Access to a large database of high quality images is vital for the development of valid models. The NCSSM is actively using the Virtualskeleton database as a data hub. Currently, more than 100 high quality full-body CT-scans are available.


Within the Co-Me scientific network, the need for a image repository with focus on the statistical shape model community was identfied. Taking the lead on the development, the Institute for Surgical Technology and Biomechanic built in collaboration with the University of Applied Sience in Biel the Virtual Skeleton Database as a Webrepository which is tailored to the particular need of scientist working with medical image data.


Statistical shape models are widely used in biomedical research which require the access to a large training datasets. The Virtual Skeleton Database is as a centralized storage system where the data necessary to build statistical shape models can be stored and are shared (article JMIR). The repository is tailored to the needs of the statistical shape modeling community by offering storage and organisation of data required for model building. Dicom, ITK based and statistical models in the statismo format are stored on the Virtual Skeleton Database.



Storage of Dicom, Analyze, Niftii, MetaImage and HDF5 files.

VSD data types


Organize the data into folder and sharing thereof

folder structure of the VSD


Semantically driven search option to find datasets containing the required anatomical structures.

ontology representation of the limb in the FMA

Related data

Linking of datasets to find related data

linking of related data


Demographics of the available full body dataset
GenderCT (n) Age (y)Height (cm)Weight (kg)BMI (kg/m^2)
Female 33 mean (SD)56 (19) 167 (6.41) 61.72 (16.55) 22.18 (5.54)
range 19-95 152-18932-90 11.54-31.89
Male67 mean (SD)54 (21) 177 (8.35) 78.52 (15.92)25.08 (4.27)
range 17-95 155-195 49-11115.67-35.83
Total 100 mean (SD) 54 (20) 173 (9.05) 72.97 (17.90)24.13 (4.90)
range 17-95 152-195 32-111 11.54-35.83

Browse the VSD