---
license: other
license_name: nlst-data-use-agreement
license_link: https://cdas.cancer.gov/learn/nlst/images/NLST_CDAS_Participant_Data_Use_Agreement.pdf
language:
- en
pretty_name: OpenM3Chest Labels v2
size_categories:
- 10K<n<100K
task_categories:
- visual-question-answering
- image-classification
tags:
- medical
- radiology
- chest-ct
- ct-scan
- nlst
- openm3chest
- cardiology
- oncology
dataset_info:
- config_name: CVD_diagnosis
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: int64
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 1856650
    num_examples: 1500
  - name: test
    num_bytes: 4629904
    num_examples: 3759
  download_size: 3963880
  dataset_size: 6486554
- config_name: CVD_mortality
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: int64
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 1874650
    num_examples: 1500
  - name: test
    num_bytes: 4675012
    num_examples: 3759
  download_size: 3963506
  dataset_size: 6549662
- config_name: chest_abn_54
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: int64
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 443639
    num_examples: 348
  - name: test
    num_bytes: 16949120
    num_examples: 13346
  download_size: 10492475
  dataset_size: 17392759
- config_name: chest_abn_55
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: int64
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 456515
    num_examples: 348
  - name: test
    num_bytes: 16629981
    num_examples: 12722
  download_size: 10512788
  dataset_size: 17086496
- config_name: chest_abn_56
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: int64
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 671231
    num_examples: 348
  - name: test
    num_bytes: 25332920
    num_examples: 13168
  download_size: 19210477
  dataset_size: 26004151
- config_name: chest_abn_57
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: int64
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 550475
    num_examples: 348
  - name: test
    num_bytes: 21035508
    num_examples: 13347
  download_size: 14700073
  dataset_size: 21585983
- config_name: chest_abn_58
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: int64
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 400139
    num_examples: 348
  - name: test
    num_bytes: 15237836
    num_examples: 13303
  download_size: 8742055
  dataset_size: 15637975
- config_name: chest_abn_59
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: int64
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 393179
    num_examples: 348
  - name: test
    num_bytes: 13263221
    num_examples: 11787
  download_size: 7555441
  dataset_size: 13656400
- config_name: chest_abn_61
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: int64
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 599195
    num_examples: 348
  - name: test
    num_bytes: 19589111
    num_examples: 11402
  download_size: 14304138
  dataset_size: 20188306
- config_name: lung_cancer_risk
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: string
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 5544344
    num_examples: 2000
  - name: test
    num_bytes: 21597086
    num_examples: 10308
  download_size: 11905555
  dataset_size: 27141430
- config_name: nodule_attenuation
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: int64
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 404321
    num_examples: 336
  - name: test
    num_bytes: 13942312
    num_examples: 11620
  download_size: 7736241
  dataset_size: 14346633
- config_name: nodule_location
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: int64
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 394577
    num_examples: 336
  - name: test
    num_bytes: 13605332
    num_examples: 11620
  download_size: 6507433
  dataset_size: 13999909
- config_name: nodule_margin
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: int64
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 385841
    num_examples: 336
  - name: test
    num_bytes: 13303212
    num_examples: 11620
  download_size: 6529548
  dataset_size: 13689053
- config_name: nodule_presence
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: int64
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 543859
    num_examples: 486
  - name: test
    num_bytes: 17317435
    num_examples: 15519
  download_size: 9570035
  dataset_size: 17861294
- config_name: nodule_size
  features:
  - name: keys
    dtype: string
  - name: pids
    dtype: string
  - name: clinical_data
    dtype: string
  - name: questions
    list: string
  - name: labels
    dtype: int64
  - name: answer_dict
    dtype: string
  splits:
  - name: train
    num_bytes: 356945
    num_examples: 336
  - name: test
    num_bytes: 12231085
    num_examples: 11552
  download_size: 5578301
  dataset_size: 12588030
configs:
- config_name: CVD_diagnosis
  data_files:
  - split: train
    path: CVD_diagnosis/train-*
  - split: test
    path: CVD_diagnosis/test-*
- config_name: CVD_mortality
  data_files:
  - split: train
    path: CVD_mortality/train-*
  - split: test
    path: CVD_mortality/test-*
- config_name: chest_abn_54
  data_files:
  - split: train
    path: chest_abn_54/train-*
  - split: test
    path: chest_abn_54/test-*
- config_name: chest_abn_55
  data_files:
  - split: train
    path: chest_abn_55/train-*
  - split: test
    path: chest_abn_55/test-*
- config_name: chest_abn_56
  data_files:
  - split: train
    path: chest_abn_56/train-*
  - split: test
    path: chest_abn_56/test-*
- config_name: chest_abn_57
  data_files:
  - split: train
    path: chest_abn_57/train-*
  - split: test
    path: chest_abn_57/test-*
- config_name: chest_abn_58
  data_files:
  - split: train
    path: chest_abn_58/train-*
  - split: test
    path: chest_abn_58/test-*
- config_name: chest_abn_59
  data_files:
  - split: train
    path: chest_abn_59/train-*
  - split: test
    path: chest_abn_59/test-*
- config_name: chest_abn_61
  data_files:
  - split: train
    path: chest_abn_61/train-*
  - split: test
    path: chest_abn_61/test-*
- config_name: lung_cancer_risk
  data_files:
  - split: train
    path: lung_cancer_risk/train-*
  - split: test
    path: lung_cancer_risk/test-*
- config_name: nodule_attenuation
  data_files:
  - split: train
    path: nodule_attenuation/train-*
  - split: test
    path: nodule_attenuation/test-*
- config_name: nodule_location
  data_files:
  - split: train
    path: nodule_location/train-*
  - split: test
    path: nodule_location/test-*
- config_name: nodule_margin
  data_files:
  - split: train
    path: nodule_margin/train-*
  - split: test
    path: nodule_margin/test-*
- config_name: nodule_presence
  data_files:
  - split: train
    path: nodule_presence/train-*
  - split: test
    path: nodule_presence/test-*
- config_name: nodule_size
  data_files:
  - split: train
    path: nodule_size/train-*
  - split: test
    path: nodule_size/test-*
---

# OpenM3Chest Labels v2

Stratified label subset of [OpenM3Chest](https://openm3chest.org) for LoRA fine-tuning of [MedGemma 1.5-4B-IT](https://huggingface.co/google/medgemma-1.5-4b-it).

**CT scan images (NPY):** [UngLong/openm3chest-npy-v2](https://huggingface.co/datasets/UngLong/openm3chest-npy-v2)

## Dataset Summary

| Agent | Tasks | Train samples | Test samples |
|---|---|---|---|
| **Radiology** | chest_abn_54–61, nodule_presence/location/attenuation/margin/size | ~350/task | full |
| **Cardiology** | CVD_diagnosis, CVD_mortality | 1,500/task | 3,759/task |
| **Oncology** | lung_cancer_risk | 2,000 | 10,308 |

**Training subset**: stratified sampling — binary tasks ~50/50, all categorical classes covered.  
**Test split**: full original test set (unfiltered) for proper evaluation.

## Using the Dataset

```python
from datasets import load_dataset

# Radiology — chest abnormality detection
ds = load_dataset("UngLong/openm3chest-labels-v2", "chest_abn_54")
print(ds["train"][0])

# Cardiology — CVD diagnosis
ds = load_dataset("UngLong/openm3chest-labels-v2", "CVD_diagnosis")

# Oncology — lung cancer risk
ds = load_dataset("UngLong/openm3chest-labels-v2", "lung_cancer_risk")
```

## Schema

Each record contains:

| Field | Type | Description |
|---|---|---|
| `keys` | string | SeriesInstanceUID — maps to `{pids}/{keys}.npy` in NPY repo |
| `pids` | string | Patient ID |
| `clinical_data` | string | JSON-serialized clinical information |
| `questions` | list[string] | 10 candidate questions for this task |
| `labels` | int64 / string | Task label (see below) |
| `answer_dict` | string | JSON-serialized dict mapping label index → human-readable answer text |

## Tasks & Labels

### Radiology

| Config | Label type | Classes |
|---|---|---|
| `chest_abn_54` | int (0/1) | 0 = No, 1 = Yes |
| `chest_abn_55` | int (0/1) | 0 = No, 1 = Yes |
| `chest_abn_56` | int (0/1) | 0 = No, 1 = Yes |
| `chest_abn_57` | int (0/1) | 0 = No, 1 = Yes |
| `chest_abn_58` | int (0/1) | 0 = No, 1 = Yes |
| `chest_abn_59` | int (0/1) | 0 = No, 1 = Yes |
| `chest_abn_61` | int (0/1) | 0 = No, 1 = Yes |
| `nodule_presence` | int (0/1) | 0 = No nodule, 1 = Nodule present |
| `nodule_location` | int (0–4) | 0=RUL, 1=RML, 2=RLL, 3=LUL, 4=LLL |
| `nodule_attenuation` | int (0–2) | 0=Solid, 1=Ground Glass, 2=Others |
| `nodule_margin` | int (0–3) | 0=Spiculated (Stellate), 1=Smooth, 2=Poorly defined, 3=Unable to determine |
| `nodule_size` | int (0–5) | 0=≤4mm, 1=4–6mm, 2=6–8mm, 3=8–15mm, 4=15–30mm, 5=>30mm |

### Cardiology

| Config | Label type | Classes |
|---|---|---|
| `CVD_diagnosis` | int (0/1) | 0 = No, 1 = Yes |
| `CVD_mortality` | int (0/1) | 0 = Low risk, 1 = High risk |

### Oncology

| Config | Label type | Description |
|---|---|---|
| `lung_cancer_risk` | string (JSON) | Dict with `y` (bool), `y_seq` (6-year sequence), `y_mask` |

```python
import json
record = ds["train"][0]
label = json.loads(record["labels"])
print(label["y"])      # True/False — cancer within follow-up
print(label["y_seq"])  # [0,0,0,1,0,0] — year-by-year occurrence
```

## Loading NPY Files

```python
import json
import numpy as np
from huggingface_hub import hf_hub_download

record = ds["train"][0]
npy_path = hf_hub_download(
    repo_id   = "UngLong/openm3chest-npy-v2",
    filename  = f"{record['pids']}/{record['keys']}.npy",
    repo_type = "dataset",
)
volume = np.load(npy_path)  # shape: [Z, H, W], dtype: float16
```

## Sequential Inference (Radiology)

The Radiology agent uses **sequential inference**:

```
Step 1 — Screening:  chest_abn_54–61 + nodule_presence
Step 2 — If nodule_presence = Yes:
          nodule_location, nodule_attenuation, nodule_margin, nodule_size
```

Training data reflects this:
- **Pool A** (198 scans): all 12 tasks, `nodule_presence = 1`
- **Pool B** (150 scans): screening tasks only, `nodule_presence = 0`

## Source

Built from [OpenM3Chest](https://openm3chest.org) — NLST-based chest CT dataset with clinical labels.  
CT images available via [IDC (Imaging Data Commons)](https://imaging.datacommons.cancer.gov/).