Understanding Stain Separation Improves Cross-Scanner Adenocarcinoma Segmentation with Joint Multi-Task Learning

TL;DR

The paper addresses domain shift in digital pathology caused by organ variability, tissue processing, and scanner color differences, by evaluating cross-scanner adenocarcinoma segmentation on six scanners (COSAS Task 2). It proposes a joint multi-task learning framework with a multi-decoder autoencoder that performs unsupervised stain separation, learning the stain matrix $W$ and stain density $H$ from $I \in \mathbb{R}^{m \times n}$; the per-pixel logits $\hat{y}$ for segmentation are obtained by feeding the concatenation of $\hat{H}$ and $\hat{W}$ into a classifier. The objective combines reconstruction loss $\mathcal{L}_{recon}$ and segmentation loss $\mathcal{L}_{seg}$ as $\mathcal{L}_{total} = \alpha \mathcal{L}_{recon} + \mathcal{L}_{seg}$, with $\alpha$ optimized (found to be $0.3$) and a mix of stain augmentations including RandStainNA and SPCN-based augmentation to enhance domain generalization. Results show COSAS 0.846 with Dice 0.887 and IoU 0.805 in four-fold CV; these findings suggest improved cross-scanner robustness and practical potential for reliable digital pathology diagnostics.

Abstract

Digital pathology has made significant advances in tumor diagnosis and segmentation, but image variability due to differences in organs, tissue preparation, and acquisition - known as domain shift - limits the effectiveness of current algorithms. The COSAS (Cross-Organ and Cross-Scanner Adenocarcinoma Segmentation) challenge addresses this issue by improving the resilience of segmentation algorithms to domain shift, with Task 2 focusing on adenocarcinoma segmentation using a diverse dataset from six scanners, pushing the boundaries of clinical diagnostics. Our approach employs unsupervised learning through stain separation within a multi-task learning framework using a multi-decoder autoencoder. This model isolates stain matrix and stain density, allowing it to handle color variation and improve generalization across scanners. We further enhanced the robustness of the model with a mixture of stain augmentation techniques and used a U-net architecture for segmentation. The novelty of our method lies in the use of stain separation within a multi-task learning framework, which effectively disentangles histological structures from color variations. This approach shows promise for improving segmentation accuracy and generalization across different histopathological stains, paving the way for more reliable diagnostic tools in digital pathology.

Figures (1)

• Figure 1: Multi-decoder Unet architecture for joint multi-task learning