Abstract #3187

Tsao J, Razavi R, G. Hill D, Kozerke S, Baltes C, Muthurangu V, Hansen M

In this abstract we present a method for k-t SENSE reconstruction of undersampled data acquired on non-Cartesian trajectories in k-t space. This is an extension of a previously presented method to include complementary information from multiple receive coils. The method is applied to real-time radial imaging of speech production. The radial trajectory is ideal for this application since it provides integrated training data. In simulations and in vivo data, we show that the radial k-t SENSE reconstruction is superior to a modified sliding window reconstruction which has full temporal bandwidth for the central part of k-space.