RWTHmediTEC / VSDFullBodyBoneModels

Surface models of bones created from CT datasets of the open source VSDFullBody database hosted at www.smir.ch.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

VSDFullBodyBoneModels

3D surface models of the bones of the lower body created from CT datasets of the open access VSDFullBody collection. subject-selection

How to cite

Publication

Fischer, M. C. M. Database of segmentations and surface models of bones of the entire lower body created from cadaver CT scans. Sci. Data 10, 763; 10.1038/s41597-023-02669-z (2023).

Releases

  • v3.0 DOI The database contains the bones of the lower extremities of 30 subjects. One duplicate subject (z024) was removed from the database that was part of the previous versions.
  • v2.0 DOI additionally contains the femora and manually selected femoral landmarks of five experienced raters.
  • v1.0 DOI contains only the bones of the pelvis of 20 subjects of the VSDFullBody database. Additionally, it contains manually selected pelvic landmarks of five experienced raters.

Usage

In the Bones folder a MATALB MAT file is stored for each subject containing the triangle meshes of the bones. Clone with the recursive option to get the submodules. Run the MATLAB or Python example script to plot one of the subjects: plotBoneModels_example.m or plotBoneModels_example.py.

Segmentation and reconstruction process

The surface of each bone was semi-automatically reconstructed by thresholding using 200 Hounsfield units as the lower threshold and the maximum Hunsfield unit value present in the volume data as the upper threshold. The thresholding was followed by a manual post-processing procedure using the software 3D Slicer (slicer.org) with default smoothing settings. The bones were manually segmented at the joint spaces if necessary. Subsequently, holes in the outer surface of the bones were manually closed. The reconstructions were exported as mesh files in the PLY format (see 'Related data') and imported into MATLAB using a conservative decimation and remeshing procedure. The decimator restricted the Hausdorff distance between input and output mesh to 0.05 mm. The adaptive remesher permitted a maximum deviation of 0.05 mm from the input mesh with a minimum edge length of 0.5 mm and a maximal edge length of 100 mm. Number of iterations was set to 10 and back-projection enabled. The decimator and remesher are plugins of the software OpenFlipper (openflipper.org).

Related data

  • DOI CT volume data, segmentations, reconstructions and raw PLY mesh files of each subject linked by a project file (MRML scene file) that can be opened with 3D Slicer (slicer.org).
  • DOI Mirror of the full VSDFullBody database as hosted originally by Michael Kistler at smir.ch.

Licenses

  • License: CC BY-NC-SA 4.0 for MAT and XLSX files.
  • License: EUPL v1.2 for MATLAB and Python source code.

Related publications

  • v2.0 was used in: Fischer, M. C. M. et al. A robust method for automatic identification of femoral landmarks, axes, planes and bone coordinate systems using surface models. Sci. Rep. 10, 20859; 10.1038/s41598-020-77479-z (2020).
  • v1.0 was used in: Fischer, M. C. M. et al. A robust method for automatic identification of landmarks on surface models of the pelvis. Sci. Rep. 9, 13322; 10.1038/s41598-019-49573-4 (2019).

About

Surface models of bones created from CT datasets of the open source VSDFullBody database hosted at www.smir.ch.


Languages

Language:MATLAB 95.4%Language:Python 4.6%