Mind the Gap: Continuous Magnification Sampling for Pathology Foundation Models
cs.CV, cs.LG
Submitted: 2026-01-05
Updated: 2026-01-05
Journal ref: Medical Image Analysis 2026, Volume 116
DOI: 10.1016/j.media.2026.104327
Code: https://github.com/bifold-pathomics/continuous-magnification-sampling
License: http://creativecommons.org/licenses/by/4.0/
The gist: In histopathology, pathologists examine both tissue architecture at low magnification and fine-grained morphology at high magnification.
Terminology
Abstract
In histopathology, pathologists examine both tissue architecture at low magnification and fine-grained morphology at high magnification. Yet, the performance of pathology foundation models across magnifications and the effect of magnification sampling during training remain poorly understood. We model magnification sampling as a multi-source domain adaptation problem and develop a simple theoretical framework that reveals systematic trade-offs between sampling strategies. We show that the widely used discrete uniform sampling of magnifications (0.25, 0.5, 1.0, 2.0 mpp) leads to degradation at intermediate magnifications. We introduce continuous magnification sampling, which removes gaps in magnification coverage while preserving performance at standard scales. Further, we derive sampling distributions that optimize representation quality across magnification scales. To evaluate these strategies, we introduce two new benchmarks (TCGA-MS, BRACS-MS) with appropriate metrics. Our experiments show that continuous sampling substantially improves over discrete sampling at intermediate magnifications, with gains of up to 4 percentage points in balanced classification accuracy, and that optimized distributions can further improve performance. Finally, we evaluate current histopathology foundation models, finding that magnification is a primary driver of performance variation across models. Our work paves the way towards future pathology foundation models that perform reliably across magnifications.
Sources
- Multi-Resolution Pathology-Language Pre-training Model with Text-Guided Visual Representation
- Atlas: A Novel Pathology Foundation Model by Mayo Clinic, Charit'e, and Aignostics
- RudolfV: A Foundation Model by Pathologists for Pathologists
- Phikon-v2, A large and public feature extractor for biomarker prediction
- PLUTO: Pathology-Universal Transformer
- Towards Large-Scale Training of Pathology Foundation Models
- Towards Robust Foundation Models for Digital Pathology
- DINOv2: Learning Robust Visual Features without Supervision
- Molecular-driven Foundation Model for Oncologic Pathology
- Accelerating Data Processing and Benchmarking of AI Models for Pathology
- iBOT: Image BERT Pre-Training with Online Tokenizer
- Adapting Self-Supervised Learning for Computational Pathology
- Virchow2: Scaling Self-Supervised Mixed Magnification Models in Pathology
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs