One-by-One Stainer: A Fast and Hallucination Resilient Domain Adaptation Method for Histopathology

Karel Moens, Jonas De Vylder, Tinne Tuytelaars, Matthew B. Blaschko

Keywords: domain adaptation, medical image analysis, histopathology, information loss

TL;DR: Histological stain normalization corrects data drift. AI based methods risk image hallucinations that affect the diagnosis. Constraints on the network lead to resilience to hallucinations.

Abstract: Histological staining is a crucial step in the analysis of tissue samples, enabling pathologists to identify and diagnose diseases. However, variations in staining protocols and equipment can lead to inconsistencies in image quality. As a result, histological stain normalization remains an active area of research with increasing demand for AI driven diagnosis support. Unlike other domain adaptation problems, in pathology, the consequences of image hallucinations, obscuring or inserting information, are much greater. We propose a method that mitigates the risk of hallucinations by simplifying the network architecture and training process. Our fully 1x1 convolutional architecture prevents textural modifications and we show that a residual connection combined with weight regularization effectively suppresses color information loss. The model does not need supervision from CycleGAN based models. It is trained directly and leverages target domain color distribution information for better convergence without requiring any paired images. As a result, our method can be trained concurrently on images at varying scales, showing differing anatomical structures or dyes. This simplifies the dataset collection, facilitates the adoption at new centers, and reduces the number of models needed at inference.

Primary Subject Area: Transfer Learning and Domain Adaptation

Secondary Subject Area: Safe and Trustworthy Learning-assisted Solutions for Medical Imaging

Registration Requirement: Yes

Visa & Travel: No

Read CFP & Author Instructions: Yes

Originality Policy: Yes

Single-blind & Not Under Review Elsewhere: Yes

LLM Policy: Yes

Midl Latex Submission Checklist: Ensure no LaTeX errors during compilation., Replace NNN with your OpenReview submission ID., Includes \documentclass{midl}, \jmlryear{2026}, \jmlrworkshop, \jmlrvolume, \editors, and correct \bibliography command., Did not override options of the hyperref package., Did not use the times package., Use the correct spelling and format, avoid Unicode characters, and use LaTeX equivalents instead., Any math in the title and abstract must be enclosed within $...$., Did not override the bibliography style defined in midl.cls and did not use \begin{thebibliography} directly to insert references., Avoid using \scalebox; use \resizebox when needed., Included all necessary figures and removed *unused* files in the zip archive., Removed special formatting, visual annotations, and highlights used during rebuttal., All special characters in the paper and .bib file use LaTeX commands (e.g., \'e for é)., No separate supplementary PDF uploads., Acknowledgements, references, and appendix must start after the main content.

Submission Number: 121