Michael Herbst¹, Julian Maclaren¹, Matthias Weigel¹, Jan Gerrit Korvink^2,3, and Maxim Zaitsev^1
1Medical Physics, University Medical Center Freiburg, Freiburg, Germany, ²Dept. of Microsystems Engineering - IMTEK, University of Freiburg, Freiburg, Germany,³Freiburg Institute of Advanced Studies (FRIAS), University of Freiburg, Germany

Diffusion weighted imaging (DWI) has become indispensable in clinical routine, especially due to its sensitivity to early stages of brain ischemia. Patient motion during measurements is a major source of artifacts. There are different approaches to correct for rigid body motion depending on the type of motion. However, intrascan motion remains problematic for all of these correction methods. Even for single-shot EPI intrascan motion can lead to severe signal dropouts. This work presents a method to continuously apply prospective motion correction to correct for intra- and interscan motion with six degrees of freedom.

Murat Aksoy¹, Christoph Forman², Daniel Kopeinigg¹, Matus Straka¹, Rafael O'Halloran¹, Samantha Holdsworth¹, Stefan Skare^1,3, and Roland Bammer^1
1Radiology, Stanford University, Stanford, CA, United States, ²Computer Science, Friedrich-Alexander-University Erlangen-Nuremberg, Erlangen, Germany, ³Clinical Neuroscience, Karolinska Institute, Stockholm, Sweden

Correction of motion is important in diffusion tensor imaging (DTI) due to long acquisition times and the increased likelihood of involuntary motion. Single-shot echo-planar imaging (sshEPI), which is the most frequently used sequence in DTI, allows retrospective motion correction that includes realigning individual diffusion-weighted volumes. However, this correction method is considerably flawed due to its inability to deal with intra-volume motion or spin history effects. In this study, we used a monovision-based optical tracking system to provide real-time motion correction capabilities for sshEPI-DTI. In-vivo results demonstrate that even for single-shot methods, prospective motion correction yields more accurate and precise FA maps and fiber tracts than retrospective motion correction.

Rita Gouveia Nunes^1,2, Shaihan J. Malik², and Joseph V. Hajnal^2
1Institute of Biophysics and Biomedical Engineering, Faculty of Sciences, University of Lisbon, Lisbon, Portugal, ²Robert Steiner MRI Unit, Imaging Sciences Department, MRC Clinical Sciences Centre, Hammersmith Hospital, Imperial College London, London, United Kingdom

Echo-planar imaging is widely used for diffusion-weighted imaging despite its high sensitivity to field in-homogeneities. Although single-shot fast spin-echo images do not have this limitation, the signal is highly sensitive to the phase of the magnetization prior to the start of the refocusing train (CPMG condition). Brain pulsation and bulk patient motion during diffusion sensitization lead to unpredictable phase patterns, while scanner vibrations produce reproducible non-linear phase modulations. A prospective method for correcting for spatially non-linear phase structures using tailored RF pulses is presented and shown to be effective in phantoms.

Alexandru Vlad Avram^1,2, Trong-Kha Truong², Arnaud Guidon^1,2, Chunlei Liu², and Allen W. Song^2
1Biomedical Engineering Department, Duke University, Durham, NC, United States, ²Brain Imaging and Analysis Center, Duke University Medical Center, Durham, NC, United States

We present a novel stimulated echo (STE) based Diffusion Tensor Imaging (DTI) spiral sequence with inherent capability to correct for off-resonance effects (blurring) due to spatial and temporal variations in the main field B0 (e.g. due to tissue susceptibility, eddy currents, system instabilities, etc). Combined with the self-navigated interleaved spiral acquisition (SNAILS) this technique provides dynamic and non-linear corrections for phase errors (e.g. due to motion) and off-resonance effects, which may find broad applications in diffusion weighted MRI where both of these artifacts are significant.

Jürgen Finsterbusch^1,2
1Department of Systems Neuroscience, University Medical Center Hamburg-Eppendorf, Hamburg, Germany, ²Neuroimage Nord, University Medical Centers Hamburg-Kiel-Lübeck, Hamburg-Kiel-Lübeck, Germany

Inner-field-of-view EPI based on 2D-selective RF excitations (2DRF) has been shown to be a promising tool for high-resolution diffusion-weighted imaging, e.g. in the spinal cord. In this study, it is shown that tilting the excitation plane to position the side excitations in the dead corner between the slice stack to acquire and the current image section represents a simple and robust method to suppress the unwanted signal contributions. This approach can reduce the 2DRF pulse durations and the echo time considerably and, thus, increase the SNR significantly as is demonstrated in the human spinal cord in vivo.

Anh Tu Van¹, Joseph Holtrop², and Bradley P Sutton^2
1Electrical and Computer Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ²Bioengineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States

The complicated structure of many neuronal regions places an ever-increasing demand on the imaging resolution. By combining 3D encoding with multiecho acquisition on a localized region of interest, the current work targets diffusion tensor imaging with 0.8 x 0.8 x 1 mm3 resolution. Penalized iterative reconstruction is also proposed for minimizing the effects of imperfect realization of localized imaging. In vivo results are shown on the human pons.

Sangwoo Lee¹, and Gaohong Wu^1
1GE Healthcare, Waukesha, WI, United States

Spectro-spatial RF pulses (SPSP RF pulses) for clinical 2D imaging have been widely used for B1-insensitive excitation with excellent lipid suppression compared to the conventional fat saturation methods [1,2]. However, at 3T, pulse design is challenging due to the reduced sub-pulse width, and often results in compromise in the minimum slice thickness and poor slice excitation profile. In this work, we introduce a new spectral-spatial design scheme which allows very thin slice excitation with superior slice profile for single spin echo diffusion imaging.

Mathias Engström¹, Roland Bammer², and Stefan Skare^1
1Clinical Neuroscience, Karolinska Institute, Stockholm, Sweden, ²Radiological Sciences Laboratory, Stanford University, Palo Alto, CA, United States

A Stejskal-Tanner diffusion preparation with a GRAPPA-accelerated vGRASE acquisition is proposed. Compared to conventional ss-EPI, vGRASE is less sensitive to off-resonance effects despite being a single-shot technique and hence also robust to motion. With sequence modifications and phase correction techniques we adress problems induced by the diffusion gradients.

Thorsten Feiweier^1
1Siemens AG, Healthcare Sector, Erlangen, Germany

High-resolution DTI requires both high SNR and precise spatial alignment of images acquired with different diffusion encodings. While the bipolar (twice refocused) diffusion-encoding scheme effectively reduces eddy-current-induced distortions as compared to the monopolar Stejskal-Tanner approach, it involves increased TE due to the need for additional spoiler gradients. A new bipolar diffusion-encoding variant is discussed here, omitting the need for explicit spoiling. This approach allows for a markedly reduced TE and correspondingly increased SNR with negligible impact on eddy current suppression efficiency.

Zhiqiang Li¹, James G Pipe², Chu-Yu Lee^2,3, Josef P Debbins^2,3, John P Karis⁴, and Donglai Huo^1,2
1MR Engineering, GE Healthcare, Waukesha, WI, United States, ²Neuroimaging Research, Barrow Neurological Institute, Phoenix, AZ, United States, ³Electrical Engineering, Arizona State University, Tempe, AZ, United States, ⁴Radiology, Barrow Neurological Institute, Phoenix, AZ, United States

DW-PROPELLER techniques have the advantages such as no susceptibility artifacts and the capability for high-resolution imaging. In TurboPROP, gradient and spin echoes are grouped together to form a wider blade, providing robustness to motion but leading to off-resonance artifact. A variant of split-blade TurboPROP was proposed by separating the gradient and spin echoes into individual blades. However, the robustness to motion is lost due to the smaller overlapping area in the center of k-space. To address this issue, X-PROP is proposed by spreading the blades from one TR uniformly in k-space and removing the motion-induced phase error.