Room 355 EF

13:30 - 15:30

Moderators:

Hua Guo, Maxim Zaitsev

Maximilian Haeberlin¹, Lars Kasper¹, Christoph Barmet², David Otto Brunner¹, and Klaas P. Pruessmann^1
1Institute for Biomedical Engineering, University and ETH Zurich, Zurich, Switzerland, ²Skope Magnetic Resonance Technologies, Zurich, Switzerland

Gradient-based 3D positioning of an NMR field probe relies on independent information on all 3 gradient axes. Position encoding using sinusoidal gradient tones in the kHz range suffers from inductive coupling among 0^th and 1^st order fields. Field decoupling is demonstrated, the improvement is quantified (accuracy < 100 µm ) and the method is applied to in-vivo real-time prospective motion correction in the brain.

Melvyn B. Ooi¹, Murat Aksoy¹, Julian R. Maclaren¹, Ronald D. Watkins¹, and Roland Bammer^1
1Radiology, Stanford University, Stanford, CA, United States

The ability to track the positions of miniature RF coils in the MRI scanner has been the foundation of several recent advances in prospective motion correction. The current work introduces the use of multiple “wireless markers” for rigid-body motion tracking. Each wireless marker is a miniature RF coil that is not connected to the MR receiver via traditional coaxial cables, but rather transmits its signal wirelessly via inductive coupling with the nearby imaging head-coil. Wire-free prospective real-time motion correction is demonstrated in a moving phantom and brain.

Maryna Babayeva^1,2, Tobias Kober^2,3, Michael Herbst⁴, Jürgen Hennig⁴, Matthias Seeger⁵, Rolf Gruetter^3,6, Maxim Zaitsev⁴, and Gunnar Krueger^2,3
1CIBM-AIT, École Polytechnique Fédérale de Lausanne and University of Lausanne, Lausanne, Switzerland, ²Advanced Clinical Imaging Technology, Siemens Healthcare IM S AW, Lausanne, Switzerland, ³CIBM-AIT, École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, ⁴Department of Radiology, University Medical Center Freiburg, Freiburg, Germany, ⁵Laboratory for Probabilistic Machine Learning, École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland,⁶Departments of Radiology, Universities of Lausanne and Geneva, Lausanne, Switzerland

This work investigates the ability of free induction decay (FID) navigators to provide information on rigid head motion. FID navigators were incorporated in a gradient-echo sequence. In parallel, optical tracking data was acquired and served as the ground truth. Three subjects were scanned at 3T with a 32-channel head coil while performing complex head movements. A linear model was trained with FID and optical tracking data and verified by cross-validation. Following the linear assumption, it can be shown that FID signal changes can quantify all six motion parameters with sub-millimeter and sub-degree precision.

Daniel Giese¹, Martin Buehrer², Constantin von Deuster^1,2, Tobias Schaeffter¹, and Sebastian Kozerke^1,2
1Division of Imaging Sciences and Biomedical Engineering, King's College London, London, United Kingdom, ²Institute for Biomedical Engineering, University and ETH Zurich, Zurich, Switzerland

A spatially invariant 2D navigator acquired simultaneously using multi-band excitation pulses is proposed and demonstrated for multi-slice free-breathing liver imaging. Using the 2D navigator 3D reconstructions of the liver at different respiratory positions are possible hence allowing assessment of respiratory-induced tissue deformations to aid for instance radio-therapy planning.

Axel Hartwig¹, Magnus Mårtensson², and Stefan Skare^1
1Neuroradiology, Karolinska University Hospital, Stockholm, Sweden, ²EMEA Research and Collaboration, Applied Science Laboratory, GE Healthcare, Stockholm, Sweden

Patient head motion is one of the leading sources of artifacts in brain MRI. Motion in the 'nodding direction' is particular difficult to restrain and will, in 2D axial and coronal scans, cause spin-history effects through the plane which cannot be addressed retrospectively. In this work, we propose a 2D fat-only (FatNav) navigator image for prospective correction of head nodding motion. FatNav is advantageous over water-based navigators as it leaves the brain water magnetization unaffected. We have shown that the motion estimates using the FatNav images reflects the true 'nodding motion' of a moving subject.

Daniel Gallichan¹, José P. Marques², and Rolf Gruetter^3,4
1LIFMET, Ecole Polytéchnique Fédérale de Lausanne, Lausanne, Vaud, Switzerland, ²Dept. of Radiology, University of Lausanne, Lausanne, Vaud, Switzerland,³LIFMET, École Polytechnique Fédérale de Lausanne, Lausanne, Vaud, Switzerland, ⁴Depts. of Radiology, Universities of Lausanne and Geneva, Lausanne, Vaud, Switzerland

We tested the feasibility of using the fat signals in the head for high-resolution motion-navigation at very high acceleration factors. 6 subjects were scanned at high resolution with both fat and water excitations at 3 time-points during a single session where there was no intentional movement. The registration of the water images was used as the reference for motion, and the parameters extracted from the fat images, both fully sampled and accelerated to R=50, were compared. Although consistent results were obtained from the fat images, there was a tendency to slightly underestimate the true motion – which requires further investigation.

Holger Eggers¹, Peter Boernert², and Tim Leiner^3
1Philips Research Laboratories, Hamburg, Germany, ²Philips Research, Hamburg, Germany, ³Department of Radiology, University Medical Center Utrecht, Utrecht, Netherlands

Due to stringent scan time constraints, flow artifact suppression by moment nulling of the gradients is usually not applied in subtraction first-pass peripheral angiography, since it prolongs echo and repetition times. In this work, a recently proposed subtractionless approach to first-pass peripheral angiography based on dual-echo Dixon imaging is shown to allow an efficient, retrospective elimination of especially high intensity ghosting from pulsatile flow, without requiring changes to the bipolar dual-gradient-echo acquisition employed for chemical shift encoding.

Onur Afacan¹, Ali Gholipour¹, Erez Nevo², and Simon K. Warfield^1
1Computational Radiology Laboratory, Department of Radiology, Boston Children's Hospital and Harvard Medical School, Boston, Massachusetts, United States,²Robin Medical, Inc, Baltimore, MD, United States

In this work we propose a novel motion robust high-resolution FLASH acquisition scheme to be used for cortical lesion detection in uncooperative subjects. This scheme is based on acquiring multiple low-resolution images with different orientations and then combining these acquisitions into an isotropic high-resolution image using super-resolution reconstruction. We use a motion sensor to identify the low-resolution images corrupted with motion to reacquire them, as well as to correct for the motion between the low-resolution acquisitions. With this method whole brain images with isotropic resolution of 0.6 mm were reconstructed with a total scan time of 20 minutes.

Joseph Y. Cheng¹, Martin Uecker², Marcus T. Alley³, Shreyas S. Vasanawala³, John M. Pauly¹, and Michael Lustig^2
1Electrical Engineering, Stanford University, Stanford, California, United States, ²Electrical Engineering and Computer Sciences, University of California, Berkeley, California, United States, ³Radiology, Stanford University, Stanford, California, United States

In pediatric imaging, the patient must often be placed under deep anesthesia and put into breath-holds. This procedure adds time to prepare/finish the studies, complication to the exams, and overall risk to the patient. We developed a novel scheme to eliminate the need for deep anesthesia and breath-holds while achieving high-resolution motion-free images from a free-breathing 3D scan. First, Butterfly navigation provides motion information and data-consistency weights. Next, these weights are incorporated into a compressed-sensing & parallel-imaging reconstruction, wESPIRiT. Lastly, autofocusing uses the localized motion measurements to remove residual motion-artifacts. We demonstrate the effectiveness of our method in patient studies.

Zhongnan Fang¹ and Jin Hyung Lee^1,2
1Electrical Engineering, University of California, Los Angeles, Los Angeles, CA, United States, ²Department of Neurology and Neurological Sciences, Department of Bioengineering, Stanford University, Stanford, CA, United States

A novel GPU based parallel motion correction method for high throughput ofMRI study is presented. With ultra-high processing speed, this method enables real-time motion correction while allowing future integration of computationally intense processing steps including iterative reconstruction and automatic segmentation for high-throughput interactive brain circuit analysis. The algorithm utilizes an iterative inverse update strategy, which dramatically reduces computational cost. GPU specific features such as texture caching and hard-wired interpolation are also utilized for the highest efficiency. Compared to currently available methods, the proposed algorithm shows the lowest RMS error rate and highest speed.