Optimal scan design for tomographic imaging of time varying distributions

The authors investigate reductions in the required temporal sampling rates for tomographic imaging of time-varying distributions. These rates can be lowered significantly if the temporal variation of the distribution being imaged is spatially localized, as is the case for some medical applications. The reduction in the temporal sampling rate is achieved by a nonconventional angular sampling order, minimizing a worst-case aliasing error. Computer simulation results demonstrating good reconstruction with temporal undersampling by a factor of two and three are shown