OzzyChen97/TC-SSA

TC-SSA: Token Compression via Semantic Slot Aggregation for Gigapixel Pathology Reasoning

    

Links: Project homepage | arXiv paper | Code

Authors: Zhuo Chen^1,2, Xiaoyu Yang¹, and Lijian Xu^1,*

¹ Shenzhen University of Advanced Technology, Shenzhen, Guangdong, China ² University of Nottingham Ningbo China, FoSE, Ningbo, Zhejiang, China ^* Corresponding author: xulijian@suat-sz.edu.cn

TC-SSA WSI Feature Bags

This public repository contains pre-extracted whole-slide image patch features used by TC-SSA and SSAMIL experiments. Binary feature bags are organized by feature source and encoder. The Dataset Viewer intentionally displays the file index in metadata/file_index.csv; it does not attempt to interpret the HDF5 or PyTorch feature files as tabular datasets.

Inventory

CONCHv1.5 features

Each HDF5 file contains a features dataset with shape (num_patches, 768). Some files also contain patch coordinates.

CPathPatchFeature TCGA-BRCA subset

The four encoder directories have identical slide filename sets. The 1,125 slides used by the formal TCGA-BRCA experiment splits were validated for full coverage and expected feature dimensions. Missing files from the upstream BRCA subset were locally re-extracted and merged with the released collection.

Layout

features/
  conch_v15/
    BRACS/*.h5
    CAMELYON17/*.h5
    TCGA-BRCA/*.h5
    TCGA-LUAD/*.h5
  cpath_patch_feature/
    brca/
      chief/pt_files/*.pt
      gigap/pt_files/*.pt
      r50/pt_files/*.pt
      uni/pt_files/*.pt
      patches/*.h5
metadata/
  file_index.csv

Dataset Viewer and access

The Viewer exposes a searchable index rather than loading hundreds of gigabytes of heterogeneous binary tensors:

from datasets import load_dataset

index = load_dataset("OzzyChen97/TC-SSA", "file-index", split="train")
print(index[0])

Download selected files with hf download, then load HDF5 features with h5py.File(...) or PyTorch tensors with torch.load(..., map_location="cpu").

Provenance and references

The CPath encoder features and patch coordinates are derived from Dearcat/CPathPatchFeature, released with the work Revisiting End-to-End Learning with Slide-level Supervision in Computational Pathology. Please cite that repository and its associated publications when using the CPath subset.

The CONCH feature bags use CONCH, a pathology vision-language foundation model. The CPath collection uses UNI, CHIEF, GigaPath, and ResNet-50.

Original WSI datasets and access pages:

• —BRACS dataset and paper
• —CAMELYON17 challenge data
• —TCGA-BRCA at the NCI Genomic Data Commons
• —TCGA-LUAD at the NCI Genomic Data Commons

The upstream CPathPatchFeature repository additionally contains derived features for TCGA-BLCA, TCGA-NSCLC, CPTAC-NSCLC, CAMELYON, and PANDA. Those additional cohorts are referenced upstream and are not duplicated in this TC-SSA repository.

Citation

@inproceedings{chen2026tcssa,
  title     = {TC-SSA: Token Compression via Semantic Slot Aggregation for Gigapixel Pathology Reasoning},
  author    = {Chen, Zhuo and Yang, Xiaoyu and Xu, Lijian},
  booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026},
  publisher = {Springer Nature},
  series    = {Lecture Notes in Computer Science},
  year      = {2026},
  doi       = {10.48550/arXiv.2603.01143},
  url       = {https://arxiv.org/abs/2603.01143}
}

Please also cite the original WSI dataset and encoder papers relevant to the files used in your study.