Three of the most important forms of medical imaging use reconstruction from erojections - catscan, pet, and spect. We have been primarily on catscan but the results relate to all three. Over the last few years we have worked out mathematically methods which give significantly better reconstructions and are tolerant of irregular source and detector geometries and have verified these in simulations. Now we are trying to use our programs on data from commercial catscans, but computation speed and memory space of our local machines are inadequate. The short run objective is simply better catscan images, but long run goals include portable 3-D imagers and imagers for the heart in motion.
Showing the most recent 10 out of 292 publications