Purpose: To develop a multishot magnetic resonance imaging (MRI) pulse sequence and reconstruction algorithm for diffusion-weighted imaging (DWI) in the brain with submillimeter in-plane resolution. Materials and Methods:A self-navigated multishot acquisition technique based on variable-density spiral kspace trajectory design was implemented on clinical MRI scanners. The image reconstruction algorithm takes advantage of the oversampling of the center k-space and uses the densely sampled central portion of the k-space data for both imaging reconstruction and motion correction. The developed DWI technique was tested in an agar gel phantom and three healthy volunteers.Results: Motions result in phase and k-space shifts in the DWI data acquired using multishot spiral acquisitions. With the two-dimensional self-navigator correction, diffusion-weighted images with a resolution of 0.9 ϫ 0.9 ϫ 3 mm 3 were successfully obtained using different interleaves ranging from 8 -32. The measured apparent diffusion coefficient (ADC) in the homogenous gel phantom was (1.66 Ϯ 0.09) ϫ 10 -3 mm 2 /second, which was the same as measured with single-shot methods. The intersubject average ADC from the brain parenchyma of normal adults was (0.91 Ϯ 0.01) ϫ 10 -3 mm 2 /second, which was in a good agreement with the reported literature values. Conclusion:The self-navigated multishot variable-density spiral acquisition provides a time-efficient approach to acquire high-resolution diffusion-weighted images on a clinical scanner. The reconstruction algorithm based on motion correction in the k-space data is robust, and measured ADC values are accurate and reproducible.

No abstract

MRI signal dropout in gradient recalled echo acquisitions limits the capability of blood-oxygen-level-dependent functional magnetic resonance imaging (fMRI) to study activation tasks that involve the orbitofrontal, temporal, and basal areas of the brain where significant macroscopic magnetic susceptibility differences exist. Among the various approaches aimed to address this issue, the acquisition method based on spiral in/out trajectories is one of the most time-efficient and effective techniques. In this study, we extended further the spiral in/out approach into 3D acquisition and compared the effectiveness of the different spiral in/out trajectory combinations in reducing signal dropout. The activation results from whole brain fMRI studies using complex finger tapping and breath-holding tasks demonstrate that the acquisition method based on dual-echo spiral in/in (DSPIN) trajectories is the most favorable. The DSPIN acquisition method has the following advantages: (1) It reduces most effectively signal dropout in the brain where magnetic susceptibility inhomogeneity is problematic and significantly improves the sensitivity to detect functional activations in those regions. Single-shot spiral and echo planar imaging (EPI) have been widely employed in brain functional MRI (fMRI) to produce neuronal activation maps due to their high time efficiency and good sensitivity to blood-oxygen-level-dependent (BOLD) contrast. BOLD function contrast is mainly derived from gradient recalled echo (GRE) images owing to their sensitivity to the modulation of T* 2 by changes in the oxygenation level of the activated brain regions (1). When the transverse magnetization decay is assumed to be exponential, BOLD functional sensitivity is maximized by choosing a echo time (TE) equal to the local T* 2 . In homogeneous brain regions, the T* 2 for gray matter is about 70 ms at 1.5 T; thus, in optimizing the BOLD contrast from the microvasculature and its surroundings, the GRE acquisition is also made very sensitive to intravoxel dephasing resulting from macroscopic field inhomogeneity near the air-tissue interface. The field inhomogeneity associated with susceptibility difference between brain tissue and air can cause severe image distortions and signal losses. Problematic areas are the orbitofrontal cortex, the medial and inferior temporal lobes, and central brain regions close to the brain stem.The susceptibility artifacts limit severely the applicability of BOLD fMRI technique in many cognitive and sensory experiments, which are significant for psychiatric and basic neuroscientific research. The development of imaging methods that can reduce susceptibility artifacts and still preserve sensitivity to BOLD functional contrast, high SNR efficiency, and adequate spatial resolution is of importance. This has been the topic of numerous recent studies. In addition to hardware improvements involving the development of elaborated shimming methods (2) and susceptibility-matched inserts for the mouth and ears (3), four types of data sampling ...

When volume coils are used for 1H imaging of the human head at 7T, wavelength effects in tissue cause a variation in intensity, that is typically brighter at the center of the head and darker in the periphery. Much of this image nonuniformity can be attributed to variation in the effective transmit B1 field, which falls by ∼ 50% to the left and right of center at mid‐elevation in the brain. Because most of this B1 loss occurs in the periphery of the brain, we have explored use of actively controlled, off‐resonant loop elements to locally enhance the transmit B1 field in these regions. When tuned to frequencies above the NMR frequency, these elements provide strong local enhancement of the B1 field of the transmit coil. Because they are tuned off‐resonance, some volume coil detuning results, but resistive loading of the coil mode remains dominated by the sample. By digitally controlling their frequency offsets, the field enhancement of each inner element can be placed under active control. Using an array of eight digitally controlled elements placed around a custom‐built head phantom, we demonstrate the feasibility of improving the B1 homogeneity of a transmit/receive volume coil without the need for multiple radio frequency transmit channels. Magn Reson Med, 2011. © 2011 Wiley‐Liss, Inc.