Abstract
Fibrosis gradings are a valuable indicator to provide diagnostic information for chronic kidney disease. The assessment of the percentage of renal fibrosis by physicians is based mainly on visual estimation, which is highly subjective and varies widely between physicians, hence the need for objective and reliable morphological assessment algorithms. For ultra-high resolution images, acquiring patch-level labels imposes a heavy annotation burden; conversely, indiscriminately assigning WSI-level labels to each patch poses a significant label noise problem. In this paper, we propose a weakly supervised two-stage framework. In the first stage (Patches Selection Stage), patches with low uncertainty, i.e., strongly correlated with WSI labels, are screened using approximate Bayesian inference. In the subsequent second stage (Decision Aggregation Stage), low uncertainty patches are merged into a large map and fed into a classification network to obtain WSI-level diagnostic results. The uncertainty estimation efficiently targets local regions of interest in high-resolution pathology slice images and excludes noise unrelated to WSI labels. We compared this method with previous methods for grading renal fibrosis in a self-constructed dataset of renal pathology provided by the Institute of Nephrology, Southeast University. We used the quadratic weighted kappa coefficient as a grading consistency evaluation index. The results show that this method is superior in accuracy and kappa consistency.
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
Chen, R.J., et al.: Scaling vision transformers to gigapixel images via hierarchical self-supervised learning. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 16144–16155 (2022)
Chen, W., Jiang, Z., Wang, Z., Cui, K., Qian, X.: Collaborative global-local networks for memory-efficient segmentation of ultra-high resolution images. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp. 8924–8933 (2019)
Cheng, J., Wang, Z., Pollastri, G.: A neural network approach to ordinal regression. In: 2008 IEEE International Joint Conference on Neural Networks (IEEE world congress on computational intelligence), pp. 1279–1284. IEEE (2008)
Courtiol, P., et al.: Deep learning-based classification of mesothelioma improves prediction of patient outcome. Nat. Med. 25(10), 1519–1525 (2019)
Farris, A.B., et al.: Morphometric and visual evaluation of fibrosis in renal biopsies. J. Am. Soc. Nephrol. 22(1), 176–186 (2011)
Farris, A.B., et al.: Banff digital pathology working group: going digital in transplant pathology. Am. J. Transplant. 20(9), 2392–2399 (2020)
Farris, A.B., Vizcarra, J., Amgad, M., Cooper, L.A.D., Gutman, D., Hogan, J.: Image analysis pipeline for renal allograft evaluation and fibrosis quantification. Kidney Int. Reports 6(7), 1878–1887 (2021)
Gal, Y., Ghahramani, Z.: Dropout as a Bayesian approximation: representing model uncertainty in deep learning. In: International Conference on Machine Learning, pp. 1050–1059. PMLR (2016)
Ginley, B., et al.: Automated computational detection of interstitial fibrosis, tubular atrophy, and glomerulosclerosis. J. Am. Soc. Nephrol. 32(4), 837–850 (2021)
Hou, L., Samaras, D., Kurc, T.M., Gao, Y., Davis, J.E., Saltz, J.H.: Patch-based convolutional neural network for whole slide tissue image classification. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 2424–2433 (2016)
Kendall, A., Gal, Y.: What uncertainties do we need in Bayesian deep learning for computer vision? In: Advances in Neural Information Processing Systems 30 (2017)
Lu, M.Y., Williamson, D.F., Chen, T.Y., Chen, R.J., Barbieri, M., Mahmood, F.: Data-efficient and weakly supervised computational pathology on whole-slide images. Nature Biomed. Eng. 5(6), 555–570 (2021)
Mehta, S., et al.: End-to-end diagnosis of breast biopsy images with transformers. Med. Image Anal. 79, 102466 (2022)
Tan, M., Le, Q.: Efficientnet: rethinking model scaling for convolutional neural networks. In: International Conference on Machine Learning, pp. 6105–6114. PMLR (2019)
Yu, L., Wang, S., Li, X., Fu, C.-W., Heng, P.-A.: Uncertainty-aware self-ensembling model for semi-supervised 3D left atrium segmentation. In: Shen, D., et al. (eds.) MICCAI 2019. LNCS, vol. 11765, pp. 605–613. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-32245-8_67
Zheng, Y., et al.: Deep-learning-driven quantification of interstitial fibrosis in digitized kidney biopsies. Am. J. Pathol. 191(8), 1442–1453 (2021)
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
Tang, K., Hu, X., Chen, P., Xia, S. (2023). Fibrosis Grading Methods for Renal Whole Slide Images Based on Uncertainty Estimation. In: Lu, H., Blumenstein, M., Cho, SB., Liu, CL., Yagi, Y., Kamiya, T. (eds) Pattern Recognition. ACPR 2023. Lecture Notes in Computer Science, vol 14407. Springer, Cham. https://doi.org/10.1007/978-3-031-47637-2_30
Download citation
DOI: https://doi.org/10.1007/978-3-031-47637-2_30
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-47636-5
Online ISBN: 978-3-031-47637-2
eBook Packages: Computer ScienceComputer Science (R0)