Abstract
Deep learning-based medical image segmentation is widely used and has achieved the state-of-the-art segmentation performance, in which nnU-Net is a particularly successful pipeline due to its pre-processing and auto-configuration features. However, the output predicted probabilities from neural networks are generally not properly calibrated and don’t necessarily indicate segmentation errors, which are problematic for clinical use. Bayesian deep learning is a promising way to address these problems by improving the probability calibration and error localisation ability. In this paper, we proposed a novel Bayesian approach based on posterior bootstrap theory to sample the neural network parameters from a posterior distribution. Based on nnU-Net, we implemented our method and other Bayesian approaches, and evaluated their uncertainty estimation quality. The results show that the proposed posterior bootstrap method provides improvement on uncertainty estimation with equivalent segmentation performance. The proposed method is easy to implement, compatible with any deep learning-based image segmentation pipeline, and doesn’t require additional hyper-parameter tuning, enabling it to totally preserve nnU-Net’s auto-configuration feature.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Antonelli, M., et al.: The medical segmentation decathlon. Nat. Commun. 13(1), 4128 (2022)
Bakas, S., et al.: Identifying the best machine learning algorithms for brain tumor segmentation, progression assessment, and overall survival prediction in the brats challenge. arXiv preprint arXiv:1811.02629 (2018)
Blei, D.M., Kucukelbir, A., McAuliffe, J.D.: Variational inference: a review for statisticians. J. Am. Stat. Assoc. 112(518), 859–877 (2017)
Drozdzal, M., Vorontsov, E., Chartrand, G., Kadoury, S., Pal, C.: The importance of skip connections in biomedical image segmentation. In: Carneiro, G., et al. (eds.) LABELS/DLMIA -2016. LNCS, vol. 10008, pp. 179–187. Springer, Cham (2016). https://doi.org/10.1007/978-3-319-46976-8_19
Filipović, M., Dautremer, T., Comtat, C., Stute, S., Barat, É.: Reconstruction, analysis and interpretation of posterior probability distributions of PET images, using the posterior bootstrap. Phys. Med. Biol. 66(12), 125018 (2021)
Fong, E., Lyddon, S., Holmes, C.: Scalable nonparametric sampling from multimodal posteriors with the posterior bootstrap. In: Chaudhuri, K., Salakhutdinov, R. (eds.) Proceedings of the 36th International Conference on Machine Learning. Proceedings of Machine Learning Research, vol. 97, pp. 1952–1962. PMLR (2019)
Gal, Y., Ghahramani, Z.: Dropout as a Bayesian approximation: representing model uncertainty in deep learning. In: Proceedings of the 33rd International Conference on International Conference on Machine Learning, vol. 48, pp. 1050–1059. ICML’16, JMLR.org (2016)
Guo, C., Pleiss, G., Sun, Y., Weinberger, K.Q.: On calibration of modern neural networks. In: Proceedings of the 34th International Conference on Machine Learning, vol. 70, pp. 1321–1330. ICML’17, JMLR.org (2017)
Isensee, F., Jaeger, P.F., Kohl, S.A.A., Petersen, J., Maier-Hein, K.H.: nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nat. Methods 18(2), 203–211 (2021)
Jungo, A., Balsiger, F., Reyes, M.: Analyzing the quality and challenges of uncertainty estimations for brain tumor segmentation. Front. Neurosci. 14, 282 (2020)
Lakshminarayanan, B., Pritzel, A., Blundell, C.: Simple and scalable predictive uncertainty estimation using deep ensembles. In: Proceedings of the 31st International Conference on Neural Information Processing Systems, pp. 6405–6416. NIPS’17, Curran Associates Inc., Red Hook, NY, USA (2017)
Lyddon, S., Walker, S., Holmes, C.: Nonparametric learning from Bayesian models with randomized objective functions. In: Proceedings of the 32nd International Conference on Neural Information Processing Systems, pp. 2075–2085. NIPS’18, Curran Associates Inc., Red Hook, NY, USA (2018)
Naeini, M.P., Cooper, G.F., Hauskrecht, M.: Obtaining well calibrated probabilities using Bayesian binning. Proc. AAAI Conf. Artif. Intell. 2015, 2901–2907 (2015)
Newton, M.A., Polson, N.G., Xu, J.: Weighted Bayesian bootstrap for scalable posterior distributions. Can. J. Stat. 49(2), 421–437 (2021)
Ovadia, Y., et al.: Can you trust your model’s uncertainty? Evaluating predictive uncertainty under dataset shift. In: Proceedings of the 33rd International Conference on Neural Information Processing Systems, pp. 14003–14014. Curran Associates Inc. (2019)
Ronneberger, O., Fischer, P., Brox, T.: U-Net: convolutional networks for biomedical image segmentation. In: Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F. (eds.) MICCAI 2015. LNCS, vol. 9351, pp. 234–241. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-24574-4_28
Rubin, D.B.: The Bayesian bootstrap. Ann. Stat. 9(1), 130–134 (1981)
Wilson, A.G., Izmailov, P.: Bayesian deep learning and a probabilistic perspective of generalization. In: Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M., Lin, H. (eds.) Advances in Neural Information Processing Systems, vol. 33, pp. 4697–4708. Curran Associates, Inc. (2020)
Zhao, Y., Yang, C., Schweidtmann, A., Tao, Q.: Efficient Bayesian uncertainty estimation for nnU-Net. In: Wang, L., Dou, Q., Fletcher, P.T., Speidel, S., Li, S. (eds.) Medical Image Computing and Computer Assisted Intervention - MICCAI 2022, pp. 535–544. Springer Nature Switzerland, Cham (2022). https://doi.org/10.1007/978-3-031-16452-1_51
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Wang, S., Nuyts, J., Filipovic, M. (2023). Uncertainty Estimation in Liver Tumor Segmentation Using the Posterior Bootstrap. In: Sudre, C.H., Baumgartner, C.F., Dalca, A., Mehta, R., Qin, C., Wells, W.M. (eds) Uncertainty for Safe Utilization of Machine Learning in Medical Imaging. UNSURE 2023. Lecture Notes in Computer Science, vol 14291. Springer, Cham. https://doi.org/10.1007/978-3-031-44336-7_19
Download citation
DOI: https://doi.org/10.1007/978-3-031-44336-7_19
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-44335-0
Online ISBN: 978-3-031-44336-7
eBook Packages: Computer ScienceComputer Science (R0)
