L-Diffusion: Laplace Diffusion for Efficient Pathology Image Segmentation

Weihan Li, Linyun Zhou, Jian Yang, Shengxuming Zhang, Xiangtong Du, Xiuming Zhang, Jing Zhang, Chaoqing Xu, Mingli Song, Zunlei Feng
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:36954-36973, 2025.

Abstract

Pathology image segmentation plays a pivotal role in artificial digital pathology diagnosis and treatment. Existing approaches to pathology image segmentation are hindered by labor-intensive annotation processes and limited accuracy in tail-class identification, primarily due to the long-tail distribution inherent in gigapixel pathology images. In this work, we introduce the Laplace Diffusion Model, referred to as L-Diffusion, an innovative framework tailored for efficient pathology image segmentation. L-Diffusion utilizes multiple Laplace distributions, as opposed to Gaussian distributions, to model distinct components—a methodology supported by theoretical analysis that significantly enhances the decomposition of features within the feature space. A sequence of feature maps is initially generated through a series of diffusion steps. Following this, contrastive learning is employed to refine the pixel-wise vectors derived from the feature map sequence. By utilizing these highly discriminative pixel-wise vectors, the segmentation module achieves a harmonious balance of precision and robustness with remarkable efficiency. Extensive experimental evaluations demonstrate that L-Diffusion attains improvements of up to 7.16%, 26.74%, 16.52%, and 3.55% on tissue segmentation datasets, and 20.09%, 10.67%, 14.42%, and 10.41% on cell segmentation datasets, as quantified by DICE, MPA, mIoU, and FwIoU metrics. The source are available at https://github.com/Lweihan/LDiffusion.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-li25ea, title = {L-Diffusion: {L}aplace Diffusion for Efficient Pathology Image Segmentation}, author = {Li, Weihan and Zhou, Linyun and Yang, Jian and Zhang, Shengxuming and Du, Xiangtong and Zhang, Xiuming and Zhang, Jing and Xu, Chaoqing and Song, Mingli and Feng, Zunlei}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {36954--36973}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/li25ea/li25ea.pdf}, url = {https://proceedings.mlr.press/v267/li25ea.html}, abstract = {Pathology image segmentation plays a pivotal role in artificial digital pathology diagnosis and treatment. Existing approaches to pathology image segmentation are hindered by labor-intensive annotation processes and limited accuracy in tail-class identification, primarily due to the long-tail distribution inherent in gigapixel pathology images. In this work, we introduce the Laplace Diffusion Model, referred to as L-Diffusion, an innovative framework tailored for efficient pathology image segmentation. L-Diffusion utilizes multiple Laplace distributions, as opposed to Gaussian distributions, to model distinct components—a methodology supported by theoretical analysis that significantly enhances the decomposition of features within the feature space. A sequence of feature maps is initially generated through a series of diffusion steps. Following this, contrastive learning is employed to refine the pixel-wise vectors derived from the feature map sequence. By utilizing these highly discriminative pixel-wise vectors, the segmentation module achieves a harmonious balance of precision and robustness with remarkable efficiency. Extensive experimental evaluations demonstrate that L-Diffusion attains improvements of up to 7.16%, 26.74%, 16.52%, and 3.55% on tissue segmentation datasets, and 20.09%, 10.67%, 14.42%, and 10.41% on cell segmentation datasets, as quantified by DICE, MPA, mIoU, and FwIoU metrics. The source are available at https://github.com/Lweihan/LDiffusion.} }
Endnote
%0 Conference Paper %T L-Diffusion: Laplace Diffusion for Efficient Pathology Image Segmentation %A Weihan Li %A Linyun Zhou %A Jian Yang %A Shengxuming Zhang %A Xiangtong Du %A Xiuming Zhang %A Jing Zhang %A Chaoqing Xu %A Mingli Song %A Zunlei Feng %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-li25ea %I PMLR %P 36954--36973 %U https://proceedings.mlr.press/v267/li25ea.html %V 267 %X Pathology image segmentation plays a pivotal role in artificial digital pathology diagnosis and treatment. Existing approaches to pathology image segmentation are hindered by labor-intensive annotation processes and limited accuracy in tail-class identification, primarily due to the long-tail distribution inherent in gigapixel pathology images. In this work, we introduce the Laplace Diffusion Model, referred to as L-Diffusion, an innovative framework tailored for efficient pathology image segmentation. L-Diffusion utilizes multiple Laplace distributions, as opposed to Gaussian distributions, to model distinct components—a methodology supported by theoretical analysis that significantly enhances the decomposition of features within the feature space. A sequence of feature maps is initially generated through a series of diffusion steps. Following this, contrastive learning is employed to refine the pixel-wise vectors derived from the feature map sequence. By utilizing these highly discriminative pixel-wise vectors, the segmentation module achieves a harmonious balance of precision and robustness with remarkable efficiency. Extensive experimental evaluations demonstrate that L-Diffusion attains improvements of up to 7.16%, 26.74%, 16.52%, and 3.55% on tissue segmentation datasets, and 20.09%, 10.67%, 14.42%, and 10.41% on cell segmentation datasets, as quantified by DICE, MPA, mIoU, and FwIoU metrics. The source are available at https://github.com/Lweihan/LDiffusion.
APA
Li, W., Zhou, L., Yang, J., Zhang, S., Du, X., Zhang, X., Zhang, J., Xu, C., Song, M. & Feng, Z.. (2025). L-Diffusion: Laplace Diffusion for Efficient Pathology Image Segmentation. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:36954-36973 Available from https://proceedings.mlr.press/v267/li25ea.html.

Related Material