cs.LGFeb 26, 2026

A Proper Scoring Rule for Virtual Staining

Authors: Samuel Tonks, Steve Hood, Ryan Musso, Ceridwen Hopely, Steve Titus, Minh Doan, Iain Styles, Alexander Krull

Organizations: School of Computer Science, University of Birmingham, Birmingham, UK · GSK Drug Metabolism & Pharmacokinetics, GSK Medicines Research Centre, Stevenage, UK · GSK Genome Biology, Collegeville, PA, USA · School of Electronics, Electrical Engineering and Computer Science, Queen’s University Belfast, UK

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Generative virtual staining (VS) models for high-throughput screening (HTS) can provide an estimated posterior distribution of possible biological feature values for each input and cell. However, when evaluating a VS model, the true posterior is unavailable. Existing evaluation protocols only check the accuracy of the marginal distribution over the dataset rather than the predicted posteriors. We introduce information gain (IG) as a cell-wise evaluation framework that enables direct assessment of predicted posteriors. IG is a strictly proper scoring rule and comes with a sound theoretical motivation allowing for interpretability, and for comparing results across models and features. We evaluate diffusion- and GAN-based models on an extensive HTS dataset using IG and other metrics and show that IG can reveal substantial performance differences other metrics cannot.

Figures & tables

Figure 1: How to evaluate a predicted posterior distribution with a single target sample? A traditional HTS pipeline (top) uses fluorescence microscopy and extracts feature values Yi,j\mathcal{Y}_{i,j} for each cell jj in image ii . A cell’s true feature value Yi,j\mathcal{Y}_{i,j} can be seen as a single sample from an inaccessible posterior P(Yi,j∣xi,j)P(\mathcal{Y}_{i,j}|\textbf{x}_{i,j}) . Although a generative VS model can produce many samples (bottom) Y^i,j1,…,K\hat{\mathcal{Y}}_{i,j}^{1,\ldots,K} , its learned distribution must be evaluated against the single true value.
Figure 2: Qualitative results for single input image . For an unseen bright-field, we show three model samples and the sample corresponding to the median posterior result. Alongside, we display the predicted posterior distribution for the ’upper quartile intensity’ feature (F7), highlighting the median values and the target feature value. Pix2PixHD underestimates cell intensity, while cDDPM exhibits greater variability; its posterior is wider and closer to the target value.
Figure 3: Evaluation metrics for upper quartile intensity (F7) feature. Marginal KLD and rank distributions are inconclusive and mask differences in model performance. Log-likelihood distribution reveals superior performance of cDDPM.
Figure 4: Comparing evaluation metrics across features (F1-F18). Metrics averaged for F10-F18 due to space limitations. Best model for each feature and metric is highlighted as bold/italic. The proposed information gain shows a clear advantage for cDDPM across features, while the other metrics mask the performance difference.
More in this topic

Information Gain

See all 13 papers

Explore similar work

CardsList

  1. Towards Reliable AI-Based Histological Staining: A Systematic Study of Scaling and Uncertainty in Unpaired Generative Models

    Aug 25, 2026Qasim Siddiqui, Adrian Friebel, Maiju Myllys +4Virtual StainingImage-to-Image Translation

    Interdisciplinary Centre of Bioinformatics (IZBI) Universität Leipzig Leipzig, Germany · Center for Scalable Data Analytics and Artificial Intelligence (ScaDS.AI) University of Leipzig Leipzig, Germany · Department of Toxicology Technical University Dortmund Dortmund, Germany +1

  2. UNIStainNet: Foundation-Model-Guided Virtual Staining of H&E to IHC

    Mar 13, 2026Jillur Rahman Saurav, Thuong Le Hoai Pham, Pritam Mukherjee +3Virtual StainingComputational Pathology

    Department of Computer Science, University of Texas at Arlington, Arlington, Texas, USA · Department of Radiology, St. Jude Children’s Research Hospital, Memphis, Tennessee, USA · Department of Pathology, St. Jude Children’s Research Hospital, Memphis, Tennessee, USA

  3. HAPS: Rethinking Image Similarity for Virtual Staining

    May 19, 2026Fedor Gubanov, Svetlana Illarionova, Vlad Kozlovskiy +6Computational PathologyMedical Image Analysis

    1*Skolkovo Institute of Science and Technology, Moscow, 121205, Russia. · 2BIMAI-Lab, Biomedically Informed Artificial Intelligence Laboratory,2026 University of Sharjah, Sharjah, United Arab Emirates. · 3National Medical Research Radiological Centre of the Ministry of Health of the Russian Federation, Moscow, Russia.