News Column

Patent Issued for Diffusion-Weighted Magnetic Resonance Imaging Using 3D Mosaic Segmentation and 3D Navigator Phase Correction

July 7, 2014

By a News Reporter-Staff News Editor at Cardiovascular Week -- Isis Innovation Limited (Oxford, GB) has been issued patent number 8760163, according to news reporting originating out of Alexandria, Virginia, by NewsRx editors (see also Isis Innovation Limited).

The patent's inventors are Frost, Robert (Colwyn Bay, GB); Jezzard, Peter (Oxford, GB); Miller, Karla (Oxford, GB); Porter, David Andrew (Poxdorf, DE).

This patent was filed on May 10, 2011 and was published online on June 24, 2014.

From the background information supplied by the inventors, news correspondents obtained the following quote: "The present invention relates to diffusion-weighted magnetic resonance imaging, and in particular to a method and magnetic resonance apparatus wherein phase errors in the diffusion-weighted magnetic resonance data are corrected.

"Multi-shot, diffusion-weighted imaging (DWI) of the brain is sensitive to spatially dependent, non-linear phase errors that are caused by the effect of CSF pulsation and cardiac-related pulsatile motion on the brain during the diffusion-encoding gradients. The phase images in FIG. 1 shows how these phase errors vary substantially from one shot to the next, as illustrated for exemplary shots S1, S2, S3, S4 and S5. This leads to severe image artefacts in the final image if no correction is applied.

"For 2D acquisitions, in which each image plane is excited by a separate slice-selective radiofrequency (RF) pulse, a number of techniques have been demonstrated that allow the correction of these 2D non-linear, phase errors. These methods use signals from a `2D navigator`, which maps the spatial phase variation in two dimensions within the plane of the excited slice, and use this information to correct the associated imaging data. These 2D methods are extensions of the original concept of navigator correction, which was a 1D technique that corrected for a linear phase variation along the readout direction.

"In a number of applications of DWI it would be beneficial to be able to perform a diffusion-encoded 3D image acquisition. This would make it feasible to acquire data sets with an isotropic resolution that would offer an improved visualisation of small lesions and allow easy comparison with 3D anatomical data sets, acquired with SPACE (Sampling Perfect with Application optimized Contrast using different flip angle Evaluations) or MPRAGE (Magnetization Prepared Rapid Gradient Echo Imaging) sequence types. These 3D data sets would allow interactive and automatic post-processing procedures, such as volume rendering or the reconstruction of 2D slices with arbitrary orientation. 3D diffusion-weighted data would also be valuable for tractography studies in diffusion tensor imaging (DTI).

"However, when performing 3D data acquisition, the entire imaging volume is excited at each shot and not just a single imaging slice as in the 2D case. This means that it is now necessary to apply a correction for the spatial phase variation in all three dimensions, which is not possible using the existing acquisition and navigator correction schemes used in 2D multi-shot imaging. There is thus a need for a 3D acquisition scheme for imaging and navigator data that is suitable for performing a true 3D phase correction.

"A number of sequences have been proposed for performing 3D DWI.

"One approach is to add a diffusion preparation module before each train of low-flip-angle excitations in the 3D MPRAGE sequence, as described in (Numano T, Homma K, Hirose T. Diffusion-weighted three-dimensional MP-RAGE MR imaging. Magn Reson Imaging. 2005 April; 23(3):463-8). As there is no navigator phase correction applied in this sequence, it is susceptible to phase variations from one echo-train to the next. This method appears to have only been used in anaesthetized animal studies.

"Another method, which has been used in human studies, is a single-shot 3D acquisition from a restricted volume (Jeong E-K, Kim S-E, Kholmovski E G, Parker D L. High resolution DTI of a localized volume using 3D single-shot diffusion-weighted stimulated echo-planar imaging (3D ssDWSTEPI). Magn. Reson. Med. 2006; 56:1173-1181.) The sequence uses a single diffusion preparation for each volume acquisition, so there are no problems with shot-to-shot phase variation. The signal in a selected volume is spatially encoded using a stimulated echo approach with a series of small flip angles to create a train of stimulated echoes, each of which is sampled with an EPI readout to provide encoding in two (in-plane) dimensions. Spatial encoding in the third dimension is produced by applying a different slice-direction, phase-encoding gradient for each individual stimulated echo signal. To avoid distortion and blurring problems, associated with the single-shot readout, application of the sequence is restricted to small volumes of interest. Also the combination of the stimulated echo acquisition and the low flip-angle detection pulses is likely to result in a relatively low signal-to-noise ratio (SNR).

"One approach to navigator-correction in 3D DWI has been proposed in the TURBINE method (McNab J A, Gallichan D, Miller K L. 3D steady-state diffusion-weighted imaging with trajectory using radially batched internal navigator echoes (TURBINE). Magn Reson Med. 2010 January; 63(1):235-42.0). This technique uses a diffusion-weighted, steady-state free-precession signal, which is formed when a monopolar, diffusion-encoding gradient is applied after each RF excitation pulse in a train of low-flip-angle pulses with short TR. Each RF pulse excites magnetisation from the entire measurement volume and an EPI readout is used to spatially encode the signals in two dimensions. To provide spatial encoding in three dimensions the plane of EPI spatial encoding is rotated about one of its central axes from one excitation to the next. Because each EPI readout samples a plane which passes through the centre of k-space, 2D phase correction data are available directly from the imaging data and no separate navigator signal is required. However, there is no information about phase variation perpendicular to the plane of the EPI encoding. This is addressed by acquiring an ECG signal during the measurement and combining data from different RF excitations, but which were acquired at similar points in the cardiac cycle. In this way, a multi-shot 3D navigator signal is generated for a given point in the cardiac cycle. In addition, data acquired during early systole, when pulsatile effects are most significant, were rejected and not used during image reconstruction.

"The idea of using 2D data to correct phase errors in a 3D diffusion-weighted acquisition has also been described in relation to a turbo-spin-echo (TSE) sequence in (von Mengershausen M, Norris D G, Driesel W. 3D diffusion tensor imaging with 2D navigated turbo spin echo. MAGMA. 2005; 18(4)).

"A 2D navigator-corrected, readout-segmented EPI (rs-EPI) technique described in U.S. Pat. No. 7,205,763, and is briefly reviewed as follows before describing the new method in the following section.

"As shown in FIG. 2, at each shot, two spin echoes are sampled using an EPI readout with a limited coverage in the k.sub.x direction and full coverage in the k.sub.y direction. For the imaging echo, a variable dephase gradient G.sub.R is applied along the readout direction to provide a k-space offset along k.sub.x, so that a `segment` of k-space is acquired for a subset of the k.sub.x points required for the image. For the navigator echo a fixed dephase gradient G.sub.R is applied, so that the readout segment in this case is always located at the centre of k-space. During image reconstruction, the navigator data are used to perform an image-domain phase correction on the corresponding readout segment sampled by the imaging echo. Critically, both imaging and navigator acquisitions sample contiguous regions of k-space, so that the Nyquist sampling condition is fulfilled in both k.sub.x and k.sub.y directions. This allows navigator phase correction to be applied without the complications of aliased signal contributions.

"The reduced k.sub.x coverage allows a much shorter spacing between the echoes in the EPI readout than single-shot EPI, thereby reducing susceptibility artefacts. The corresponding reduction in EPI echo-train length reduces blurring due to T.sub.2* decay and allows high spatial resolutions that are not possible with single-shot EPI."

Supplementing the background information on this patent, NewsRx reporters also obtained the inventors' summary information for this patent: "The method in accordance with the invention provides a way to acquire true 3D navigator data, which can be used to correct the spatially varying phase errors in the corresponding imaging data at each shot. To make this possible, a `mosaic` sampling scheme is used that samples (at each shot) a contiguous set of k-space points in 3D k-space, which form a subset of the full range of k-space required. The navigator data are acquired from a fixed region at the center of k-space and the imaging data are acquired from a region of k-space that may be offset from the k-space centre along all three axes. At each shot the imaging data offsets are varied, so that the whole of k-space is sampled by the multi-shot acquisition. Because the data points are contiguous, the Nyquist sampling condition is now fulfilled in all three dimensions and 3D phase correction can be applied without problems from aliased signals.

"The extent of the sampled region in each dimension would typically be less than that of the full range of k-space points required for the chosen image resolution. However, with some scanning protocols it might be possible to sample the complete set of k-space points along one or two of the k-space axes at each shot. In some cases, this approach might reduce image artefacts or allow faster scan times."

For the URL and additional information on this patent, see: Frost, Robert; Jezzard, Peter; Miller, Karla; Porter, David Andrew. Diffusion-Weighted Magnetic Resonance Imaging Using 3D Mosaic Segmentation and 3D Navigator Phase Correction. U.S. Patent Number 8760163, filed May 10, 2011, and published online on June 24, 2014. Patent URL:

Keywords for this news article include: Cardiology, Isis Innovation Limited.

Our reports deliver fact-based news of research and discoveries from around the world. Copyright 2014, NewsRx LLC

For more stories covering the world of technology, please see HispanicBusiness' Tech Channel

Source: Cardiovascular Week

Story Tools Facebook Linkedin Twitter RSS Feed Email Alerts & Newsletters