Out-of-distribution generalization for segmentation of lymph node metastasis in breast cancer
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 Metastas...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-01-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-024-80495-y |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841544843758665728 |
|---|---|
| author | Yiannis Varnava Kiran Jakate Richard Garnett Dimitrios Androutsos Pascal N. Tyrrell April Khademi |
| author_facet | Yiannis Varnava Kiran Jakate Richard Garnett Dimitrios Androutsos Pascal N. Tyrrell April Khademi |
| author_sort | Yiannis Varnava |
| collection | DOAJ |
| description | 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 . |
| format | Article |
| id | doaj-art-c04e1d54c78c4ee98e91356c63bf2fc9 |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-c04e1d54c78c4ee98e91356c63bf2fc92025-01-12T12:15:16ZengNature PortfolioScientific Reports2045-23222025-01-0115111610.1038/s41598-024-80495-yOut-of-distribution generalization for segmentation of lymph node metastasis in breast cancerYiannis Varnava0Kiran Jakate1Richard Garnett2Dimitrios Androutsos3Pascal N. Tyrrell4April Khademi5Department of Electrical, Computer, and Biomedical Engineering, Toronto Metropolitan UniversityDepartment of Pathology, Unity Health TorontoDepartment of Electrical, Computer, and Biomedical Engineering, Toronto Metropolitan UniversityDepartment of Electrical, Computer, and Biomedical Engineering, Toronto Metropolitan UniversityDepartment of Medical Imaging, University of TorontoDepartment of Electrical, Computer, and Biomedical Engineering, Toronto Metropolitan UniversityAbstract 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 .https://doi.org/10.1038/s41598-024-80495-yHistopathologyLymph nodeBreast cancerDeep learningSegmentationGeneralization |
| spellingShingle | Yiannis Varnava Kiran Jakate Richard Garnett Dimitrios Androutsos Pascal N. Tyrrell April Khademi Out-of-distribution generalization for segmentation of lymph node metastasis in breast cancer Scientific Reports Histopathology Lymph node Breast cancer Deep learning Segmentation Generalization |
| title | Out-of-distribution generalization for segmentation of lymph node metastasis in breast cancer |
| title_full | Out-of-distribution generalization for segmentation of lymph node metastasis in breast cancer |
| title_fullStr | Out-of-distribution generalization for segmentation of lymph node metastasis in breast cancer |
| title_full_unstemmed | Out-of-distribution generalization for segmentation of lymph node metastasis in breast cancer |
| title_short | Out-of-distribution generalization for segmentation of lymph node metastasis in breast cancer |
| title_sort | out of distribution generalization for segmentation of lymph node metastasis in breast cancer |
| topic | Histopathology Lymph node Breast cancer Deep learning Segmentation Generalization |
| url | https://doi.org/10.1038/s41598-024-80495-y |
| work_keys_str_mv | AT yiannisvarnava outofdistributiongeneralizationforsegmentationoflymphnodemetastasisinbreastcancer AT kiranjakate outofdistributiongeneralizationforsegmentationoflymphnodemetastasisinbreastcancer AT richardgarnett outofdistributiongeneralizationforsegmentationoflymphnodemetastasisinbreastcancer AT dimitriosandroutsos outofdistributiongeneralizationforsegmentationoflymphnodemetastasisinbreastcancer AT pascalntyrrell outofdistributiongeneralizationforsegmentationoflymphnodemetastasisinbreastcancer AT aprilkhademi outofdistributiongeneralizationforsegmentationoflymphnodemetastasisinbreastcancer |