Abstract
As deep learning classifiers become ever more widely deployed for medical image analysis tasks, issues of predictive calibration need to be addressed. Mis-calibration is the deviation between predictive probability (confidence) and classification correctness. Well-calibrated classifiers enable cost-sensitive and selective decision-making. This paper presents an empirical investigation of calibration methods on two medical image datasets (multi-class dermatology and binary histopathology image classification). We show the effect of temperature scaling with temperature optimized using various measures of calibration replacing the standard negative log-likelihood. We do so not only for networks trained using one-hot encoding and cross-entropy loss, but also using focal loss and label smoothing. We compare these with two Bayesian methods. Results suggest little or no advantage to the use of alternative calibration metrics for tuning temperature. Temperature scaling of networks trained using focal loss (with appropriate hyperparameters) provided strong results in terms of both calibration and accuracy across both datasets.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
GitHub Repository: https://github.com/UoD-CVIP/Medical_Calibration.
References
Bejnordi, B.E., et al.: Diagnostic assessment of deep learning algorithms for detection of lymph node metastases in women with breast cancer. J. Am. Med. Assoc. 318(22), 2199–2210 (2017)
Blundell, C., Cornebise, J., Kavukcuoglu, K., Wierstra, D.: Weight uncertainty in neural networks. In: Proceedings of the 32nd International Conference on Machine Learning, pp. 1613–1622. PMLR (2015)
Botev, A., Ritter, H., Barber, D.: Practical Gauss-Newton optimisation for deep learning. In: Proceedings of the 34th International Conference on Machine Learning, pp. 557–565. PMLR (2017)
Carse, J., McKenna, S.: Active learning for patch-based digital pathology using convolutional neural networks to reduce annotation costs. In: Reyes-Aldasoro, C.C., Janowczyk, A., Veta, M., Bankhead, P., Sirinukunwattana, K. (eds.) ECDP 2019. LNCS, vol. 11435, pp. 20–27. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-23937-4_3
Carse, J., et al.: Robust selective classification of skin lesions with asymmetric costs. In: Sudre, C.H., Licandro, R., Baumgartner, C., Melbourne, A., Dalca, A., Hutter, J., Tanno, R., Abaci Turk, E., Van Leemput, K., Torrents Barrena, J., Wells, W.M., Macgowan, C. (eds.) UNSURE/PIPPI -2021. LNCS, vol. 12959, pp. 112–121. Springer, Cham (2021). https://doi.org/10.1007/978-3-030-87735-4_11
Codella, N.C., et al.: Skin lesion analysis toward melanoma detection: a challenge at the 2017 international symposium on biomedical imaging (ISBI), hosted by the international skin imaging collaboration (ISIC). In: 2018 IEEE 15th International Symposium on Biomedical Imaging, pp. 168–172. IEEE (2018)
Combalia, M., et al.: BCN20000: dermoscopic lesions in the wild. arXiv preprint arXiv:1908.02288 (2019)
Dai, Z., Low, B.K.H., Jaillet, P.: Federated Bayesian optimization via Thompson sampling. Adv. Neural Inf. Process. Syst. 33, 9687–9699 (2020)
Daxberger, E., Kristiadi, A., Immer, A., Eschenhagen, R., Bauer, M., Hennig, P.: Laplace redux-effortless Bayesian deep learning. Adv. Neural Inf. Process. Syst. 34, 20089–20103 (2021)
Deng, J., Dong, W., Socher, R., Li, L.J., Li, K., Fei-Fei, L.: ImageNet: a large-scale hierarchical image database. In: 2009 IEEE Conference on Computer Vision and Pattern Recognition, pp. 248–255. IEEE (2009)
Frenkel, L., Goldberger, J.: Network calibration by class-based temperature scaling. In: 2021 29th European Signal Processing Conference, pp. 1486–1490. IEEE (2021)
Gawlikowski, J., et al.: A survey of uncertainty in deep neural networks. arXiv preprint arXiv:2107.03342 (2021)
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, pp. 1321–1330. PMLR (2017)
Hendrycks, D., Mazeika, M., Dietterich, T.: Deep anomaly detection with outlier exposure. arXiv preprint arXiv:1812.04606 (2018)
Hendrycks, D., Mu, N., Cubuk, E.D., Zoph, B., Gilmer, J., Lakshminarayanan, B.: AugMix: a simple data processing method to improve robustness and uncertainty. arXiv preprint arXiv:1912.02781 (2019)
Islam, M., Glocker, B.: Spatially varying label smoothing: capturing uncertainty from expert annotations. In: Feragen, A., Sommer, S., Schnabel, J., Nielsen, M. (eds.) IPMI 2021. LNCS, vol. 12729, pp. 677–688. Springer, Cham (2021). https://doi.org/10.1007/978-3-030-78191-0_52
Kwon, Y., Won, J.H., Kim, B.J., Paik, M.C.: Uncertainty quantification using Bayesian neural networks in classification: application to biomedical image segmentation. Comput. Stat. Data Anal. 142, 106816 (2020)
Liang, G., Zhang, Y., Wang, X., Jacobs, N.: Improved trainable calibration method for neural networks on medical imaging classification. arXiv preprint arXiv:2009.04057 (2020)
Lin, T.Y., Goyal, P., Girshick, R., He, K., Dollár, P.: Focal loss for dense object detection. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 2980–2988 (2017)
Liu, D.C., Nocedal, J.: On the limited memory BFGS method for large scale optimization. Math. Program. 45, 503–528 (1989)
MacKay, D.J.: Bayesian interpolation. Neural Comput. 4(3), 415–447 (1992)
Maron, R.C., et al.: Systematic outperformance of 112 dermatologists in multiclass skin cancer image classification by convolutional neural networks. Eur. J. Cancer 119, 57–65 (2019)
Mukhoti, J., Kulharia, V., Sanyal, A., Golodetz, S., Torr, P., Dokania, P.: Calibrating deep neural networks using focal loss. Adv. Neural Inf. Process. Syst. 33, 15288–15299 (2020)
Müller, R., Kornblith, S., Hinton, G.E.: When does label smoothing help? In: Advances in Neural Information Processing Systems, vol. 32 (2019)
Parzen, E.: On estimation of a probability density function and mode. Ann. Math. Stat. 33(3), 1065–1076 (1962)
Platt, J., et al.: Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods. Adv. Large Margin Classifiers 10(3), 61–74 (1999)
Roelofs, R., Cain, N., Shlens, J., Mozer, M.C.: Mitigating bias in calibration error estimation. In: International Conference on Artificial Intelligence and Statistics, pp. 4036–4054. PMLR (2022)
Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., Wojna, Z.: Rethinking the inception architecture for computer vision. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 2818–2826 (2016)
Tan, M., Le, Q.: EfficientNet: rethinking model scaling for convolutional neural networks. In: Proceedings of the 36th International Conference on Machine Learning International Conference on Machine Learning, pp. 6105–6114. PMLR (2019)
Tschandl, P., Rosendahl, C., Kittler, H.: The HAM10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions. Sci. Data 5(1), 1–9 (2018)
Ulmer, D., Meijerink, L., Cinà , G.: Trust issues: uncertainty estimation does not enable reliable OOD detection on medical tabular data. In: Machine Learning for Health, pp. 341–354. PMLR (2020)
Veeling, B.S., Linmans, J., Winkens, J., Cohen, T., Welling, M.: Rotation equivariant CNNs for digital pathology. In: Frangi, A.F., Schnabel, J.A., Davatzikos, C., Alberola-López, C., Fichtinger, G. (eds.) MICCAI 2018. LNCS, vol. 11071, pp. 210–218. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-00934-2_24
Zhang, J., Kailkhura, B., Han, T.Y.J.: Mix-n-Match: ensemble and compositional methods for uncertainty calibration in deep learning. In: Proceedings of the 37th International Conference on Machine Learning, pp. 11117–11128. PMLR (2020)
Acknowledgments
J. Carse was supported by the UK Engineering and Physical Sciences Research Council (EPSRC Training Grant EP/N509632/1). This paper also reports independent research partly funded by the National Institute for Health Research (Artificial Intelligence, Deep learning for effective triaging of skin disease in the NHS, AI AWARD01901) and NHSX. The views expressed in this publication are those of the authors and not necessarily those of the National Institute for Health Research, NHSX or the Department of Health and Social Care.
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
Carse, J., Alvarez Olmo, A., McKenna, S. (2022). Calibration of Deep Medical Image Classifiers: An Empirical Comparison Using Dermatology and Histopathology Datasets. In: Sudre, C.H., et al. Uncertainty for Safe Utilization of Machine Learning in Medical Imaging. UNSURE 2022. Lecture Notes in Computer Science, vol 13563. Springer, Cham. https://doi.org/10.1007/978-3-031-16749-2_9
Download citation
DOI: https://doi.org/10.1007/978-3-031-16749-2_9
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-16748-5
Online ISBN: 978-3-031-16749-2
eBook Packages: Computer ScienceComputer Science (R0)