This talk describes the project I worked on at the workshop at
the IMA (Institute of Mathematics and its Applications) from July 19
to July 28, 2000.  During a 10 day stay we were working on solving Fritz
John's ultrahyperbolic PDE for volumetric computerized tomography. The
most interesting aspect of this project was to apply results from
Numerical Analysis to a real problem from industry and to explore the
difficulties that arose from this, like the limit of computer memory
and run time.  At my talk I will go more into detail about these aspects.