Implementation of alternating minimization algorithms for fully 3D CT imaging

摘要: Algorithms based on alternating minimization (AM) have recently been derived for computing maximum-likelihood images in transmission CT, incorporating accurate models of the transmission-imaging process. In this work we report first fully three-dimensional implementation these algorithms, intended use with multi-row detector spiral CT systems. The most demanding portion computations, projections and backprojections, are calculated using a precomputed lookup table containing discretized version point-spread function that maps between measurement image spaces. This accounts details scanner. A cylindrical phantom spherical inserts known attenuation was scanned Siemens Sensation 16, which employed rapid, acquisition mode 16 active rows. These data were downsampled reconstructed monoenergetic our AM algorithm. estimated coefficients closely match cylinder embedded objects. We investigating methods further accelerating computations by combination techniques reduce time each iteration increase convergence log-likelihood from to iteration.