Section: Software

Medical Image Extractor

Participants : Stéphane Lanteri, Julien Wintz [SED team] .

Medical Image Extractor provides functionalities needed to extract meshes from labeled MR or PET-CT medical images. It puts the emphasis on consistency, by generating both boundary surfaces, and volume meshes for each label (ideally identifying a tissue) of the input image, using the very same tetreahedrization. As this process requires user interaction, images and meshes are visualized together with tools allowing navigation and both easy and accurate refinement of the generated meshes, that can then be exported to serve as an input for other tools, within a multidisciplinar software toolchain. Using both DTK and NUM3SIS SDKs, Medical Image Extractor comes within NUM3SIS' framework. Using cutting edge research algorithms developed by different teams at Inria, spread among different research topics, namely, visualization algorithms from medical image processing, meshing algorithms from algorithmic geometry, it illustrates the possibility to bridge the gap between software that come from different communities, in an innovative and highly non invasive development fashion.

Figure 2. Medical Image Extractor tool graphical user interface.