Constrained ARMA method for magnetic resonance imaging

Quantitative measurement on medical phantoms has shown that the transient error reconstruction approach (TERA), successfully reintroduces 40% to 50% of truncated energy. This indicates a considerable increase in resolution and reduction in image artifacts. The algorithm is not as successful on normal medical data. This paper suggests an approach to constrain the modeling algorithm to include a priori information and improve the modeling. Preliminary results are presented for a constraint that introduces additional edge information into the TERA algorithm