Abstract
Trachoma is the leading bacterial infectious cause of blindness worldwide. Examination for clinical signs of trachoma involves careful inspection of the lashes, cornea, eversion of the upper lid, and the tarsal conjunctiva. In this paper, we present a system for automatic detection and grading of trachoma using deep convolutional network. Salient texture features that account for the symptom of the disease are extracted from the eye image using Gabor filters. Then, a texture feature based deep convolutional neural network is used for classification. A 4-way Softmax is used for grading into a specific class (normal, trachomatous scarring, trachomatous trichiasis, and corneal opacity). Although deep learning systems are known to extract and learn features from raw image, we also show that extracting characteristic features still improves the learning capability of deep learning systems. Our model is found to be faster to train and has smaller model size as compared to state-of-the-art models such as AlexNet and GoogLeNet. Furthermore, the model achieved a diagnosis accuracy of 97.9% for detecting and grading trachoma, which improves the accuracies obtained by AlexNet and GoogLeNet by 10% and 3%, respectively.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Data Availability
The image dataset utilized for training, validation, and test was acquired from St. Paul Hospital and Carter Center Ethiopia, Ethiopia. This dataset is not publicly available, and restrictions apply to their use.
Code Availability
The code for feature extraction and learning includes intellectual property and cannot be released publicly.
References
Adrian R (2017) Deep Learning for Computer Vision with Python. PyImageSearch
Alemayehu M, Gail D (2005) Assessing The Prevalance of Active Trachoma Among Young Children in Relation to The Implementation of SAFE Strategy in Ebinat and East Belesa woreda, North West Ethiopia. Addis Ababa University, School of Graduate Studie, pp 1–61
Alex K, Ilya S, Geoffrey EH (2012) Imagenet classification with deep convolutional neural networks. Advances in Neural Information Processing Systems, pp 1097–1105
Amit AB, Manu J, Sushila S (2016) Automated Detection of eye diseases. In: 016 international conference on wireless communications, signal processing and networking (WiSPNET), Chennai
Anish Singh Walia (2019) Activation functions and it’s types-Which is better?. https://www.towardsdatascience.com/activation-functions-and-its-types-which-is-better-a9a5310cc8f. Accessed 12 November
Baidaa AB, Waleed AN, Majid AT, Yalin Z (2017) Automated glaucoma diagnosis using deep learning approach. In: 2017 14th international multi-conference on systems, signals & devices (SSD), Marrakech
Berhane Y, Worku A, Bejiga A, Adamu L (2008) National Survey on blindness, low vision and trachoma in ethiopia: methods and study clusters profile. in ethiopian journal of health development
Brian C Eye Infections. https://www.allaboutvision.com/conditions/eye-infections.htm. Accessed 20 Nov 2018
CDC (2008) Guidelines For management of trachoma in the northeren territory. Alice Springs
Darshit D, Aniket S, Deep S, Prachi G (2016) Diabetic retinopathy detection using deep convolutional neural networks. In: 2016 International conference on computing, analytics and security trends (CAST), Pune
Gad AF (2018) Practical computer vision applications using deep learning with CNNs. Apress
Gheisari S, Sharifou S, Phu J, Kennedy PJ, Agar A, Kalloniatis M, Golzan SM (2021) A combined convolutional and recurrent neural network for enhanced glaucoma detection. Scientific Reports
Gullì A (2017) Deep learning with Keras : implement neural networks with Keras on Theano and TensorFlow
He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. In: 2016 IEEE conference on computer vision and pattern recognition (CVPR), Las Vegas, NV, pp 770–778. https://doi.org/10.1109/CVPR.2016.90
Ian G, Yoshua B, Aaron C (2016) Deep learning. Cambridge Massachusetts, MIT Press
Ioffe S, Szegedy C (2015) Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift. arXiv:1502.031167v3
Juneja M, Singh S, Agarwal N, Bali S, Gupta S, Thakur N, Jindal P (2019) Automated detection of Glaucoma using deep learning convolution network (G-net), Springer, Multimedia Tools and Applications
Li M, Yunhong W, Tieniu T (2002) Iris recognition based on multichannel gabor filtering. In: Proc Fifth Asian Conf Computer Vision Vol 1 Australia
Matthew CK, Kazunori O, Alexander MR, Abdou A, Zerihun T, Sun YC, el et (2019) Sensitivity and specificity of computer vision classification of eyelid photographs for programmatic trachoma assessment. PLOS ONE 14:1–12
Mohammad A, Miad F (2016) Determination For glaucoma disease based on red area percentage. In: 2016 IEEE long island systems, applications and technology conference (LISAT), Farmingdale, NY
Mrunalini DM, Krishna KW (2017) Histogram of oriented gradient based automated detection of eye diseases. In: Third international conference on computing, communication, control and automation
Oh K, Kang HM, Leem D, Lee H, Seo KY, Yoon S (2021) Early detection of diabetic retinopathy based on deep learning and ultra-wide-feld fundus images. Scientific Reports
Saadia M, Muhammed YJ (2009) Iris Feature extraction using gabor filter. In: 2009 international conference on emerging technologies (Islamabad), Pakistan
Samar KB (2013) Atlas of clinical ophthalmology. Jaypee brothers Medical Publishers, New Delhi
Saul NR, Richard OCJ, Matthew JB (2012) Trachomatous Trichiasis and its Management in Endemic Countries. Survey of Ophthalmology, pp 105–135
Sheila KW (2004) Trachoma: new assault on an ancient disease. Prog Retin Eye Res, pp 381–401
Simonyan K, Zisserman A (2014) Very deep convolutional networks for Large-Scale image recognition. arXiv:abs/1409.1556
Szegedy C et al (2015) Going deeper with convolutions. In: 2015 IEEE conference on computer vision and pattern recognition (CVPR), Boston, MA, pp 1–9. https://doi.org/10.1109/CVPR.2015.7298594
Thylefors B, Dawson CR, Jones BR, West SK, Taylor HR (1987) A Simple System for the Assessment of Trachoma and its Complications. Buletin of the World Health Organization 65:477–483
WHO Blindness: Vision 2020, WHO Media centre, https://www.who.int/blindness/causes/trachoma/en/. Accessed 13 Nov 2018
WHO Trachoma, 18 November 2018. http://www.who.int/news-room/fact-sheets/detail/trachoma. Accessed 20 Nov 2018
WHO Trachoma, World Health Organization, https://www.who.int/trachoma/diagnosis/en/. Accessed 17 Mar 2019
Xiangyu C, Yanwu X, Damon WKW, Tien YW, Jiang L (2015) Glaucoma detection based on deep convolutional neural network. In: 2015 37th annual international conference of the IEEE engineering in medicine and biology society (EMBC), Milan
Yagmur FD, Karlik B, Okatan A (2008) Automatic recognition of retinopathy diseases by using wavelet based neural network. IEEE
Yanyan D, Qinyan Z, Zhiqiang Q, Ji-Jiang Y (2017) Classification of cataract fundus image based on deep learning. In: 2017 IEEE international conference on imaging systems and techniques (IST), Beijing
Yashal SK, Bhargav S, Savita C (2017) Detecting diabetic retinopathy using deep learning. In: 2017 2nd IEEE international conference on recent trends in electronics, information & communication technology (RTEICT), Bangalore
Yemane B, Alemayehu W, Wondu A, Abebe B, Liknaw A, Amir B, Zegeye H, Allehone A, Yilikal A, Teshome G, Tewodros DK, Emily W, Sheila W (2007) Prevalence and causes of blindness and low vision in ethiopia. Ethiop J Health Dev, pp 204–210
Author information
Authors and Affiliations
Contributions
B.Y.: Writing original draft, research and experimental planning, deep learning system design, experiments, results analysis, manuscript writing, and revised manuscript; Y.A.: Writing: review and editing, texture feature based deep learning design, writing parts of the manuscript, and revised manuscript
Corresponding author
Ethics declarations
Conflict of Interests
The authors declare no competing interests.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix: Sample Dataset for CO, TT, TS, and Normal cases from Top to Bottom, respectively
Appendix: Sample Dataset for CO, TT, TS, and Normal cases from Top to Bottom, respectively
Rights and permissions
About this article
Cite this article
Yenegeta, B., Assabie, Y. TrachomaNet: Detection and grading of trachoma using texture feature based deep convolutional neural network. Multimed Tools Appl 82, 4209–4234 (2023). https://doi.org/10.1007/s11042-022-13214-2
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-022-13214-2