Zijing Dong^1,2, Fuyixue Wang^1,3, Jie Xiang⁴, and Kawin Setsompop^5,6
1Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States, ²Department of Electrical Engineering and Computer Science, MIT, Cambridge, MA, United States, ³Harvard-MIT Health Sciences and Technology, MIT, Cambridge, MA, United States, ⁴Tsinghua University, Beijing, China, ⁵Department of Radiology, Stanford University, Stanford, CA, United States, ⁶Department of Electrical Engineering, Stanford University, Stanford, CA, United States

A motion-correction method is developed for the recently proposed 3D-EPTI to achieve fast and motion-robust quantitative imaging of the human brain. A 4D-navigator (x-y-z-echoes) is inserted into the relaxation-recovery dead-time of the sequence to provide accurate estimations of 3D-motion and B₀-inhomogeneity changes at every TR (~2-3s), which are incorporated into a motion-and-phase corrected subspace reconstruction. The navigator utilizes an optimized spatiotemporal encoding to acquire central 3D k-space for accurate motion-estimation using just 4 small-flip-angle excitations, resulting in negligible signal-recovery reduction (<1%) to the 3D-EPTI acquisition. Simulation and in-vivo experiments preliminarily validate the accuracy of estimation and effectiveness of the reconstruction.

Emil Ljungberg¹, Tobias Wood¹, Ana Beatriz Solana², Steven C.R. Williams¹, Gareth J. Barker¹, and Florian Wiesinger^1,2
1Neuroimaging, Institute of Psychiatry, Psychology, and Neuroscience, King's College London, London, United Kingdom, ²ASL Europe, GE Healthcare, Munich, Germany

In this work we present MERLIN (Motion Elimination in Radial acquisition Leveraging Interleaved Navigators): a new method for silent, motion insensitive, MRI using self-navigated zero echo time (ZTE) imaging. Using T₁w ZTE neuroimaging as an example, we demonstrate that MERLIN can correct for rigid body motion and markedly improve image quality. Such a silent and motion insensitive neuroimaging protocol can save time and money in both clinical and research settings.

Nikos Priovoulos¹, Mads Andersen², Vincent O Boer³, and Wietske van der Zwaag^1
1Spinoza Center, Amsterdam, Netherlands, ²Philips Healthcare, Copenhagen, Denmark, ³Danish Research Centre for Magnetic Resonance, Centre for Functional and Diagnostic Imaging and Research, Copenhagen University Hospital, Hvidovre, Denmark

The cerebellar cortical microstructure is important for several functions, like motor control, but remain largely underexplored in vivo due to its small thickness. Here, we combined a susceptibility-weighted acquisition with in-plane resolution of just 0.2x0.2mm² at 7Tesla along with a fat-navigator-based prospective motion correction. Using this setup, the cerebellar layers could be seen in the phase images of all 4 MRI-naive participants. Our results show that the cerebellar layers can be consistently visualized opening new neuroscientific and clinical dimensions.

Tom Wilkinson^1,2, Felipe Godinez^1,2, Yannick Brackenier^1,2, Raphael Tomi-Tricot^1,2,3, Lucilio Cordero-Grande^1,2,4, Philippa Bridgen^1,2, Sharon Giles^1,2, Joseph V Hajnal^1,2, and Shaihan J Malik^1,2
1Biomedical Engineering Department, School of Biomedical Engineering and Imaging Sciences, Kings College London, London, United Kingdom, ²Centre for the Developing Brain, School of Biomedical Engineering and Imaging Sciences, Kings College London, London, United Kingdom, ³MR Research Collaborations, Siemens Healthcare Limited, Frimley, United Kingdom, ⁴Biomedical Image Technologies, ETSI Telecomunicación, Universidad Politécnica de Madrid and CIBER-BNN, Madrid, Spain

A ‘pilot-tone’ implementation for 7T head MRI was constructed by broadcasting RF at the appropriate frequency into the scanner room during data acquisition. This signal was demonstrated to enable motion estimation, when calibrated first by correlating measurements with motion estimates from image registration. Subsequently these estimates were compared with others obtained from the iterative DISORDER joint motion estimation & reconstruction method. These independent methods of motion estimation can potentially improve or replace other methods of motion correction at ultra-high field where motion-correction is particularly relevant. 

Florian Wiesinger¹, Timothy Deller², Floris Jansen², Jose de Arcos Rodriguez¹, Ronny R Buechel³, Philipp A Kaufmann³, and Edwin EGW ter Voert^3
1GE Healthcare, Munich, Germany, ²GE Healthcare, Waukesha, WI, United States, ³Department of Nuclear Medicine, University Hospital Zurich, Zurich, Switzerland

Respiratory motion correction is a long-standing problem in hybrid PET/MR imaging with many partial solutions.  Here we present a novel method based on extracting respiratory motion information from the PET data using the ultra-fast listmode reconstruction framework.  Doing so a highly accurate respiratory waveform derived from the inside of the body (i.e. lung liver interface) is obtained for free without requiring an extra motion sensor or complicating the PET/MR imaging workflow.  

Daniel Polak^1,2, Daniel Nicolas Splitthoff¹, Berkin Bilgic^2,3,4, Lawrence L. Wald^2,3,4, Kawin Setsompop⁵, and Stephen F. Cauley^2,3,4
1Siemens Healthcare GmbH, Erlangen, Germany, ²Department of Radiology, A. A. Martinos Center for Biomedical Imaging, Charlestown, MA, United States, ³Department of Radiology, Harvard Medical School, Boston, MA, United States, ⁴Harvard-MIT Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge, MA, United States, ⁵Department of Radiology, Stanford, Stanford, CA, United States

SAMER is a navigation-free retrospective motion-correction technique which achieves rapid motion estimation using an ultra-fast, low-resolution scout scan as an image prior. In this work, the SAMER framework is extended to 3D volumetric reconstructions of 2D TSE imaging data. The optimized 3D volumetric scout scan is combined with a distributed 2D TSE slice ordering for fully separable motion estimation with negligible added scan time. The motion correction performance was evaluated in-vivo for representative motion trajectories and compatibility to highly accelerated Simultaneous Multi-Slice acquisitions is demonstrated.

Wen Shi^1,2,3, Jiwei Sun¹, Yamin Li³, Cong Sun⁴, Tianshu Zheng¹, Yi Zhang¹, Guangbin Wang⁴, and Dan Wu^1
1Key Laboratory for Biomedical Engineering of Ministry of Education, Department of Biomedical Engineering, College of Biomedical Engineering & Instrument Science, Zhejiang University, Hangzhou, China, ²2. Department of Biomedical Engineering, Johns Hopkins University School of Medicine, Baltimore, MD, United States, ³School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China, ⁴Department of Radiology, Shandong Medical Imaging Research Institute, Cheeloo College of Medicine, Shandong University, Jinan, China

Prenatal MRI of fetal brain is vulnerable to unpredictable fetal motion and maternal movement. The conventional registration-based motion correction methods sometimes fail in excessive motion. In this work, we proposed a learning-based scheme to estimate fetal brain motion using a deep recursive framework, which replicated the iterative slice-to-volume registration and 3D volumetric reconstruction process. The network outperformed the previous learning-based methods and with good computational efficiency compared to traditional method. It also achieved high super-resolution reconstruction accuracy on simulated motion-corrupted slices, and therefore, is promising for fetal brain MRI analysis. 

Thomas Küstner^1,2, Jiazhen Pan³, Haikun Qi², Gastao Cruz², Kerstin Hammernik^3,4, Christopher Gilliam⁵, Thierry Blu⁶, Sergios Gatidis¹, Daniel Rueckert^3,4, René Botnar², and Claudia Prieto^2
1Department of Radiology, Medical Image and Data Analysis (MIDAS), University Hospital of Tübingen, Tübingen, Germany, ²School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom, ³AI in Medicine and Healthcare, Klinikum rechts der Isar, Technical University of Munich, München, Germany, ⁴Department of Computing, Imperial College London, London, United Kingdom, ⁵RMIT, University of Melbourne, Melbourne, Australia, ⁶Chinese University of Hong Kong, Hong Kong, Hong Kong

Estimation of non-rigid motion is an important task in respiratory and cardiac motion correction. Usually, this problem is formulated in image space via diffusion, parametric-spline or optical flow methods. However, image-based registration can be impaired by aliasing artefacts or by estimating in low image resolution in cases of highly accelerated acquisitions. In this work, we propose a novel deep learning-based non-rigid motion estimation directly in k-space, named LAPNet. The proposed method, inspired by optical flow, is compared against registration in image space and tested for respiratory and cardiac motion as well as different acquisition trajectories providing a generalizable diffeomorphic registration.

Kwang Eun Jang^1,2, Mario O. Malavé¹, Dwight G. Nishimura¹, and Shreyas S. Vasanawala^3
1Magnetic Resonance Systems Research Lab (MRSRL), Department of Electrical Engineering, Stanford University, Stanford, CA, United States, ²Department of Bioengineering, Stanford University, Stanford, CA, United States, ³Department of Radiology, Stanford University, Stanford, CA, United States

Motion remains a major challenge in MRI. Many motion-corrected reconstruction methods are available, yet models are often simplified. We propose image-space gridding that resamples images onto arbitrary grids, which provides a pair of operators that represents the forward and adjoint of a nonrigid transform. This allows existing nonrigid image registration techniques to be incorporated into model-based reconstructions. We apply this method to correct for respiratory motion in free-breathing cardiac MRI. Data from individual heartbeats are binned to reconstruct image-based self-navigators. Nonrigid motion is estimated using a diffeomorphic demons algorithm, and corrected by solving an optimization problem with image-space gridding operators.

Sihao Chen¹, Cihat Eldeniz¹, Weijie Gan¹, Ulugbek Kamilov¹, Tyler Fraum¹, and Hongyu An^1
1Washington University in St. Louis, Saint Louis, MO, United States

Dynamic contrast-enhanced MRI (DCE-MRI) of the liver offers structural and functional information for assessing the contrast uptake visually. However, respiratory motion and the requirement of high temporal resolution make it difficult to generate high-quality DCE-MRI. In this study, we proposed a novel forward-Fourier motion-corrected reconstruction utilizing deep learning based 3D motion information on severely undersampled DCE-MRI. With no need to use view-sharing or DCE contrast smoothness constraint, this approach avoids enhancement spillover from adjacent DCE contrasts and reconstructs high-quality motion-free DCE images with reduced artifacts and enhanced sharpness.