All the software and source code on studierfenster.at (or studierfenster.tugraz.at) is licensed under the GNU General Public License (GPL) version 3. The rights of third-party components are hold by their producers, with their respective license.

Text and multimedia material, such as images and videos, are released under the Creative Common Attribution-Sharealike 3.0 Unported license (CC-BY-SA) and the GNU Free Documentation License (GFDL) (unversioned, with no invariant sections, front-cover texts, or back-cover texts).

Text, multimedia and source code can be re-used and modified, given that the original sources are cited. Namely:

• Studierfenster.at
• Graz University of Technology
• Weber M., Wild D., Wallner J. and Egger J. “A Client/Server based Online Environment for the Calculation of Medical Segmentation Scores”. IEEE Engineering in Medicine and Biology Conference, pp. 3463-3467, 2019.
• Prutsch A., Pepe A. and Egger J. “Design and Development of a Web-based Tool for Inpainting of Dissected Aortae in Angiography Images“. The 24th Central European Seminar on Computer Graphics (CESCG), pp. 1-8, May 2020.

The “Dicom Browser” allows parsing a local folder with DICOM files. Afterwards, the whole folder can be converted to compressed .NRRD files and downloaded as a single .zip file. It is also possible to select specific Studies or Series, and only convert these.

We chose the compressed .NRRD format, because it contains only the image data. All patient DICOM tags will be removed. The compressed .NRRD files can then be opened with the “Medical 3D Viewer”, for visualization in 2D, 3D and further processing.

Note: The “Dicom Browser” is completely client-sided. No data is transferred to our server during the conversion!

Here the Dice Similarity Coefficient (DSC) and the Hausdorff Distance (HD) between two segmentation masks (in .nrrd format) can be calculated.

The File Converter, converts a medical volume file (e.g. a non-compressed NRRD) to a (compressed/binary) NRRD. The compressed NRRD can then be used for example with our Medical 3D Viewer.

The 3D Skull Reconstruction performs an automatic cranial defect reconstruction via our cloud platform with Deep Learning. The main steps are:

• Choosing a defected skull dataset
• Converting the image dataset (NRRD) to a mesh
• Choosing a defected skull dataset again
• Perform the skull reconstruction with Deep Learning on our sever
• Perform the implant generation (subtracting the defected skull from the reconstructed)

Afterwards, the implant can be downloaded and opened for example in MeshLab for further post-processing, like removing noise.

The Medical VR Viewer allows viewing medical data in Virtual Reality (VR), for example with the Google Cardboard or the HTC Vive (via the WebVR App). However, you need to convert your (medical) data (e.g., in NRRD format) to the VTI (.vti) format. This can be done with ParaView.

The 3D Face Reconstruction module allows registering two facial meshes.

The Medical 3D Viewer allows viewing of medical data (CT, MRI, etc.) in 2D and 3D (Volume Rendering) directly in a standard Web Browser. More enhanced functionalities of the 3D Viewer are the manual slice-by-slice outlining (segmentation) of structures in medical images (brush and contour modes available) and the manual placing of Landmarks in medical images. More sophisticated features of the 3D Viewer are the inpainting of Aortic Dissections with a Generative Adversarial Network (GAN) and the automatic aortic landmark detection with Deep Learning in CTA (Computed Tomography Angiography) scans.