Datasets:

Modalities:
Tabular
Text
Formats:
csv
Size:
< 1K
ArXiv:
Libraries:
Datasets
pandas
License:
RR-Findings / README.md
YYama0's picture
Update README.md
6a41d5e verified
metadata
license: cc-by-4.0

RR-Findings: Chest CT Findings Labels for Japanese Radiology Reports

RR-Findings is a labeled dataset of Japanese chest CT radiology reports derived from the MedNLP-SC Radiology Report TNM Staging (RR-TNM) Dataset, released as part of the NTCIR-17 MedNLP-SC shared task.

A total of 243 Japanese radiology reports (covering 27 lung cancer cases, annotated independently by nine board-certified radiologists) were further enriched with multi-label chest CT findings assigned by radiology residents. All labels follow the schema of the CT-RATE dataset.


Labels

Each report is annotated with 18 clinically relevant chest CT findings:

Medical material  
Arterial wall calcification  
Cardiomegaly  
Pericardial effusion  
Coronary artery wall calcification  
Hiatal hernia  
Lymphadenopathy  
Emphysema  
Atelectasis  
Lung nodule  
Lung opacity  
Pulmonary fibrotic sequela  
Pleural effusion  
Mosaic attenuation pattern  
Peribronchial thickening  
Consolidation  
Bronchiectasis  
Interlobular septal thickening

The labeling schema is based on CT-RATE: https://huggingface.co/datasets/ibrahimhamamci/CT-RATE


Source Dataset

RR-Findings is built upon:

MedNLP-SC Radiology Report TNM Staging (RR-TNM) Dataset NTCIR-17 MedNLP-SC Shared Task 243 PHI-free Japanese radiology reports https://sociocom.naist.jp/download/mednlp-sc-rr-tnm/


Related Research and Citation

This dataset was developed as part of the following study:

Yosuke Yamagishi, Tomohiro Kikuchi, Shouhei Hanaoka, Takeharu Yoshikawa, Osamu Abe.
ModernBERT is More Efficient than Conventional BERT for Chest CT Findings Classification in Japanese Radiology Reports.
arXiv:2503.05060 (2025).
https://arxiv.org/abs/2503.05060

If you use this dataset in your research or publication, please cite the paper above.


License

CC BY 4.0 You may use, modify, and redistribute the dataset with appropriate attribution.