Datasets:
license: cc-by-nc-sa-4.0
language:
- en
task_categories:
- text-classification
tags:
- radiology
- chest-ct
- ct-rate
- multi-label
- report-labeling
size_categories:
- 10K<n<100K
dataset_info:
features:
- name: original_report
dtype: string
- name: refined_report
dtype: string
- name: split
dtype: string
- name: Consolidation
dtype: int8
- name: Ground-glass opacity (GGO)
dtype: int8
- name: Crazy-paving pattern
dtype: int8
- name: Mosaic attenuation / air-trapping
dtype: int8
- name: Tree-in-bud
dtype: int8
- name: Centrilobular nodules / bronchiolitis pattern
dtype: int8
- name: Pulmonary nodule (solid / PSN / GGN)
dtype: int8
- name: Pulmonary mass (>3 cm)
dtype: int8
- name: Cavitary nodule / mass
dtype: int8
- name: Emphysema
dtype: int8
- name: Bullae / giant bulla
dtype: int8
- name: Pulmonary cysts / cystic lung disease
dtype: int8
- name: Reticulation / intralobular thickening
dtype: int8
- name: Interlobular septal thickening
dtype: int8
- name: Traction bronchiectasis / bronchiolectasis
dtype: int8
- name: Honeycombing
dtype: int8
- name: Parenchymal scarring / fibrotic band
dtype: int8
- name: Tracheal stenosis / malacia
dtype: int8
- name: Tracheal / bronchial wall thickening
dtype: int8
- name: Bronchiectasis
dtype: int8
- name: Mucoid impaction / plugging
dtype: int8
- name: Tracheal diverticulum
dtype: int8
- name: Endotracheal tube
dtype: int8
- name: Tracheostomy tube
dtype: int8
- name: Lobar / segmental atelectasis
dtype: int8
- name: Subsegmental / linear atelectasis
dtype: int8
- name: Post-lobectomy / segmentectomy
dtype: int8
- name: Post-pneumonectomy
dtype: int8
- name: Lung transplant
dtype: int8
- name: Lungs & Airways_others
dtype: int8
- name: Pleural effusion
dtype: int8
- name: Loculated pleural effusion
dtype: int8
- name: Hemothorax
dtype: int8
- name: Chest tube / pleural drain
dtype: int8
- name: Pneumothorax
dtype: int8
- name: Tension pneumothorax
dtype: int8
- name: Pleural thickening
dtype: int8
- name: Pleural plaques
dtype: int8
- name: Pleural nodule / mass
dtype: int8
- name: Pleura_others
dtype: int8
- name: Mediastinal lymphadenopathy
dtype: int8
- name: Hilar lymphadenopathy
dtype: int8
- name: Calcified mediastinal / hilar lymph nodes
dtype: int8
- name: Anterior mediastinal mass
dtype: int8
- name: Middle / posterior mediastinal mass or cyst
dtype: int8
- name: Thymic remnant / hyperplasia
dtype: int8
- name: Esophageal wall thickening / mass
dtype: int8
- name: Hiatal hernia
dtype: int8
- name: Esophageal dilation
dtype: int8
- name: Nasogastric / orogastric tube
dtype: int8
- name: Pneumomediastinum
dtype: int8
- name: Mediastinal hematoma / fluid collection
dtype: int8
- name: Mediastinum & Hila_others
dtype: int8
- name: Cardiomegaly
dtype: int8
- name: Pericardial effusion
dtype: int8
- name: Pericardial thickening / calcification
dtype: int8
- name: Coronary artery calcification
dtype: int8
- name: Coronary stent or bypass graft
dtype: int8
- name: Thoracic aortic calcification
dtype: int8
- name: Thoracic aortic ectasia / dilation (non-aneurysmal)
dtype: int8
- name: Thoracic aortic aneurysm
dtype: int8
- name: Aortic dissection / intramural hematoma
dtype: int8
- name: Main pulmonary artery enlargement
dtype: int8
- name: Pulmonary embolism
dtype: int8
- name: Aortic valve calcification
dtype: int8
- name: Mitral annular calcification
dtype: int8
- name: Pacemaker / ICD leads
dtype: int8
- name: Central venous catheter / PICC
dtype: int8
- name: LVAD / other cardiac assist device
dtype: int8
- name: Cardiovascular_others
dtype: int8
- name: Chest wall soft tissue edema / hematoma
dtype: int8
- name: Subcutaneous emphysema
dtype: int8
- name: Chest wall mass
dtype: int8
- name: Post-thoracotomy change
dtype: int8
- name: Chest wall tumor invasion
dtype: int8
- name: Chest Wall_others
dtype: int8
- name: Acute rib fracture
dtype: int8
- name: Non-acute / healed rib fracture
dtype: int8
- name: Sternal fracture
dtype: int8
- name: Vertebral compression fracture
dtype: int8
- name: Degenerative spine changes
dtype: int8
- name: Osteolytic bone lesion
dtype: int8
- name: Osteosclerotic bone lesion
dtype: int8
- name: Mixed osteolytic-osteosclerotic lesion
dtype: int8
- name: Osteopenia
dtype: int8
- name: Scoliosis / kyphosis
dtype: int8
- name: Vertebral hemangioma
dtype: int8
- name: Postoperative spine change / hardware
dtype: int8
- name: Bones / Spine_others
dtype: int8
- name: Hepatic steatosis
dtype: int8
- name: Focal liver lesion (nodule / mass)
dtype: int8
- name: Hepatomegaly
dtype: int8
- name: Liver contour irregularity / cirrhosis features
dtype: int8
- name: Hepatic calcification
dtype: int8
- name: Cholelithiasis / gallstones
dtype: int8
- name: Post-cholecystectomy (gallbladder operated / absent)
dtype: int8
- name: Gallbladder wall thickening
dtype: int8
- name: Hydropic gallbladder / distension
dtype: int8
- name: Biliary sludge
dtype: int8
- name: Biliary stent / catheter / drain
dtype: int8
- name: Splenomegaly
dtype: int8
- name: Accessory spleen / splenule / polysplenia
dtype: int8
- name: Focal splenic lesion (nodule / mass)
dtype: int8
- name: Pancreatic mass / focal lesion
dtype: int8
- name: Pancreatic lipomatosis
dtype: int8
- name: Adrenal nodule / mass
dtype: int8
- name: Adrenal thickening / hyperplasia
dtype: int8
- name: Adrenal calcification
dtype: int8
- name: Simple renal cyst
dtype: int8
- name: Complex renal cyst / solid renal mass
dtype: int8
- name: Hydronephrosis
dtype: int8
- name: Renal calculi / nephrolithiasis
dtype: int8
- name: Renal atrophy / decreased renal size
dtype: int8
- name: Nephrectomy (kidney absent / operated)
dtype: int8
- name: Ascites
dtype: int8
- name: Pneumoperitoneum
dtype: int8
- name: Bowel wall thickening / inflammation
dtype: int8
- name: Diverticulosis
dtype: int8
- name: Omental caking / peritoneal carcinomatosis
dtype: int8
- name: Abdominal lymphadenopathy
dtype: int8
- name: Abdominal aortic aneurysm (partially imaged)
dtype: int8
- name: Abdominal aortic calcification / atherosclerosis (partially imaged)
dtype: int8
- name: IVC filter
dtype: int8
- name: Upper Abdomen_others
dtype: int8
- name: Thyroid enlargement (goiter)
dtype: int8
- name: Thyroid nodule
dtype: int8
- name: Cervical / supraclavicular lymphadenopathy
dtype: int8
- name: Neck soft tissue mass
dtype: int8
- name: Lower Neck_others
dtype: int8
- name: Breast mass / focal asymmetry
dtype: int8
- name: Post-lumpectomy / post-mastectomy change
dtype: int8
- name: Breast implant (intact or present)
dtype: int8
- name: Axillary lymphadenopathy
dtype: int8
- name: Motion artifact / suboptimal study
dtype: int8
- name: Study limitation / limited evaluation (non-motion)
dtype: int8
- name: No significant intrathoracic abnormality
dtype: int8
- name: Others_others
dtype: int8
splits:
- name: train
num_bytes: 58124167
num_examples: 20648
- name: valid
num_bytes: 4525829
num_examples: 1483
- name: test
num_bytes: 4196521
num_examples: 1483
download_size: 18881153
dataset_size: 66846517
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: valid
path: data/valid-*
- split: test
path: data/test-*
CT-RATE Findings — Chest Imaging Leaf Labels
Chest-CT findings from the CT-RATE dataset. Each row maps an original findings report → a section-structured refined version, plus a 137-label ternary multi-label annotation over a chest-imaging taxonomy.
- Rows: 23,614 unique CT-RATE findings reports (one row per report)
- Splits (report-text-level de-duplicated): train 20,648 / valid 1,483 / test 1,483
- Labels: 137 leaf labels (106 clinical + 31 other). Full taxonomy, definitions and per-split counts in
LABEL_HIERARCHY.md.
Columns
| Column | Type | Description |
|---|---|---|
original_report |
string | Original CT-RATE findings text (input) |
refined_report |
string | Section-structured / cleaned findings (target; empty for 281 rows) |
split |
string | train / valid / test |
| 137 label columns | int8 | Per-label status — see encoding below |
Label encoding (ternary)
| Value | Meaning |
|---|---|
1 |
positive |
0 |
negative |
-1 |
uncertain |
null |
not assessed for this report |
Notes
- Splits are de-duplicated across each other at the report-text level: no
original_reportorrefined_reporttext appears in more than one split. - Breast & Axilla were originally intended as their own top-level section, but because
such findings are relatively infrequent in CT-RATE they were folded into the Others
section. The individual breast/axilla leaf labels are still present (under
Others). - 281 rows have an empty
refined_report(labels are still provided). - 601 reports (2.5%) have no positive on any label.
IVC filteris present for taxonomy completeness but is entirely unlabeled (allnull— no positive, negative, or uncertain in any row).- Label names exactly match the hierarchy in
LABEL_HIERARCHY.md.
Radiologist validation (test set)
A radiologist manually reviewed 966 of the 1,483 test reports (65.1%), comparing the assigned labels against the report:
| Verdict | Reports | Share |
|---|---|---|
| Fully accepted | 857 | 88.7% |
| Imperfect / uncertain but acceptable | 60 | 6.2% |
| Failed | 49 | 5.1% |
| Reviewed | 966 | 100% |
Acceptable (accepted + borderline): 917 / 966 = 94.9%. This is a spot-check of the weak-label quality on the held-out test split, not a re-annotation — the published labels are the original pipeline output, unchanged.
Provenance & caveats
- Weak labels, not radiologist ground truth. Labels were generated by an LLM labeling pipeline from the report text (not from images), then validated against the fixed taxonomy.
refined_reportis an LLM-cleaned, section-structured rewrite oforiginal_report.
License & attribution
This dataset is a derivative of CT-RATE
and is released under CC-BY-NC-SA-4.0, the same license as CT-RATE. The report text
(original_report) originates from CT-RATE; the refined_report rewrite and the 137 leaf
labels are added by this work.
If you use this dataset you must cite the original CT-RATE paper (a requirement of the CC-BY-NC-SA attribution terms), in addition to this dataset:
@misc{hamamci2024ctrate,
title = {A foundation model utilizing chest CT volumes and radiology reports
for supervised-level zero-shot detection of abnormalities},
author = {Hamamci, Ibrahim Ethem and others},
year = {2024},
eprint = {2403.17834},
archivePrefix = {arXiv},
primaryClass = {cs.CV}
}
Please confirm the canonical CT-RATE citation on the official CT-RATE page.
Usage
from datasets import load_dataset
ds = load_dataset("chest2vec/chest2vec_labels")
ds["test"][0]["original_report"]
[k for k, v in ds["test"][0].items() if v == 1] # positive labels for the first test report