Towards real-time multi-modality 3-D medical image registration

Netsch, Thomas; Rösch, Peter; Muiswinkel, Arianne van; Weese, Jürgen

doi:10.1109/iccv.2001.937595

Cited by 41 publications

(43 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…If w varies in space, the right-hand side of Eq. [9] cannot be written simply in primed coordinates, and hence image reconstruction via FFT is no longer exact. Similar problems may be encountered for B 0 , B 1 , and gradient inhomogeneities.…”

Section: Theorymentioning

confidence: 99%

“…A detailed description of the registration algorithm is given in Ref. 9. To allow independent scan plane positions and angulations during the calibration scan and subsequent diagnostic scan, the registration is performed in magnet coordinates.…”

Section: Calibration Of Motion Modelmentioning

confidence: 99%

“…However, in the strict sense this argument holds only under the implicit assumption of spatially constant receive coil sensitivities. The acquired signal S, which is used for image reconstruction, may be written as S͑t͒ ϭ ͵ v w ⅐ M xy ͑r,t͒dr [9] where w denotes the sensitivity of the receive coil. If w varies in space, the right-hand side of Eq.…”

Section: Theorymentioning

confidence: 99%

See 2 more Smart Citations

Prospective correction of affine motion for arbitrary MR sequences on a clinical scanner

Nehrke

Börnert

2005

Magnetic Resonance in Med

View full text Add to dashboard Cite

The concept of prospective 3D affine motion correction was generalized, based on the Bloch equations, for signal excitation and sampling using arbitrary MR sequences. The technique was implemented on a clinical MRI scanner for Cartesian, radial, and spiral imaging sequences, as well as for 2D spatially selective RF excitation pulses. A patient-specific motion model steered by real-time navigators was employed to account for the additional degrees of freedom provided by the affine motion model. Different navigator concepts (multiple spatial and temporal navigators, quadratic navigators and other motion sensors) were investigated, with the aim of improving the correlation between navigator information and the motion model. Many MR imaging (MRI) and spectroscopy (MRS) applications suffer from motion artifacts caused by, e.g., respiratory and cardiac motion. Currently employed real-time gating techniques, such as EKG triggering and navigatorbased respiratory gating (1), reduce motion artifacts considerably but result in a strong increase in scan time. Hence, motion-correction techniques, which allow the acquisition of MR data in the presence of motion, are highly desirable. Slice-tracking approaches based on respiratory navigators have been shown to improve image quality in coronary MRA (2). However, due to the simplicity of the underlying 1D translatory motion model, only small gating windows (typically 5 mm) can be applied (3). Recently a patient-specific calibration approach was introduced for coronary MRA that allows the prediction of affine transformation parameters, including 3D translation, rotation, scale, and shear motion, from navigator measurements (4). In addition, a prospective correction of affine motion was proposed. However, the technical complexity of such an approach limited the implementation to prospective correction of 3D translation and 1D expansion for Cartesian imaging sequences. Furthermore, additional offline processing was required, which challenged the practical feasibility of the method in a clinical workflow.In the current work, prospective motion correction is discussed in the framework of the Bloch equations, which conceptually overcomes the restriction to Cartesian imaging sequences. Thus, the approach is generalized to signal excitation and sampling for arbitrary MRI sequences, such as multidimensional excitation pulses and non-Cartesian imaging sequences. Following the theoretical treatment, the technique is fully implemented on a clinical scanner in a sequence-independent manner, which enables the employment of a full 3D affine transformation for all available pulse and imaging sequences, including Cartesian, radial, and spiral scanning, as well as 2D excitation pulses. The results of phantom experiments are presented that prove the technical feasibility of the approach, whereas initial in vivo results demonstrate the potential for coronary MRA and abdominal imaging. THEORYThe principles of affine motion correction may be derived directly from the Bloch equations. The Bloch equati...

show abstract

Section: Theorymentioning

confidence: 99%

Section: Calibration Of Motion Modelmentioning

confidence: 99%

Section: Theorymentioning

confidence: 99%

See 1 more Smart Citation

Prospective correction of affine motion for arbitrary MR sequences on a clinical scanner

Nehrke

Börnert

2005

Magnetic Resonance in Med

View full text Add to dashboard Cite

show abstract

“…Retrospective interscan motion correction was achieved using a 3D-rigid coregistration algorithm (25). Subsequently, eddy current-induced image warping was reduced with a correlation-based 2D-affine registration algorithm described in Ref.…”

Section: Postprocessing and Visualizationmentioning

confidence: 99%

SENSE‐DTI at 3 T

Jaermann

Crelier

Pruessmann

et al. 2004

Magnetic Resonance in Med

Self Cite

202

169

View full text Add to dashboard Cite

While holding vast potential, diffusion tensor imaging (DTI) with single-excitation protocols still faces serious challenges. Limited spatial resolution, susceptibility to magnetic field inhomogeneity, and low signal-to-noise ratio (SNR) may be considered the most prominent limitations. It is demonstrated that all of these shortcomings can be effectively mitigated by the transition to parallel imaging technology and high magnetic field strength. Using the sensitivity encoding (SENSE) technique at 3 T, brain DTI was performed in nine healthy volunteers. Despite enhanced field inhomogeneity, parallel acquisition permitted both controlling geometric distortions and enhancing spatial resolution up to 0.8 mm in-plane. Heightened SNR requirements were met in part by high base sensitivity at 3 T. Diffusion tensor imaging (1,2) is a promising noninvasive method for studying white matter structure of the human brain in vivo. Based on the concept of anisotropic water diffusion across tissue, the measurement of 3D diffusion properties, as described by a local diffusion tensor, allows the characterization of the axonal architecture of white matter networks. For that purpose, a 3D tracking of axonal projections, known as fiber tracking (3-7), is required. However, the low SNR and the limited spatial resolution (8 -10) of the method severely impair its application. A serious resolution limit stems from the strong link between voxel size and SNR, the latter being inherently low due to diffusion weighting. Only improving the SNR of the initial diffusion-weighted (DW) images will enable better spatial resolution. Therefore, the use of high magnetic fields and the related SNR gain could considerably enhance the performance of DTI and fiber tracking.The calculation of the local diffusion tensor requires a set of DW images, acquired with diffusion gradients applied in at least six noncollinear directions, plus a reference image without diffusion weighting. The sequence most commonly used for DTI is spin-echo single-shot EPI (SE-sshEPI). It allows for whole brain coverage in an acceptable scan time and is insensitive to bulk motion due to its speed. Critical shortcomings of sshEPI are image blurring due to T* 2 decay during the EPI readout interval and off-resonance effects, caused by the long EPI echo train (11,12). Both effects scale with the main magnetic field B 0 , making the transition to higher field strength challenging. At 3 T, signal alteration and geometric distortion due to static resonance offset effects, e.g., in the vicinity of airtissue interfaces, are a serious problem when using sshEPIbased protocols.Recently, the potential of parallel imaging techniques, such as simultaneous acquisition of spatial harmonics (SMASH) (13) and sensitivity encoding (SENSE) (14), has been demonstrated for sshEPI in general (15), as well as for diffusion-weighted MRI (DWI) (16 -18) and DTI (19) at 1.5 T. Parallel imaging techniques were shown to significantly reduce EPI-related artifacts as a result of shortening the echo train by facto...

show abstract

“…It has become an important tool in radiotherapy treatment. Brain surgery and neurosciences have been the driving applications for developing an abundant number of techniques as discussed in [2] and [3]. Despite the significant progress that has been made, deformable registration remains a challenging problem in the field of radiotherapy.…”

Section: Introductionmentioning

confidence: 99%