Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.12929 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916055087054848 |
|---|---|
| author | Ma, Yingzhe Yang, Xiao Yin, Yuguo Wang, Zheyu |
| author_facet | Ma, Yingzhe Yang, Xiao Yin, Yuguo Wang, Zheyu |
| contents | Retinal diagnosis is inherently bilateral: clinicians compare homologous structures across eyes (e.g., optic disc asymmetry), yet most deep models operate on monocular representations. We investigate whether explicit structural correspondence improves diagnosis, and propose Anatomy-Slot to operationalize this hypothesis. Anatomy-Slot introduces an unsupervised anatomical bottleneck by decomposing patch tokens into a set of emergent, structurally-coherent slots that correspond to anatomical regions, then aligning these slots across eyes via bidirectional cross-attention. On ODIR-5K with $n=10$ seeds, the method improves AUC by $4.2$ points over a matched ViT-L baseline (95% CIs; Wilcoxon signed-rank test, $W=0$, $p=0.002$). Pairing disruption and stress testing under Gaussian noise provide controlled tests of correspondence dependence and robustness under corruption. We further report quantitative optic disc grounding on REFUGE and cross-attention localization analysis. Beyond the reported gains, these results indicate that object-centric anatomical correspondence offers a principled path toward interpretable diagnostic systems aligned with clinical bilateral comparison. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_12929 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Anatomy-Slot: Unsupervised Anatomical Factorization for Homologous Bilateral Reasoning in Retinal Diagnosis Ma, Yingzhe Yang, Xiao Yin, Yuguo Wang, Zheyu Computer Vision and Pattern Recognition Artificial Intelligence Retinal diagnosis is inherently bilateral: clinicians compare homologous structures across eyes (e.g., optic disc asymmetry), yet most deep models operate on monocular representations. We investigate whether explicit structural correspondence improves diagnosis, and propose Anatomy-Slot to operationalize this hypothesis. Anatomy-Slot introduces an unsupervised anatomical bottleneck by decomposing patch tokens into a set of emergent, structurally-coherent slots that correspond to anatomical regions, then aligning these slots across eyes via bidirectional cross-attention. On ODIR-5K with $n=10$ seeds, the method improves AUC by $4.2$ points over a matched ViT-L baseline (95% CIs; Wilcoxon signed-rank test, $W=0$, $p=0.002$). Pairing disruption and stress testing under Gaussian noise provide controlled tests of correspondence dependence and robustness under corruption. We further report quantitative optic disc grounding on REFUGE and cross-attention localization analysis. Beyond the reported gains, these results indicate that object-centric anatomical correspondence offers a principled path toward interpretable diagnostic systems aligned with clinical bilateral comparison. |
| title | Anatomy-Slot: Unsupervised Anatomical Factorization for Homologous Bilateral Reasoning in Retinal Diagnosis |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2605.12929 |