Abstract
Over recent years, deep learning algorithms have gained prominence in medical image analysis research. Like other connectionist systems, such networks have been found to be prone to catastrophic forgetting effects. This makes generalization a challenge as new additions to prediction requirements at runtime would invariably require retraining on not only the new dataset, but also substantial portions of older task data. This is a difficult task in clinical imaging where retention of datasets over extended time is challenged by legal and infrastructure constraints. Thus, there is a requirement of algorithmic designs that address for-getting as a part of base and incremental task learning. This has been cast as an incremental learning problem recently. We propose a novel approach to the incremental class addition problem, where a retention of limited numbers of exemplars of old classes helps reduce forgetting instead of large scale data storage, using a strategy of incremental time augmentation with Mobius transformations and weighted distillation objectives to correct evolving class imbalance effects.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Goodfellow, I.J., Mirza, M., Xiao, D., Courville, A., Bengio, Y. An empirical investigation of catastrophic forgetting in gradient-based neural networks. arXiv:1312.6211 (2013)
Ravishankar, H., et al.: Understanding the mechanisms of deep transfer learning for medical images. In: Carneiro, G., et al. (eds.) LABELS/DLMIA -2016. LNCS, vol. 10008, pp. 188–196. Springer, Cham (2016). https://doi.org/10.1007/978-3-319-46976-8_20
Patra, A., Chakraborti, T.: Learn more, forget less: cues from human brain. In: Proceedings of the Asian Conference on Computer Vision (2020)
Li, Z., Hoiem, D.: Learning without forgetting. IEEE Trans. Pattern Anal. Mach. Intell. 40(12), 2935–2947 (2017)
Kemker, R., Kanan, C.: Fearnet: brain-inspired model for incremental learning. arXiv:1711.10563 (2017)
Patra, A., Cai, Y., Chatelain, P., Sharma, H., Drukker, L., Papageorghiou, A.T., Noble, J.A.: Multimodal continual learning with sonographer eye-tracking in fetal ultrasound. In: International Workshop on Advances in Simplifying Medical Ultrasound, pp. 14–24 (2020)
Kirkpatrick, J., et al.: Overcoming catastrophic forgetting in neural networks. In: Proceedings of the National Academy of Sciences, pp. 3521–3526 (2017)
Hinton, G., Vinyals, O., Dean, J.: Distilling the knowledge in a neural network. In: NIPS 2014 Deep Learning Workshop (2014)
Patra, A., Noble, J.A.: Multi-anatomy localization in fetal echocardiography videos. In: 2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019), pp. 1761–1764 (2019)
Rebuffi, S.A., Kolesnikov, A., Sperl, G., Lampert, C.H.: iCaRl: incremental classifier and representation learning. In: Proceedings of the IEEE CVPR, pp. 2001–2010 (2017)
Patra, A., et al.: Efficient ultrasound image analysis models with sonographer gaze assisted distillation. In: Shen, D., et al. (eds.) MICCAI 2019. LNCS, vol. 11767, pp. 394–402. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-32251-9_43
Hou, S., Pan, X., Loy, C.C., Wang, Z., Lin, D.: Lifelong learning via progressive distillation and retrospection. In: Ferrari, V., Hebert, M., Sminchisescu, C., Weiss, Yair (eds.) ECCV 2018. LNCS, vol. 11207, pp. 452–467. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-01219-9_27
Ozdemir, F., Fuernstahl, P., Goksel, O.: Learn the new, keep the old: Extending pretrained models with new anatomy and images. In: Frangi, A., Schnabel, J., Davatzikos, C., Alberola-López, C., Fichtinger, G. (eds.) International Conference on Medical Image Computing and Computer-Assisted Intervention. MICCAI 2018, LNIP, vol. 11073. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-00937-3_42
Zhang, J.,Wang, Y.: Continually modeling Alzheimer’s disease progression via deep multi- order preserving weight consolidation. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. LNIP, vol. 11765, pp. 850–859. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-32245-8_94
Kim, H.E., Kim, S., Lee, J.: Keep and learn: continual learning by constraining the latent space for knowledge preservation in neural networks. In: International Conference on Medical Image Computing and Computer-Assisted Intervention, LNCS. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-00928-1_59
Patra, A., Noble, J.A.: Hierarchical class incremental learning of anatomical structures in fetal echocardiography videos. IEEE J. Biomed. Health Inf. 24 (2020)
Cubuk, E. D., Zoph, B., Mane, D., Vasudevan, V. and Le, Q. V., Autoaugment: Learning augmentation strategies from data. In: Proceedings of the IEEE Conference on Computer ViSion and Pattern Recognition, pp. 113–123 (2019)
Ahlfors, L., Möbius Transformations in Several Dimensions. University of Minnesota, 1989
Özdemir, N., Iskender, B.B., Özgür, N.Y.: Complex valued neural network with Mö-bius activation function. Commun. Nonlinear Sci. Numer. Simul. 16(12), 4698–4703 (2011)
Zammit-Mangion, A., Ng, T.L.J., Vu, Q., Filippone, M.: Deep compositional spatial models. arXiv preprint arXiv:1906.02840 (2019)
Zhou, S., Zhang, J., Jiang, H., Lundh, T., Ng, A.Y.: Data augmentation with Mobius transformations. arXiv preprint arXiv:2002.02917 (2020)
Ganea, O., Bécigneul, G., Hofmann, T.: Hyperbolic neural networks. In: Advances in Neural Information Processing Systems, pp. 5345–5355 (2018)
Islam, M.A., Anderson, D.T., Pinar, A., Havens, T.C., Scott, G., Keller, J.M.: Enabling explainable fusion in deep learning with fuzzy integral neural networks. IEEE Trans. Fuzzy Syst 28 (2019)
Kather, J., Weis, C., Bianconi, F., Melchers, S.M., Schad, L.R., Gaiser, T., Marx, A., Zöllner, F.G.: Multi-class texture analysis in colorectal cancer histology.’ In Scientific re-ports (2016)
Liouville, J.: Extension au cas des trois dimensions de la question du tracé géographique. Note VI, pp. 609–617 (1850)
Kingma, D.P., Adam, J.B.: A method for stochastic optimization. arXiv:1412.6980
DeVries, T., Taylor, G.W.: Improved regularization of convolutional neural networks with cutout. arXiv preprint arXiv:1708.04552 (2017)
Tang, Z., Peng, X., Li, T., Zhu, Y., Metaxas, D.N.: Adatransform: adaptive data transfor-mation. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 2998–3006 (2019)
Ho, D., Liang, E., Stoica, I., Abbeel, P., Chen, X.: Population based augmentation: efficient learning of augmentation policy schedules. arXiv preprint arXiv:1905.05393 (2019)
Cubuk, E.D., Zoph, B., Shlens, J., Le, Q.V.: Randaugment: practical data augmentation with no separate search. arXiv preprint arXiv:1909.13719 (2019)
Omar, H.A., Patra, A., Domingos, J.S., Leeson, P., Noblel, A.J.: Automated myocardial wall motion classification using handcrafted features vs a deep cnn-based mapping. In: 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), pp. 3140–3143. IEEE (2018)
Lee, C.S., Lee, A.Y.: Clinical applications of continual learning in machine learning. In: The Lancet Digital Health 2.6, pp. 279–e281 (2020)
Patra, A., Noble, J.A.: Sequential anatomy localization in fetal echocardiography videos. arXiv preprint arXiv:1810.11868 (2019)
Patra, A., Noble, J.A.: Incremental learning of fetal heart anatomies using interpretable saliency maps. In: Zheng, Y., Williams, B., Chen, K. (eds.) Medical Image Understanding and Analysis. MIUA 2019. Communications in Computer and Information Science, CCIS. vol. 1065, pp. 129–141. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-39343-4_11
Patra, A., Huang, W., Noble, J.A.: Learning spatio-temporal aggregation for fetal heart analysis in ultrasound video. In: Cardoso, M.J., et al. (eds.) DLMIA/ML-CDS -2017. LNCS, vol. 10553, pp. 276–284. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-67558-9_32
Chakraborti, T., Patra, A., Noble, J.A.: Contrastive fairness in machine learning. IEEE Lett. Comput. Soc. 3(2), 38–41 (2020)
Omar, H.A., Domingos, J.S., Patra, A., Leeson, P., Noble, J.A.: Improving visual detection of wall motion abnormality with echocardiographic image enhancing methods. In: 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), pp. 1128–1131. IEEE (2018)
Food, Drug Administration, et al.: Proposed regulatory framework for modifications to arti-ficial intelligence/machine learning (AI/ML)-based software as a medical device (SaMD) - discussion paper (2019)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Patra, A. (2022). Real Time Data Augmentation Using Fractional Linear Transformations in Continual Learning. In: Zamzmi, G., Antani, S., Bagci, U., Linguraru, M.G., Rajaraman, S., Xue, Z. (eds) Medical Image Learning with Limited and Noisy Data. MILLanD 2022. Lecture Notes in Computer Science, vol 13559. Springer, Cham. https://doi.org/10.1007/978-3-031-16760-7_13
Download citation
DOI: https://doi.org/10.1007/978-3-031-16760-7_13
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-16759-1
Online ISBN: 978-3-031-16760-7
eBook Packages: Computer ScienceComputer Science (R0)