Reconocimiento de imágenes médicas mediante aprendizaje automático

Thumbnail Image
Official URL
Full text at PDC
Publication Date
Advisors (or tutors)
Journal Title
Journal ISSN
Volume Title
Google Scholar
Research Projects
Organizational Units
Journal Issue
Este proyecto trata de aplicar técnicas de deep learning para solucionar un problema de categorización con múltiples etiquetas en imágenes médicas. En concreto, se utiliza un corpus de radiografías de pulmón con cuatro etiquetas: Normal, COVID-19, Opacidad de Pulmón y Neumonía Viral.El trabajo consta de un estudio inicial de las técnicas de reconocimiento de imagen usando deep learning, tratando los tipos de redes neuronales más comunes y los conceptos claves en el entrenamiento de estas. También se tratan conceptos más avanzados de las redes convolucionales que permitirán mejorar algunos puntos el rendimiento del modelo. Además, se realiza una experimentación sencilla con redes neuronales que identifican letras a modo de introducción a la tecnología. Finalmente, se proponen una serie de modelos para solventar el problema, se entrenan y se evalúan los resultados. Se ha llegado a conseguir una exactitud de 92,91% en el conjunto de validación, lo cual es un resultado bastante satisfactorio, teniendo en cuenta el problema planteado.
The purpose of this project is to apply deep learning techniques to solve a multi label classification problem with medical images. For that, a corpus of lung radiographies with four labels (Normal, COVID-19, Lung Opacity and Viral Pneumonia) will be used. The initial part of the project will consist in an study of the image recognition deep learning techniques. Most common types of neural networks will be explained as well as the fundamental concepts of their training. More advanced topics will also be treated such as separable convolutional layers or augmentation and batch normalization techniques. After that, a letter recognition problem will be solved using convolutional neural networks to get familiarized with the technology. Finally, several models will be proposed and trained for the medical image problems, following with an evaluation of the results. An accuracy of 92.91% was acquired in the validation set, which is definitely a good result taking into account the proposed problem.
[1] F. Chollet, Deep Learning with Python. Manning Publications Co., 2018. [2] I. Goodfellow, Y. Bengio, and A. Courville, Deep Learning. MIT Press, 2016. http: // [3] “Why tensorflow.” Última consulta: 2021-05-02. [4] “Pytorch.” Última consulta: 2021-05-02. [5] D. Silver, A. Huang, C. J. Maddison, A. Guez, L. Sifre, G. van den Driessche, J. Schrittwieser, I. Antonoglou, V. Panneershelvam, M. Lanctot, S. Dieleman, D. Grewe, J. Nham, N. Kalchbrenner, I. Sutskever, T. Lillicrap, M. Leach, K. Kavukcuoglu, T. Graepel, and D. Hassabis, “Mastering the game of go with deep neural networks and tree search,” Nature, vol. 529, no. 7587, pp. 484–489, 2016. [6] P. Lakhani and B. Sundaram, “Deep learning at chest radiography: Automated classification of pulmonary tuberculosis by using convolutional neural networks,” Radiology, vol. 284, no. 2, pp. 574–582, 2017. [7] S. Minaee, N. Kalchbrenner, E. Cambria, N. Nikzad, M. Chenaghlu, and J. Gao, “Deep learning–based text classification: A comprehensive review,” ACM Comput. Surv., vol. 54, Apr. 2021. [8] S. Grigorescu, B. Trasnea, T. Cocias, and G. Macesanu, “A survey of deep learning techniques for autonomous driving,” Journal of Field Robotics, vol. 37, no. 3, pp. 362–386, 2020. [9] J. B. Heaton, N. G. Polson, and J. H. Witte, “Deep learning for finance: deep portfolios,”Applied Stochastic Models in Business and Industry, vol. 33, no. 1, pp. 3–12, 2017. [10] L. Rice, E. Wong, and Z. Kolter, “Overfitting in adversarially robust deep learning,” in Proceedings of the 37th International Conference on Machine Learning (H. D. III and A. Singh, eds.), vol. 119 of Proceedings of Machine Learning Research, pp. 8093–8104, PMLR, 13–18 Jul 2020. [11] Sklearn, “Metrics and scoring: quantifying the quality of predictions.” classification-metrics. Última consulta: 2021-05-02. [12] “Tensorflow, dense layer.” layers/Dense. Última consulta: 2021-05-02. [13] S. S. S Sharma and A. Athaiya, “Activation functions in neural networks,” International Journal of Engineering Applied Sciences and Technology, vol. 4, no. 12, pp. 310–316, 2020. [14] S. Bock and M. Weiß, “A proof of local convergence for the adam optimizer,” in 2019 International Joint Conference on Neural Networks (IJCNN), pp. 1–8, 2019. [15] M. Sundermeyer, R. Schlüter, and H. Ney, “Lstm neural networks for language modeling,”in Thirteenth annual conference of the international speech communication association, 2012. [16] S. Albawi, T. A. Mohammed, and S. Al-Zawi, “Understanding of a convolutional neural network,” in 2017 International Conference on Engineering and Technology (ICET), pp. 1–6, 2017. [17] L. Perez and J.Wang, “The effectiveness of data augmentation in image classification using deep learning,” arXiv preprint arXiv:1712.04621, 2017. [18] J. Bjorck, C. P. Gomes, and B. Selman, “Understanding batch normalization,” CoRR, vol. abs/1806.02375, 2018. [19] F. Chollet, “Xception: Deep learning with depthwise separable convolutions,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), July 2017. [20] I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y. Bengio, “Generative adversarial networks,” Commun. ACM, vol. 63, p. 139–144, Oct. 2020. [21] Y. Wang, F. Li, H. Sun, W. Li, C. Zhong, X. Wu, H. Wang, and P. Wang, “Improvement of MNIST image recognition based on CNN,” IOP Conference Series: Earth and Environmental Science, vol. 428, p. 012097, jan 2020. [22] C. Kusuma, A. Fadhilah, and A. Afiahayati, “Convolutional neural networks for handwritten javanese character recognition,” IJCCS (Indonesian Journal of Computing and Cybernetics Systems), vol. 12, p. 83, 01 2018. [23] “Covid-19 radiography database.” covid19-radiography-database. Última consulta: 2021-05-02. [24] Tensorflow, “Classification on imbalanced data.” Última consulta: 2021-05-02. [25] H. Farhat, G. E. Sakr, and R. Kilany, “Deep learning applications in pulmonary medical imaging: recent updates and insights on covid-19,” Machine Vision and Applications, vol. 31, no. 6, p. 53, 2020. [26] Y. Dong, Y. Pan, J. Zhang, and W. Xu, “Learning to read chest x-ray images from 16000+ examples using cnn,” in 2017 IEEE/ACM International Conference on Connected Health: Applications, Systems and Engineering Technologies (CHASE), pp. 51–57, 2017. [27] “Colab pro.” Última consulta: 2021-06-28. [28] “Nvidia t4 tensor core gpu for ai inference.” data-center/tesla-t4/. Última consulta: 2021-06-28. [29] A. Choi, O. Stephen, M. Sain, U. J. Maduh, and D.-U. Jeong, “An efficient deep learning approach to pneumonia classification in healthcare,” Journal of Healthcare Engineering, vol. 2019, p. 4180949, 2019.