Correction strategies for segmented spiral imaging at 7 Tesla

Purpose/Introduction: Short acquisition times and short effective echo times are of prime importance for functional MR imaging studies of the brain to reduce distortion artifacts and increase SNR. The spiral sequence, specifically in the segmented version, offers these features and presents itself in specific applications as an alternative to conventional EPI. Spiral imaging is also less prone to flow and motion artifacts. But this method is more demanding on the exact gradient performance and correction mechanisms need to be implemented to achieve necessary readout gradient corrections. Subjects and Methods: To test and illustrate the performance of the implemented corrections, a geometry phantom was measured with a spin echo sequence as a reference and compared with a segmented spiral-out sequence. The measurements were performed on a 7T/ 60 cm Bruker Biospec vertical wide bore monkey MR system. A saddle coil was used in transmit and receive mode. Measurements of the k-space trajectories and images were performed[1]. With the k-space trajectory of the readout gradient the deviation from the theoretical course were calculated. From these the k-space offset, the time delay and the slope and baseline of the gradient courses were derived. The latter is done by comparing trajectories with positive and negative amplitudes. These corrections were then applied for spiral image acquisition and reconstruction. Results: The image where no corrections are applied (Fig.1, middle) shows that the edges of the rectangle inside the phantom are reproduced sharp and at the first look not distorted. It reveals although that the geometry is not at all correct compared to the reference image gained by the MSME sequence (Fig.1, left). Regarding the edges of the phantom in the lower part of the image a strong blurring occurs. Performing now the corrections before acquiring and before reconstructing the image the geometry is reproduced satisfactorily (Fig.1, right). The blurring artifact at the lower edge of the phantom is nearly fully eliminated. Discussion/Conclusion: The measurements of the trajectories and images on a geometry phantom show the efficiency of the corrections. Blurring artifacts are reduced. A misleading geometry like for an image measured with an uncorrected spiral-out EPI sequence can be avoided by applying the corrections. The corrections enable the spiral imaging sequence to be an alternative to conventional EPI in specific applications utilizing short echo times. [1] Takahashi A. et al. MagnResonMed 1995;34:446-456