Ultra Fast Registration of Multiple MR Volumes Using MOPED
Mark E. Bastin¹, Benjamin D. Panter², Robert J. Tweedie², William J. Hossack³, Alan F. Heavens^2
1Medical Physics, University of Edinburgh, Edinburgh, Midlothian, United Kingdom; ²Institute for Astronomy, University of Edinburgh, Edinburgh, United Kingdom; ³Physics, University of Edinburgh, Edinburgh, United Kingdom

Registration is a critical step in the calculation of imaging biomarkers derived from functional, diffusion, perfusion and permeability MRI. These datasets typically comprise many tens of volumes, and contain up to 100 individual images, registration of which leads to a significant computational overhead in the processing pipeline. In this abstract we present initial results from the application of a novel registration method based on the MOPED algorithm, developed in the field of astronomy, which has the potential to reduce significantly the time taken to align high dimensional MRI data.

Multi-Modal Structural Networks:  Mapping of Connectivity Through Diffusion, Functional, and Structural Assessment of Intervening Pathways
John A. Bogovic¹, Min Chen¹, Aaron Carass¹, Pierre-Louis Bazin², Dzung Pham², Susan M. Resnick³, Jerry L. Prince^1,4, Bennett Allan Landman^4,5
1Electrical and Computer Engineering, Johns Hopkins University, Baltimore, MD, United States; ²Radiology, Johns Hopkins University, Baltimore, MD, United States; ³Laboratory of Personality and Cognition, National Institute on Aging, Baltimore, MD, United States; ⁴Biomedical Engineering, Johns Hopkins University, Baltimore, MD, United States; ⁵Electrical Engineering, Vanderbilt University, Nashville, TN, United States

Understanding anatomical connectivity and multivariate relationships in neuroimaging data may be essential to elucidate multiple small changes across the brain that combine to manifest in observable phenotypes. While there are powerful tools to assess connectivity through graphs using diffusion weighted MRI (DW-MRI), association of DW-MRI metrics with connectivity necessitates ad hoc choices. Herein, we show how connectivity can be interpreted by multimodal characterization of the tissues through which estimated tracts pass (in addition to metrics on the DW-MRI tracts). We define and compute multi-modal structural networks, which are multivariate graphs representing connectivity among structural regions.

MR-Based Whole-Body PET Attenuation Correction in Hybrid PET/MRI: A Computationally Inexpensive Algorithm for T1, T2, and Proton Density Weighted Images
Harry Robert Marshall^1,2, Robert Z. Stodilka, ¹², Benoit Lewden², Jean Theberge^1,2, Eric Sabondjian^1,2, Alexandre G. Legros², Andrea J. Mitchell², Lela Deans², Jane M. Sykes², R. Terry Thompson^1,2, Frank S. Prato^1,2
1Medical Biophysics, The University of Western Ontario, London, ON, Canada; ²Imaging, Lawson Health Research Institute, London, ON, Canada

Whole-body attenuation correction of PET images remains a crucial unsolved problem in hybrid PET/MRI. We present an algorithm capable of taking any of T1, T2, or proton density weighted MRI images as input and producing a PET attenuation map of comparable quality to a gold standard CT-derived attenuation map. The idea is that no “special” MRI sequences need to be acquired solely for the purposes of attenuation correction. The algorithm was tested on nine low resolution canine images with significant motion artefacts to ensure robustness.

The algorithm ran to completion in under one minute making it practical for clinical use.

MRI Measurement of Ischemic Brain Penumbra Using an Inelastic Collision Model
Hassan Bagher-Ebadian^1,2, Panayiotis D. Mitsias¹, Mohammad Hossein Asgari¹, Michael Chopp^1,2, James Russel Ewing^1,2
1Department of Neurology, Henry Ford Hospital, Detroit, MI, United States; ²Department of Physics, Oakland University, Rochester, MI, United States

Experimental and clinical studies indicate that the likelihood for progression to infarction in the penumbra of physiologically impaired but potentially salvageable tissue surrounding the central core of focal cerebral ischemia is an important factor in evaluating treatment efficacy. Thus, a multi-parametric analysis that increases the ability of investigators to detect and characterize ischemic penumbra in the early stages of stroke have a profound clinical significance. In this study, a mechanical model of inelastic collision is recruited and adapted to information theory for constructing a model-based algorithm for multi-parametric analysis of MR information in acute stroke to detect ischemic brain penumbra.

A General Framework for the Analysis of Vessel Encoded Arterial Spin Labelling
Michael A. Chappell^1,2, Tom W. Okell¹, Peter Jezzard¹, Mark W. Woolrich^1
1FMRIB Centre, University of Oxford, Oxford, United Kingdom; ²Institute of Biomedical Engineering, University of Oxford, Oxford, United Kingdom

Vessel Encoded ASL offers non-invasive vascular territory images. By spatially modulating the ASL label over a series of acquisitions blood from individual arteries is uniquely encoded such that its contribution can subsequently be extracted in the analysis. We propose a framework for the analysis of VE-ASL that combines the advantages of the two leading analysis approaches and is able to estimate perfusion even in areas supplied by multiple arteries in the face of limited data quality and quantity.

In Vivo Myelin Water Imaging Using 3D Multi-Gradient-Echo Pulse Sequences
Claudia Lenz¹, Klaus Scheffler¹, Markus Klarhöfer^1
1Radiological Physics, University of Basel Hospital, Basel, Switzerland

Quantitative imaging of the myelin water fraction (MWF) is able to show demyelinating processes and therefore provides insight into the pathology of white matter diseases such as multiple sclerosis. So far, mapping of the MWF most often was performed using single-slice multi-echo spin-echo sequences. Lately, a different approach, using multi-gradient-echo pulse sequences, was introduced by one study measuring formalin-fixed brains and has been adapted to in vivo measurements by different groups since then. In this work, we present a solution for 3D in vivo myelin water imaging with whole brain coverage by applying multi-gradient-echo pulse sequences and using a non-negative least squares algorithm to analyze the T2* decay.

Inferring Axon Properties with Double-PGSE MRI Using Analytical Water Diffusion Model
Wenjin Zhou¹, David H. Laidlaw^1
1Computer Science, Brown University, Providence, RI, United States

We present an analytical water diffusion model for inferring axon properties using double-PGSE MRI accounting for finite gradient pulses. Our estimation results demonstrate the feasibility of revealing axon properties including axon caliber using this approach. The model utilizes the signal intensity dependency on two gradient-pair direction variation to compensate for high-q requirement in single-PGSE experiments. Since many gradient directions can be acquired in rather short time on the current MRI scanner, this approach may suggest potential for clinical axonal-property estimation.

A Rapid, Robust, Anatomy and Atlas Guided Lesion Quantification Framework from Diffusion Weighted MR Images - not available
Sumit K. Nath¹, Dattesh Dayanand Shanbhag¹, Rakesh Mullick¹, Uday Patil¹, Marie Luby², Katherine D. Ku², Lawrence L. Latour², Steven Warach², NINDS Natural History of Stroke Investigators^2
1Imaging Technologies, GE Global Research, Bangalore, Karnataka, India; ²NINDS, NIH, Bethesda, MD, United States

A novel anatomical and atlas guided split-and-merge algorithm is presented for quantifying potential lesions in diffusion weighted MR images. Compared with a conventional non split-and-merge method, our approach leads to highly improved results when analyzed with ground truth.

Robust Automatic Rodent Brain Extraction Using Pulse-Coupled Neural Networks in 3D
Nigel Chou¹, Jolena Tan¹, Asad Abu Bakar Md Ali¹, Kai-Hsiang Chuang^1
1Laboratory of Molecular Imaging, Singapore Bioimaging Consortium, Agency for Science, Technology and Research (A*STAR), Singapore, Singapore

We present an automatic brain-extraction algorithm optimized for rodents, based on a pulse-coupled neural network (PCNN) operating in 3D. PCNN ‘links’ pixels with similar intensity, then a morphological operation is used to separate regions, of which the largest is selected as the brain mask.  Using Jaccard index and True-positive Rate as a measures of similarity to a manual gold-standard, this method showed improved performance compared to an existing algorithm (Brain Surface Extraction) and a PCNN algorithm operating in 2D mode (on slices).  Additional advantages include reduced user intervention and accurate segmentation of the olfactory bulb and paraflocculus of cerebellum.

Non-Invasive and Temporally Resolved Measurement of Ischaemic Tissue Damage in Acute Stroke Using Quantitative ²³Na Magnetic Resonance Microscopy at 7 T
Friedrich Wetterling¹, Lindsay Gallagher², Mhairi I. Macrae³, Sven Junge⁴, Andrew John Fagan^5
1School of Physics, Trinity College Dublin, Dublin, Ireland; ²Glasgow Experimental MRI Centre, Division of Clinical Neuroscience, Faculty of Medicine, University of Glasgow, Glasgow, Scotland, United Kingdom; ³Glasgow Experimental MRI Centre, Division of Clinical Neuroscience, Faculty of Medicine, University of Glasgow, Glasgow, Scotland, United Kingdom; ⁴Bruker BioSpin GmbH, Ettlingen, Germany; ⁵Centre for Advanced Medical Imaging, St. James’s Hospital, Dublin, Ireland

In the current study, quantitative ²³Na Magnetic Resonance Microscopy (qNa MRM) was used to measure the time course of Tissue Sodium Concentration (TSC) in order to investigate regional variations in TSC behavior in the first 8 hours after stroke in a rodent model.  The timecourse of the TSC evolution was reproducible (n=5) with similar regional delays evident in the timepoint at which the TSC increased during the first hours after MCAO in each rat.  The delay time parameter could be used as a measure of ischaemic core tissue growth, non-invasively and temporally resolved, thereby offering an alternative method to post-mortem histology.