Electronic poster

Tuesday 13:30-15:30 Computer 122

13:30 5043. A Three-Dimensional Magnetic Resonance Imaging Geometric Distortion Correction Algorithm for Radiotherapy

Scott Hanvey¹, John Foster<sup>2

1</sup>Radiotherapy Physics, Beatson West of Scotland Cancer Centre, Glasgow, Lanarkshire, United Kingdom; ²MRI Physics, Beatson West of Scotland Cancer Centre, Glasgow, United Kingdom

The following article describes the testing of a three-dimensional geometric distortion correction algorithm, with applications in radiotherapy planning. Magnetic resonance imaging is well known to suffer from geometric distortion and although in-plane distortion is now routinely corrected, in general, through-plane distortion remains uncorrected. This study investigates a three-dimensional and a two geometric distortion correction algorithm using a variety of test objects to establish their accuracy in the axial and coronal planes. An acceptable clinical range for radiotherapy is established for the two-dimensional and three-dimensional distortion correction algorithms using three slices and many slices.

14:00 5044. Evaluation of Geometrical Distortion in a Head-Sized Phantom at Ultra-High-Field (7 Tesla) MRI

Philipp Dammann¹, Oliver Kraff, Stefan Maderwald, Elke Gizewski², Mark Ladd, Thomas Gasser<sup>3

1</sup>Neurosurgery, University of Duisburg-Essen, Germany and Erwin L. Hahn Institute for Magnetic Resonance Imaging, Essen, NRW, Germany; ²Department for Diagn. and Interv. Radiology and Neuroradiology, University of Duisburg-Essen, Germany; ³Department of Neurosurgery, University of Duisburg-Essen, Germany

In this study we assess geometrical distortion over a head-sized phantom at ultra-high-field MRI (7 Tesla). As field inhomogenity is proportional to B0, system-related distortions of spatial encoding are expected to increase with the field strength. This may interfere with potential image-guided applications in ultra-high-field MRI. This study revealed that distortion is lower than expected, showing no significant difference in distortion at 7 Tesla compared to lower field strength within a range of 91 mm from the magnetic center. Image distortion correction reduced local distortion maxima but didn`t significantly reduce mean distortion over the whole volume.

14:30 5045. Accuracy the Spatial Position of Gold Markers in MR Images

Tufve Nyholm¹, Joakim Jonsson¹, Magnus G. Karlsson¹, Mikael Karlsson<sup>1

1</sup>Umeå University, Umeå, Sweden

MR has potential to replace CT as basis for radiotherapy of prostate cancer. To be able to use the MR images for patient positioning at treatment, the position of the internal gold markers needs to be correct in the MR images. A phantom was created and scanned to evaluate the dependence between the apparent position and relevant sequence parameters. It was found that use of 2D sequences introduces errors in the position, and that the apparent marker position is dependent on the tissue type surrounding the marker. The apparent position is not dependent on band-width or frequency encoding direction.

15:00 5046. Attenuation Correction of a Simple Phantom from Simultaneous SPECT and MR Imaging

Mark Jason Hamamura¹, Seunghoon Ha¹, Werner W. Roeck¹, Lufti Tugan Muftuler¹, Douglas J. Wagenaar², Dirk Meier², Bradley E. Patt², Orhan Nalcioglu<sup>1

1</sup>Tu & Yuen Center for Functional Onco-Imaging, University of California, Irvine, CA, United States; ²Gamma Medica-Ideas, Inc., Northridge, CA, United States

Accurate SPECT imaging requires attenuation correction (AC) of the nuclear projection data. In this study, we simultaneously acquired SPECT and MRI data of a simple cylindrical phantom using a novel MR-SPECT system. This MRI data was then used to perform the AC of the SPECT data. The results demonstrate the feasibility of performing AC using data acquired from simultaneous MR and SPECT imaging.

Wednesday 13:30-15:30 Computer 122

13:30 5047. Application of IDEAL for the Correction of Chemical Shift Artifacts in MREIT

Mark Jason Hamamura¹, Orhan Nalcioglu¹, Lufti Tugan Muftuler<sup>1

1</sup>Tu & Yuen Center for Functional Onco-Imaging, University of California, Irvine, CA, United States

Chemical shift artifacts in magnetic resonance electrical impedance tomography (MREIT) degrade the accuracy of the reconstructed conductivity. In this study, we investigated the use of the iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL) algorithm to remove these artifacts in a simple fat/water phantom. The results demonstrate that this technique can be used to correct for chemical shift artifacts in MREIT.

14:00 5048. SSFP Banding Artefact Removal in Large FOV Images at 3T

Sonia I. Gonçalves¹, Maria L.W. Ziech¹, Jaap Stoker¹, Aart J. Nederveen<sup>1

1</sup>Radiology, AMC, Amsterdam, Netherlands

Banding artefacts are a serious obstacle to the use of B-FFE sequences in large FOV images and (ultra-)high field strengths. It is more so because the shortening of TR to minimize this type of artefacts is often not possible because of SAR constraints. In this work, it is shown that by combining scans with different phase cycling schemes one is able to correct for banding artefacts in large FOV abdominal images at 3 T, with as few as 6 different phase cycling schemes, independently of the chosen TR.

14:30 5049. Ramp Sampling Strageties for High Resolution Single-Pass Dixon Imaging at 3T

Ken-Pin Hwang¹, Basak E. Dogan², Zachary W. Slavens³, Anthony T. Vu³, Wei Tse Yang², Jingfei Ma<sup>2

1</sup>MR Applied Science Laboratory, General Electric Healthcare, Houston, TX, United States; ²Department of Diagnostic Radiology, University of Texas MD Anderson Cancer Center, Houston, TX, United States; ³GE Healthcare, Waukesha, WI, United States

Errors in Dixon fat-water separation may occur when acquired echo times deviate far from those expected by the separation algorithm. Single pass, dual-echo sequences are particularly vulnerable when pursuing high resolution at higher field strengths, where the increased frequency shift of lipid demands shorter in- and out-of-phase echo times. This study examines the effect of improper echo times on a Dixon algorithm and corrects them with the use of ramp sampling methods. Suppression is improved and artifacts are reduced by aligning the echo times closer to those expected by the algorithm, with no observable degradation of image quality.

15:00 5050. SERA: A Technique to Improve the Performance of the 3D Sequence by Reducing Aliasing Artifacts in Edge Slices

Yanle Hu<sup>1

1</sup>Imaging Research Center, University of Texas at Austin, Austin, TX, United States

When the object being imaged is larger than the field of view in slice-selection direction (zFOV), wrap-around aliasing artifacts will be observed in 3D sequences even with the use of a high performance excitation pulse. Although throwing away a couple of edge slices can solve the problem, it reduces the efficiency of the 3D method. In this work, a new technique is introduced. It excites and saturates spins in two thin slices immediately outside of zFOV before the slab excitation. As a result, aliasing artifacts in edge slices can be suppressed and the efficiency of 3D acquisition can be preserved.

Thursday 13:30-15:30 Computer 122

13:30 5051. An A-Priori Supported Image Correction Method for Severe Intensity Non-Uniformities at 3 T

Christian Würslin¹, Fritz Schick<sup>1

1</sup>Department of Diagnostical and Interventional Radiology, Section on Experimental Radiology, University Hospital Tübingen, 72076 Tübingen, Germany

Images, acquired at high field strengths usually suffer from a high amount of image non-uniformities (INUs), which cause a large amount of automatic post-processing techniques (e.g. for quantification) to fail. This applies especially for abdominal image slices, where INUs are particularly strong and common INU correction schemes do not apply. The authors therefore propose a correction algorithm which incorporates anatomic information for the compensation of heavily corrupted images. The algorithm was validated on real and simulated image data and showed a high potential in the reduction of INUs, enabling further post-processing procedures, such as thresholding, at high field strengths.

14:00 5052. A Robust and Simple Technique for the Suppression of Artifacts Arising from Long T1 Species in Segmented Inversion Recovery Sequences

Wolfgang Gerhard Rehwald¹, Pooja Aggarwal², Igor Klem², Han Kim², Raymond J. Kim<sup>2

1</sup>Siemens Healthcare, Chicago, IL, United States; ²Duke Cardiovascular MR Center, United States

This is only a Test.

14:30 5053. Improvement of the Arterial Input Function Considering B₁-Inhomogeneities

Robert Merwa¹, Karin Kapp², Franz Ebner³, Thorsten Feiweier⁴, Gernot Reishofer³, Rudolf Stollberger<sup>5

1</sup>Medical Engineering, FH OÖ - Upper Austria University of Applied Sciences, Linz, Austria; ²Department of Radiation Therapy, Medical University of Graz, Graz, Austria; ³Department of Radiology, Medical University of Graz, Graz, Austria; ⁴Healthcare, Siemens AG, Germany; ⁵Institute of Medical Engineering, Graz University of Technology, Graz, Austria

DCE MRI was performed at 3 T in combination with a special sequence in order to determine B₁ inhomogenities. AIF and tissue concentrations were calculated and the kinetic parameters K^trans and V_e were determined with a generalized kinetic model. The absolute deviation of the maximum value of two comparable AIFs can be improved by a factor up to 70 and the root mean square deviation concerning the two AIFs can be decreased by a factor up to 30 if B₁ inhomogeneities are corrected. Also the deviations of Ktrans and Ve in respect of the two AIFs are significantly lower.

15:00 5054. Stent Imaging Using Metal Artifact Reduction Sequence

Sang-Young Zho¹, Min Oh Ghim¹, Dong Joon Kim², Dong-Hyun Kim<sup>1,2

1</sup>Electrical and Electronic Engineering, Yonsei University, Seoul, Korea, Republic of; ²Radiology, Yonsei University College of Medicine, Seoul, Korea, Republic of

We examine a method for high-resolution stent imaging using metal artifact correction sequence and parallel imaging technique.

EPI Correction

Hall B Monday 14:00-16:00 Computer 123

14:00 5055. Robust Elimination of EPI Nyquist Ghosts Via Spatial and Temporal Encoding

W Scott Hoge¹, Huan Tan², Robert A. Kraft<sup>2

1</sup>Radiology, Brigham and Women's Hospital, Boston, MA, United States; ²Virgina-Tech Wake Forest School of Biomedical Engineering, Winston-Salem, NC, United States

Nyquist ghosts are an inherent artifact in EPI acquisitions. We propose here a method that fuses ghost correctuion methods based on spatial encoding (via multiple coils) and temporal encoding (via cyclic variations in the acquisition sequence). Post acquisition, PLACE is employed to cancel ghosting artifacts in data used to estimate self-referenced parallel MR imaging reconstruction coefficients. The improved pMRI reconstruction coefficients are then employed on each frame, to reconstruct a ghost free image. We demonstrate that this self-referenced approach significantly reduces Nyquist ghosts, and is robust to temporal variations such as magnetic field drift with minimal latency.

14:30 5056. Optimized Acquisition Strategy for Reference-Free Reduction of Nyquist Ghosting in EPI at 7 T

Benedikt A. Poser^1,2, Pål Erik Goa, ^1,3, Markus Barth, <sup>1,2

1</sup>Erwin L Hahn Institute for Magnetic Resonance Imaging, University Duisburg-Essen, Essen, Germany; ²Donders Institute for Brain, Cognition and Behaviour, Radboud University Nijmegen, Nijmegen, Netherlands; ³Department of Medical Imaging, St. Olavs University Hospital, Trondheim, Norway

Nyquist (N/2) ghosting in EPI tends to become particularly problematic at ultra-high field. Strongly dependent on imaging parameters –especially echo spacing and readout bandwidth– ghosting may pose considerable practical limitations when setting up fMRI protocols at 7 T and above.

We here show that residual ghosting is caused by an often appreciable mismatch between phase correction and imaging data. We propose a small but powerful sequence modification, namely an optimized timing of the phase-correction navigators, to overcome this problem and thereby remove the practical restrictions due to ghosting. Phantom and in vivo results demonstrate ghost reductions by up to factor four.

15:00 5057. Robust Method for EPI Ghost Correction

Frank Godenschweger¹, Myung-Ho In¹, Oliver Speck<sup>1

1</sup>Biomedical Magnetic Resonance, Institute for Experimental Physics, Otto-von-Guericke University, Magdeburg, Germany

A Nyquist ghost correction of EPI is propose, which determines the phase correction differences between channels and slices with high accuracy in a preparation step from a set of navigator echoes. Taking this calibration into account, only one single correction needs to be determined dynamically for all slices and channels during the EPI series, greatly improving stability.

The robustness of the proposed technique was tested on phantom and in-vivo data. The proposed technique for EPI ghost correction dramatically improves the image quality.

15:30 5058. Reducing Ghosting in EPI Using Trajectory Based Reconstruction with Dixon Method Fat Suppressed Navigator Echoes at 7T

Oliver Josephs¹, Chloe Hutton¹, Joerg Stadler², Johannes Bernarding³, Oliver Speck³, Nikolaus Weiskopf<sup>1

1</sup>Wellcome Trust Centre for Neuroimaging, University College London, London, United Kingdom; ²Leibniz Institute for Neurobiology, Magdeburg; ³Otto-von-Guericke University, Magdeburg

At 7T, navigator echoes, acquired at short TE, and used in EPI to reduce Nyquist ghosting, can be significantly compromised by fat signal. Usually, in EPI, fat is suppressed by applying a fat saturation pulse before slice selective excitation but at 7T this significantly increases the required SAR. We present a two point Dixon technique for suppressing the fat signal in the navigator echoes and demonstrate its effectiveness in human brain imaging. The new technique is an efficient alternative for improving phase navigators and can be used in additon to fat saturation and other artifact suppression methods.

Tuesday 13:30-15:30 Computer 123

13:30 5059. Navigator-Free Dynamic Phase Correction for Echo-Planar Imaging Based Functional MRI

Dan Xu¹, R. Scott Hinks¹, Bruce D. Collick

¹Applied Science Laboratory, GE Healthcare, Waukesha, WI, United States

In echo-planar imaging based functional MRI, non-phase-encoded navigator echoes are sometimes collected to enable correction of temporal frame dependent even-odd-echo phase modulation. However, the navigator-based method assumes that the additional modulation that the center echoes experience is the same as that predicted by navigator echoes, which is not true when there is additional modulation building up across echoes. Therefore, the modulation of the center echoes would not be well corrected, leading to ghost drift. We propose a method to use scan data itself to more faithfully estimate the per-temporal-frame modulation than the navigator-based method, which significantly reduces ghost drift.

14:00 5060. Robust 2D Phase Correction for Echo-Planar Imaging Under a Tight Field-Of-View

Dan Xu¹, Kevin F. King¹, Yuval Zur², R. Scott Hinks<sup>1

1</sup>Applied Science Laboratory, GE Healthcare, Waukesha, WI, United States; ²GE Healthcare, Haifa, Israel

The existing 2D phase correction methods to reduce Nyquist ghost in echo-planar imaging (EPI) have several unaddressed issues that largely affect their practicality. These issues include uncharacterized noise behavior, image artifact due to unoptimized phase estimation, and most seriously a new image artifact under tight FOV. We propose a modified, more robust method that addresses all the abovementioned issues. Various EPI results show that the proposed method can robustly generate images free of Nyquist ghost and some other image artifacts even in oblique scans or when cross-term eddy current terms are significant.

14:30 5061. Comparison of Applying 1D Phase and 2D Phase N/2 Ghost Correction Prior to PROPELLER-EPI Reconstruction

Hing-Chiu Chang^1,2, Chun-Jung Juan³, Tzu-Chao Chuang⁴, Yi-Jui Liu⁵, Chao-Chun Lin^2,6, Hsiao-Wen Chung<sup>2

1</sup>Applied Science Laboratory, GE Healthcare Taiwan, Taipei, Taiwan; ²Institute of Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei, Taiwan; ³Department of Radiology, Tri-Service General Hospital, Taipei, Taiwan; ⁴Electrical Engineering, National Sun Yat-sen University, Kaohsiung, Taiwan; ⁵Department of Automatic Control Engineering, Feng Chia University, Taichung, Taiwan; ⁶Department of Radiology, China Medical University Hospital, Taichung, Taiwan

PROPELLER-EPI consists of EPI signal readout with alternative echoes, thereby the phase inconsistencies between odd and even echoes generate N/2 ghost artifact in each rotating blade as well as conventional EPI imaging. The 1D correction method fails in oblique scan (rotating blades) because the phase inconsistencies along both readout and phase direction. A 2D phase correction method can overcome this problem by modifying the reference scan manner. In this work, we compare the quality of reconstructed PROPELLER-EPI images by applying 1D phase and 2D phase N/2 ghost correction prior to PROPELLER-EPI reconstruction.

15:00 5062. In FMRI a Dual Echo Time EPI Pulse Sequence Can Induce Sources of Error in Dynamic Magnetic Field Maps

Andrew Hahn¹, Andrew Nencka¹, Daniel Rowe, <sup>1,2

1</sup>Biophysics, Medical College of Wisconsin, Milwaukee, WI, United States; ²Mathematics, Statistics, and Computer Science, Marquette University, Milwaukee, WI

Estimations of main magnetic field inhomogeneity are often acquired for correction of image warping in echo planar images (EPI) resulting from vulnerability to off-resonance effects of EPI. Many established methods exist for field estimation, one of which involves two EPI acquisitions with different echo times. The method is fast, easily implemented and can be performed in-line with fMRI experiments. However, inconsistencies in the MRI scanner hardware, specifically with the RF pulse, as well as physiologic phenomena that alter the off-resonance characteristics between image acquisitions such as motion or respiration can induce errors in field maps estimated in this manner.

Wednesday 13:30-15:30 Computer 123

13:30 5063. Jacobian Weighting of Distortion Corrected EPI Data

Stefan Skare^1,2, Roland Bammer<sup>1

1</sup>Radiology, Stanford University, Stanford, CA, United States; ²MR Center, Clinical Neuroscience, Karolinska Institute, Stockholm, Sweden

By acquiring EPI data both with positive and negative phase encoding blips one obtains two oppositely distorted images. The reversed gradient polarity (RGPM) method can be used to correct these images by searching for a displacement field that explains their difference. However, even if the estimated displacement field is adequate, the two corrected EPI images have a very low resolution in anatomical regions that have been too compressed. In this work, we use a Jacobian weighting scheme to make an informative choice about the combination of the two images that avoids the inclusion of signals from very compressed regions.

14:00 5064. Using PLACE for EPI Distortion Correction of Diffusion Weighted Images (DWIs)

Sofia Chavez¹, Elizabeth Ramsay¹, Donald Plewes^1,2, Greg Stanisz^1,2, Q-San Xiang<sup>3

1</sup>Imaging Research, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada; ²Medical Biophysics, University of Toronto, Toronto, Ontario, Canada; ³Department of Radiology, University of British Columbia, Vancouver, B.C., Canada

A feasibility study for the application of PLACE, an EPI distortion correction, to diffusion weighted images (DWIs) is presented. PLACE requires a minimum of two input images which differ by an extra “blip” along the phase encode (PE) direction. The phase relation between the images encodes the PE coordinate, allowing for correction of the EPI-based distortion along the PE direction. Results show successful distortion correction for DWIs of a phantom despite the lower SNR, partial k-space and ramp sampling typical of standard DWI sequences. In vivo application of PLACE to DWI is currently under investigation.

14:30 5065. Accelerating Phase Modulation for Correcting EPI Geometry Distortion by Modern GPGPU Parallel Computation.

Yao-Hao Yang¹, Teng-Yi Huang, Fu-Nien Wang², Nan-Kuei Chen<sup>3

1</sup>National Taiwan University of Science and Technology, Taipei, Taiwan; ²Department of Biomedical Engineering and Environmental Sciences, National Tsing Hua university; ³Brain Imaging and Analysis Center, Duke University Medical Center

Phase modulation combined with field mapping can correct the EPI geometry distortion but it is a time-consuming algorithm. We proposed to incorporate the GPGPU technique into phase-modulation calculation to reduce the whole computation time. Applying on the PROPELLER EPI data set, the parallel algorithm reduced the computation time from ~1750 seconds to ~100 seconds. We conclude that the GPU computing is a promising method to accelerate EPI geometry correction.

15:00 5066. Probabilistic Reconstruction of Undistorted EPI Images Using a Rician Noise Model

Jesper Leif Roger Andersson¹, Mark Jenkinson<sup>1

1</sup>fMRIB, Oxford University, Oxford, Oxfordshire, United Kingdom

We have developed a method for estimating and correcting distortions from reverse-blip data with poor SNR. It is based on a forward model that allows us to make predictions about the images and a Rician noise model that enables us to calculate the probability of observed images. Bayesian inversion is used to find the most probable distortion-free image and field. It performs well even on data with very poor SNR.