Improved image reconstruction by combining ARMA modeling with wavelet decomposition: preliminary results

We present a novel technique to combine algorithms applicable to wavelet decomposition and magnetic resonance imaging (MRI) reconstruction to remove artifacts present in both approaches. Convolution and edge effects in wavelet decomposition can be removed by applying autoregressive moving average (ARMA) algorithms, currently used to remove truncation artifacts in MRI. Conversely, decreased complexity of the complex domain magnetic resonance data can be achieved using wavelet decomposition prior to image reconstruction. Preliminary results of combining these algorithms have been shown.