Optimization of Basis Functions for Both Reconstruction and Visualization

AbstractAlgebraic Reconstruction Techniques (ART) for the reconstruction of distributions from projections have yielded improvements in diverse fields such as medical imaging and electron microscopy. An important property of these methods is that they allow the use of various basis functions. Recently spherically symmetric functions (blobs) have been introduced as efficacious basis functions for reconstruction. However, basis functions whose parameters were found to be appropriate for use in reconstruction are not necessarily good for visualization. We propose a method of selecting blob parameters for both reconstruction and visualization