Adaptive Multi-Scale Integration Unlocks Robust Cell Annotation in Histopathology Images

2025-11-17

AI Summary

Cell-type identification in histopathological images faces two key challenges: (1) the disconnection between local nuclear morphological features and global tissue microenvironmental context, and (2) the scarcity and severe class imbalance of fine-grained subtype annotations. To address these, we propose NuClass—a novel framework that adaptively fuses multi-scale morphological and contextual information via a learnable gating mechanism, and enhances robustness and interpretability through uncertainty-guided learning. NuClass further incorporates Xenium spatial transcriptomics–informed weak supervision for annotation, Grad-CAM–based visual interpretability, and multi-scale ensemble inference. Evaluated across eight organ types encompassing over 2 million cells, NuClass achieves a peak F1-score of 96%, substantially outperforming state-of-the-art methods. It delivers both high-accuracy and interpretable fine-grained cellular phenotyping, advancing computational pathology toward clinically actionable cell-level analysis.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models
• Computer Vision: Interpretability, Explainability, and Transparency

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics

Abstract

Identifying cell types and subtypes from routine histopathology images is essential for improving the computational understanding of human disease. Existing tile-based models can capture detailed nuclear morphology but often fail to incorporate the broader tissue context that influences a cell's function and identity. In addition, available human annotations are typically coarse-grained and unevenly distributed across studies, making fine-grained subtype-level supervision difficult to obtain. To address these limitations, we introduce NuClass, a pathologist workflow inspired framework for cell-wise multi-scale integration of nuclear morphology and microenvironmental context. NuClass includes two main components: Path local, which focuses on nuclear morphology from 224-by-224 pixel crops, and Path global, which models the surrounding 1024-by-1024 pixel neighborhood. A learnable gating module adaptively balances local detail and contextual cues. To encourage complementary learning, we incorporate an uncertainty-guided objective that directs the global path to prioritize regions where the local path is uncertain. We also provide calibrated confidence estimates and Grad-CAM visualizations to enhance interpretability. To overcome the lack of high-quality annotations, we construct a marker-guided dataset from Xenium spatial transcriptomics assays, yielding single-cell resolution labels for more than two million cells across eight organs and 16 classes. Evaluated on three fully held-out cohorts, NuClass achieves up to 96 percent F1 for its best-performing class, outperforming strong baselines. Our results show that multi-scale, uncertainty-aware fusion can bridge the gap between slide-level pathological foundation models and reliable, cell-level phenotype prediction.

Problem

• Identifying cell types from histopathology images without sufficient fine-grained annotations
• Balancing nuclear morphology details with broader tissue context in cell analysis
• Overcoming limitations of tile-based models that lack microenvironmental integration

Innovation

• Multi-scale integration of nuclear morphology and context
• Adaptive gating balances local details and global cues
• Uncertainty-guided objective prioritizes complementary learning regions