Classification of the Lipid Layer of the Tear Film using K-Means and Deep Learning
Abstract
Dry eye syndrome, one of the most common ophthalmic conditions, poses diagnostic challenges due to its multifactorial nature. Traditional diagnosis, involving manual classification of images of the tear film, is limited by the instability of this film. This study proposes an innovative approach that combines unsupervised learning for region of interest segmentation and deep learning for patch classification. The results are promising, with an accuracy of 99.23% and an F1-score of 99.16%, surpassing other techniques and studies in the literature. It is believed that this methodology could be a valuable tool to assist experts in disease diagnosis.
References
Abadi, M., Agarwal, A., Barham, P., Brevdo, E., Chen, Z., Citro, C., Corrado, G. S., Davis, A., Dean, J., Devin, M., et al. (2015). Tensorflow: Large-scale machine learning on heterogeneous systems. Software available from tensorflow.org.
Burman, P. (1989). A comparative study of ordinary cross-validation, v-fold crossvalidation and the repeated learning-testing methods. Biometrika, 76(3):503–514.
Chollet, F. et al. (2015). Keras. [link].
Cruz, L. B., Araújo, J. D. L., Sousa, J. A., Almeida, J. D. S., Júnior, G. B., Silva, A. C., and Paiva, A. C. (2018a). Classificação do filme lacrimal usando a função k de ripley como descritor de textura. In Anais do XVIII Simpósio Brasileiro de Computação Aplicada à Saúde, Porto Alegre, RS, Brasil. SBC.
Cruz, L. B. d., Araújo, J. D. L., Souza, J. C., de Sousa, J. A., de Almeida, J. D. S., Junior, G. B., Silva, A. C., and de Paiva, A. C. (2018b). Tear film classification using phylogenetic diversity indexes as texture descriptor. In 2018 IEEE Symposium on Computers and Communications (ISCC), pages 00853–00858.
da Cruz, L. B., Souza, J. C., de Paiva, A. C., de Almeida, J. D. S., Junior, G. B., Aires, K. R. T., Silva, A. C., and Gattass, M. (2020). Tear film classification in interferometry eye images using phylogenetic diversity indexes and ripley’s k function. IEEE Journal of Biomedical and Health Informatics, 24(12):3491–3498.
da Cruz, L. B., Souza, J. C., de Sousa, J. A., Santos, A. M., de Paiva, A. C., de Almeida, J. D. S., Silva, A. C., Junior, G. B., and Gattass, M. (2020). Interferometer eye image classification for dry eye categorization using phylogenetic diversity indexes for texture analysis. Computer Methods and Programs in Biomedicine, 188:105269.
Davidson, I. (2002). Understanding k-means non-hierarchical clustering. Computer Science Department of State University of New York (SUNY), Albany.
Doane, M. G. (1989). An instrument for in vivo tear film interferometry. Optometry and Vision Science, 66(6):383–388.
Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., Uszkoreit, J., and Houlsby, N. (2020). An image is worth 16x16 words: Transformers for image recognition at scale. CoRR, abs/2010.11929.
G. Villaverde, D., Remeseiro, B., Barreira, N., G. Penedo, M., and Mosquera, A. (2014). Feature selection applied to human tear film classification. In Proceedings of the 6th International Conference on Agents and Artificial Intelligence Volume 1, ICAART 2014, pages 395–402, Portugal. SCITEPRESS Science and Technology Publications, Lda.
Gedraite, E. S. and Hadad, M. (2011). Investigation on the effect of a gaussian blur in image filtering and segmentation. In Proceedings ELMAR-2011, pages 393–396.
Huang, G., Liu, Z., and Weinberger, K. Q. (2016). Densely connected convolutional networks. CoRR, abs/1608.06993.
Júnior, D. D., Cruz, L., Diniz, J., Júnior, G. B., and Silva, A. (2021). Classificação automática de glóbulos brancos usando descritores de forma e textura e extreme gradient boosting. In Anais do XXI Simpósio Brasileiro de Computação Aplicada à Saúde, pages 95–106, Porto Alegre, RS, Brasil. SBC.
Koonce, B. and Koonce, B. (2021). Resnet 50. Convolutional Neural Networks with Swift for Tensorflow: Image Recognition and Dataset Categorization, pages 63–72.
Leite, F., Almeida, J., Teixeira, J., Cruz, L., Junior, G., and Paiva, A. (2020). Planejamento cirúrgico de estrabismo horizontal utilizando Árvore de regressão de múltiplas saídas. In Anais do XX Simpósio Brasileiro de Computação Aplicada à Saúde, pages 356–367, Porto Alegre, RS, Brasil. SBC.
Marculino, L. G. C., Hazarbassanov, R. M., Hazarbassanov, N. G. T. d. Q., Hirai, F., Milhomens Filho, J. A. P., Wakamatsu, T. H., and Gomes, J. A. (2022). Prevalence and risk factors for dry eye disease: the sao paulo dry eye study. Arquivos Brasileiros de Oftalmologia, 85:549–557.
Messmer, E. M. (2015). The Pathophysiology, Diagnosis, and Treatment of Dry Eye Disease. Dtsch Arztebl International, 112(5):71–82.
Panetta, K., Zhou, Y., Agaian, S., and Jia, H. (2011). Nonlinear unsharp masking for mammogram enhancement. IEEE Transactions on Information Technology in Biomedicine, 15(6):918–928.
Pereira, R., Matos, C., Diniz, J., Junior, G. B., de Almeida, J., Silva, A., and de Paiva, A. (2016). Abordagem deep learning para classificação de lesões mamárias. In Anais do XVI Workshop de Informática Médica, pages 2597–2600, Porto Alegre, RS, Brasil. SBC.
Quintana, G. I., Li, Z., Vancamberg, L., Mougeot, M., Desolneux, A., and Muller, S. (2023). Exploiting patch sizes and resolutions for multi-scale deep learning in mammogram image classification. Bioengineering, 10(5):534.
Remeseiro, B., Oliver, K. M., Tomlinson, A., Martin, E., Barreira, N., and Mosquera, A. (2015). Automatic grading system for human tear films. Pattern Analysis and Applications, 18(3):677–694.
Simonyan, K. and Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556.
Stapleton, F., Alves, M., Bunya, V. Y., Jalbert, I., Lekhanont, K., Malet, F., Na, K.-S., Schaumberg, D., Uchino, M., Vehof, J., Viso, E., Vitale, S., and Jones, L. (2017). Tfos dews ii epidemiology report. The Ocular Surface, 15(3):334 – 365. TFOS International Dry Eye WorkShop (DEWS II).
Tan, M. and Le, Q. V. (2021). Efficientnetv2: Smaller models and faster training. CoRR, abs/2104.00298.
VOPTICAL GCU, V. (2017). optical dataset acquired and annotated by optometrists from the department of life sciences, glasgow caledonian university (uk), 2013. [link].
Burman, P. (1989). A comparative study of ordinary cross-validation, v-fold crossvalidation and the repeated learning-testing methods. Biometrika, 76(3):503–514.
Chollet, F. et al. (2015). Keras. [link].
Cruz, L. B., Araújo, J. D. L., Sousa, J. A., Almeida, J. D. S., Júnior, G. B., Silva, A. C., and Paiva, A. C. (2018a). Classificação do filme lacrimal usando a função k de ripley como descritor de textura. In Anais do XVIII Simpósio Brasileiro de Computação Aplicada à Saúde, Porto Alegre, RS, Brasil. SBC.
Cruz, L. B. d., Araújo, J. D. L., Souza, J. C., de Sousa, J. A., de Almeida, J. D. S., Junior, G. B., Silva, A. C., and de Paiva, A. C. (2018b). Tear film classification using phylogenetic diversity indexes as texture descriptor. In 2018 IEEE Symposium on Computers and Communications (ISCC), pages 00853–00858.
da Cruz, L. B., Souza, J. C., de Paiva, A. C., de Almeida, J. D. S., Junior, G. B., Aires, K. R. T., Silva, A. C., and Gattass, M. (2020). Tear film classification in interferometry eye images using phylogenetic diversity indexes and ripley’s k function. IEEE Journal of Biomedical and Health Informatics, 24(12):3491–3498.
da Cruz, L. B., Souza, J. C., de Sousa, J. A., Santos, A. M., de Paiva, A. C., de Almeida, J. D. S., Silva, A. C., Junior, G. B., and Gattass, M. (2020). Interferometer eye image classification for dry eye categorization using phylogenetic diversity indexes for texture analysis. Computer Methods and Programs in Biomedicine, 188:105269.
Davidson, I. (2002). Understanding k-means non-hierarchical clustering. Computer Science Department of State University of New York (SUNY), Albany.
Doane, M. G. (1989). An instrument for in vivo tear film interferometry. Optometry and Vision Science, 66(6):383–388.
Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., Uszkoreit, J., and Houlsby, N. (2020). An image is worth 16x16 words: Transformers for image recognition at scale. CoRR, abs/2010.11929.
G. Villaverde, D., Remeseiro, B., Barreira, N., G. Penedo, M., and Mosquera, A. (2014). Feature selection applied to human tear film classification. In Proceedings of the 6th International Conference on Agents and Artificial Intelligence Volume 1, ICAART 2014, pages 395–402, Portugal. SCITEPRESS Science and Technology Publications, Lda.
Gedraite, E. S. and Hadad, M. (2011). Investigation on the effect of a gaussian blur in image filtering and segmentation. In Proceedings ELMAR-2011, pages 393–396.
Huang, G., Liu, Z., and Weinberger, K. Q. (2016). Densely connected convolutional networks. CoRR, abs/1608.06993.
Júnior, D. D., Cruz, L., Diniz, J., Júnior, G. B., and Silva, A. (2021). Classificação automática de glóbulos brancos usando descritores de forma e textura e extreme gradient boosting. In Anais do XXI Simpósio Brasileiro de Computação Aplicada à Saúde, pages 95–106, Porto Alegre, RS, Brasil. SBC.
Koonce, B. and Koonce, B. (2021). Resnet 50. Convolutional Neural Networks with Swift for Tensorflow: Image Recognition and Dataset Categorization, pages 63–72.
Leite, F., Almeida, J., Teixeira, J., Cruz, L., Junior, G., and Paiva, A. (2020). Planejamento cirúrgico de estrabismo horizontal utilizando Árvore de regressão de múltiplas saídas. In Anais do XX Simpósio Brasileiro de Computação Aplicada à Saúde, pages 356–367, Porto Alegre, RS, Brasil. SBC.
Marculino, L. G. C., Hazarbassanov, R. M., Hazarbassanov, N. G. T. d. Q., Hirai, F., Milhomens Filho, J. A. P., Wakamatsu, T. H., and Gomes, J. A. (2022). Prevalence and risk factors for dry eye disease: the sao paulo dry eye study. Arquivos Brasileiros de Oftalmologia, 85:549–557.
Messmer, E. M. (2015). The Pathophysiology, Diagnosis, and Treatment of Dry Eye Disease. Dtsch Arztebl International, 112(5):71–82.
Panetta, K., Zhou, Y., Agaian, S., and Jia, H. (2011). Nonlinear unsharp masking for mammogram enhancement. IEEE Transactions on Information Technology in Biomedicine, 15(6):918–928.
Pereira, R., Matos, C., Diniz, J., Junior, G. B., de Almeida, J., Silva, A., and de Paiva, A. (2016). Abordagem deep learning para classificação de lesões mamárias. In Anais do XVI Workshop de Informática Médica, pages 2597–2600, Porto Alegre, RS, Brasil. SBC.
Quintana, G. I., Li, Z., Vancamberg, L., Mougeot, M., Desolneux, A., and Muller, S. (2023). Exploiting patch sizes and resolutions for multi-scale deep learning in mammogram image classification. Bioengineering, 10(5):534.
Remeseiro, B., Oliver, K. M., Tomlinson, A., Martin, E., Barreira, N., and Mosquera, A. (2015). Automatic grading system for human tear films. Pattern Analysis and Applications, 18(3):677–694.
Simonyan, K. and Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556.
Stapleton, F., Alves, M., Bunya, V. Y., Jalbert, I., Lekhanont, K., Malet, F., Na, K.-S., Schaumberg, D., Uchino, M., Vehof, J., Viso, E., Vitale, S., and Jones, L. (2017). Tfos dews ii epidemiology report. The Ocular Surface, 15(3):334 – 365. TFOS International Dry Eye WorkShop (DEWS II).
Tan, M. and Le, Q. V. (2021). Efficientnetv2: Smaller models and faster training. CoRR, abs/2104.00298.
VOPTICAL GCU, V. (2017). optical dataset acquired and annotated by optometrists from the department of life sciences, glasgow caledonian university (uk), 2013. [link].
Published
2024-06-25
How to Cite
DIAS JUNIOR, Domingos A.; CRUZ, Luana B. da; DINIZ, João O. B..
Classification of the Lipid Layer of the Tear Film using K-Means and Deep Learning. In: BRAZILIAN SYMPOSIUM ON COMPUTING APPLIED TO HEALTH (SBCAS), 24. , 2024, Goiânia/GO.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2024
.
p. 1-12.
ISSN 2763-8952.
DOI: https://doi.org/10.5753/sbcas.2024.1500.
