README.md

---
license:
  - cc-by-4.0
  - cc-by-nc-4.0
pretty_name: FleXray Data
viewer: false
task_categories: [image-segmentation]
size_categories: [100K<n<1M]
tags: [medical, x-ray, segmentation, anatomy]
---

# FleXray data release

- Website: [FleXray project page](https://flexray.csail.mit.edu/)
- Paper: [FleXray: Universal Clinical X-ray Segmentation](https://arxiv.org/abs/2609.26756)
- Code: [github.com/VictorButoi/FleXray](https://github.com/VictorButoi/FleXray)

Training and evaluation data for **FleXray**, a pan-anatomy X-ray segmentation model.
This repository holds every real X-ray source whose license permits redistribution,
repackaged as image/mask pairs with `fxr-dataset` manifests and dataset-native
label names that FleXray maps into its common protocol (CC BY 4.0), plus two more parts of the data described in the paper:

- **FluXray** (synthetic, CC BY-NC 4.0): `FluXray/` — 138,063 generatively edited
  digitally reconstructed radiographs rendered from the 1,597 MOOSE CTs at 90 poses
  each, quality-filtered, with exact overlapping masks in 63 channels (62 structures plus background).
  Shipped as the training database itself (`FluXray/thunder_dbs/1.0/data.mdb`, an LMDB
  of float16 256 x 256 images and 63-channel binary masks that `flexray` reads directly)
  together with `samples.csv` (pose, split, MOOSE subject and per-sample license for
  every image), `protocol.yml` (label order and mask thresholds), and the
  `filter_*.csv` quality-control scores and thresholds. `FluXray/README.md` documents
  every file; `FluXray/LICENSE` summarizes the license.
- **MURA forearm/humerus annotations** (masks only, CC BY 4.0): bundled with the
  FleXray GitHub repository and mirrored here under `mura_forearm_humerus_annotations/`

Licenses therefore differ by folder: the redistributed real X-ray folders, `splits/`,
and the MURA annotations are CC BY 4.0; `FluXray/` is CC BY-NC 4.0 as a collection,
with each image inheriting the license of its MOOSE source site.

## Data access

This release contains **4,232 real image/mask pairs** across seven datasets,
**138,063 FluXray samples**, and **100 MURA annotation masks** whose source
images must be obtained separately. Browse the dataset folders below or use
the [download and packaging instructions](#usage).

The automatic Hugging Face image-folder viewer is disabled because it only
recognizes a subset of the images and does not represent the paired masks or
the FluXray database. Use each `dataset.yml` for image/mask paths and split
assignments; FluXray's `samples.csv` records its samples and splits.

- [Real image/mask datasets](#redistributed-datasets-cc-by-40): PNG images,
  PNG or NPY masks, and a packaging manifest per dataset.
- [FluXray](FluXray/README.md): a ThunderDB database plus protocol and sample
  metadata. The database download is about **18.8 GB**.
- [MURA annotations](mura_forearm_humerus_annotations/README.md): masks and a
  script that joins them to your own MURA download.
- [Splits and exclusions](splits/README.md): source-relative records for
  reconstructing the published partitions.

## Redistributed datasets (CC BY 4.0)

| Dataset | Role in paper | Samples (train/val/test) | Labels | Source |
|---|---|---|---|---|
| [HandBones](HandBones/README.md) | training | 93 (65/15/13) | carpals, phalanges (distal/intermediate/proximal), metacarpals, radii, ulnae | https://universe.roboflow.com/boneage-x90qt/-hand-bones-mdjkr |
| [FootBones](FootBones/README.md) | training | 571 (400/86/85) | metatarsals (1-5), toes | https://universe.roboflow.com/monchbot1/foot_op |
| [ElbowLat](ElbowLat/README.md) | evaluation | 601 (419/91/91) | humeri, radii, ulnae | https://universe.roboflow.com/ionspace/elbow_lat-lnn0s-pmycd |
| [HipRay](HipRay/README.md) | evaluation | 139 (97/21/21) | femurs, hips | https://data.mendeley.com/datasets/zm6bxzhmfz/1 |
| [LowerLimbs](LowerLimbs/README.md) | evaluation | 56 (39/9/8) | femurs, tibiae, fibulae | https://universe.roboflow.com/orthopedicstitching/bone-identifier-1rey5 |
| [MTDDH](MTDDH/README.md) | evaluation | 905 (632/138/135) | ilium, pubis, ischium, femoral head, femur | https://doi.org/10.57760/sciencedb.24372 |
| [BTXRD](BTXRD/README.md) | evaluation (finetuning) | 1867 (1305/281/281) | tumor | https://doi.org/10.6084/m9.figshare.27865398 |

Each `<Dataset>/` folder contains `dataset.yml`, `images/` (16-bit PNG), `labels/`
(PNG index maps or NPY channel masks), a `README.md` with preprocessing and label details,
and a `LICENSE` with attribution. Images were min-max normalized per image, zero-padded to
a square and resized to 256 x 256; splits are the ones used in the paper.

## Datasets referenced by pointer only

These sources are used by FleXray but not redistributed here. Their FleXray label
specifications (native label names, protocol aliases and drops) are shipped with the
`flexray` package under `fxr/configs/datasets/<Name>.yml`.

| Dataset | Role | Why not redistributed | Where to get it |
|---|---|---|---|
| MOOSE / ENHANCE-PET 1.6k | training (CT, DRR rendering) | CT sources are not redistributed; already public | https://registry.opendata.aws/enhance-pet-1-6k/ |
| ElbowCT | training (CT) | CT sources are not redistributed | https://figshare.com/articles/dataset/3D_models_of_elbow_joints_along_with_corresponding_CT_data_from_Chinese_individuals/28245599 |
| PedsCT | training (CT) | CT sources are not redistributed | https://www.cancerimagingarchive.net/collection/pediatric-ct-seg/ |
| HaN-Seg | training (CT) | CC BY-NC-ND 4.0 (no derivatives) | https://han-seg2023.grand-challenge.org/ |
| RSNAFrac | training (CT) | Kaggle competition rules forbid redistribution | https://www.kaggle.com/competitions/rsna-2022-cervical-spine-fracture-detection/ |
| Shoulder-CT | training (CT) | no license granted by the uploader | https://www.kaggle.com/datasets/syxlicheng/automatically-transform-ct-datasets-into-drrs |
| MURA | training (images) | Stanford Research Use Agreement | https://stanfordmlgroup.github.io/competitions/mura/ (our masks: see above) |
| AASCE | evaluation | license undetermined | https://aasce19.github.io/ |
| DarwinCVD19 | evaluation | mixed per-image image licenses | https://darwin.v7labs.com/v7-labs/covid-19-chest-x-ray-dataset |
| DeepFluoro | evaluation | CC BY-NC 4.0; already hosted on Hugging Face | https://huggingface.co/datasets/eigenvivek/xvr-data |
| RAM-W600 | evaluation | CC BY-NC-SA 4.0; already hosted on Hugging Face | https://huggingface.co/datasets/TokyoTechMagicYang/RAM-W600 |
| VinDr-Rib | evaluation | signed data use agreement required | https://vindr.ai/ribcxr |
| PedsTorso | evaluation | upstream project no longer available; license cannot be verified | https://universe.roboflow.com/monchbot1/thoracoabdominal |

## Splits and exclusions

`splits/<Dataset>/splits.csv` lists the train/val/test assignment of every image FleXray
trained or evaluated on, and `splits/<Dataset>/exclusions.csv` lists every image removed
during quality control together with the reason, for all fifteen real X-ray sources above
(redistributed or not). Paths are relative to each source's original download, so the
paper's partitions can be rebuilt exactly; see `splits/README.md` for the schema.

## Usage

Dataset packaging and training require the training extra:

```bash
python -m pip install "flexray[train]"
```

Download only the dataset you need. For example, fetch HipRay and its split
records without downloading the FluXray database:

```python
from huggingface_hub import snapshot_download

snapshot_download(
    repo_id="VictorButoi/flexray-data",
    repo_type="dataset",
    local_dir="./flexray-data",
    allow_patterns=["README.md", "HipRay/*", "splits/HipRay/*"],
)
```

Validate and pack it into the layout consumed by FleXray:

```bash
fxr-dataset validate ./flexray-data/HipRay/dataset.yml
fxr-dataset pack ./flexray-data/HipRay/dataset.yml /data/flexray/HipRay
export XRAY_DATAPATH=/data/flexray
```

Repeat with another folder name to package another real X-ray dataset.
To include FluXray, download its approximately 18.8 GB database and sidecars
into the same download directory:

```python
snapshot_download(
    repo_id="VictorButoi/flexray-data",
    repo_type="dataset",
    local_dir="./flexray-data",
    allow_patterns=["FluXray/*"],
)
```

```bash
export GENERATED_DATAPATH="$PWD/flexray-data"
```

FluXray is already packaged at `FluXray/thunder_dbs/1.0/`. Its historical
`GENERATED_DATAPATH` variable selects the storage root; it is configured as
an `Xray` source:

```yaml
# training config excerpt
data:
  Xray:
    HipRay: {}
    FluXray: {version: "1.0"}
```

See [dataset documentation](https://github.com/VictorButoi/FleXray/blob/main/docs/datasets.md)
and [training configuration](https://github.com/VictorButoi/FleXray/blob/main/docs/training.md)
for complete examples. Source datasets retain their roles in the paper;
adding one to a training config does not change its published evaluation split.

## Citation

If you use either **FluXray** or our **MURA annotations**, please cite the
FleXray paper:

```bibtex
@misc{butoi2026flexray,
      title={FleXray: Universal Clinical X-ray Segmentation},
      author={Victor Ion Butoi and Vivek Gopalakrishnan and John V. Guttag and Adrian V. Dalca and Neel Dey},
      year={2026},
      eprint={2609.26756},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2609.26756},
}
```

**If you use any of the other datasets, please cite the original dataset sources
and comply with their copyright and license terms.** Citations and licensing
details are listed in the `README.md` and `LICENSE` files within each dataset folder.