A few pictures from my 2001 PhD thesis (feel free to visit my medical imaging page too).

Assorted screenshots

Overview of the whole process.

3D reconstructed polygonal iso-surfaces. Left: transparent skin and bone. Right: bone tissues (139K points, 280K polygons) and 4 contoured slices (blue) with labels at randomly selected locations (yellow). Rendered in real-time.

3D reconstruction. Left: 5 randomly selected CT slices (256x256), stacked. Center: reconstructed skin model (36K points, 72K polygons) and same CT slices. Right: bone reconstruction.

Latest interface (Windows). 2D/3D viewports. Left head : opacity texture is used to interactively set a specific viewing area through the skin (skin color proportional to the distance to the center of the "hole"). Right head : same model, with 360° texture reconstructed from photographs (see texture on Page 3 )

Earlier interface (Windows, Linux, Solaris). 2D/3D viewports. First attempt to compute 3D cephalometry on the polygonal bone reconstruction.

User interface. Left: setting up a cutting plane to perform an osteotomy. Right: moving (excessively) a previously cut part. The 3D model is rendered at interactive frame-rate.

User interface. Objects are listed in the left part of the interface. Left: hard and soft tissues, transparency and cut-off. Right: interactive development and testing through the scripting console (setting up a convex hull).

Sébastien Barré