Out-of-distribution Generalization for Segmentation of Lymph Node Metastasis in Breast Cancer

Overview

Journal Sci Rep

Specialty Science

Date 2025 Jan 8

PMID 39775089

Authors

Yiannis Varnava

Kiran Jakate

Richard Garnett

Dimitrios Androutsos

Pascal N Tyrrell

April Khademi

Affiliations

Soon will be listed here.

Abstract

Pathology provides the definitive diagnosis, and Artificial Intelligence (AI) tools are poised to improve accuracy, inter-rater agreement, and turn-around time (TAT) of pathologists, leading to improved quality of care. A high value clinical application is the grading of Lymph Node Metastasis (LNM) which is used for breast cancer staging and guides treatment decisions. A challenge of implementing AI tools widely for LNM classification is domain shift, where Out-of-Distribution (OOD) data has a different distribution than the In-Distribution (ID) data used to train the model, resulting in a drop in performance in OOD data. This work proposes a novel clustering and sampling method to automatically curate training datasets in an unsupervised manner with the aim of improving model generalization abilities. To evaluate the generalization performance of the proposed models, we applied a novel use of the Two One-sided Tests (TOST) method. This method examines whether the performance on ID and OOD data is equivalent, serving as a proxy for generalization. We provide the first evidence for computing equivalence margins that are data-dependent, which reduces subjectivity. The proposed framework shows the ensembled models constructed from models that generalized across both tumor and normal patches enhanced performance, achieving an F1 score of 0.81 for LNM classification on unseen ID and OOD samples. Interactive viewing of slide-level segmentations can be accessed on PathcoreFlow through https://web.pathcore.com/folder/18555?s=QTJVHJuhrfe5 . Segmentation models are available at https://github.com/IAMLAB-Ryerson/OOD-Generalization-LNM .

References

Davies D, Bouldin D . A cluster separation measure. IEEE Trans Pattern Anal Mach Intell. 2011; 1(2):224-7. View

Razavi S, Khameneh F, Nouri H, Androutsos D, Done S, Khademi A . MiNuGAN: Dual Segmentation of Mitoses and Nuclei Using Conditional GANs on Multi-center Breast H&E Images. J Pathol Inform. 2022; 13:100002. PMC: 8860738. DOI: 10.1016/j.jpi.2022.100002. View

Jose L, Liu S, Russo C, Nadort A, Ieva A . Generative Adversarial Networks in Digital Pathology and Histopathological Image Processing: A Review. J Pathol Inform. 2021; 12:43. PMC: 8609288. DOI: 10.4103/jpi.jpi_103_20. View

Bandi P, Balkenhol M, van Ginneken B, van der Laak J, Litjens G . Resolution-agnostic tissue segmentation in whole-slide histopathology images with convolutional neural networks. PeerJ. 2019; 7:e8242. PMC: 6924324. DOI: 10.7717/peerj.8242. View

Deng S, Zhang X, Yan W, Chang E, Fan Y, Lai M . Deep learning in digital pathology image analysis: a survey. Front Med. 2020; 14(4):470-487. DOI: 10.1007/s11684-020-0782-9. View

Metter D, Colgan T, Leung S, Timmons C, Park J . Trends in the US and Canadian Pathologist Workforces From 2007 to 2017. JAMA Netw Open. 2019; 2(5):e194337. PMC: 6547243. DOI: 10.1001/jamanetworkopen.2019.4337. View

Ho D, Yarlagadda D, DAlfonso T, Hanna M, Grabenstetter A, Ntiamoah P . Deep Multi-Magnification Networks for multi-class breast cancer image segmentation. Comput Med Imaging Graph. 2021; 88:101866. PMC: 7975990. DOI: 10.1016/j.compmedimag.2021.101866. View

Kleppe A, Skrede O, de Raedt S, Liestol K, Kerr D, Danielsen H . Designing deep learning studies in cancer diagnostics. Nat Rev Cancer. 2021; 21(3):199-211. DOI: 10.1038/s41568-020-00327-9. View

Houvenaeghel G, de Nonneville A, Cohen M, Chopin N, Coutant C, Reyal F . Lack of prognostic impact of sentinel node micro-metastases in endocrine receptor-positive early breast cancer: results from a large multicenter cohort. ESMO Open. 2021; 6(3):100151. PMC: 8314870. DOI: 10.1016/j.esmoop.2021.100151. View

10.

Ding Y, Yang F, Han M, Li C, Wang Y, Xu X . Multi-center study on predicting breast cancer lymph node status from core needle biopsy specimens using multi-modal and multi-instance deep learning. NPJ Breast Cancer. 2023; 9(1):58. PMC: 10345095. DOI: 10.1038/s41523-023-00562-x. View

11.

Litjens G, Bandi P, Bejnordi B, Geessink O, Balkenhol M, Bult P . 1399 H&E-stained sentinel lymph node sections of breast cancer patients: the CAMELYON dataset. Gigascience. 2018; 7(6). PMC: 6007545. DOI: 10.1093/gigascience/giy065. View

12.

Das R, Bose S, Chowdhury R, Maulik U . Dense Dilated Multi-Scale Supervised Attention-Guided Network for histopathology image segmentation. Comput Biol Med. 2023; 163:107182. DOI: 10.1016/j.compbiomed.2023.107182. View

13.

Ruifrok A, Johnston D . Quantification of histochemical staining by color deconvolution. Anal Quant Cytol Histol. 2001; 23(4):291-9. View

14.

Bandi P, Geessink O, Manson Q, van Dijk M, Balkenhol M, Hermsen M . From Detection of Individual Metastases to Classification of Lymph Node Status at the Patient Level: The CAMELYON17 Challenge. IEEE Trans Med Imaging. 2019; 38(2):550-560. DOI: 10.1109/TMI.2018.2867350. View

15.

Budginaite E, Magee D, Kloft M, Woodruff H, Grabsch H . Computational methods for metastasis detection in lymph nodes and characterization of the metastasis-free lymph node microarchitecture: A systematic-narrative hybrid review. J Pathol Inform. 2024; 15:100367. PMC: 10918266. DOI: 10.1016/j.jpi.2024.100367. View

16.

Khened M, Kori A, Rajkumar H, Krishnamurthi G, Srinivasan B . A generalized deep learning framework for whole-slide image segmentation and analysis. Sci Rep. 2021; 11(1):11579. PMC: 8172839. DOI: 10.1038/s41598-021-90444-8. View

17.

Pontalba J, Gwynne-Timothy T, David E, Jakate K, Androutsos D, Khademi A . Assessing the Impact of Color Normalization in Convolutional Neural Network-Based Nuclei Segmentation Frameworks. Front Bioeng Biotechnol. 2019; 7:300. PMC: 6838039. DOI: 10.3389/fbioe.2019.00300. View

18.

Khan A, Janowczyk A, Muller F, Blank A, Nguyen H, Abbet C . Impact of scanner variability on lymph node segmentation in computational pathology. J Pathol Inform. 2022; 13:100127. PMC: 9577043. DOI: 10.1016/j.jpi.2022.100127. View

19.

Jarkman S, Karlberg M, Poceviciute M, Boden A, Bandi P, Litjens G . Generalization of Deep Learning in Digital Pathology: Experience in Breast Cancer Metastasis Detection. Cancers (Basel). 2022; 14(21). PMC: 9659028. DOI: 10.3390/cancers14215424. View

20.

Sammouda R, El-Zaart A . An Optimized Approach for Prostate Image Segmentation Using K-Means Clustering Algorithm with Elbow Method. Comput Intell Neurosci. 2021; 2021:4553832. PMC: 8608531. DOI: 10.1155/2021/4553832. View