Research map / Label-efficient and robust learning / Self-supervised and contrastive
Cyto-SSL: A Self-Supervised Pretraining Framework for Cytology Foundation Model
Yiming Zhang, Rui Yan, Xiaohua Wan, Yifan Zhao, Shuang Feng, Zhetao Xu, Ying Wang, Fa Zhang, Bin Hu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add Nuclei-Centered Perturbation to DINO (ablation) | DINO ViT-tiny · Cervical Cytologic WSI (ABMIL) | accuracy (%): 91.77 -> 93.07 | Improved |
| add Nuclei-Centered Perturbation to DINO (ablation) | DINO ViT-tiny · Cervical Cytologic WSI (LESS) | accuracy (%): 92.20 -> 93.50 | Improved |
| add sparse self-attention to DINO (ablation) | DINO ViT-tiny · Cervical Cytologic WSI (ABMIL) | accuracy (%): 91.77 -> 93.51 | Improved |
| add sparse self-attention to DINO (ablation) | DINO ViT-tiny · Cervical Cytologic WSI (LESS) | AUC: 0.963 -> 0.952 | Mixed |
| add iRPE to DINO (ablation) | DINO ViT-tiny · Cervical Cytologic WSI (ABMIL) | accuracy (%): 91.77 -> 92.21 | Improved |
| add iRPE to DINO (ablation) | DINO ViT-tiny · Cervical Cytologic WSI (LESS) | accuracy (%): 92.20 -> 93.07 | Improved |
| combine all three components (full Cyto-SSL) versus DINO (ablation) | DINO ViT-tiny · Cervical Cytologic WSI (LESS) | accuracy (%): 92.20 -> 95.67 | Improved |
| add Cyto-SSL to DINO | DINO ViT-tiny · Cervical Cytologic WSI (ABMIL) | accuracy (%): 91.77 -> 94.37 | Improved |
| add Cyto-SSL to SimCLR | SimCLR ViT-tiny · Cervical Cytologic WSI | AUC: 0.932 -> 0.915 | Mixed |
| add Cyto-SSL to MoCo | MoCo ViT-tiny · Cervical Cytologic WSI | accuracy (%): 87.88 -> 88.74 | Improved |
| pretrain from scratch on cytology instead of ImageNet ResNet-50 features | ABMIL · Cervical Cytologic WSI | accuracy (%): 83.04 -> 94.37 | Improved |
| pretrain from scratch with Cyto-SSL instead of LoRA-tuned Prov-GigaPath | ABMIL · Cervical Cytologic WSI | accuracy (%): 92.50 -> 94.37 | Improved |
Closest papers by meaning
- Unsupervised Foundation Model-Agnostic Slide-Level Representation Learning (CVPR 2025 · similarity 0.87)
- Nonparametric Distribution Matching for Self-Supervised Whole-Slide Image Condensation (NeurIPS 2026 · similarity 0.87)
- MLLM-HWSI: A Multimodal Large Language Model for Hierarchical Whole Slide Image Understanding (CVPR 2026 · similarity 0.86)
- PathVQ: Reforming Computational Pathology Foundation Model for Whole Slide Image Analysis via Vector Quantization (NeurIPS 2025 · similarity 0.86)
- Towards Scalable Context-Aware Single-Cell Spatial Transcriptomics Prediction from Histology Images (NeurIPS 2026 · similarity 0.86)
- Sequential Attention-based Sampling for Histopathological Analysis (NeurIPS 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
