MRS Acquisition Methods
 

Wednesday 3 June 2015

Eulalia Serés Roig¹ and Rolf Gruetter^1,2
1Laboratory of Functional and Metabolic Imaging (LIFMET), Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Vaud, Switzerland, ²Department of Radiology, Universities of Lausanne and Geneva, Vaud, Switzerland

Natural abundance 13C MRS allows the detection of a wide range of 13C signals, such as glycogen in human muscle at 7T. The energy required for muscular contraction can be assessed by dynamic 13C MRS while measuring glycogen and lipid levels before and after exercise. The large chemical shift dispersion comprising glycogen and lipid resonances requires uniform 13C-excitation and broadband 1H-decoupling for a reliable assessment of 13C signal intensities. In this study, changes of glycogen and lipid levels were monitored in the human calf before and after exercise by 13C MRS at 7T using uniform 13C-excitation and broadband 1H-decoupling.

Lucas Lindeboom^1,2, Robin A de Graaf³, Christine I Nabuurs^2,4, Matthijs KC Hesselink⁴, Joachim E Wildberger², Patrick Schrauwen¹, and Vera B Schrauwen-Hinderling^1,2
1Department of Human Biology, Maastricht University Medical Center, Maastricht, Netherlands, ²Department of Radiology, Maastricht University Medical Center, Maastricht, Netherlands, ³Department of Diagnostic Radiology, Magnetic Resonance Research Center, Yale University School of Medicine, New Haven, CT, United States, ⁴Department of Human Movement Sciences, Maastricht University Medical Center, Maastricht, Netherlands

¹³C-enriched lipid tracers are an excellent candidate for in vivo MRS tracer studies on lipid metabolism. Indirect detection of ¹³C-lipid signals can be useful to enhance the detection sensitivity. Single shot methods are preferred in tissue where movement artifacts and/or low fractional enrichments are expected. We here compared the use of modified versions of the conventional ge-HSQC and ge-HMQC to get non phase distorted 1D 1H-[¹³C]-edited spectra. We show that the ge-HSQC sequence yields the highest signal and is therefore a candidate to be used for studies following the incorporation of ¹³C-lipid tracers in human tissue.

Amandine COUM^1,2, Lobna OULDAMER^3,4, Laurent BARANTIN⁵, Fanny NOURY^1,2, Anne VILDE⁶, Aymeric SAINT-HILAIRE⁶, Philippe BOUGNOUX^4,7, and Giulio GAMBAROTA^1,2
1LTSI, Université de Rennes 1, Rennes, France, ²INSERM UMR 1099, Rennes, France, ³Department of Gynecology, CHU Tours, Tours, France, ⁴INSERM U1069, Université François-Rabelais, Tours, France, ⁵INSERM U930, Université François-Rabelais, Tours, France, ⁶Department of Radiology, CHU Tours, Tours, France, ⁷Department of Oncology, CHU Tours, Tours, France

Feasibility of an in vivo fatty acids quantification in breast tissue on clinical settings (3 T).

Hai Lu¹ and Shumin Wang^1
1Auburn University, Auburn, AL, United States

RF coils for multinuclear spectroscopy are resonant structures simultaneously tuned at discrete Larmor frequencies. Although popular, coils need to be designed careful to avoid the interference of the different modes for different nuclei species. They also need to be tuned to a drastically different frequency if the nuclei species changes. A novel broadband RF transmitter is developed that utilizes the frequency-independent transverse electromagnetic (TEM) mode of a parallel-plate waveguide. Without using any reactive tuning components, the simple coil structure can function from 100 to 300 MHz. It was applied to 31P spectroscopy of human forearm at 7 Tesla.

Carolina Campanha Fernandes¹, Emma Louise Hall¹, Chen Chen¹, Peter Gordon Morris¹, and Carlos Garrido Salmon^1,2
1Sir Peter Mansfield Imaging Centre, University of Nottingham, Nottingham, United Kingdom, ²Department of Physics, University of Sao Paulo, Ribeirao Preto, Brazil

The aim of this study is to derive an analytical model that takes into account the effects of the chemical shift displacement artefact (in three dimensions) on J-modulation of weakly coupled spins in the STEAM sequence. The signal modulation obtained theoretically was compared to the one obtained experimentally using a lactate phantom. This work demonstrated that significant changes in the coupled-spin response arise due to the chemical shift displacement artefact, especially at ultra-high magnetic field strengths. This model provides a means to predict the resulting lineshapes of metabolites of AXn form and is a useful tool for optimization of sequence timing parameters.

Veronika Rackayova¹, Cristina Cudalbu², Lijing Xin¹, Nicolas Kunz³, Jana Starcukova⁴, Zenon Starcuk, Jr.⁴, and Rolf Gruetter^1,2
1Laboratory of Functional and Metabolic Imaging, Center for Biomedical Imaging, Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Vaud, Switzerland, ²Centre d'Imagerie Biomedicale (CIBM), Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Vaud, Switzerland, ³Centre d'Imagerie Biomedicale (CIBM-AIT), Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Vaud, Switzerland, ⁴Institute of Scientific Instruments of the Academy of Sciences of the Czech Republic, Brno, Czech Republic

With increasing magnetic field all metabolites have tendency to shorten their apparent T2 relaxation time and thus the use of ultra-short TE (<10ms) is recommended. The aim was to test how TE affects the SNR, the precision and the number of reliable quantifiable metabolites. With increasing TE, we could observe decreased number of quantifiable metabolites as well as increase in mean CRLBs together with decreased SNR. As expected, TE has an important effect on the reliability of metabolites quantification. Ultra-short TE combined with a reliable quantification software is a preferential choice.

Ariane Fillmer¹ and Anke Henning^1,2
1Institute for Biomedical Engineering, UZH and ETH Zurich, Zurich, Switzerland, ²Max Planck Institute for Biological Cybernetics, Tuebingen, Germany

The transition of Magnetic Resonance Spectroscopy to ultra-high fields promises major advantages with respect to SNR and spectral resolution. In order to gain meaningful results, excellent B₀ shimming is of outermost importance. Naturally, the amplitudes of the shim fields are limited by hardware specifications. However, an optimal B₀ inhomogeneity compensation might not be possible within these hardware constraints, especially for regions with large inhomogeneities, such as the frontal cortex. This work investigates the hardware requirements for optimal B₀ inhomogeneity compensation and the effect of shim hardware constraints on B₀ shimming for a spectroscopy voxel in the frontal cortex.

Clark Lemke¹, Aaron Hess², Jamie Near³, Stuart Clare¹, Peter Jezzard¹, and Uzay Emir^1
1FMRIB, University of Oxford, Oxford, Oxfordshire, United Kingdom, ²OCMR, University of Oxford, Oxford, Oxfordshire, United Kingdom, ³Douglas Institute, McGill University, Verdun, Quebec, Canada

An in-vivo implementation of a novel technique to acquire long echo time spectroscopy without J-modulation has been compared with conventional PRESS at short (30 ms) and long (60 ms) echo times. The novel technique, entitled PRESS-JR, repeats a so-called “perfect-echo” refocusing unit in order to remove J-modulation at arbitrary echo time. At long echo time, PRESS-JR produces desirable spectra without baseline contamination or J-modulation. This technique could be very useful for disease conditions such as tumors or epilepsy since larger and more complex baseline signals result in decreased quantification accuracy for coupled metabolites at short echo times.

Liangjie Lin¹, Zhiliang Wei¹, Jian Yang¹, Yanqin Lin¹, and Zhong Chen^1
1Electronic Science, Xiamen University, Xiamen, Fujian, China

The inherent heterogeneity of the samples or living organisms can lead to magnetic field fluctuations and losses of local homogeneity. Intermolecular multiple quantum coherences (iMQCs) can be used to obtain high resolution magnetic resonance spectroscopy (MRS) under field inhomogeneity, while intrinsic low signal-to-noise ratio (SNR) of iMQC signals hampers their practical applications. Here, a pulse sequence is designed to acquire 2D high-resolution spectra under inhomogeneous fields with high SNRs. The high-resolution chemical-shift difference information and J-coupling splitting information are separately shown in two independent dimensions.

Hidehiro Watanabe¹, Nobuhiro Takaya¹, and Fumiyuki Mitsumori^1
1Center for Environmental Measurement and Analysis, National Institute for Environmental Studies, Tsukuba, Ibaraki, Japan

High signal to noise ratio and good peak resolution are significant features in ¹H MRS at high B₀ field. However, large chemical displacement and B₁ inhomogeneity may weight a peak on a localized spectrum by different reception sensitivity due to chemical shift. For example, magnitudes of peaks of NAA and Cr are weighed by the different values and this may lead to quantitation error. In this work, we will demonstrate this error in phantom experiments at 4.7 T. We will also demonstrate correction method using the profile of reception sensitivity measured by water signals.

Ashley D Harris^1,2, Nicolaas AJ Puts^1,2, Peter B Barker^1,2, and Richard A. E. Edden^1,2
1The Russell H Morgan Department of Radiology and Radiological Sciences, The John Hopkins School of Medicine, Baltimore, Maryland, United States, ²F.M. Kirby Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, Maryland, United States

-

Felix Raschke¹, Ralph Noeske², Dorothee P Auer¹, and Dineen Rob^1
1Sir Peter Mansfield Imaging Centre, School of Medicine, University of Nottingham, Nottingham, Nottinghamshire, United Kingdom, ²GE Healthcare, Berlin, Germany MR Application and Workflow Development, Berlin, Germany

In this study we use 1H MR spectroscopy to investigate if a conventional short echo STEAM sequence can be used to measure glutathione in the brains of 14 healthy controls at 3T. The STEAM results are compared to that of MEGA-PRESS spectral editing. There is a strong albeit non-significant correlation between glutathione measured by MEGA-PRESS and STEAM. Additionally, glutathione measured in the STEAM sequence shows a low intra- and inter-subject variability of 5.4% and 10.8% respectively compared to the inter-subject variability of around 17% using MEGA-PRESS. Further work is now needed to validate these initial results.

Faezeh Sanaei Nehzad¹, Adriana Anton², Bill Deakin², and Stephen Williams^1
1Center for Imaging Science, University of Manchester, Manchseter, United Kingdom, ²Neuroscience and Psychiatry Unit, University of Manchester, Manchester, United Kingdom

Glutathione (GSH) is commonly quantified in the brain using spectral editing (e.g. MEGA-PRESS), but a number of studies report GSH measurement using short echo PRESS. In this study, quantification of GSH using PRESS and MEGA-PRESS is evaluated at 3T. Two identical phantoms containing brain metabolites, one with and one without GSH, were scanned using PRESS and MEGA-PRESS. Also, seven healthy volunteers were scanned with the same protocol. The results from PRESS spectra, quantified with QUEST and from MEGA-PRESS, quantified with AMARES, were compared. We conclude that GSH cannot be reliably quantified using short echo PRESS.

Felix Raschke¹, Antonio Napolitano², Ralph Noeske³, Dineen Rob¹, and Dorothee P Auer^1
1Sir Peter Mansfield Imaging Centre, School of Medicine, University of Nottingham, Nottingham, Nottinghamshire, United Kingdom, ²Enterprise Risk Management, Unity of Imaging Research, Bambino Gesù Children’s Hospital, Rome, Italy, ³GE Healthcare, Berlin, Germany MR Application and Workflow Development, Berlin, Germany

We compare three MRS sequences to measure GABA at 3T using 13 healthy controls. The GABA signal of the popular MEGA-PRESS sequence with TE=68ms correlates with that of macromolecule suppressed MEGA-PRESS with TE=80ms and with a previously published optimised PRESS sequence using a TE and TE1 of 105ms and 15ms respectively. The PRESS105 may be a suitable sequence to measure GABA reliably with the benefit of reduced scan time, no macro-molecule contribution and less susceptibility to subject motion compared to MEGA-PRESS.

Lijing Xin^1,2 and Rolf Gruetter^1,3
1Laboratory for Functional and Metabolic Imaging (LIFMET), École polytechnique fédérale de Lausanne, Lausanne, Vaud, Switzerland, ²Department of Psychiatry, Lausanne University Hospital, Lausanne, Vaud, Switzerland, ³Department of Radiology, University of Lausanne and Geneva, Vaud, Switzerland

This study aims to evaluate the accuracy and precision of GSH measurements at 3T with short-TE MRS under various experimental conditions using synthesized spectra with known metabolite concentrations. An underestimation of GSH concentrations was observed and the measurement accuracy was found to depend on the experimental conditions. The precision of the measurement is largely influenced by SNR regardless of the linewidth. We conclude that GSH concentrations can be measured with high precision using short-TE MRS methods at 3T when SNR is sufficient. GSH changes can also be accurately measured if experimental conditions are carefully matched between groups.

Dimitri Martel¹, Jean Baptiste Langlois², Denis Friboulet¹, Olivier Beuf¹, and Helene Ratiney^1
1CREATIS; CNRS UMR 5220; INSERM U1044; Université Lyon 1; INSA Lyon, Villeurbanne, France, ²CERMEP- Imagerie du Vivant, Bron, France

Localized Correlation Spectroscopy is used to assess evolution of liver fatty acid in an ob/ob mice model at two time points. A quantitative study is performed using a dedicated time domain quantification procedure using a simplified quantum mechanically basis set. This later is constituted by the sub spin systems involved in the 2D spectrum of triglycerides. Fatty acid Indexes and T2 values are obtained in agreement with expected model evolution.Indexes and T2 values are obtained in agreement with model evolution.

Song Chen¹, Meng Chen¹, Congyu Liao¹, Linfei Wen¹, Darong Zhu², Xu Yan³, Keith Heberlein⁴, and Jianhui Zhong^1
1Center for Brain Imaging Science and Technology, Department of Biomedical Engineering, Zhejiang University, Hangzhou, Zhejiang, China, ²Hangzhou First People's Hospital, Hangzhou, Zhejiang, China, ³MR Collaboration NE Asia, Siemens Healthcare, Shanghai, China, ⁴Siemens Medical Solutions USA, Inc, Malvern, PA, United States

Gamma-aminobutyric acid (GABA) is one of important inhibitory neurotransmitters, closely related to mental illness such as depression and schizophrenia. MEGA-PRESS, a spectral editing technique in MRS, is a non-invasive method for detecting GABA concentration in vivo. In this study, data from volunteers was acquired using a robust prototype with MEGA-PRESS pulse sequence, and the residual bootstrap analysis was used to assess uncertainty in different regions, variability from repeated scans, and natural variation among different subjects.

Georg Oeltzschner^1,2 and Pallab K. Bhattacharyya^3,4
1Department of Diagnostic and Interventional Radiology, Medical Faculty, University Dusseldorf, Düsseldorf, Germany, ²Institute of Clinical Neuroscience and Medical Psychology, Medical Faculty, Heinrich-Heine-University Düsseldorf, Düsseldorf, Germany, ³Imaging Institute, Cleveland Clinic, Cleveland, OH, United States, ⁴Cleveland Clinic Lerner College of Medicine - CWRU, Cleveland, OH, United States

Editing efficiency (EE) of J-edited GABA spectroscopy is an important parameter for GABA quantification. Different methods of calculating EE were applied to spectra from a GABA phantom, including a macromolecule (MM) suppressed and a MM unsuppressed editing scheme. It was found that EE may exceed the expected maximum of 0.5 which can be attributed to the central peak of the 3.01 ppm GABA resonance. Size and shape of the central peak are shown to differ across editing schemes. Simulation of J-edited spectra indicate that this effect is driven by the OFF-resonance spectra, suggesting peak shape and size are influenced by the set of editing frequencies and sequence properties.

Jianfeng Bao^1,2, Yuchuan Zhuang¹, Yanqin Lin², Zhong Chen², and Jianhui Zhong^1
1University of Rochester, Rochester, NY, United States, ²Xiamen University, Xiamen, Fujian, China

Trabecular bone structure leads to serious local inhomogeneous B0 field, rendering it difficult to obtain high resolution fatty acid MRS in red bone marrow. Although intermolecular double-quantum coherences (iDQC) 2D MRS has been shown to overcome macroscopic inhomogeneity induced by biological tissues, this advantage has not been studied in red bone marrow in humans. In this study, a localized module iDQC was used to obtain high resolution fatty acid spectra on red bone marrow in human knee on a 3.0 T human scanner. This technique has potential in providing new biomarker information for diagnosis purposes.

Xiaobo Qu¹, Maxim Mayzel², Jian-Feng Cai³, Zhong Chen¹, and Vladislav Orekhov^2
1Department of Electronic Science, Xiamen University, Xiamen, Fujian, China, ²Swedish NMR Centre, University of Gothenburg, Gothenburg, Sweden, ³Department of Mathematics, University of Iowa, Iowa City, Iowa, United States

Accelerated multi-dimensional NMR spectroscopy is a prerequisite for studying short-lived molecular systems, monitoring chemical reactions in real-time, high-throughput applications, etc. Non-uniform sampling is a common approach to reduce the measurement time. Here, we introduce a new method for high quality spectra reconstruction from non-uniformly sampled data, which is based on recent developments in the field of signal processing theory and utilizes the so far unexploited general property of the NMR signal, its low-rank. Using experimental and simulated data, we demonstrate that the low-rank reconstruction is a viable alternative to the current state-of-the-art technique compressed sensing. In particular, the low-rank approach is good in preserving of low intensity broad peaks, and thus increases the effective sensitivity in the reconstructed spectra.

Scott Gregory Quadrelli¹, Alexander Lin², Saadallah Ramadan¹, and Carolyn Mountford^1,3
1Centre for MR in Health, The University of Newcastle, Callaghan, NSW, Australia, ²Center for Clinical Spectroscopy, Brigham & Women’s Hospital - Harvard Medical School, Boston, MA, United States, ³Center for Clinical Spectroscopy, Brigham & Women’s Hospital - Harvard Medical School, Boston, NSW, Australia

A growing literature indicates that Fucose7α(1−2)7galactose sugars are implicated in the molecular mechanisms that underlie neuronal development, learning, and memory in the human brain. An understanding of the in-vivo roles played by these terminal fucose residues, has been hampered by lack of technology to non-invasively monitor their levels in the human brain. We have implemented the in-vivo two dimensional magnetic resonance spectroscopy technology to examine the human brain in a three Tesla clinical MR scanner and report that six Fucose-α(1−2)-galactose residues, and free alpha-fucose, are available for inspection. Fucose7α(1−3)-galactose residues cannot yet be assigned using this technology as they resonate under the water resonance. This new application offers an unprecedented insight into the molecular mechanisms by which fucosylated sugars contribute to neuronal processes and how they alter during development, ageing and disease.

MRS Processing & Quantification
 

Wednesday 3 June 2015

Kyu-Ho Song¹, Sang-Young Kim¹, Do-Wan Lee¹, Jin-Young Jung¹, Hyeon-Man Baek², and Bo-Young Choe^1
1Department of Biomedical Engineering, Research Institute of Biomedical Engineering, Seoul, Seoul, Korea, ²Center for Magnetic Resonance Research, Korea Basic Science Institute, Chungbuk, Korea

Design of MRI-MRS Fused Phantom for Quantitative Evaluation

Lotte C Houtepen¹, Remmelt R Schür¹, Vincent O Boer², Bart van de Bank³, Tom Scheenen³, Anouk Marsman⁴, Christiaan H Vinkers¹, and Dennis W.J. Klomp^2
1Psychiatry, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands, ²Radiology, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands, ³Radiology, Radboud University Nijmegen Medical Centre, Nijmegen, Gelderland, Netherlands, ⁴Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore, MD, United States

Multiple software packages are available to analyze 1H Magnetic resonance spectroscopy (MRS) data. Analyzing the same scans with two software packages significantly affects the outcomes resulting from the analysis pipeline used to process the MRS data. This illustrates the necessity to exactly homogenize the details of these pipelines.

Meriem Taous Laleg¹, Zineb Kaisserli¹, Rick Achten^2,3, and Hacene Serrai^2,3
1King Abdullah University of Sciences and Engineering, Jeddah, Saudi Arabia, ²University of Gent, Gent, Belgium, ³universitair Ziukenhuis Gent, Gent, Belgium

The semi-classical signal analysis method (SCSA) is a powerful post-processing technique, which uses the discrete spectrum of the Schrödinger operator where the signal is considered as potential of this operator. It is used to separate between the useful signal and noise by means of selecting eigenfunctions belonging to the signal and discarding the noise ones. Applied here, the method is able to differentiate between the eigenfunctions of the magnetic resonance spectroscopy (MRS) signal and noise. As a result, the SNR of the MRS data is improved allowing for accurate data quantification.

Keyvan Ghassemi^1,2, Mohammadreza Khanmohammadi Khorami¹, and Hamidreza Saligheh Rad^2,3
1Chemistry Department, Faculty of Science, Imam Khomeini International University, Qazvin, Iran, ²Quantitative MR Imaging and Spectroscopy Group, Research Center for Molecular and Cellular Imaging, Tehran University of Medical Sciences, Tehran, Iran, ³Department of Medical Physics and Biomedical Engineering, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran

Low signal to noise ratio (SNR), baseline distortions, large line-widths and asymmetric line-shapes caused by poor shimming, as well as contaminations caused by significant chemical shift displacement effects produce complicated MRS signals. Totally 139 spectra from healthy and tomure glial brains –10 healthy cases,11 grade II, 6 grade III, as well as 9 grade IV brain gliomas were collected. SIMCA was used by application of PCA in common rule and by using of the NMF. Results of robust SIMCA showed significant modification in percentage of correct classified samples after application of NMF for better decomposition of noisy measurements.

Rebecca Birch^1,2, Andrew C Peet^2,3, Hamid Dehghani⁴, and Martin Wilson^2,3
1PSIBS Doctoral Training Centre, University of Birmingham, Birmingham, West Midlands, United Kingdom, ²Department of Oncology, Birmingham Children's Hospital NHS Foundation Trust, Birmingham, West Midlands, United Kingdom, ³School of Cancer Sciences, University of Birmingham, Birmingham, West Midlands, United Kingdom, ⁴School of Computer Science, University of Birmingham, Birmingham, West Midlands, United Kingdom

Accurate and reproducible MR Spectroscopic Imaging (MRSI) data is desired. Echo time (TE) is important as a trade-off between signal and fitting accuracy has to be established. Macromolecular (MM) signals form a spectral baseline; which introduces fitting error as signal overlap occurs with metabolites of interest. MM signals are predominant in Short TE spectra; therefore a well-defined MM estimation would be beneficial. The effect of echo time and experimentally acquired MM fitting on 2D MRSI reproducibility has been assessed. TE=80ms was found to be the most reproducible and short TE accuracy increased with the inclusion of an experimental acquired baseline.

Ashley D Harris^1,2, Nicolaas AJ Puts^1,2, and Richard A. E. Edden^1,2
1The Russell H Morgan Department of Radiology and Radiological Sciences, The John Hopkins School of Medicine, Baltimore, Maryland, United States, ²F.M. Kirby Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, MD, United States

Correcting GABA-edited MRS for tissue fraction is a recommended procedure; however, its implementation is inconsistent. Different corrections are compared and a recommended correction presented.

Danilo Mendes Dias Delfino da Silva¹, Thales Sinelli Lima¹, Alberto Tannús¹, Claudio José Magon¹, and Fernando Fernandes Paiva^1
1Department of Physics and Interdisciplinary Science, Sao Carlos Institute of Physics, University of Sao Paulo, Sao Carlos, Sao Paulo, Brazil

KBDM is a parametric non-linear method that allows fitting and spectral analysis of experimentally measured transient time signals. The method has already shown its potential in the MR field and its characteristics may prove valuable for clinical MRS data processing. However, the method accuracy appears to be correlated to the noise level, which may impose a limitation for this application due to typical noise present at in vivo data. Thus, the main goal of this work is to show a strategy to reduce noise impact in order to enable the method to be used on clinical MRS data processing.

Boyu Jiang¹, Xiaoping Hu², and Hao Gao^1,3
1School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, Shanghai, China, ²Department of Biomedical Engineering, Emory University and Georgia Institute of Technology, Atlanta, GA, United States, ³Department of Mathematics, Shanghai Jiao Tong University, Shanghai, Shanghai, China

A new MRS reconstruction method has been proposed using the Lorentzian-function-based sparsity, with significantly reduced number of unknown variables. The new method can achieve significantly better MRS reconstruction results than FFT method or L1-based sparsity method, e.g., even with 1% k-space data.

Evita C. Wiegers¹, Bart Philips¹, Hanne M. M. Rooijackers², Alan J. Wright³, Arend Heerschap¹, and Marinette van der Graaf^1,4
1Radiology and Nuclear Medicine, Radboud University Medical Center, Nijmegen, Gelderland, Netherlands, ²Internal Medicine, Radboud University Medical Center, Nijmegen, Gelderland, Netherlands, ³Cancer Research UK Cambridge Institute, University of Cambridge, Cambridge, United Kingdom, ⁴Pediatrics, Radboud University Medical Center, Nijmegen, Gelderland, Netherlands

When subtracting MRS data, which is commonly done in J-difference editing techniques, it is of great importance that spectra are accurately phase and frequency aligned with respect to each other. To allow accurate phase and frequency alignment, we present a simple, robust and fast alignment algorithm, based on maximizing the normalized scalar product between two spectra.

Jan Willem van der Veen¹, Stefano Marenco², and Jun Shen^1
1Magnetic Resonance Spectroscopy Core, NIH, NIMH, Bethesda, Maryland, United States, ²CTNB, NIH, NIMH, Bethesda, Maryland, United States

A new algorithm to improve eddy current correction and residual water removal was proposed and tested on in vivo single-voxel MRS data acquired from 141 healthy subjects. From the water reference scan the metabolite signal was subtracted to reveal water sidebands. Together with HSVD fit this method improves spectral baseline and allows accurate quantification of single voxel MRS.

Xin Liu¹, Yuqian Li¹, Yiming Pi¹, Sofie Van Cauter², Yi Yao^3,4, and Jiunjie Wang^5
1School of Electronic Engineering, University of Electronic Science and Technology of China, Chengdu, China, ²Department of Radiology, University Hospitals Leuven, Belgium,³School of Communication and Information Engineering, University of Electronic Science and Technology of China, Chengdu, China, ⁴National Key Laboratory of Science and Technology on Communications, China, ⁵Department of Medical Imaging and Radiological Sciences, ChangGung University, Taiwan

Magnetic resonance imaging (MRI) is widely used for the diagnosis of brain tumour. However, it failed to provide information of metabolism. Magnetic resonance spectroscopy imaging (MRSI) has shown great potentials to reveal the tumour heterogeneity, yet, with relatively poor resolution. The current study developed an unsupervised data fusion method for brain abnormality diagnosis. The accurate tissue sources from MRSI and the high resolution spatial distribution from MRI can be integrated and differentiated by using non-negative matrix factorization. The in vivo experiments on patients with brain tumour of low grade glioma and glioblastoma multiforme demonstrated successful fusion between MRSI data and MRI in brain tumour.

Qiang Ning¹, Chao Ma², and Zhi-Pei Liang^1,2
1Electrical and Computer Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ²Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, United States

Spectral quantitation is a key problem in MR spectroscopic imaging (MRSI), and a large body of work has been done to develop robust solutions for practical applications. Existing methods have effectively exploited the prior knowledge about the spectral structures of metabolites. However, very limited work has been done to exploit the spatial characteristics of metabolite distributions. As a result, state-of-the-art methods often have large estimation variations. This work addresses this problem by jointly estimating the concentration map of each metabolite, incorporating spatial regularization. Experimental results show that the proposed method significantly improved the estimates of spectral parameters over state-of-the-art methods.

Michal Považan¹, Bernhard Strasser¹, Gilbert Hangel¹, Stephan Gruber¹, Siegfried Trattnig¹, and Wolfgang Bogner^1
1MRCE, Department of Biomedical Imaging and Image-guided therapy, Medical University Vienna, Vienna, Austria

To fully take an advantage of MRSI techniques and to ensure the quality control of data, we have developed an automated post-processing tool capable of handling different MR modalities. This software creates metabolic maps, SNR, FWHM and CRLB maps and exports the results into spreadsheet for further statistical analysis. The prior knowledge obtained from MRI is used for partial volume correction of spectroscopic data as well as for the registration to the common reference frame. The use of the software led to decrease of processing time and to minimization of user-induced bias.

Ahmad Seif Kanaan^1,2, André Pampel¹, Kirsten Müller-Vahl², and Harald E. Möller^1
1Nuclear Magnetic Resonance, Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Sachsen, Germany, ²Department of Psychiatry, Social Psychiatry and Psychotherapy, Medical School of Hannover, Hannover, Niedersachsen, Germany

The quantitation of absolute metabolite concentrations via Magnetic Resonance Spectroscopy using water as an internal concentration standard, requires the accurate determination of the compartmentation within the localized region of interest. Previous work has shown that different segmentation approaches yield different estimates of metabolite levels in Grey matter [1]. In this study, we investigated the test-retest reliability of absolute metabolite quantitation using two commonly used segmentation algorithms.

Alfredo Liubomir Lopez Kolkovsky¹ and Fawzi Boumezbeur^1
1Neurospin, I2BM, Commissariat à l’Energie Atomique, Gif-sur-Yvette, Essonne, France

³¹P MRS allows the study of in vivo tissue energy metabolism by directly measuring key high-energy phosphate compounds closely coupled to the brain energetics such as ATP and PCr. In this study, ³¹P MR spectra acquired in the rat brain in vivo at 17.2 T were analyzed in the frequency domain using LCModel. ³¹P metabolite quantification was done by using brain tissue water as an internal concentration reference by employing a BISTRO outer volume suppression scheme for localization. Quantification of 11 metabolites was achieved and were in agreement with previously reported values (PCr: 4.01 ± 0.54 mM).

Roland Kreis¹ and Sreenath Pruthviraj Kyathanahally^1
1Depts. Radiology and Clinical Research, University Bern, Bern, Switzerland

It is demonstrated that the use of relative Cramer-Rao Lower Bounds (CRLB) for quality filtering can easily lead to wrong conclusions either to believe in metabolic alterations where there are none, or to miss significant effects. This is illustrated with synthetic examples. In addition, consequent application of rejection of data with high relative error prevents the clinical use of MRS in any disease leading to very low metabolite levels. CRLB are valuable to judge the trust one can have in a MRS-based measurement, but it has to be judged either as an absolute value or relative to the normal/control levels.

Niklaus Zoelch¹, Andreas Hock^1,2, Milan Scheidegger^1,2, Lea Hulka², Boris Quednow^2,3, and Anke Henning^1,4
1Institute for Biomedical Engineering, UZH and ETH Zurich, Zurich, Switzerland, ²Department of Psychiatry, Psychotherapy and Psychosomatics Hospital of Psychiatry, University of Zurich, Zurich, Switzerland, ³Zurich Center for Integrative Human Physiology, University of Zurich, Zurich, Switzerland, ⁴Max Planck Institute for Biological Cybernetics, Tuebingen, Germany

The goal of this work was to investigate absolute quantification of brain metabolites in heterogeneous voxels using internal references. To obtain reliable concentrations in heterogeneous voxels, it is necessary to include corrections for the different concentrations as well as different relaxation properties in grey matter, white matter and cerebrospinal fluid of the reference compound. Absolute concentrations were determined in 18 healthy volunteers either by using the internal water as reference or internal creatine and applying the commonly used corrections. A good agreement between the mean concentrations obtained with both reference standards could be only observed when the commonly used corrections are applied. However the correlation between these concentrations compared to correlation of the simple peak area ratios was decreased.

Ioannis Angelos Giapitzakis¹ and Anke Henning^1,2
1Max Planck Institute for Biological Cybernetics, Tuebingen, Baden-Wuerttemberg, Germany, ²Institute for Biomedical Engineering, UZH and ETH Zurich, Zurich, Switzerland

Single voxel spectroscopy benefits from the use of multi-channel coils. Several methods have been proposed in order to increase the signal to noise ratio of the combined spectrum. In particular, three methods (Brown’s method, singular value decomposition method and generalized least square) were studied to access their performance and robustness. For this purpose, simulated data were created mimicking spectrum under different decoupling conditions, in vivo data were also used. The results demonstrated that the methods produces combined spectra with similar values of SNR. However, their robustness varies in a high degree.

MRS Animal Cells
 

Wednesday 3 June 2015

Aline Seuwen¹, Aileen Schröter¹, and Markus Rudin^1,2
1Institute for Biomedical Engineering, ETH & University of Zürich, Zürich, Switzerland, ²Institute for Pharmacology and Toxicology, University of Zürich, Zürich, Switzerland

BOLD fMRI is an indirect readout for neuronal activity, which relies on hemodynamic signal changes. In mice, peripheral stimulation is accompanied by strong changes in cardiovascular parameters eventually confounding specific BOLD response. Non-vascular readouts could constitute an alternative to study stimulus-evoked brain activation. Here we propose to use spectroscopic imaging in mice to visualize more specifically neurotransmitter and metabolite changes associated to neural activation in the respective brain region. Improved spatial and temporal resolution allowed monitoring short term changes occurring when applying e.g. electrical stimuli to the mouse paw.

Ivan Tkac^1
1Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States

C57BL/6 is the most widely used mouse strain used in different research areas including developmental biology and neuroscience. In addition, C57BL/6 mice are commonly used as a background strain in the production of transgenic mouse model of human neurodegenerative diseases. But these mice exhibit a sporadic congenital portosystemic shunt resulting in abnormally high levels of brain glutamine. This study we demonstrate how ¹H MRS can help to identify these mice to eliminate potential bias in studies using this mouse strain.

Veronika Rackayova¹, Bernard Lanz¹, Corina Berset², Rolf Gruetter^1,2, Valérie A. McLin³, Olivier Braissant⁴, and Cristina Cudalbu^2
1Laboratory of Functional and Metabolic Imaging, Center for Biomedical Imaging, Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Vaud, Switzerland, ²Centre d'Imagerie Biomedicale (CIBM), Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Vaud, Switzerland, ³Swiss Center for Liver Disease in Children, Department of Pediatrics, University Hospitals Geneva, Geneva, Switzerland, ⁴Service of Biomedicine, University Hospital of Lausanne, Lausanne, Vaud, Switzerland

1H MRS and 31P MRS were used to study brain (energy) metabolism in Hepatic Encephalopathy (HE) in animal model of Chronic liver disease. We could see 2-fold increase of Gln and decrease in other brain osmolytes but only non-significant decrease in ATP suggesting that mild brain edema, present in HE, is unlikely due to energy disturbances but rather high concentrations of the osmotically active Gln look, for the moment, as one of the principal causes.

Do-Wan Lee^1,2, Jung-Whan Min³, Jung-Hoon Lee^1,4, Kyu-Ho Song¹, and Bo-Young Choe^1
1Department of Biomedical Engineering, and Research Institute of Biomedical Engineering, The Catholic University of Korea College of Medicine, Seoul, Seoul, Korea, ²Asan Institute for Life Sciences, Asan Medical Center, Seoul, Seoul, Korea, ³Department of Radiological Science, The Shingu University College of Korea, Seongnam, Korea, ⁴Department of Radiology, Kyunghee Medical Center, Seoul, Korea

This study aimed to determine the influence of the time-dependent effects of SIE exposure on cerebral neurochemical differences and responses among control (CNTL) rats and rat groups at 6-h (SIE-06H), 3-days (SIE-03D), and 7-days (SIE-07D) after the last gavage procedure, using ex vivo high-resolution spectra. Our results showed significantly higher Lac, mIns, GSH, and GABA signals were significantly differed among the CNTL, SIE-06H, SIE-03D, and SIE-07D. Our results indicate several potential metabolic markers for time-dependent influence of short-term intermittent ethanol exposure and abstinence on frontal cortex.

Jan Weis¹, Lina Carlbom¹, Lars Johansson¹, Alireza Biglarnia², Olle Korsgren³, and Håkan Ahlström^1
1Department of Radiology, Oncology and Radiation Science, Uppsala University, Uppsala, Sweden, ²Department of Surgical Sciences, Uppsala University, Uppsala, Sweden,³Department of Immunology, Genetics and Pathology, Uppsala University, Uppsala, Sweden

1H-MRS of human pancreas grafts was performed using a 1.5 T scanner. The aim of the study was estimation of pancreatic water, tCho, and lipid (-CH2-)n relaxation times T1, T2, and absolute concentration of tCho and fat. Spectra were processed by MRUI software package. T1 and T2 values were obtained by mono-exponential fitting of spectral intensities versus repetition or echo times. Knowledge of the relaxation times enables quantification of pancreas metabolite concentrations using water as the internal concentration reference. The pancreatic fat content is useful for evaluation the pancreas graft quality prior to transplantation.

Santosh Kumar Bharti¹ and Zaver Bhujwalla^1
1Div. of Cancer Imaging Research, The Russell H. Morgan Dept. of Radiolog and Radiological science, Johns Hopkins University, School of Medicine, Baltimore, Maryland, United States

Quantitative lipid analysis helps in understanding role of lipid in the cancer processes including invasion, metastasis, proliferation and many other normal and pathological conditions with clinical interest. Quantitative NMR analysis of lipids is largely affected by sample preparation and referencing methods. Evaporation of TMS during sample preparation significantly reduces the quantitative accuracy. Optimized sample preparation protocol and storage showed improved accuracy and reproducibility. QUANTAS and Stem co-axial insert containing TSP shows improved qNMR accuracy over TMS referencing method. QUANTAS offers easy and rapid quantitative lipids analysis when dealing with large no of sample and scaling the data for statistical analysis.

Verena Hoerr^1,2, Gavin E. Duggan³, Lori Zbytnuik⁴, Karen K.H. Poon⁴, Bettina Löffler², and Hans J. Vogel^3
1Department of Clinical Radiology, University Hospital Muenster, Muenster, Germany, ²Institute of Medical Microbiology, Jena University Hospital, Jena, Germany, ³Department of Biological Sciences, University of Calgary, Calgary, Alberta, Canada, ⁴Department of Physiology and Pharmacology, University of Calgary, Calgary, Alberta, Canada

In our study we used 1H NMR metabolomics to investigate the metabolic response of Escherichia coli cultures to antibiotics with different mode of action. Within this frame multivariate statistical pattern recognition analysis was applied on intracellular fingerprints and extracellular footprints to obtain complementary information. PLS-DA analysis of metabolic fingerprint profiles clearly separated intra- from extracellular antibiotic action. In contrast metabolic footprint analysis could distinguish between different antibiotics that all inhibit cell wall biosynthesis. Using a training set of E. coli fingerprints, extracted after treatment with different antibiotic classes, we could predict the mode of action of corresponding antibiotic analogs.

Gaëlle Diserens¹, Christina Precht², Martina Vermathen³, Anna Oevermann⁴, Chris Boesch¹, and Peter Vermathen^1
1Depts. Radiology and Clinical Research, University Bern, Bern, Switzerland, ²Dept. of Clinical Veterinary Medicine, University Bern, Bern, Switzerland, ³Dept. of Chemistry & Biochemistry, University Bern, Bern, Switzerland, ⁴Dept. of Clinical Research and Veterinary Public Health, University Bern, Bern, Switzerland

Proton HR-MAS NMR can assess mobile lipids from different tissues. The study purpose was to better understand the origin of visible lipids and the cause of this lipid visibility, by investigating the effect of spinning speed variation in HR-MAS NMR I) in different organs, comprising liver, cardiac muscle, breast, kidney and adipose tissue, and II) in white and grey matter accompanied by histological lipid visualization. The spinning speed significantly affects the lipid visibility. Histology results support the hypothesis that lipids from myelinated membranes are NMR-visible at high spinning speed, while lipid droplets may be less affected by the spinning speed.

Gaëlle Diserens¹, Martina Vermathen², Ilche Gjuroski², Sandra Eggimann³, Christina Precht⁴, Chris Boesch¹, and Peter Vermathen^1
1Depts. Radiology and Clinical Research, University Bern, Bern, Switzerland, ²Dept. of Chemistry & Biochemistry, University Bern, Bern, Switzerland, ³University Institute of Clinical Chemistry, Bern University Hospital, Bern, Switzerland, ⁴Dept. of Clinical Veterinary Medicine, University Bern, Bern, Switzerland

The study aim was to unambiguously assign nucleotide sugars, mainly UDP-X that are known to be important in glycosylation processes, and glucose-phosphates that are important intermediate metabolites for storage and transfer of energy, directly in spectra of intact cells, as well as in skeletal and heart muscle biopsies by ¹H HR-MAS NMR. A second aim was to study the kinetic of Glc-1P, identified in cardiac muscle. The results demonstrate that sugar phosphates can be determined quickly and non-destructively in cells and biopsies by HR-MAS, which may prove valuable considering the importance of phosphate-sugars in cell metabolism for nucleic acid synthesis.

Naranamangalam R Jagannathan¹, Deepti Upadhyay¹, Uma Sharma¹, Kamini Dangat², Anita Kilari², Savita Mehendale³, Sanjay Lalwani⁴, and Sadhana Joshi^5
1Department of NMR & MRI Facility, All India Institute of Medical Sciences, New Delhi, Delhi, India, ²Department of Nutritional Medicine, Bharati Vidyapeeth University, Pune, Maharashtra, India, ³Department of Obstetrics and Gynecology, Bharati Vidyapeeth University, Pune, Maharashtra, India, ⁴Department of Pediatrics, Bharati Vidyapeeth University, Pune, Maharashtra, India, ⁵Bharati Vidyapeeth University, Pune, Maharashtra, India

Preeclampsia (PE) is a pregnancy specific syndrome and is associated with vascular pathology which may affect the lactogenesis. Proton NMR based metabolomics of milk of PE mothers showed lower concentration of glycine and glycerophosphocholine as compared to control women. Glycine and glycerophosphocholine are known to be associated with neurodevelopment of the infant and alteration in levels of these metabolites may affect the infant’s growth and particularly brain development.

SONIA GANDHI¹, HEMANTH KUMAR B S¹, SUNIL KOUNDAL¹, SHUBHRA CHATURVEDI², RAJENDRA P TRIPATHI¹, and SUBASH KHUSHU^1
1NMR Research Centre, INMAS, DELHI, DELHI, India, ²Division and Cyclotron & Radiopharmaceutical Sciences, INMAS, DELHI, India

The effects of cold stress can be detected in all aspects of body functioning resulting in multisystem stress such as cardiovascular and respiratory diseases, cognitive deficits and neurological changes. Present study investigates the changes in metabolic profiles of spleen tissue in rats due to prolonged cold stress using NMR & multivariate statistical analysis. Results showed changes in metabolites such as taurine, choline, creatine, myo-inositol and aspartic acid on cold exposure up to 15 days indicating osmotic imbalance, altered reabsorption, changes in cell integrity and signaling role for cell membrane. Correlation with other parameters can be helpful in detecting early biomarkers for cold stress injuries in humans and organ specific dosimetry.

Savita Singh¹, Sujeet Mewar², Deepti Upadhyay², Uma Sharma², Anand Jaiswal³, Rohit Sarin³, Naranamangalam R Jagannathan², and H K Prasad^1
1Department of Biotechnology, All India Institute of Medical Sciences, New Delhi, Delhi, India, ²Department of NMR and MRI Facility, All India Institute of Medical Sciences, New Delhi, Delhi, India, ³Department of TB and Respiratory Diseases, National Institute of Tuberculosis and Respiratory Diseases, New Delhi, Delhi, India

In this study, 1H NMR spectroscopy was employed to identify metabolites present in sera of untreated pulmonary tuberculosis patients (PTB =15); household healthy contacts (HHC=13); healthy controls (HC=14); and treated pulmonary tuberculosis patients (TPTB=13). One–way analysis of variance followed by Bonferroni test was used for determining the statistical significance of estimated metabolite concentrations. Of the 13 metabolites that showed significant difference between various groups, citrate was found to be reduced in PTB patients compared to all other groups. Our study revealed that NMR based metabonomics may provide an insight to understand the biochemical alteration(s) that occur in PTB patients.

Omkar B Ijare¹, Tedros Bezabeh^1,2, Tom Blydt-Hansen³, Martin Reed⁴, and Ian C.P. Smith^1
1Chemistry, University of Winnipeg, Winnipeg, MB, Canada, ²Human Nutritional Sciences, University of Manitoba, Winnipeg, MB, Canada, ³Pediatrics, University of British Columbia, Vancouver, BC, Canada, ⁴Radiology, University of Manitoba, Winnipeg, MB, Canada

Urinary tract infection (UTI) is the most common bacterial infection in adults (especially women) and children. The majority of UTIs are caused by uropathogenic Escherichia coli (UPEC), accounting for 70 – 90% of the community acquired UTIs. Due to the longer diagnostic wait time required for the gold standard (the culture method), dipstick methods are commonly used for the quick diagnosis of UTI. Given the high false negative and false positive rates of dipstick methods, other instrumental methods such as ¹H NMR-based metabolic profiling of urine samples have been proposed. This ¹H NMR study on urine samples revealed the presence of elevated levels of acetate in urine samples from pediatric UTI patients compared to the control group.

Xiaohong Cui¹, Yali Jin¹, Honghao Cai¹, Yulan Lin¹, and Zhong Chen^1
1Department of Electronic Science, Fujian Provincial Key Laboratory of Plasma and Magnetic Resonance, Xiamen University, Xiamen, Fujian, China

NMR spectroscopy has become an important technique to study biological tissues. Since tissues are in semisolid-state, high-resolution (HR) spectra of tissues cannot be obtained by conventional NMR spectroscopy. In this study, the feasibility of HR ex vivo spectroscopy based on ultrafast intermolecular single-quantum coherence (UF iSQC) technique is explored using fish muscle and eggs. Furthermore, the metabolites variations of fresh and decayed tissues are explored. The experimental results indicate that UF iSQC technique is a fast feasible way for HR NMR spectrum of fish tissues and suitable for quality control of semisolid food and inspections on biological tissues in vivo.

Christina Precht¹, Gaëlle Diserens², Martina Vermathen³, Anna Oevermann⁴, Torsten Seuberlich⁴, Josiane Lauper¹, Daniela Gorgas¹, Chris Boesch², and Peter Vermathen^2
1Dept. of Clinical Veterinary Medicine, University Bern, Bern, Bern, Switzerland, ²Depts. Radiology and Clinical Research, University Bern, Bern, Switzerland, ³Dept. of Chemistry and Biochemistry, University Bern, Bern, Switzerland, ⁴Dept. of Clinical Research and Veterinary Public Health, University Bern, Bern, Switzerland

Listeria brainstem encephalitis is associated with high mortality rates in ruminants and humans. The purpose of our study was to investigate metabolic changes in diseased small ruminants as a model for an inflammatory disease. HR-MAS NMR spectroscopy identified clear differences in the metabolic profile of biopsies from the primarily affected brainstem. In the thalamus, which commonly did show no or mild inflammatory infiltrates, no clear separation but a trend towards a clustering of the control and diseased group could be observed, and may indicate that NMR spectroscopy is able to detect metabolic changes before histopathologically observable inflammatory infiltrates occur.

Naranamangalam R Jagannathan¹, Deepti Upadhyay¹, Uma Sharma¹, Prasenjit Das², Siddharth Dutta Gupta², and Govind K Makharia^3
1Department of NMR and MRI Facility, All India Institute of Medical Sciences, New Delhi, Delhi, India, ²Department of Pathology, All India Institute of Medical Sciences, New Delhi, Delhi, India, ³Department of Gastroenterology & Human Nutrition, All India Institute of Medical Sciences, New Delhi, Delhi, India

Celiac disease (CeD) is an autoimmune enteropathy caused by ingestion of gluten and related prolamines present in cereals like wheat, rye, and barley in genetically predisposed individuals. Proton NMR based metabolomics of urine of CeD patients showed higher concentration of trans-aconitate suggesting the alteration in energy metabolism. Urine of CeD patients were also characterized by higher level of fucose. This may be due to the liver abnormalities associated with CeD.The findings of the present study suggested the utility of NMR based metabonomics in determining the biomarker/s for the differentiation of CeD from controls.

Yuqing Huang¹, Chunhua Tan¹, Shuhui Cai¹, and Zhong Chen^1
1Electronic Science, Xiamen University, Xiamen, Fujian, China

2D J-resolved spectroscopy (JRES) is an alternative to 1D NMR for the metabolite analysis due to its ability on separating congested signals along two frequency dimensions. However, conventional 2D JRES approaches are generally subjected to field inhomogeneous effects from macroscopic magnetic susceptibility in biological samples. In this report, an NMR acquisition method based on intermolecular single-quantum coherences (iSQCs) is presented to achieve high-resolution 2D J-resolved information directly from intact biological samples. This method is merely based on pulse sequence design and can work on standard NMR spectrometers without any hardware requirement.

Non Proton MRI
 

Wednesday 3 June 2015

Nadia Benkhedah¹ and Armin M. Nagel^1
1Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany

The intracellular sodium concentration is of great interest since it is a potential measure for small changes of the sodium ion homeostasis. Methods which allow a weighting toward intracellular sodium are based on the generation a weighted subtraction image and are thus highly influenced by inhomogeneities of the magnetic field. In the two pulse sequence two different pathways contribute to the long echo time image, so destructive interference can also lead to signal voids in regions prone to field inhomogeneities. In this study, the pathways of the two pulse sequence have been acquired separately to obtain an intracellular weighted ²³Na image which is more robust to variations in the main magnetic field.

Yifan Zhang¹, Bernadette O Erokwu², Yuchi Liu¹, George W Farr^3,4, Walter F Boron⁴, Chris A Flask^5,6, and Xin Yu^4,5
1Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States, ²Radiology, Case Western Reserve University, Cleveland, OH, United States, ³Aeromics, LLC, Cleveland, OH, United States, ⁴Physiology and Biophysics, Case Western Reserve University, Cleveland, OH, United States, ⁵Biomedical Engineering and Radiology, Case Western Reserve University, Cleveland, OH, United States, ⁶Pediatrics, Case Western Reserve University, Cleveland, OH, United States

In this study, oxygen-17 chemical shift imaging (¹⁷O-CSI) was combined with a keyhole acquisition strategy to delineate the kinetics of H₂¹⁷O uptake and washout in mouse brains with high temporal resolution. The utility of the method was evaluated in a study comparing H₂¹⁷O uptake and retention in the brains of AQP4-knockout (AQP4-KO) and wildtype mice. The results showed that AQP4-KO led to decreased H₂¹⁷O uptake and retention but an unaltered rate of H₂¹⁷O washout.

Yongxian Qian¹ and Luke C. Henry^2
1Qian's Lab for MRI, General Labs Cloud LLC, Pittsburgh, PA, United States, ²Department of Orthopaedic Surgery, University of Pittsburgh, Pittsburgh, PA, United States

This work presents a preliminary study of a novel methodology for noninvasive detection of membrane disruption of nerve cells (neurons) and stretching tears of nerve fibers (axons) in concussed patients. Sodium (23Na) MRI was used to detect cell membrane disruption through the change of intracellular sodium concentration while high resolution (HR) ultrashort echo time (UTE) proton (1H) MRI was used to detect axonal tears through thin fissuring or tearing in brain tissues. Both imaging modalities were performed on a clinical scanner at 3T. The results showed promise of the proposed method for the detection of nerve cell injuries in concussed patients.

Shingo Matsumoto^1,2, Keita Saito¹, Jeeva P Munasinghe³, Nallathamby Devasahayam¹, James B Mitchell¹, Robert J Gillies⁴, and Murali C Krishna^1
1Radiation Biology Branch, National Cancer Institute, NIH, Bethesda, MD, United States, ²Hokkaido University, Sapporo, Hokkaido, Japan, ³Mouse Imaging Facility, NINDS, NIH, Bethesda, MD, United States, ⁴Imaging and Metabolism, H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL, United States

Pancreatic cancer is a malignant neoplasm with an extremely poor prognosis. The TH-302 is a kind of drugs known as hypoxia-activated prodrug (HAP), which activated under hypoxia (low concentration of oxygen) and exhibits anti-tumor effect. A question derived is how clinicians select patients who receive most benefit from treatment with this hypoxia targeting new drug. Electron paramagnetic resonance imaging (EPRI) can non-invasively provide 3D absolute oxygen images. Here, we investigated if the EPR oxygen imaging can predict treatment benefit of oxygen dependent or independent therapies in three different pancreatic cancer xenografts.

Sebastian C. Niesporek¹, Aaron S. Kujawa¹, Nadia Benkhedah¹, and Armin M. Nagel^1
1Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany

Chloride regulates fundamental physiological processes which makes quantitative (³⁵Cl)-MRI interesting for biomedical research applications. Partial volume effects caused by larger voxel sizes due to a much lower MR-sensitivity and fast relaxation times strongly reduce the accuracy of concentration measurements. In this work partial volume correction was applied on ³⁵Cl-data of phantom and in vivo measurements. Quantification was performed using an external reference cushion as well as B₀- and B₁-correction. Good correction capability was seen for phantom measurements and in vivo application allowed quantification of chlorine tissue concentration in the human brain (CSF: 998, brain matter: 273).

Nicolas G. R. Behl¹, Christine Gnahm¹, Peter Bachert¹, and Armin M. Nagel^1
1Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany

An asymmetrically undersampled 3D-radial k-space sampling is proposed for ²³Na-MRI in combination with a 3D-Dictionary-Learning Compressed Sensing iterative reconstruction. The asymmetrical undersampling takes advantage of the point symmetry in k-space; 60% of the data is acquired with an undersampling factor (USF) of 10 and the remaining 40% with USF = 40. The results from simulated data show reduced artifacts and increased SSIM and PSNR for the 3D-DLCS reconstruction in the case of asymmetric undersampling. Reconstructions from in-vivo data show an improved delineation of small structures and reduced blurring.

MRS Normal & Ageing Brain
 

Wednesday 3 June 2015

Huijun Liao¹, Kristin Heaton², Praveen Merugumala¹, Jessica Saurman², Xi Long¹, Irina Orlovsky², Sai Merugumala¹, Kelly Rudolph², Nicole Murphy², Benjamin Rowland¹, and Alexander P. Lin^1
1Center for Clinical Spectroscopy, Brigham and Women's Hospital, Boston, MA, United States, ²Military Performance Division, US Army Research Institute of Environmental Medicine, Natick, MA, United States

The objective of this ¹H MRS study was to investigate whether healthy age and gendered-matched civilian subjects and military subjects show significant differences in MRS quantifiable metabolites. Compared to the healthy civilian group, lower Glu and NAA were found in the healthy military group across four voxels and were significant in posterior cingulate gyrus and parietal white matter. These findings could not be explained by differences in neurocognitive test performance or education levels which did not significantly differ between the two groups. Our results demonstrate the importance of utilizing appropriate military controls for military studies.

Bijaya Thapa¹, Marjanna Dahl², Deborah Frank², Phillip Burch³, and Eun-Kee Jeong^1,4
1Dept. of Physics and Astronomy, Utah Center for Advanced Imaging Research, University of Utah, Salt Lake City, Utah, United States, ²Dept. of Pediatrics, University of Utah, Utah, United States, ³Dept. of Surgery, University of Utah, Utah, United States, ⁴Dept. of Radiology, Korea University, Seoul, Korea

Oxandrolone is supposed to improve cardiac energy utilization in the infants with hypoplastic left heart syndrome (HPLS). HPLS infants have high mortality and are unable to gain weight may be due to high energy requirement from their shunt dependent physiology.We evaluate creatine kinase (CK) rate of the heart of the HLHS modeled lamb using 31P-MT-MRS. For this we developed a 1H/31P double tuned coil, TR switch and 1D-MT-CSI pulse sequence.The hardware were optimized for the 31P signal. Our initial studies shows an increase in the CK rate of oxandrolone-treated shunted heart compared to shunted heart.

Manuela B. Rösler¹, Nadia Benkhedah¹, Armin M. Nagel¹, Tanja Platt¹, Peter Bachert¹, and Reiner Umathum^1
1Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany

Potassium ions are of fundamental importance in the physiology of living organisms. In human thigh muscle oscillations in the T₂^* decay of ³⁹K MR-signal have been observed. To investigate this effect, in vivo spectra of potassium and for comparison of sodium were acquired. The ³⁹K spectrum of a human thigh muscle shows triplet splitting. Its peaks were quantitatively analyzed with the AMARES algorithm and led to results in agreement with the theory of quadrupolar splitting due to anisotropy.

Darong Zhu¹, Song Chen², Xu Yan³, Linfei Wen², Congyu Liao², Meng Chen², Keith Heberlein⁴, and Jianhui Zhong^2
1Hangzhou First People's Hospital, Hangzhou, Zhejiang, China, ²Center for Brain Imaging Science and Technology, Department of Biomedical Engineering, Zhejiang University, Hangzhou, Zhejiang, China, ³MR Collaboration NE Asia, Siemens Healthcare, Shanghai, China, ⁴Siemens Medical Solutions USA, Inc, Malvern, PA, United States

Glutamate (Glu) and gamma-aminobutyric acid (GABA) are very important neurotransmitters, and they play critical roles in mental illness such as depression and schizophrenia. However, neurological bases and mechanisms of GABA for quantifying specific diseases or characterizing regional neural function remain poorly understood. Metabolites such as GABA are difficult to be detected owing to their lower concentration and J-coupling effect. MEGA-PRESS, a sequence based on J-difference spectral editing, can detect the regional variations of GABA concentration, which will be useful in investigating the concentration of GABA in different brain regions and its correlation with the function of the nervous system.

Pui Wai Chiu¹, Peiying Liu², Queenie Chan³, Raymond Chuen Chung Chang⁴, Leung Wing Chu⁵, Hanzhang Lu², and Henry Ka Fung Mak^1
1Diagnostic Radiology, The University of Hong Kong, Hong Kong, Hong Kong, Hong Kong, ²Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Texas, United States, ³Philips Healthcare, Hong Kong, Hong Kong, ⁴Laboratory of Neurodegenerative Disease, Department of Anatomy, The University of Hong Kong, Hong Kong, Hong Kong, ⁵Division of Geriatric Medicine, Department of Medicine, Queen Mary Hospital, Hong Kong, Hong Kong

In this study, we evaluated the age-related change in glutamatergic metabolism and its role in neurovascular coupling via quantitiave proton magnetic resonance spectroscopy and pseudo-continuous arterial spin labeling in anterior cingulate cortex (ACC). Glutamine concentration was found to increase with age, whereas as both global and regional cerebral blood flow (CBF) in ACC showed age-related decline. However, no direct relationship was found between CBF and glutamatergic synaptic activity in aging suggesting the age-related alteration in glutamatergic metabolism might not directly affect CBF.

John Jensen^1,2, Stephanie Licata^1,2, Lisa Nickerson^1,2, Marisa Silveri^1,2, Carolyn Caine², Kristina Wang², Rosemond Villefuerte², Kevin Hill^1,2, and David Olson^1,2
1Harvard Medical School, Boston, MA, United States, ²McLean Hospital, Belmont, MA, United States

Multi-voxel imaging of pharmaco-dynamic changes in GABA is possible by combining MRSI with the well-known MEGAPRESS-based difference-editing acquisition. Our “functional” MRSI (fMRSI) technique is demonstrated during a double-blind crossover design drug challenge study of placebo, stimulant and sedative drug effects. We image brain GABA levels, both spatially and temporally, in order to assess the effects of sedatives and stimulants on brain GABA over time. Refinements of this methodology should place fMRSI in the same functional realm of other functional imaging modalities, e.g., positron-emission tomography (PET) with while being completely non-invasive.

MRS Neurological Diseases
 

Wednesday 3 June 2015

Yazhuo Kong¹, Uzay Emir¹, George Tackley¹, Lucy Matthews², Charlotte Stagg¹, Irene Tracey¹, and Jacqueline Palace^2
1FMRIB Centre, Nuffield Department of Clinical Neuroscience, University of Oxford, Oxford, Oxfordshire, United Kingdom, ²Nuffield Department of Clinical Neuroscience, University of Oxford, Oxford, Oxfordshire, United Kingdom

The periaqueductal gray (PAG) is a key component of descending pain modulatory system that powerfully modulates nociceptive inputs. Here, we specifically assess the relationship between increased descending inhibition, as reflected by an increased excitatory glutamate level within the PAG, and the degree of neuropathic pain, using 1H MR spectroscopy and painDETECT questionnaire with Neuromyelitis Optica patients. Glutamate was found to be negatively correlated to the degree of neuropathic pain. Our finding suggests that glutamate levels may reflect the tone of inhibitory activity established in the descending pain inhibition system.

Praveen Dev Merugumala¹, April Long¹, Huijun Liao¹, Yvonne Robles², Nina Lin³, and Alexander P Lin^1
1Center for Clinical Spectroscopy, Brigham and Women's Hospital, Boston, MA, United States, ²Infectious Disease Clinic, Brigham and Women's Hospital, Boston, MA, United States,³Infectious Disease Clinical Research Unit, Boston University School of Medicine, Boston, MA, United States

The most prominent antiretroviral treatment for HIV-infected individuals is the drug Efaverenz even though it is known to cause some neurocognitive side effects. Another newer antiretroviral treatment are integrase inhibitors, which are equally effective but have unknown relative neurotoxicity. Seven HIV Patients on Efaverenz were recruited to switch treatments to an integrase inhibitor and undergo monitoring via magnetic resonance spectroscopy to measure brain chemistry. Each participant was scanned at baseline and again eight weeks after switching the treatment. What resulted were increased presences of GABA and choline, which may have neurocognitive implications.

Brenda Bartnik-Olson¹, Daniel Ding², John Howe², Amul Shah², and Travis Losey^3
1Radiology, Loma Linda University Medical Center, Loma Linda, CA, United States, ²School of Medicine, Loma Linda University, Loma Linda, CA, United States, ³Neurology, Loma Linda University Medical Center, Loma Linda, CA, United States

Several studies suggest that glucose metabolism and glutamate-glutamine cycle activity may be impaired in temporal lobe epilepsy (TLE). Glutamate is the primary excitatory neurotransmitter in the brain and increased levels of Glu have been reported in the epileptogenic hippocampus. However, reduced Glu concentrations have also been reported. In this study we used 1H MR spectroscopy (MRS) to measure the Glu concentration and rate of Glu synthesis in the medial temporal lobe of TLE patients following an infusion of [U-13C] glucose. Our findings show reduced Glu concentrations in the involved mesial temporal lobe of TLE subjects while the rate of Glu synthesis remained similar to healthy controls. These findings suggest that reduced Glu concentrations in the interictal state are not due to reduced oxidative metabolism.

Reggie Taylor^1,2, Betsy Schaefer³, Elizabeth Osuch^2,3, Maria Densmore², Nagalingam Rajakumar³, Jean Theberge^1,2, and Peter Williamson^2,3
1Medical Biophysics, Western University, London, ON, Canada, ²Imaging, Lawson Health Research Institute, London, ON, Canada, ³Psychiatry, Western University, London, ON, Canada

MRS voxels were placed in the anterior cingulate (AC) and the thalamus (TH) of volunteers with schizophrenia, volunteers with unipolar mood disorder, and healthy controls using a 7T MRI. The schizophrenic group and mood group act as psychiatric controls for each other. Significant glutamate increases were observed in the AC, and glutamine increases in the TH, of the schizophrenic patients relative to controls. Glycine concentrations were also significantly decreased in the schizophrenic group relative to both control and mood groups in the TH. Myo-inositol concentrations were significantly decreased in both AC and TH relative to the controls and schizophrenic groups.

Marinette van der Graaf¹, Marleen CDG Huigen¹, Eva Morava^1,2, A Carin M Dassel³, Maurice AM van Steensel^4,5, Marieke MB Seyger¹, Ron A Wevers¹, and Michèl A Willemsen^1
1Radboud University Medical Center, Nijmegen, Netherlands, ²Tulane University Medical School, New Orleans, LA, United States, ³Deventer Hospital, Deventer, Netherlands,⁴Maastricht University Medical Center, Maastricht, Netherlands, ⁵Institute of Medical Biology, Immunos, Singapore

A patient with congenital ichthyosis suspected of having Sjögren-Larsson syndrome (SLS) underwent cerebral MRS measurements in which the typical pattern of SLS was expected with a high lipid resonance in white matter. However, this particular patient showed overall cerebral lipid accumulation not restricted to only white matter. Further metabolic and genetic screening led to the diagnosis of Chanarin-Dorfman Syndrome (CDS) caused by a defect in the CGI-58/ABHD5 gene. The finding of cerebral lipid accumulation in CDS was reproduced in another patient with genetically proven CDS. To our knowledge these MRS results demonstrate for the first time cerebral involvement in CDS.

Simona Nikolova¹, Jorge Burneo², and Robert Bartha^3
1Robarts Research Institute, London, ON, Canada, ²Schulich School of Medicine and Dentistry, University of Western Ontario, London, Canada, ³Medical Biophysics, University of Western Ontario, London, Ontario, Canada

The purpose of this preliminary 7T study was to characterize metabolite levels in the hippocampus of 1.5T- MRI negative temporal lobe epilepsy (TLE) patients. Eight MRI negative TLE patients (mean age 26  8 years) and eight healthy controls (mean age 32  8 years) were studied. Using short echo time MRS Semi-LASER spectroscopy at 7T significant differences were detected in hippocampal metabolite concentrations in patients with normal 1.5T MRI scans compared to healthy controls. The identification of structural and metabolic abnormalities using high field MRI could allow improved surgical selection of patients with temporal lobe epilepsy.

Dominic Carlin^1,2, Ben Babourina-Brooks^1,2, Martin Wilson^1,2, and Andrew C Peet^1,2
1School of Cancer Sciences, University of Birmingham, Birmingham, United Kingdom, ²Birmingham Children's Hospital, Birmingham, United Kingdom

Metabolite concentrations are fundamental prognostic biomarkers, however measurement is known to be affected by T2 relaxation. Single-voxel MRS was performed using PRESS (TR 1500ms, TE 30/135ms) on a 1.5T scanner. Thirty-three childhood brain tumour patients and twenty-four healthy controls were analysed using TARQUIN for metabolite concentrations and T2 relaxation times were calculated. Concentrations were corrected for relaxation effects using metabolite and water T2 values and the importance of accurate T2 values was assessed. Significant differences in the T2 relaxation of water were found between tumour and normal brain and between tumour types which should be accounted for in metabolite quantification.

MRSI
 

Wednesday 3 June 2015

Kristian Rink¹, Nadia Benkhedah¹, Moritz C. Berger¹, Peter Bachert¹, and Armin M. Nagel^1
1German Cancer Research Center (DKFZ), Heidelberg, Germany

Phosphorus and particularly phosphocreatine play a crucial role in physiological processes such as the human energy metabolism. Since the in vivo MR signal of ³¹P is four orders of magnitude smaller compared to hydrogen, new acquisition strategies to improve the SNR are required. In this work, the human calf was examined applying a radially sampled and density adapted screening combined with several fbSSFP sequences. The highest SNR was achieved for a positive followed by a negative radially sampled gradient acquiring two contrasts.

Gilbert Hangel¹, Bernhard Strasser¹, Michal Považan¹, Stephan Gruber¹, Marek Chmelik¹, Siegfried Trattnig¹, and Wolfgang Bogner^1
1MCRE, Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Wien, Vienna, Austria

This work presents a method of lipid and macromolecule suppression using double inversion recovery (DIR) for accelerated FID MRSI in the brain at 7T. DIR suppression greatly reduces artifacts caused by extracranial lipids and macromolecules, caused by fold-in due to parallel imaging and non-optimal PSF. Two 10th-order WURST pulses with 80 ms pulse duration and their timing were optimized for 7T with a TI1/TI2 of 160/75 ms and a total TR of 760 ms. Loss of metabolite SNR was quantified in phantom measurements while volunteer measurements showed in vivo feasibility and successful lipid artifact reduction.

Ipshita Bhattacharya¹ and Mathews Jacob^1
1The University of Iowa, Iowa City, IA, United States

The leakage of extra cranial lipids to the brain and unsuppressed water is one of the biggest obstacles in acquiring reliable metabolite estimates in MR spectroscopic imaging (MRSI). Current methods to suppress lipids require either considerable expertise, provides only partial brain coverage, or result in loss of metabolite SNR. We introduce a variable density spiral sequence to considerably improve k-space coverage with minimal loss in metabolite SNR. A two step recovery strategy is used to suppress the nuisance signals and obtain good quality short echo-time MRSI spectra without any fat suppression methods.

Zhiyong Zhang^1,2, Noam Shemesh^1,3, and Lucio Frydman^1
1Chemical Physics Department, Weizmann Institute of Science, Rehovot, Israel, ²Department of Electronic Science, Xiamen University, Xiamen, Fujian, China, ³Champalimaud Neuroscience Programme, Champalimaud Centre for the Unknown, Lisbon, Portugal

Recently, a “relaxation-enhanced” (RE) selective-excitation MRS approach to acquire in vivo localized spectra with flat baselines and very good signal-to-noise ratios (SNR) has been proposed. As RE MRS targets a priori known resonances, MRSI information may be acquired in a faster, more efficient manner. Hereby we present such Relaxation-Enhanced Chemical-shift-Encoded Spectroscopically-Separated (RECESS) MRI method, targeting images of various metabolite by collecting a series of conventional imaging data echoes where the chemical shifts of the different resonances are encoded by coherent phase modulations, and disentangled by Fourier-based inversions. The RECESS method thus accelerates the acquisition of quality MRSI data.

Zohaib Iqbal¹, Neil E. Wilson¹, Brian L. Burns¹, Margaret A. Keller¹, and Michael Albert Thomas^1
1University of California - Los Angeles, Los Angeles, California, United States

This is a pilot demonstration of a novel technique, called accelerated TE-averaged echo-planar Spectroscopic Imaging (accelerated TEA-EPSI), which uses a combination of an echo-planar readout and Compressed Sensing reconstruction to achieve multi-slice TE-averaged spectroscopic images in a clinically feasible time. The data had non-uniform sampling applied along the ky,kz,t1 volume and were reconstructed using minimization of total variation. Afterwards, the data were averaged over the TE increments to obtain TE-averaged spectra. Brain phantom data and in vivo brain spectra show good reproducibility and high spectral quality, demonstrating the potential usefulness of this technique for investigating metabolic changes in different pathologies.

Zohaib Iqbal¹, Neil E. Wilson¹, Rajakumar Nagarajan¹, Daniel A. Margolis¹, Robert E. Reiter², Steven S. Raman¹, and Michael Albert Thomas^1
1Radiological Sciences, University of California - Los Angeles, Los Angeles, California, United States, ²Urology, University of California - Los Angeles, Los Angeles, California, United States

The purpose of this study was to validate a five dimensional spectroscopic imaging technique (3 spatial and 2 spectral dimensions) utilizing semi-Laser 180º pulses, called semi-Laser 5D Echo-Planar J-Resolved Spectroscopic Imaging (5D EP-JRESI), in the prostate of healthy volunteers and prostate cancer patients. This technique provides full coverage of the prostate. The results show that qualitative differences in the two-dimensional spectra (differences in citrate and choline) can be seen between a healthy location and an unhealthy location in a prostate cancer patient. This technique will continue to be evaluated in a larger number of healthy volunteers and cancer patients.