Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Beer-Lambert Autoencoder for Unsupervised Stain Representation Learning and Deconvolution in Multi-immunohistochemical Brightfield Histology Images

Mark Eastwood, Thomas McKee, Zedong Hu, Sabine Tejpar, Fayyaz Minhas

TL;DR

The paper tackles the underdetermined problem of separating $K>3$ chromogenic stains in brightfield RGB histology by learning cohort-specific stain representations. It introduces a compact encoder-decoder with a differentiable Beer-Lambert forward model and a learnable stain matrix $S_\phi$, optimized via an unsupervised, priors-based objective that promotes sparse, nonoverlapping stain maps. Applied to a colorectal cancer panel with stains $H$, CDX2, MUC2, MUC5, and CD8, it achieves excellent RGB reconstruction and crisper per-stain maps with reduced cross-talk compared to matrix-based deconvolution. The approach enables scalable, quantitative mIHC analysis across labs, with code available at GitHub.

Abstract

Separating the contributions of individual chromogenic stains in RGB histology whole slide images (WSIs) is essential for stain normalization, quantitative assessment of marker expression, and cell-level readouts in immunohistochemistry (IHC). Classical Beer-Lambert (BL) color deconvolution is well-established for two- or three-stain settings, but becomes under-determined and unstable for multiplex IHC (mIHC) with K>3 chromogens. We present a simple, data-driven encoder-decoder architecture that learns cohort-specific stain characteristics for mIHC RGB WSIs and yields crisp, well-separated per-stain concentration maps. The encoder is a compact U-Net that predicts K nonnegative concentration channels; the decoder is a differentiable BL forward model with a learnable stain matrix initialized from typical chromogen hues. Training is unsupervised with a perceptual reconstruction objective augmented by loss terms that discourage unnecessary stain mixing. On a colorectal mIHC panel comprising 5 stains (H, CDX2, MUC2, MUC5, CD8) we show excellent RGB reconstruction, and significantly reduced inter-channel bleed-through compared with matrix-based deconvolution. Code and model are available at https://github.com/measty/StainQuant.git.

Beer-Lambert Autoencoder for Unsupervised Stain Representation Learning and Deconvolution in Multi-immunohistochemical Brightfield Histology Images

TL;DR

The paper tackles the underdetermined problem of separating chromogenic stains in brightfield RGB histology by learning cohort-specific stain representations. It introduces a compact encoder-decoder with a differentiable Beer-Lambert forward model and a learnable stain matrix , optimized via an unsupervised, priors-based objective that promotes sparse, nonoverlapping stain maps. Applied to a colorectal cancer panel with stains , CDX2, MUC2, MUC5, and CD8, it achieves excellent RGB reconstruction and crisper per-stain maps with reduced cross-talk compared to matrix-based deconvolution. The approach enables scalable, quantitative mIHC analysis across labs, with code available at GitHub.

Abstract

Separating the contributions of individual chromogenic stains in RGB histology whole slide images (WSIs) is essential for stain normalization, quantitative assessment of marker expression, and cell-level readouts in immunohistochemistry (IHC). Classical Beer-Lambert (BL) color deconvolution is well-established for two- or three-stain settings, but becomes under-determined and unstable for multiplex IHC (mIHC) with K>3 chromogens. We present a simple, data-driven encoder-decoder architecture that learns cohort-specific stain characteristics for mIHC RGB WSIs and yields crisp, well-separated per-stain concentration maps. The encoder is a compact U-Net that predicts K nonnegative concentration channels; the decoder is a differentiable BL forward model with a learnable stain matrix initialized from typical chromogen hues. Training is unsupervised with a perceptual reconstruction objective augmented by loss terms that discourage unnecessary stain mixing. On a colorectal mIHC panel comprising 5 stains (H, CDX2, MUC2, MUC5, CD8) we show excellent RGB reconstruction, and significantly reduced inter-channel bleed-through compared with matrix-based deconvolution. Code and model are available at https://github.com/measty/StainQuant.git.
Paper Structure (5 sections, 8 equations, 1 figure, 1 table)

This paper contains 5 sections, 8 equations, 1 figure, 1 table.

Table of Contents

  1. Introduction
  2. Methods
  3. Objective and loss terms
  4. Experiments and Results
  5. Discussion & Conclusions

Figures (1)

  • Figure 1: Method and results.A: Encoder-decoder with learnable stain matrix. B: Examples of Input patches and model reconstructions from channel decomposition. C: Single-channel per-stain concentration map (H, CDX2, MUC2, MUC5, CD8) renderings and knock-out (KO) images, and comparison with single channel renders from matrix-based deconvolution (bottom row). D: Reconstruction metrics and pairwise channel similarity comparison.