Improvements in spiral MR reconstruction and imaging

Spiral magnetic resonance imaging is a newly developed fast data acquisition technique that has gained in popularity over the past decade. Spiral MRI collects data over a large portion of k-space (MRI data space in the spatial frequency domain) in a spiral fashion after a single excitation. Spiral imaging has been shown to be insensitive to flow artifacts and has found a niche in cardiac imaging and angiography. However, spiral imaging has several drawbacks: It is sensitive to main magnetic field inhomogeneity which leads to image blurring (off-resonance effects); The fat signal is usually suppressed using spatially and spectrally selective (SPSP) RF pulses in spiral imaging. However, SPSP RF pulses often lead to non-uniform fat signal suppression and undesirable water signal suppression in the presence of magnetic field inhomogeneity. Although spiral imaging is a very fast imaging technique, if a moving object is scanned, there are often observable motion artifacts in the reconstructed images. Therefore, further reduction of acquisition time is desirable as time reductions help to reduce motion artifacts.; Research in this project is aimed at overcoming the disadvantages described above and thereby advancing the current state-of-the-art condition of spiral imaging techniques. Off-resonance blurring artifacts are one primary disadvantage of spiral imaging. Hence, it is one of the most important issues throughout this project to develop efficient and effective off-resonance correction methods for several spiral data acquisition techniques. With conventional spiral acquisition techniques, new off-resonance correction algorithms have been developed that require reduced computations compared to the conventional off-resonance correction algorithms. The previously developed water-fat decomposition methods in rectilinear acquisition have been combined with the new fast off-resonance correction method. These methods provide uniform fat signal suppression with reduced acquisition time in spiral imaging. A new spiral reconstruction algorithm for parallel data acquisition has been developed. This algorithm is significantly simplified from the previously proposed comparable algorithms and thus provides a computation time advantage as they obviate the need for convolution-based gridding procedures. All of these newly developed techniques have improved the accuracy of measurements in MR cardiac applications with reduced acquisition time compared with the conventional spiral techniques.