Tomographic reconstruction from limited data

Introduction At present, image reconstruction in micro-CT imaging is typically performed using the FDK algorithm. This algorithm, which is similar to the Filtered Backprojection algorithm for singleslice tomography, is capable of computing accurate reconstructions in a relatively short time, provided that the acquired dataset satisfies certain requirements: (i) a large number of projection images must be available; (ii) the angular step between consecutive projections must be approximately constant. (iii) the projections should not be too noisy (iv) the object must be contained within the field-of-view (FOV) of the detector. Violating these requirements typically results in degraded reconstructions, which can be difficult to interpret and analyze.