medizin, 13(4), 155-160.

[38] Khalid, S., et al, (2017). Fully automated robust system to detect retinal edema, central serous chorioretinopathy, and age related macular degeneration from optical coherence tomography images. J. Biomed. Opt, 71(48), 278-293.

[39] Lu, D. et al. (2017). Retinal fluid segmentation and detection in optical coherence tomography images using fully convolutional neural network. Hindawi BioMed Research International, 17(6), 477- 489.

[40] Gangeh, M. et al, (2015). Supervised dictionary learning and sparse representation-A review. Computer Vision and Pattern Recognition, 20(3), 145-158.

[41] Guan, J., et al. (2018). Polynomial dictionary learning algorithms in sparse representations. Signal Processing, 142(7), 492-503

[42] Su, H., et al, (2018), Multifeature dictionary learning for collaborative representation classification of Hyperspectral imagery. IEEE Transactions on Geoscien, 56(4), 345-357.

[43] Aggarwal, C. C, & Reddy, C.K. (2013). Data clustering, algorithms and applications. New York: Wiley.

[44] Kafieh, R. (2014). Combination of graph based and space-frequency methods in analysis of Optical coherence Tomography (OCT) images, (Unpublished doctoral thesis), Isfahan University, Isfahan. [in Persian[.