F L 326102. m d Childrens Hospital. Los Angeles. CA 900543 onance (MR) imaging method, one echo signal corresponding to one line of the Fourier space data can be obtained in each cycle of the imaging sequence. The data on the line are sampled in t.he read out magnetic field gradient direction, which is related to one spatial frequency variable. Each echo signal is encoded by the phase encoding field gradieiit,, Abstract A magnetic resonance image m a y contain, truncation artifacts if there are not enough high frequency data when the conventional Fourier transform method is used for reconstructzon. I n this paper we propose a method f o r reducing the G Ttifacts using a multi-layer neural network. The network consists of one h e a r output layer and at least one non-linear hidden layer. I n this method the missing high frequency components are predicted based o n known low frequency components and are used to improve the resolution of the image. T h e method is tested with simulated data with good results. 1 624 0-8186-2742-5/92 $3.00 0 1992 IEEE
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.