Abstract
This paper investigates the potential of the new Weighted-Compressive Sensing approach which overcomes the major limitations of other compressive sensing and outperforms current state-of-the-art methods for low-rate reconstruction of sequences of MRI images. The underlying idea of this approach is to use the image of the previous time instance to extract an estimated probability model for the image of interest, and then use this model to guide the reconstruction process. This is motivated by the observation that MRI images are hugely sparse in Wavelet domain and the sparsity changes slowly over time.