The use of modeling as an alternative magnetic resonance image technique

M.R. Smith et al. (see IEEE Trans. Biomed. Imaging, vol.BM5-3, no.3, p.132-139, 1986) and E.M. Haacke et al. (see IEEE Trans. Acoust., Speech, and Signal Proc., to appear) have reported success with modeling techniques that overcome the problem of reduced data sets in MR data (magnetic resonance) imaging. The present authors detail the assumptions made by the Smith and Haacke reconstruction methods and their success in modeling MR data. By taking into account experimental phase distortions, they generate an improved model that makes the basic assumptions of the above models more realistic. The authors present the theory behind determining phase corrections from the gathered data and show the effect of applying the modified algorithms. Experimental results confirm that the correction of the phase distortions reduces the required model order of both the Smith and Haacke methods.<
>