Multi-label classification of retinal lesions in diabetic retinopathy for automatic analysis of fundus fluorescein angiography based on deep learning
- 18 Downloads
To automatically detect and classify the lesions of diabetic retinopathy (DR) in fundus fluorescein angiography (FFA) images using deep learning algorithm through comparing 3 convolutional neural networks (CNNs).
A total of 4067 FFA images from Eye Center at the Second Affiliated Hospital of Zhejiang University School of Medicine were annotated with 4 kinds of lesions of DR, including non-perfusion regions (NP), microaneurysms, leakages, and laser scars. Three CNNs including DenseNet, ResNet50, and VGG16 were trained to achieve multi-label classification, which means the algorithms could identify 4 retinal lesions above at the same time.
The area under the curve (AUC) of DenseNet reached 0.8703, 0.9435, 0.9647, and 0.9653 for detecting NP, microaneurysms, leakages, and laser scars, respectively. For ResNet50, AUC was 0.8140 for NP, 0.9097 for microaneurysms, 0.9585 for leakages, and 0.9115 for laser scars. And for VGG16, AUC was 0.7125 for NP, 0.5569 for microaneurysms, 0.9177 for leakages, and 0.8537 for laser scars.
Experimental results demonstrate that DenseNet is a suitable model to automatically detect and distinguish retinal lesions in the FFA images with multi-label classification, which lies the foundation of automatic analysis for FFA images and comprehensive diagnosis and treatment decision-making for DR.
KeywordsDiabetic retinopathy Fundus fluorescein angiography Deep learning Multi-label classification
This work was financially supported by Zhejiang Provincial Key Research and Development Plan (grant number 2019C03020), the Natural Science Foundation of China (grant number 81670888), and the Natural Science Foundation of China (grant number 81870635).
Compliance with ethical standards
This study was conducted in compliance with the principles of the Declaration of Helsinki and was approved by the Ethics Committees of the Second Affiliated Hospital of Zhejiang University School of Medicine. Written informed consents were obtained from all subjects for the publication of this study and accompanying images.
Conflict of interest
The authors declare that they have no conflict of interest.
- 1.Huang Y (2017) International Diabetes Federation (IDF) (2017) IDF diabetes atlas. 8th edition. http://www.diabetesatlas.org/resources/2017-atlas.html. Accessed 9 Jun 2019
- 2.Li Z, Keel S, Liu C, He Y, Meng W, Scheetz J, Lee PY, Shaw J, Ting D, Wong TY, Taylor H, Chang R, He M (2018) An automated grading system for detection of vision-threatening referable diabetic retinopathy on the basis of color fundus photographs. Diabetes Care 41:2509–2516. https://doi.org/10.2337/dc18-0147 CrossRefPubMedGoogle Scholar
- 5.Unoki N, Nishijima K, Sakamoto A, Kita M, Watanabe D, Hangai M, Kimura T, Kawagoe N, Ohta M, Yoshimura N (2007) Retinal sensitivity loss and structural disturbance in areas of capillary nonperfusion of eyes with diabetic retinopathy. Am J Ophthalmol 144:755–760. https://doi.org/10.1016/j.ajo.2007.07.011 CrossRefPubMedGoogle Scholar
- 6.Kozak I, El-Emam SY, Cheng L, Bartsch DU, Chhablani J, Freeman WR, Arevalo JF (2014) Fluorescein angiography versus optical coherence tomography-guided planning for macular laser photocoagulation in diabetic macular edema. Retina 34:1600–1605. https://doi.org/10.1097/iae.0000000000000120 CrossRefPubMedPubMedCentralGoogle Scholar
- 7.Zhang XH, Chutatape O, Ieee (2004) Detection and classification of bright lesions in color fundus images. Icip: 2004 international conference on image processing 1–5: 139–142Google Scholar
- 10.Bejnordi BE, Zuidhof G, Balkenhol M, Hermsen M, Bult P, van Ginneken B, Karssemeijer N, Litjens G, van der Laak J (2017) Context-aware stacked convolutional neural networks for classification of breast carcinomas in whole-slide histopathology images. J Med Imaging (Bellingham) 4:044504. https://doi.org/10.1117/1.jmi.4.4.044504 CrossRefGoogle Scholar
- 13.Abramoff MD, Lou Y, Erginay A, Clarida W, Amelon R, Folk JC, Niemeijer M (2016) Improved automated detection of diabetic retinopathy on a publicly available dataset through integration of deep learning. Invest Ophthalmol Vis Sci 57:5200–5206. https://doi.org/10.1167/iovs.16-19964 CrossRefPubMedGoogle Scholar
- 17.Kamel M, Belkassim S, Mendonca AM, Campilho A, Ieee; Ieee; Ieee I (2001) A neural network approach for the automatic detection of microaneurysms in retinal angiograms. Ijcnn'01: International Joint Conference on Neural Networks 1–4, Proceedings: 2695–2699Google Scholar
- 18.Hafez M, Azeem SA, Ieee; Ieee; Ieee I (2002) Using adaptive edge technique for detecting microaneurysms in fluorescein angiograms of the ocular fundus. 11th Ieee Mediterranean Electrotechnical Conference, Proceedings: 479–483Google Scholar
- 19.Tavakoli M, Shahri RP, Pourreza H, Mehdizadeh A, Banaee T, Toosi MHB (2013) A complementary method for automated detection of microaneurysms in fluorescein angiography fundus images to assess diabetic retinopathy. Pattern Recogn 46:2740–2753. https://doi.org/10.1016/j.patcog.2013.03.011 CrossRefGoogle Scholar
- 20.Frame AJ, Undrill PE, Cree MJ, Olson JA, McHardy KC, Sharp PF, Forrester JV (1998) A comparison of computer based classification methods applied to the detection of microaneurysms in ophthalmic fluorescein angiograms. Comput Biol Med 28:225–238. https://doi.org/10.1016/s0010-4825(98)00011-0 CrossRefPubMedGoogle Scholar
- 24.Gulshan V, Peng L, Coram M, Stumpe MC, Wu D, Narayanaswamy A, Venugopalan S, Widner K, Madams T, Cuadros J, Kim R, Raman R, Nelson PC, Mega JL, Webster DR (2016) Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. Jama 316:2402–2410. https://doi.org/10.1001/jama.2016.17216 CrossRefPubMedGoogle Scholar