Rapid cardiac imaging with turbo BRISK

Abstract

A variety of variable and constant rate, sparse sampling strategies have previously been proposed to rapidly image dynamically changing objects. The majority of these strategies compile a k-space data set for any given time point by substituting k-space data from the most recently sampled time positions (extracted from the sparsely sampled set). The BRISK technique, is a variable rate, sparse sampling technique which additionally incorporates an interpolation scheme to more accurately represent k-space data at positions which were not directly sampled. Here, strategies are introduced that allow turbo concepts to be incorporated with BRISK. Simulations are conducted to compare the efficacy of the turbo BRISK acquisition and processing strategy against a constant rate, sparse sampling strategy with direct substitution of the most recently acquired k-space lines. It is shown that turbo BRISK generates images of similar quality in approximately half the time as the uniform sampling rate, sparse sampling strategy. Data from turbo BRISK acquisitions of multicardiac phase image sets, obtained on a normal volunteer and cardiac patients are presented.