Abstract
Pathological myopia (PM), which results from degenerative changes in the sclera, choroid, and retinal pigment epithelium (RPE), is associated with irreversible vision loss. This study proposes automatically detecting PM or normal vision from input retina fundus image. We have experimented with various transfer learning models and the pre-processing steps using reinforcement learning (RL). The best results were achieved with our custom ResNet50 as a baseline model. It has achieved an AUC score of 0.9984 on the validation dataset provided by the PALM challenge, a Satellite Event of The IEEE International Symposium on Biomedical Imaging in Venice, Italy. This AUC score is among the top 3 performers in this challenge. As in medical domain, more accurate results are always in demand, and this score ensures that the model can be set up for a clinical application in future as a second opinion to ophthalmologists.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Wang YX, Wang S, You QS, Jonas JB, Liu HH, Xu L (2010) Prevalence and progression of myopic retinopathy in Chinese adults: the Beijing eye study. Ophthalmology 117
Ohno-Matsui K (2017) What is the fundamental nature of pathologic myopia? Retina 37
Flitcroft DI (2012) The complex interactions of retinal, optical and environmental factors in myopia aetiology. Prog Retinal Eye Res 31
Montolio FGJ, Jansonius NM, Marcus MW, de Vries MM (2011) Myopia as a risk factor for open-angle glaucoma: a systematic review and meta-analysis. Ophthalmology 118
Santhanam N, Kim HE, Cosa-Linan A et al (2022) Transfer learning for medical image classification: a literature review. BMC Med Imaging 69
Sutton RS, Barto AG (2018) Reinforcement learning: an introduction. MIT Press, Cambridge
Sun X, Liao J, Xu Y, Zhang S, Zhang X, Fu H, JosĆ© FL, Orlando I, BogunoviÄ H (2019) Palm: pathologic myopia challenge
Gilani SO, Waris A, Rauf N (2021) Automatic detection of pathological myopia using machine learning. Sci Rep 11:16570
Orlando JI, Bogunovic H, Sun X, Liao J, Xu Y, Zhang S, Zhang X, Fu H, Li F (2019) Palm: pathologic myopia challenge. IEEE Dataport
Blaschko MB, Jacob J, Stalmans I, De Boever P, Hemelings R, Elen B (2021) Pathological myopia classification with simultaneous lesion segmentation using deep learning. Comput Methods Programs Biomed 199:105920
He K, Zhang X, Ren S, Sun J (2015) Deep residual learning for image recognition
Freire CR, Moura JCC, Daniele MS et al (2020) Automatic lesion segmentation and pathological myopia classification in fundus images. arXiv: abs/2002.06382
Breda JB, van Keer K, Bathula DR, Diaz-Pinto A, Orlando JI, Fu F et al (2020) Refuge challenge: a unified framework for evaluating automated methods for glaucoma assessment from fundus photographs. Med Image Anal 59:101570
Almazroa A (2018) Retinal fundus images for glaucoma analysis: the Riga dataset. Deep blue data. University of Michigan
Kamble R, Kokare M, Porwal P, Pachade S et al (2018) Indian diabetic retinopathy image dataset (IDRID): a database for diabetic retinopathy screening research. Data 3(3)
Chollet F (2017) Xception: deep learning with depthwise separable convolutions. In: Proceedings of the IEEE conference on computer vision and pattern recognition
Yu W, Lu L, Zhou E et al (2021) Development of deep learning-based detecting systems for pathologic myopia using retinal fundus images. Commun Biol 4
Devda J, Eswari R (2019) Pathological myopia image analysis using deep learning. Procedia Comput Sci 165
Zhuo Z, Cheng J, Jiang L et al (2012) Pathological myopia detection from selective fundus image features. In: 2012 7th IEEE conference on industrial electronics and applications (ICIEA), pp 1742ā1745
Saw SM, Chua WH et al (2002) Nearwork in early onset myopia. Invest Ophthalmol Vis Sci 43
Ding C, Peng HC, Long FH (2005) Feature selection based on mutual information: criteria of max-dependency, max-relevance, and minredundancy. IEEE Trans Pattern Anal Mach Intell 27
Kong ST, Jung K-H, Son J, Kim J (2021) Leveraging the generalization ability of deep convolutional neural networks for improving classifiers for color fundus photographs. Appl Sci 591
DecenciĆØre E, Zhang X et al (2014) Feedback on a publicly distributed image database: the Messidor database. Image Anal Stereology 33(3):231ā234
Briskilal J, Kalyanasundaram A, Prabhakaran S, Senthil Kumar D (2020) Detection of pathological myopia using convolutional neural network. Int J Psychosoc Rehabil 24
Pathan S, Siddalingaswamy PC et al (2020) Automated detection of pathological and non-pathological myopia using retinal features and dynamic ensemble of classifiers. Telecommun Radio Eng 79
Hung S-K, Gan JQ (2021) Augmentation of small training data using GANs for enhancing the performance of image classification. In: 2020 25th international conference on pattern recognition (ICPR), pp 3350ā3356
Amina A, Mohammed B (2018) Peeking inside the black-box: a survey on explainable artificial intelligence (XAI). IEEE Access 6:52138ā52160
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
Ā© 2023 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Ali, S., Raut, S. (2023). Detection of Pathological Myopia fromĀ Fundus Images. In: Sharma, H., Shrivastava, V., Bharti, K.K., Wang, L. (eds) Communication and Intelligent Systems. ICCIS 2022. Lecture Notes in Networks and Systems, vol 686. Springer, Singapore. https://doi.org/10.1007/978-981-99-2100-3_17
Download citation
DOI: https://doi.org/10.1007/978-981-99-2100-3_17
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-99-2099-0
Online ISBN: 978-981-99-2100-3
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)