HAND MRI Dataset (PIANO)

The is the dataset used in PIANO: A Parametric Hand Bone Model from Magnetic Resonance Imaging and NIMBLE: A Non-rigid Hand Model with Bones and Muscles.

To learn more, please visit project website at PIANO and NIMBLE.

For comments and questions, please email us at: Yuwei Li (liyw@shanghaitech.edu.cn)


MRI Dataset Part 1 (50 vols)

  1. MRI raw volume [Google Drive]
  2. Bone mask volume [Google Drive]
  3. 3D joint annotation (in physical space) [Google Drive]

Annotation Extension

  1. Muscle mask volume (From NIMBLE) [Google Drive]

Useful code - mask2mesh.py

  • Generate mesh from volume mask
mri_mask = "00001_bonemuscle.nii"
mri_mask_vol = sitk.ReadImage(mri_mask)
bone_mesh = generate_seg_mesh(mri_mask_vol, 1)
muscle_mesh = generate_seg_mesh(mri_mask_vol, 2)

bone_mesh.export("bone.obj")
muscle_mesh.export("muscle.obj")
  • Naive fine-grained bone mask
joints_file = "00001_joints.txt"
joints3d = np.loadtxt(joints_file)
semantic_bonemesh = finegrained_bone(joints3d, bone_mesh)
semantic_bonemesh.export("sbone.obj")
  • Automatic surface segmentation
mri_raw = "00001.nii"
surf_mask_vol = naive_seg(sitk.ReadImage(mri_raw))
surf_mesh = generate_seg_mesh(surf_mask_vol)
surf_mesh.export("surf.obj")

Joint ID


If you find this data useful for your research, consider citing:

@inproceedings{li2021piano,
  title     = {PIANO: A Parametric Hand Bone Model from Magnetic Resonance Imaging},
  author    = {Li, Yuwei and Wu, Minye and Zhang, Yuyao and Xu, Lan and Yu, Jingyi},
  booktitle = {Proceedings of the Thirtieth International Joint Conference on
               Artificial Intelligence, {IJCAI-21}},
  editor    = {Zhi-Hua Zhou},
  pages     = {816--822},
  year      = {2021},
  month     = {8},
  note      = {Main Track},
  doi       = {10.24963/ijcai.2021/113},
  url       = {https://doi.org/10.24963/ijcai.2021/113}
}

@misc{li2022nimble,
  title         = {NIMBLE: A Non-rigid Hand Model with Bones and Muscles},
  author        = {Li, Yuwei and Zhang, Longwen and Qiu, Zesong and Jiang, 
                   Yingwenqi and Zhang, Yuyao and Li, Nianyi and Ma, Yuexin 
                   and Xu, Lan and Yu, Jingyi},
  year          = {2022},
  eprint        = {2202.04533},
  archiveprefix = {arXiv},
  primaryclass  = {cs.CV}
}