HistDiST: Histopathological Diffusion-based Stain Transfer

2025-05-11

AI Summary

To address training instability and poor structural fidelity in GAN-based H&E-to-IHC staining translation, this paper proposes the first dual-conditional latent diffusion framework. Methodologically, it introduces a zero-SNR noise schedule, DDIM-based structure-preserving inversion, and an η-cosine stochastic control strategy; integrates Phikon-derived morphological embeddings with VAE latent encoding; adopts v-prediction modeling and rescaled noise scheduling. A pathology-aware metric—Morphological Reconstruction Accuracy (MRA)—is defined and complemented by semantic evaluation using GigaPath. On the MIST/BCI datasets for H&E-to-Ki67 translation, the method achieves a 28% improvement in MRA, significantly enhancing molecular semantic fidelity and tissue structural consistency. This work establishes a novel paradigm for low-cost, high-fidelity IHC surrogate generation.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Hematoxylin and Eosin (H&E) staining is the cornerstone of histopathology but lacks molecular specificity. While Immunohistochemistry (IHC) provides molecular insights, it is costly and complex, motivating H&E-to-IHC translation as a cost-effective alternative. Existing translation methods are mainly GAN-based, often struggling with training instability and limited structural fidelity, while diffusion-based approaches remain underexplored. We propose HistDiST, a Latent Diffusion Model (LDM) based framework for high-fidelity H&E-to-IHC translation. HistDiST introduces a dual-conditioning strategy, utilizing Phikon-extracted morphological embeddings alongside VAE-encoded H&E representations to ensure pathology-relevant context and structural consistency. To overcome brightness biases, we incorporate a rescaled noise schedule, v-prediction, and trailing timesteps, enforcing a zero-SNR condition at the final timestep. During inference, DDIM inversion preserves the morphological structure, while an eta-cosine noise schedule introduces controlled stochasticity, balancing structural consistency and molecular fidelity. Moreover, we propose Molecular Retrieval Accuracy (MRA), a novel pathology-aware metric leveraging GigaPath embeddings to assess molecular relevance. Extensive evaluations on MIST and BCI datasets demonstrate that HistDiST significantly outperforms existing methods, achieving a 28% improvement in MRA on the H&E-to-Ki67 translation task, highlighting its effectiveness in capturing true IHC semantics.

Problem

• H&E lacks molecular specificity, needing IHC translation
• GAN-based methods face instability and fidelity issues
• Proposing HistDiST for high-fidelity H&E-to-IHC conversion

Innovation

• Latent Diffusion Model for H&E-to-IHC translation
• Dual-conditioning strategy with morphological embeddings
• Rescaled noise schedule and zero-SNR enforcement