CDNet: Causal Inference inspired Diversified Aggregation Convolution for Pathology Image Segmentation

Dawei Fan, Yifan Gao, Jiaming Yu, Changcai Yang, Riqing Chen, Lifang Wei
Proceedings of the MICCAI Workshop on Computational Pathology, PMLR 254:51-60, 2024.

Abstract

Deep learning models have shown promising performance for Nuclei segmentation in the field of pathology image analysis. However, training a robust model from multiple domains remains a great challenge for Nuclei segmentation. Additionally, the shortcomings of background noise, highly overlapping between Nuclei, and blurred edges often lead to poor performance. To address these challenges, we propose a novel framework termed CDNet, which combines Causal Inference Module (CIM) with Diversified Aggregation Convolution (DAC) techniques. The DAC module is designed which incorporates diverse downsampling features through a simple, parameter-free attention module (SimAM), aiming to overcome the problems of edge blurring. Furthermore, we introduce CIM to leverage sample weighting by directly removing the spurious correlations between features for every input sample and concentrating more on the correlation between features and labels. Extensive experiments on the MoNuSeg and GLySAC datasets yielded promising results, with mean intersection over union (mIoU) and Dice similarity coefficient (DSC) scores increasing by 3.59% and 2.61%, and 2.71% and 2.04%, respectively, outperforming other state-of-the-art methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v254-fan24a, title = {CDNet: Causal Inference inspired Diversified Aggregation Convolution for Pathology Image Segmentation}, author = {Fan, Dawei and Gao, Yifan and Yu, Jiaming and Yang, Changcai and Chen, Riqing and Wei, Lifang}, booktitle = {Proceedings of the MICCAI Workshop on Computational Pathology}, pages = {51--60}, 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/fan24a/fan24a.pdf}, url = {https://proceedings.mlr.press/v254/fan24a.html}, abstract = {Deep learning models have shown promising performance for Nuclei segmentation in the field of pathology image analysis. However, training a robust model from multiple domains remains a great challenge for Nuclei segmentation. Additionally, the shortcomings of background noise, highly overlapping between Nuclei, and blurred edges often lead to poor performance. To address these challenges, we propose a novel framework termed CDNet, which combines Causal Inference Module (CIM) with Diversified Aggregation Convolution (DAC) techniques. The DAC module is designed which incorporates diverse downsampling features through a simple, parameter-free attention module (SimAM), aiming to overcome the problems of edge blurring. Furthermore, we introduce CIM to leverage sample weighting by directly removing the spurious correlations between features for every input sample and concentrating more on the correlation between features and labels. Extensive experiments on the MoNuSeg and GLySAC datasets yielded promising results, with mean intersection over union (mIoU) and Dice similarity coefficient (DSC) scores increasing by 3.59% and 2.61%, and 2.71% and 2.04%, respectively, outperforming other state-of-the-art methods.} }
Endnote
%0 Conference Paper %T CDNet: Causal Inference inspired Diversified Aggregation Convolution for Pathology Image Segmentation %A Dawei Fan %A Yifan Gao %A Jiaming Yu %A Changcai Yang %A Riqing Chen %A Lifang Wei %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-fan24a %I PMLR %P 51--60 %U https://proceedings.mlr.press/v254/fan24a.html %V 254 %X Deep learning models have shown promising performance for Nuclei segmentation in the field of pathology image analysis. However, training a robust model from multiple domains remains a great challenge for Nuclei segmentation. Additionally, the shortcomings of background noise, highly overlapping between Nuclei, and blurred edges often lead to poor performance. To address these challenges, we propose a novel framework termed CDNet, which combines Causal Inference Module (CIM) with Diversified Aggregation Convolution (DAC) techniques. The DAC module is designed which incorporates diverse downsampling features through a simple, parameter-free attention module (SimAM), aiming to overcome the problems of edge blurring. Furthermore, we introduce CIM to leverage sample weighting by directly removing the spurious correlations between features for every input sample and concentrating more on the correlation between features and labels. Extensive experiments on the MoNuSeg and GLySAC datasets yielded promising results, with mean intersection over union (mIoU) and Dice similarity coefficient (DSC) scores increasing by 3.59% and 2.61%, and 2.71% and 2.04%, respectively, outperforming other state-of-the-art methods.
APA
Fan, D., Gao, Y., Yu, J., Yang, C., Chen, R. & Wei, L.. (2024). CDNet: Causal Inference inspired Diversified Aggregation Convolution for Pathology Image Segmentation. Proceedings of the MICCAI Workshop on Computational Pathology, in Proceedings of Machine Learning Research 254:51-60 Available from https://proceedings.mlr.press/v254/fan24a.html.

Related Material