Benchmarking Embedding Aggregation Methods in Computational Pathology: A Clinical Data Perspective

Shengjia Chen, Gabriele Campanella, Abdulkadir Elmas, Aryeh Stock, Jennifer Zeng, Alexandros D. Polydorides, Adam J. Schoenfeld, Kuan-lin Huang, Jane Houldsworth, Chad Vanderbilt, Thomas J. Fuchs
Proceedings of the MICCAI Workshop on Computational Pathology, PMLR 254:38-50, 2024.

Abstract

Recent advances in artificial intelligence (AI), in particular self-supervised learning of foundation models (FMs), are revolutionizing medical imaging and computational pathology (CPath). A constant challenge in the analysis of digital Whole Slide Images (WSIs) is the problem of aggregating tens of thousands of tile-level image embeddings to a slide-level representation. Due to the prevalent use of datasets created for genomic research, such as TCGA, for method development, the performance of these techniques on diagnostic slides from clinical practice has been inadequately explored. This study conducts a thorough benchmarking analysis of ten slide-level aggregation techniques across nine clinically relevant tasks, including diagnostic assessment, biomarker classification, and outcome prediction. The results yield following key insights: (1) Embeddings derived from domainspecific (histological images) FMs outperform those from generic ImageNet-based models across aggregation methods. (2) Spatial-aware aggregators enhance the performance significantly when using ImageNet pre-trained models but not when using FMs. (3) No single model excels in all tasks and spatially-aware models do not show general superiority as it would be expected. These findings underscore the need for more adaptable and universally applicable aggregation techniques, guiding future research towards tools that better meet the evolving needs of clinical-AI in pathology. The code used in this work are available at https://github.com/fuchs-lab-public/CPath_SABenchmark

Cite this Paper


BibTeX
@InProceedings{pmlr-v254-chen24a, title = {Benchmarking Embedding Aggregation Methods in Computational Pathology: A Clinical Data Perspective}, author = {Chen, Shengjia and Campanella, Gabriele and Elmas, Abdulkadir and Stock, Aryeh and Zeng, Jennifer and Polydorides, Alexandros D. and Schoenfeld, Adam J. and Huang, Kuan-lin and Houldsworth, Jane and Vanderbilt, Chad and Fuchs, Thomas J.}, booktitle = {Proceedings of the MICCAI Workshop on Computational Pathology}, pages = {38--50}, year = {2024}, editor = {Ciompi, Francesco and Khalili, Nadieh and Studer, Linda and Poceviciute, Milda and Khan, Amjad and Veta, Mitko and Jiao, Yiping and Haj-Hosseini, Neda and Chen, Hao and Raza, Shan and Minhas, FayyazZlobec, Inti and Burlutskiy, Nikolay and Vilaplana, Veronica and Brattoli, Biagio and Muller, Henning and Atzori, Manfredo and Raza, Shan and Minhas, Fayyaz}, volume = {254}, series = {Proceedings of Machine Learning Research}, month = {06 Oct}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v254/main/assets/chen24a/chen24a.pdf}, url = {https://proceedings.mlr.press/v254/chen24a.html}, abstract = {Recent advances in artificial intelligence (AI), in particular self-supervised learning of foundation models (FMs), are revolutionizing medical imaging and computational pathology (CPath). A constant challenge in the analysis of digital Whole Slide Images (WSIs) is the problem of aggregating tens of thousands of tile-level image embeddings to a slide-level representation. Due to the prevalent use of datasets created for genomic research, such as TCGA, for method development, the performance of these techniques on diagnostic slides from clinical practice has been inadequately explored. This study conducts a thorough benchmarking analysis of ten slide-level aggregation techniques across nine clinically relevant tasks, including diagnostic assessment, biomarker classification, and outcome prediction. The results yield following key insights: (1) Embeddings derived from domainspecific (histological images) FMs outperform those from generic ImageNet-based models across aggregation methods. (2) Spatial-aware aggregators enhance the performance significantly when using ImageNet pre-trained models but not when using FMs. (3) No single model excels in all tasks and spatially-aware models do not show general superiority as it would be expected. These findings underscore the need for more adaptable and universally applicable aggregation techniques, guiding future research towards tools that better meet the evolving needs of clinical-AI in pathology. The code used in this work are available at https://github.com/fuchs-lab-public/CPath_SABenchmark} }
Endnote
%0 Conference Paper %T Benchmarking Embedding Aggregation Methods in Computational Pathology: A Clinical Data Perspective %A Shengjia Chen %A Gabriele Campanella %A Abdulkadir Elmas %A Aryeh Stock %A Jennifer Zeng %A Alexandros D. Polydorides %A Adam J. Schoenfeld %A Kuan-lin Huang %A Jane Houldsworth %A Chad Vanderbilt %A Thomas J. Fuchs %B Proceedings of the MICCAI Workshop on Computational Pathology %C Proceedings of Machine Learning Research %D 2024 %E Francesco Ciompi %E Nadieh Khalili %E Linda Studer %E Milda Poceviciute %E Amjad Khan %E Mitko Veta %E Yiping Jiao %E Neda Haj-Hosseini %E Hao Chen %E Shan Raza %E Fayyaz MinhasInti Zlobec %E Nikolay Burlutskiy %E Veronica Vilaplana %E Biagio Brattoli %E Henning Muller %E Manfredo Atzori %E Shan Raza %E Fayyaz Minhas %F pmlr-v254-chen24a %I PMLR %P 38--50 %U https://proceedings.mlr.press/v254/chen24a.html %V 254 %X Recent advances in artificial intelligence (AI), in particular self-supervised learning of foundation models (FMs), are revolutionizing medical imaging and computational pathology (CPath). A constant challenge in the analysis of digital Whole Slide Images (WSIs) is the problem of aggregating tens of thousands of tile-level image embeddings to a slide-level representation. Due to the prevalent use of datasets created for genomic research, such as TCGA, for method development, the performance of these techniques on diagnostic slides from clinical practice has been inadequately explored. This study conducts a thorough benchmarking analysis of ten slide-level aggregation techniques across nine clinically relevant tasks, including diagnostic assessment, biomarker classification, and outcome prediction. The results yield following key insights: (1) Embeddings derived from domainspecific (histological images) FMs outperform those from generic ImageNet-based models across aggregation methods. (2) Spatial-aware aggregators enhance the performance significantly when using ImageNet pre-trained models but not when using FMs. (3) No single model excels in all tasks and spatially-aware models do not show general superiority as it would be expected. These findings underscore the need for more adaptable and universally applicable aggregation techniques, guiding future research towards tools that better meet the evolving needs of clinical-AI in pathology. The code used in this work are available at https://github.com/fuchs-lab-public/CPath_SABenchmark
APA
Chen, S., Campanella, G., Elmas, A., Stock, A., Zeng, J., Polydorides, A.D., Schoenfeld, A.J., Huang, K., Houldsworth, J., Vanderbilt, C. & Fuchs, T.J.. (2024). Benchmarking Embedding Aggregation Methods in Computational Pathology: A Clinical Data Perspective. Proceedings of the MICCAI Workshop on Computational Pathology, in Proceedings of Machine Learning Research 254:38-50 Available from https://proceedings.mlr.press/v254/chen24a.html.

Related Material