Abstract
Diabetic Retinopathy (DR) is an ocular complication of diabetes that leads to a significant loss of vision. Screening retinal fundus images allows ophthalmologists to early detect and diagnose this disease; however, the manual interpretation of images is a time-consuming task. Deep image classification models deal with this drawback and provide an efficient method to diagnose DR, but they are mainly trained with balanced dataset were all stages of DR are equally represented—a scenario that does not reflect the reality during screening. In this work, we have conducted a study of Deep Learning models to rate the severity of Diabetic Retinopathy from digital fundus images when working with a highly imbalanced dataset. Our approach, that is based on the ensemble of several ResNetRS models, achieves an agreement rate with the specialists above that of primary care physicians (\(\kappa \) of 0.81 vs. 0.75). In addition, our method can also be applied in a binary fashion to determine whether a case is derivable or non-derivable; achieving again an agreement with the specialists above that of primary care physicians (\(\kappa \) of 0.76 vs. 0.64).
This work was partially supported by Ministerio de Ciencia e Innovación [PID2020-115225RB-I00/AEI/10.13039/501100011033]. Ángela Casado-García has a FPI grant from Community of La Rioja 2020. Manuel García-Domínguez has a FPI grant from Community of La Rioja 2018. Adrián Inés has a FPU Grant [16/06903] of the Spanish Ministerio de Educación y Ciencia.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Alyoubi, W.L., Shalash, W.M., Abulkhair, M.F.: Diabetic retinopathy detection through deep learning techniques: a review. Inform. Med. Unlocked 20, 100377 (2020)
Authors not listed: Diabetic eye screening: guidance on camera approval (2020). https://www.gov.uk/government/publications/diabetic-eye-screening-approved-cameras-and-settings/diabetic-eye-screening-guidance-on-camera-approval
Bello, I., et al.: Revisiting resnets: improved training and scaling strategies. Adv. Neural Inf. Process. Syst. 34 (2021)
Dietterich, T.G.: Robust artificial intelligence and robust human organizations. Front. Comput. Sci. 13(1), 1–3 (2019). https://doi.org/10.1007/s11704-018-8900-4
He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: IEEE Conference on Computer Vision and Pattern Recognition, pp. 770–778 (2016)
Howard, J., Gugger, S.: Fastai: a layered API for deep learning. Information 11, 108 (2020)
Li, T., Gao, Y., Wang, K., Guo, S., Liu, H., Kang, H.: Diagnostic assessment of deep learning algorithms for diabetic retinopathy screening. Inf. Sci. 501, 511–522 (2019)
Meijering, E.: A bird’s-eye view of deep learning in bioimage analysis. Comput. Struct. Biotechnol. J. 18, 2312–2325 (2020)
Nagpal, D., Panda, S., Malarvel, M., Pattanaik, P.A., Khan, M.Z.: A review of diabetic retinopathy: datasets, approaches, evaluation metrics and future trends. J. King Saud Univ.-Comput. Inf. Sci. 34(9), 7138–7152 (2021)
Simard, P., Steinkraus, D., Platt, J.C.: Best practices for convolutional neural networks applied to visual document analysis. In: 12th International Conference on Document Analysis and Recognition, vol. 2, pp. 958–964 (2003)
Smith, L.: Cyclical learning rates for training neural networks. In: IEEE Winter Conference on Applications of Computer Vision, pp. 464–472 (2017)
Szegedy, C., et al.: Rethinking the inception architecture for computer vision. In: IEEE Conference on Computer Vision and Pattern Recognition, pp. 2818–2826 (2016)
Tan, M., Le, Q.V.: Efficientnet: rethinking model scaling for convolutional neural networks. In: International Conference on Machine Learning, vol. 97, pp. 6105–6114 (2019)
Vidal-Alaball, J., et al.: Artificial intelligence for the detection of diabetic retinopathy in primary care: protocol for algorithm development. JMIR Res Protoc. 1(8), e12539 (2019)
Wang, J., et al.: Deep high-resolution representation learning for visual recognition. IEEE Trans. Pattern Anal. Mach. Intell. 43(10), 3349–3364 (2020)
Zapata, M., et al.: Telemedicine for a general screening of retinal disease using nonmydriatic fundus cameras in optometry centers: Three-year results. Telemed. e-Health 23(1), 30–36 (2017)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Casado-García, Á., García-Domínguez, M., Heras, J., Inés, A., Royo, D., Zapata, M.Á. (2022). Rating the Severity of Diabetic Retinopathy on a Highly Imbalanced Dataset. In: Moreno-Díaz, R., Pichler, F., Quesada-Arencibia, A. (eds) Computer Aided Systems Theory – EUROCAST 2022. EUROCAST 2022. Lecture Notes in Computer Science, vol 13789. Springer, Cham. https://doi.org/10.1007/978-3-031-25312-6_52
Download citation
DOI: https://doi.org/10.1007/978-3-031-25312-6_52
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-25311-9
Online ISBN: 978-3-031-25312-6
eBook Packages: Computer ScienceComputer Science (R0)