Magnification-Aware Distillation (MAD): A Self-Supervised Framework for Unified Representation Learning in Gigapixel Whole-Slide Images
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866911323873345536 |
|---|---|
| author | Gokmen, Mahmut S. Klusty, Mitchell A. Nelson, Peter T. Neltner, Allison M. Cheung, Sen-Ching Samson Pearce, Thomas M. Gutman, David A Dugger, Brittany N. Bisht, Devavrat S. Flanagan, Margaret E. Bumgardner, V. K. Cody |
| author_facet | Gokmen, Mahmut S. Klusty, Mitchell A. Nelson, Peter T. Neltner, Allison M. Cheung, Sen-Ching Samson Pearce, Thomas M. Gutman, David A Dugger, Brittany N. Bisht, Devavrat S. Flanagan, Margaret E. Bumgardner, V. K. Cody |
| contents | Whole-slide images (WSIs) contain tissue information distributed across multiple magnification levels, yet most self-supervised methods treat these scales as independent views. This separation prevents models from learning representations that remain stable when resolution changes, a key requirement for practical neuropathology workflows. This study introduces Magnification-Aware Distillation (MAD), a self-supervised strategy that links low-magnification context with spatially aligned high-magnification detail, enabling the model to learn how coarse tissue structure relates to fine cellular patterns. The resulting foundation model, MAD-NP, is trained entirely through this cross-scale correspondence without annotations. A linear classifier trained only on 10x embeddings maintains 96.7% of its performance when applied to unseen 40x tiles, demonstrating strong resolution-invariant representation learning. Segmentation outputs remain consistent across magnifications, preserving anatomical boundaries and minimizing noise. These results highlight the feasibility of scalable, magnification-robust WSI analysis using a unified embedding space |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_14796 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Magnification-Aware Distillation (MAD): A Self-Supervised Framework for Unified Representation Learning in Gigapixel Whole-Slide Images Gokmen, Mahmut S. Klusty, Mitchell A. Nelson, Peter T. Neltner, Allison M. Cheung, Sen-Ching Samson Pearce, Thomas M. Gutman, David A Dugger, Brittany N. Bisht, Devavrat S. Flanagan, Margaret E. Bumgardner, V. K. Cody Image and Video Processing Artificial Intelligence Computer Vision and Pattern Recognition Machine Learning Whole-slide images (WSIs) contain tissue information distributed across multiple magnification levels, yet most self-supervised methods treat these scales as independent views. This separation prevents models from learning representations that remain stable when resolution changes, a key requirement for practical neuropathology workflows. This study introduces Magnification-Aware Distillation (MAD), a self-supervised strategy that links low-magnification context with spatially aligned high-magnification detail, enabling the model to learn how coarse tissue structure relates to fine cellular patterns. The resulting foundation model, MAD-NP, is trained entirely through this cross-scale correspondence without annotations. A linear classifier trained only on 10x embeddings maintains 96.7% of its performance when applied to unseen 40x tiles, demonstrating strong resolution-invariant representation learning. Segmentation outputs remain consistent across magnifications, preserving anatomical boundaries and minimizing noise. These results highlight the feasibility of scalable, magnification-robust WSI analysis using a unified embedding space |
| title | Magnification-Aware Distillation (MAD): A Self-Supervised Framework for Unified Representation Learning in Gigapixel Whole-Slide Images |
| topic | Image and Video Processing Artificial Intelligence Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2512.14796 |