1. Proceedings
  2. I3M
  3. 2020
  4. IWISH
  5. 10.46354/i3m.2020.iwish.004

Multitask learning for medical image classification using VGG architecture

  • a Yuan Yang, 
  • b Lin Zhang , 
  • c Lei Ren, 
  • d Yuanjun Lali 
  • a,b,c School of Automation Science and Electrical Engineering, No. 37 Xueyuan Road, Haidian District, Beijing, 100191, China
  • a,b,c Beijing Advanced Innovation Center for Big Data-Based Precision Medicine, School of Medicine and Engineering,No. 37 Xueyuan Road, Haidian District, Beijing, 100191, China
  • a,b,c Key Laboratory of Big Data-Based Precision Medicine,Ministry of Industry and Information Technology, No. 37 Xueyuan Road, Haidian District, Beijing, 100191, China
Cite as
Yang Y., Zhang L., Ren L., Lali Y. (2020). Multitask learning for medical image classification using VGG architecture. Proceedings of the 9th International Workshop on Innovative Simulation for Healthcare (IWISH 2020), pp. 20-26.
DOI: https://doi.org/10.46354/i3m.2020.iwish.004

Abstract

In order to recognize different kinds of medical images under a single network structure background, a multi-task medical image recognition model based on the combination of transfer learning and automatic path search is proposed. Based on VGG-16 model, a neural network module is designed and evolutionary algorithm is used to select the path. Experiments were conducted on ECG data sets and pneumonia data sets respectively. Finally, a joint classification test was conducted on these 2 datasets. Ultimately, joint classification experiments on the ECG and pneumonia datasets resulted in an overall accuracy of 93% and a recall rate of 88%.

References

  1. Acharya, U. R., Fujita, H., Lih, O. S., Hagiwara, Y., Tan, J. H., and Adam, M. (2017). Automated detection of arrhythmias using different intervals of tachycardia ECG segments with convolutional neural network. Information Sciences, 405:81–90.
  2. Caruana, R. (1997). Multitask Learning. Machine Learning, 28(1):41–75.
  3. Chaichulee, S., Villarroel, M., Jorge, J., Arteta, C., Green, G., McCormick, K., Zisserman, A., and Tarassenko, L. (2017). Multi-Task Convolutional Neural Network for Patient Detection and Skin Segmentation in Continuous Non-Contact Vital Sign Monitoring. Proceedings - 12th IEEE International Conference on Automatic Face and Gesture Recognition, FG 2017 - 1st International Workshop on Adaptive Shot Learning for Gesture Understanding and Production, ASL4GUP 2017, Biometrics in the Wild, Bwild 2017, Heteroge, pages 266–272.
  4. Chen, S., Bortsova, G., Juárez, A. G.-u., Tulder, G. V., and Bruijne, M. D. Learning for Medical Image Segmentation. pages 1–9.
  5. Chen, X., Ma, H., Wan, J., Li, B., and Xia, T. (2017). Multi-view 3D object detection network for autonomous driving. Proceedings - 30th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017, 2017-Janua:6526–6534.
  6. Dinov, I. D. (2016). Methodological challenges and analytic opportunities for modeling and interpreting Big Healthcare Data. GigaScience, 5(1).
  7. Fernando, C., Banarse, D., Blundell, C., Zwols, Y., Ha, D., Rusu, A. A., Pritzel, A., and Wierstra, D. (2017). PathNet: Evolution Channels Gradient Descent in Super Neural Networks.
  8. Inan, O. T., Giovangrandi, L., and Kovacs, G. T. (2006). Robust neural-network-based classification of premature ventricular contractions using wavelet transform and timing interval features. IEEE Transactions on Biomedical Engineering, 53(12):2507–2515.
  9. Liu, F., Wee, C. Y., Chen, H., and Shen, D. (2014). Intermodality relationship constrained multi-modality multi-task feature selection for Alzheimer’s Disease and mild cognitive impairment identification. NeuroImage, 84:466–475.
  10. Liu, X., He, P., Chen, W., and Gao, J. (2020). Multi-task deep neural networks for natural language understanding. ACL 2019 - 57th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference, pages 4487–4496.
  11. Simonyan, K. and Zisserman, A. (2015). Very deep convolutional networks for large-scale image recognition. 3rd International Conference on Learning Representations, ICLR 2015 - Conference Track Proceedings, pages 1–14. 
  12. Wang, X., Ju, L., Zhao, X., and Ge, Z. (2019). Retinal abnormalities recognition using regional multitask learning. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 11764 LNCS:30–38.
  13. Yan, K., Tang, Y., Peng, Y., Sandfort, V., Bagheri, M., Lu, Z., and Summers, R. M. (2019). MULAN: Multitask Universal Lesion Analysis Network for Joint Lesion Detection, Tagging, and Segmentation. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 11769 LNCS:194–202.
  14. Yin, X. and Member, X. L. Multi-Task Convolutional Neural Network for Pose-Invariant Face Recognition. pages 1–12.
  15. Zoph, B., Vasudevan, V., Shlens, J., and Le, Q. V. (2018). Learning Transferable Architectures for Scalable Image Recognition. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pages 8697–8710.