Pre-Clinical fMRI

Wednesday 3 June 2015

Rupeng Li¹, Xiping Liu², Jason W Sidabras1¹, Eric S Paulson³, Andrzej Jesmanowicz¹, and James S Hyde^1
1Biophysics, Medical College of Wisconsin, Milwaukee, WI, United States, ²Dermatology, Medical College of Wisconsin, Milwaukee, WI, United States, ³Radiation Oncology, Medical College of Wisconsin, Milwaukee, WI, United States

The aural cavity magnetic susceptibility artifact leads to significant echo planar imaging, EPI, signal dropout in rat deep brain that limits acquisition of functional connectivity fcMRI data. We provide a method that restores the EPI signal in deep brain. Needle puncture introduction of a liquid-phase fluorocarbon into the middle ear allows acquisition of rat fcMRI data without signal dropout. We demonstrate that with seeds chosen from previously unavailable areas, including the amygdala and the insular cortex, we are able to acquire whole brain scale networks, including the limbic system. This tool allows EPI-based neuroscience and pharmaceutical research in rat brain using fcMRI that was previously not feasible.

Tong Wu¹, Joanes Grandjean², Simone C. Bosshard³, Markus Rudin², David Reutens³, and Tianzi Jiang^1,4
1Queensland Brain Institute, The University of Queensland, Brisbane, Queensland, Australia, ²Institute for Biomedical Engineering, University and ETH Zurich, Zurich, Switzerland,³The Centre for Advanced Imaging, The University of Queensland, Queensland, Australia, ⁴Brainnetome Centre, Institute of Automation, Chinese Academy of Sciences, Beijing, China

Anesthesia is an integral part of most resting-state fMRI mice studies and does not influence the brain as a whole to the same degree. This study attempts to examine intra-regional alterations in spontaneous fMRI signals induced by different anaesthesia protocols. Regional homogeneity (ReHo) is a measure of the temporal similarity of a local voxel neighborhood. We observed that medetomidine showed increased ReHo in striatum and decreased ReHo in the cortex and thalamus compared to other agents. Urethane showed increased ReHo in thalamus compared to several other agents. Global signal regression did not introduce dramatic changes to ReHo in this study.

Li-Ming Hsu¹, Xia Liang¹, Hong Gu¹, Julia K. Brynildsen¹, Jennifer A. Stark², Kia Jackson³, Allison Hoffman³, Hanbing Lu¹, Elliot A. Stein¹, and Yihong Yang^1
1Neuroimaging Research Branch, National institute on drug abuse, Baltimore, MD, United States, ²Maryland Neuroimaging Center, University of Maryland, MD, United States,³Center for Tobacco Products, FDA, MD, United States

Human default mode network (DMN) has been fractionated into subcomponents based on their functional connectivity architecture and their distinct responses to different cognitive tasks. However, the partitioning of DMN in animals is much less known. Since rodents have been widely used as translational preclinical models, a thorough understanding of the architecture, and relevant functions, of the DMN in rodents would be important for interpreting resting-state fMRI (rs-fMRI) data of rodent DMN. In this study, we investigate constituents of DMN in rats using rs-fMRI and diffusion tensor imaging (DTI), and discuss their potential functional relevancy.

Xiaojia Liu^1,2, Fuqing Zhou³, Xiang Li³, Jiaolong Cui³, Mengye Lyu^1,2, Adrain Tsang^1,2, Iris Y Zhou^1,2, Ed X Wu^1,2, and Yong Hu^3
1Laboratory of Biomedical Imaging and Signal Processing, The University of Hong Kong, Hong Kong, Hong Kong, China, ²Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, Hong Kong, China, ³Department of Orthopaedics and Traumatology, The University of Hong Kong, Hong Kong, Hong Kong, China

Resting state functional organization in the human cervical cord has only been scarcely explored. In this study, we investigated the resting state functional organization using a clinically relevant 3T whole body MRI scanner from 24 healthy subjects. The correlation coefficient computed from rsfMRI images between each ventral or dorsal horn of different segments was used to generate the correlation matrix. Segment C2 demonstrated stronger correlations with other segments. Segment C2 has a stronger neural activity level than other segments. Functional organization among segments was detected, which demonstrated the functional network in the human cervical cord.

Yuncong Ma¹, Christina Hamilton¹, Pablo D. Perez¹, Zhifeng Liang¹, and Nanyin Zhang^1
1Department of Bioengineering, The Pennsylvania State University, State College, Pennsylvania, United States

The neural mechanism underlying anesthesia-induced unconsciousness (AIU) remains elusive. In this study, we aimed to investigate the system level circuit mechanism of AIU by analyzing the inter-hemispheric resting-state functional connectivity (RSFC) at the awake state and five anesthetic depths. We used pairwise T Test to reveal the connectivity changes from awake state to various anesthesia states, and Pearson correlation to compare the inter-hemispheric RSFC with behavioral data across all AIU states.

Robert Becker¹, Anke Tappe-Theodor², Ainhoa Bilbao³, Rainer Spanagel³, and Wolfgang Weber-Fahr^1
1Research group Translational Imaging, Department of Neuroimaging, Central Institute of Mental Health, Medical Faculty Mannheim / Heidelberg University, Mannheim, BW, Germany, ²Pharmacological institute, Heidelberg University, Heidelberg, BW, Germany, ³Department of Psychopharmacology, Central Institute of Mental Health, Medical Faculty Mannheim / Heidelberg University, Mannheim, BW, Germany

In a pilot rsfMRI/MRS study we investigated the effects of inflammatory pain on brain network structure in a group of 16 mice. Pain was induced by injection of Complete Freund's adjuvant (CFA). We found significant differences in local network parameters of 10 regions, mostly related with mood/stress/reward systems. The changes of local network features in these regions show their more global and less local role in the network. Furthermore quantified MR-Spectra show a trend for increased glutamate and glutamine concentrations in the prefrontal cortex of the CFA-group, which supports our working hypothesis of a hyper-glutamatergic input to the nucleus acumbens.

Christina Hamilton¹, Yuncong Ma¹, and Pablo Perez^1
1Pennsylvania State University, State College, PA, United States

Consciousness is an elusive phenomenon that has remained obscure due to its lack of definition and standardization. To better understand consciousness, resting-state functional magnetic resonance imaging (rsfMRI) can be employed to investigate systems-level mechanisms of altered states of consciousness. Here we use rsfMRI to investigate functional brain connectivity at various doses of isoflurane anesthetic and correlate this with behavioral measures of anesthetic depth. We demonstrate that increases in isoflurane result in decreased connectivity at lower correlation thresholds. This relationship allows us to further characterize different states of consciousness and may provide a novel use of rsfMRI as a biomarker for various states of consciousness.

Qasim Bukhari¹, Aileen Schröter¹, and Markus Rudin^1,2
1Institute of Biomedical Engineering, ETH and University of Zürich, Zürich, Switzerland, ²Institute of Pharmacology and Taxicology, University of Zürich, Zürich, Switzerland

In this work we have applied dual regression and network modeling using FSL Nets to the fMRI mouse data. The goal of the study was to understand network connectivity differences under the effect of different anesthesia regimens (isoflurane, propofol, urethane and medetomidine). The analysis revealed higher degree of similarity in functional networks between propofol and urethane, while isoflurane also showed similar network patterns. However medetomidine produced very different functional brain networks than other anesthesia regimens. Our study confirmed that despite the inferior signal-to-noise ratio intrinsic in mouse fMRI, it is possible to deduce significant network connectivity differences.

Alexander John Poplawsky¹, Mitsuhiro Fukuda¹, and Seong-Gi Kim^2,3
1Radiology, University of Pittsburgh, Pittsburgh, Pennsylvania, United States, ²Center for Neuroscience Imaging Research, Institute for Basic Science (IBS), Suwon, Korea,³Biomedical Engineering and Biological Sciences, Sungkyunkwan University (SKKU), Suwon, Korea

The contribution of neural spiking activity to the fMRI response is not completely understood. In the current study, we directly suppressed the spiking of output neurons in the rat olfactory bulb by simultaneous stimulation of excitatory and inhibitory neural pathways and measured the resultant cerebral blood volume-weighted (CBV) fMRI changes. We found a small, but significant, CBV increase due to spiking activity, while the majority of the increase corresponded to an inseparable combination of synaptic and spiking activities. We conclude that, although synaptic activity may be the dominant source of the fMRI response, spiking activity does contribute slightly.

Henning Matthias Reimann¹, Jaroslav Marek¹, Jan Hentschel¹, Till Huelnhagen¹, Andreas Pohlmann¹, and Thoralf Niendorf^1,2
1Berlin Ultrahigh Field Facility (B.U.F.F.), Max Delbrueck Center for Molecular Medicine, Berlin-Buch, Berlin, Germany, ²Experimental and Clinical Research Center, Charite-Universitatsmedizin, Berlin, Germany

Chronic pain is still a major clinical issue with limited treatment options. The combination of fMRI and mouse genetics provides great potential for novel targeted drug development. Still, the potential of nociceptive mouse studies using fMRI is as yet untapped due to methodological constraints. Here we present enabling methodology, which allows near-physiological nociceptive mouse fMRI studies yielding spatially discrete BOLD effects of high magnitude for mild noxious stimuli of 46°C. This is the first report on BOLD activation patterns or BOLD effects of this significance and magnitude in mouse fMRI with thermostimulation.

Damon Philip Cardenas^1,2, Eric R Muir^1,3, Shiliang Huang³, and Timothy Q Duong^1,3
1University of Texas Health Science Center at San Antonio, San Antonio, Texas, United States, ²Biomedical Engineering, University of Texas San Antonio, San Antonio, Texas, United States, ³Research Imaging Institute, San Antonio, Texas, United States

Hyperbaric oxygen (HBO) therapy is used to treat a number of ailments. Previously, during HBO at 3 atmospheres absolute fMRI studies, it was observed that BOLD response to an evoked stimulus was not “washed out” as a result of highly increased oxygen concentration. Additional experiments were performed to evaluate the potential contributions of stimulus-evoked inflow, spin density, and electrical activity to BOLD fMRI responses under HBO. CBF and BOLD fMRI of forepaw stimulation in anesthetized rats was performed under HBO and compared with normobaric air .

Georges Hankov*^1,2, Marija M Petrinovic*¹, Aileen Schroeter², Andreas Bruns¹, Markus Rudin^2,3, Markus von Kienlin¹, Basil Künnecke¹, and Thomas Mueggler^1
1Neuroscience Discovery, F. Hoffmann-La Roche Pharmaceuticals Ltd, Basel, Basel-City, Switzerland, ²Institute for Biomedical Engineering, University of Zurich and ETH, Zurich, Zurich, Switzerland, ³Institute of Pharmacology and Toxicology, University of Zurich, Zurich, Switzerland

Functional MRI applied across different genetic mouse models of neurodevelopmental disorders holds promise to reveal pathological alteration at the neurocircuitry level and thus to aid devising novel therapies and biomarkers. Prerequisite is a reliable, non-terminal anesthesia protocol that is applicable to longitudinal studies in a broad range of mouse lines. Here, a novel anesthesia protocol based on the injectable GABAergic anesthetic etomidate has been established and validated using perfusion and BOLD fMRI. Under etomidate, basal perfusion remains in a physiological range, cerebrovascular reserve capacity is preserved and regions-specific perfusion- and BOLD changes are elicited upon pharmacological and sensory stimulation, respectively.

Hirotsugu Funatsu¹, Sosuke Yoshinaga¹, Haruna Goto¹, Makoto Hirakane¹, Shigeto Iwamoto¹, and Hiroaki Terasawa^1
1Department of Structural BioImaging, Faculty of Life Sciences, Kumamoto University, Kumamoto, Japan

Odors are chemical signals that regulate a wide range of social and sexual behaviors in animals. The aim of this study is to reveal the neural circuitry from the olfactory bulb to the higher-order brain areas that recognize and discriminate odors in mice. Based on the real-time BOLD and longer time-scale MEMRI techniques, the muscone odor induced signal enhancement in the peripheral olfactory bulb, where chemical signals from the olfactory epithelium are transmitted. The BOLD signal-enhanced regions coincided with the regions identified by an immunohistochemical analysis. We successfully visualized the odor-evoked activation by the combined use of BOLD and MEMRI.

Eric R. Muir¹, Linlin Cong¹, KC Biju², William E. Rogers¹, Robert A. Clark³, and Timothy Q. Duong^1
1Research Imaging Institute, University of Texas Health Science Center, San Antonio, TX, United States, ²Department of Medicine, University of Texas Health Science Center, San Antonio, TX, United States, ³Institute for integration of Medicine & science and South Texas Veterans Health Care System, University of Texas Health Science Center, San Antonio, TX, United States

The olfactory system in most mammals consists of the main and the accessory olfactory systems. The main olfactory bulb receives input from the olfactory epithelium which senses volatile odors, while the accessory olfactory bulb receives input from the vomeronasal organ which senses pheromones. The accessory olfactory system plays an important role in rodents, but a single fMRI study investigating only the olfactory bulb of odor and pheromone stimulation has been performed to our knowledge. The aim of this study was to investigate fMRI responses to odor and urine (which contains pheromones) in the entire olfactory network of the mouse brain.

Shin-Lei Peng^1,2, Ling-Yi Huang¹, Sheng-Min Huang¹, Yi-Chun Wu³, Hanzhang Lu², Fu-Chan Wei⁴, Chih-Jen Wen⁴, Hui-Yu Cheng⁴, Chih-Hung Lin⁴, and Fu-Nien Wang^1
1Department of Biomedical Engineering and Environmental Sciences, National Tsing Hua University, Hsinchu, Taiwan, ²Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, Texas, United States, ³Molecular Imaging Center, Chang Gung Memorial Hospital, Taoyuan, Taiwan, ⁴Department of Plastic and Reconstructive Surgery, Chang Gung Memorial Hospital, Taoyuan, Taiwan

fMRI plays an role in neuroscience researches. Investigating animal fMRI studies is an essential step toward understandings of cortical reorganization after surgery. Most researches use alpha-chloralose as an anesthetic, which is improper for follow-up studies. Here, we established a new protocol of rat electric mystacial pad (MP) stimulation under isoflurane anesthesia with advantages of increased BOLD sensitivity and possibility for longitudinal studies. Results show MP stimulation exhibited reproducible activations in the sometosensory cortex. In terms of the extension of activations and induced BOLD signal changes, results from MP stimulation under isoflurane anesthesia are comparable with those under alpha-chloralose anesthesia.

Arun Niranjan¹, Jack A Wells¹, and Mark F Lythgoe^1
1Centre for Advanced Biomedical Imaging, University College London, London, United Kingdom

Mouse brain fMRI has made significant progress over the last 10 years, but remains technically challenging and requires further efforts to improve data acquisition. In this work we demonstrate the use of visual stimulation as a viable task for fMRI of the mouse brain, with robust responses in the mid-brain (superior colliculus, lateral geniculate nuclei), mapping the visual pathway. We also examine the use of GE-EPI with multiple snapshots (compressed segments) as a way of reducing image distortion due to B₀ inhomogeneities at the point of acquisition.

Giovanna Diletta Ielacqua¹, Aileen Schroeter¹, Mark Augath¹, Felix Schlegel¹, and Markus Rudin^1,2
1Institute for Biomedical Engineering, ETH and University of Zurich, Zurich, Switzerland, ²Institute of Pharmacology and Toxicology, University of Zurich, Zurich, Switzerland

Stimulus-evoked fMRI experiments in mice show widespread hemodynamic responses even to innocuous stimulation, indicating general arousal of the animal. Systemic cardiovascular responses to stimuli appear to overrule cerebral hemodynamics and mask specific stimulus-evoked responses. Measurements in rats are less confounded by such impact, resulting in predominantly contralateral activation of cortical somatosensory cortical area involved in processing of the stimulus. Here, we characterize fMRI responses in further detail and study three different paw stimulation paradigms (electrical, chemical and thermal) in rats and mice. Better understanding the interplay between specific and systemic fMRI signal contributions should help enhancing specificity in fMRI.

Klaus Scheffler¹, Philipp Ehses¹, Yi He¹, Hellmut Merkle¹, and Xin Yu^1
1MRC department, Max Planck Institute for Biological Cybernetics, Tübingen, Germany, Tübingen, Germany

Sub-millimeter fMRI at very high fields with EPI is challenging due to the rapid signal decay and B0-related distortions. Balanced SSFP offers the possibility of high-resolution (100 um) acquisitions without spatial distortions and significantly higher temporal resolution compared to FLASH. Observed signal changes largely depend on T2 rather than T2* indicating a higher spatial selectivity compared to gradient echo-based methods.

Valentin G. Kemper¹, Federico De Martino^1,2, Desmond H. Y. Tse^3,4, Benedikt A. Poser¹, Essa Yacoub², and Rainer Goebel^1
1Cognitive Neuroscience FPN, Maastricht University, Maastricht, Limburg, Netherlands, ²Center for Magnetic Resonance Research, CMRR, Radiology, University of Minnesota, Minneapolis, MN, United States, ³Neuropsychology and Psychopharmacology, FPN, Maastricht University, Maastricht, Limburg, Netherlands, ⁴Radiology, University Medical Centre, Maastricht University, Limburg, Netherlands

A zoomed, inner-volume 3D-GRASE sequence was used to acquire T2 weighted 0.8 mm and 0.6 mm isotropic resolution fMRI at 9.4 T in human early visual areas. Robust activation patterns could be observed at both resolutions. Results are compared to 7 T data of the same volunteer with the same functional paradigm and similar technical set-up. At the same resolution, temporal SNR is significantly higher at 9.4 T than at 7 T, while T2 weighted fMRI signal changes induced by a visual task were found to be similar.

Saeedeh Amirmohseni¹, Daniel Segelcke², Esther Pogatzki-Zahn², and Cornelius Faber^1
1Department of Clinical Radiology, University Hospital Muenster, Muenster, Germany, ²Department of Anaesthesiology, Intensive Care and Pain Medicine, University Hospital Muenster, Muenster, Germany

The present study investigates the BOLD-response in pain-related brain regions in an incisional and an inflammatory rat model during noxious electrical stimulation (NES) and noxious mechanical stimulation (NMS) of the injured hindpaw under the anesthesia with medetomidine. Differences in the response to NMS and NES were observed. Upon NMS, the response in the inflammation group was higher as compared to both incision and sham group. Upon NES, significantly higher responses were observed for both pain models as compared with the sham group. Different responses for the two pain models were observed in S1 and RSC.

fMRI Methods

Wednesday 3 June 2015

Maosen Wang^1,2, Yi He¹, Yaohui Tang¹, Hellmut Merkle³, and Xin Yu^1,2
1Research Group of Translational Neuroimaging and Neural Conteol,High Field Magnetic Resonance, Max Planck Institute for Biological Cybernetics, Tuebingen, Baden-Wuerttemberg, Germany, ²Graduate School of Neural & Behavioural Sciences International Max Planck Research School, University of Tuebingen, Tuebingen, Baden-Wuerttemberg, Germany, ³Laboratory of Functional and Molecular Imaging, National Institute of Neurological Disorders and Str, National Institutes of Health, Bethesda, MD, United States

It has been demonstrated that the hemodynamic signal from individual venules can be detected directly with fMRI. Here, the hemodynamic signal from BOLD and CBV fMRI was measured at high temporal(100ms) and spatial resolution(150x150µm) in layer 4/5 of the rat forepaw S1with fast gradient-echo MRI. Distinctly different voxels were activated in BOLD vs CBV fMRI. In contrast to the BOLD activated voxels primarily located at the penetrating venules, CBV activated voxels were primarily located at penetrating arterioles. This result makes it possible to directly image the CBV and BOLD response at the single-vessel level to understand neurovascular coupling.

Lydia Wachsmuth¹, Florian Schmid¹, Miriam Schwalm², Albrecht Stroh², and Cornelius Faber^1
1Department of Clinical Radiology, University of Münster, Münster, Germany, ²Institute of Microscopic Anatomy and Neurobiology, Johannes Gutenberg-University Mainz, Mainz, Germany

We combined small animal fMRI with fiber-based optogenetic excitation of CHR2 and C1V1, transduced in rat cortex and thalamus, and with cortical fluorescence detection of a Ca2+-sensitive dye (OGB-1) in order to assess the scale of optogenetic network activation in terms of the spatial and temporal patterns of network recruitment. Onset, peak time and decay characteristics of BOLD time courses and amplitude and decay characteristic of Ca2+ responses appeared similar upon optogenetic and sensory-driven, endogenous network activation, confirming that ofMRI can serve as a model for functional mapping of brain circuits.

Zhifeng Liang^1,2, Glenn D.R. Waston^2,3, Kevin D. Alloway^2,3, Gangchea Lee¹, Thomas Neuberger¹, and Nanyin Zhang^1,2
1Dept. of Biomedical Engineering, Pennsylvania State University, University Park, PA, United States, ²Center for Neural Engineering, The Huck Institutes of Life Sciences, Pennsylvania State University, University Park, PA, United States, ³Neural and Behavioral Sciences, College of Medicine, Pennsylvania State University, Hershey, PA, United States

It has been increasingly recognized that anesthesia has profound impacts on functional magnetic resonance imaging (fRMI) studies. Specifically, we have previously demonstrated the impact of anesthesia on global organization and local circuits in rodents using resting-state functional MRI (rs-fMRI). To further explore this impact in tas- based fMRI, we utilized optogenetics-fMRI (opto-fMRI) to examine the global impact of optically induced neural activation of infralimbic cortex in rodents. The results indicated that both the spatial extent and the amplitude of BOLD signal activation were reduced in the anesthetized state, compare to the awake state.

Heather K. Decot^1,2, Wei Gao^3,4, Joshua H. Jennings^1,2, Pranish A. Kantak², Yu-Chieh Jill Kao^4,5, Manasmita Das^4,5, Ilana B. Witten⁶, Karl Deisseroth⁷, Yen-Yu Ian Shih^4,5, and Garret D. Stuber^1,2
1Curriculum in Neurobiology, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, ²Departments of Psychiatry & Cell and Molecular Physiology, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, ³Department of Radiology, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, ⁴Biomedical Research Imaging Center, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, ⁵Department of Neurology, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, ⁶Princeton Neuroscience Institute & Department of Psychology, Princeton University, Princeton, NJ, United States, ⁷Department of Bioengineering, Stanford University, Stanford, CA, United States

This project aims to investigate whether sensory representation measured with fMRI is affected by selective dopaminergic activity. We first measured changes in cerebral blood volume (CBV) signals in response to a range of forepaw stimulation frequencies. We then repeatedly paired a single forepaw stimulation frequency (9 Hz) with 30 Hz optogenetic stimulation of VTA dopaminergic neurons. Following the pairing, we re-assessed changes in CBV fMRI responses to all forepaw stimuli frequencies. We found VTA dopaminergic activity paired with forepaw stimulation enhances the neuronal representation of the sensory stimulus. These data suggest that aberrant dopaminergic signaling may degrade optimal neuronal network dynamics, which in turn may shift large-scale brain network dynamics to promote maladaptive states.

E. Daniel P. Gomez¹, Jenni Schulz¹, Rasim Boyacioglu¹, David G. Norris^1,2, and Benedikt A. Poser^3
1Donders Institute for Brain, Cognition and Behaviour, Radboud University Nijmegen, Nijmegen, Gelderland, Netherlands, ²Erwin L. Hahn Institute for Magnetic Resonance Imaging, University Duisburg-Essen, Essen, Germany, ³Faculty of Psychology and Neuroscience, Maastricht University, Maastricht, Netherlands

Multiband Multiecho 2D-EPI with blipped-CAIPI is an MR imaging sequence that allows the acquisition of multiple slices simultaneously, thus dramatically reducing acquisition time and repetition time, and of multiple echoes with different TEs after a single RF excitation, improving BOLD sensitivity in fMRI studies. In this study we compared various MB-ME 2D EPI protocols - different in the number of slices excited simultaneously, in-plane acceleration factors and CAIPI factors - to maximize the BOLD CNR for fMRI studies.

Tae Kim¹, Tiejun Zhao², Yoojin Lee¹, and Kyongtae Ty Bae^1
1Department of Radiology, University of Pittsburgh, Pittsburgh, PA, United States, ²Siemens Medical Solution USA, Siemens MediCare USA, PA, United States

Multiband (MB) techniques allows many thin slices in conjunction with multi-echo within a given TR normally used for conventional EPI of the whole brain. Summation of contiguous thin slices with this technique recovered signal drop-out from susceptibility artifacts and improved BOLD sensitivity from these areas. This technique can be routinely used for fMRI studies focusing on areas with high susceptibility artifacts.

Stephanie McMains¹, R Matthew Hutchison^1,2, and Ross W Mair^1,3
1Center for Brain Science, Harvard University, Cambridge, MA, United States, ²Department of Psychology, Harvard University, Cambridge, MA, United States, ³AA Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States

We studied visual cortex response at a variety of slice accelerations and TR reductions to investigate the advantages and potential costs for task-based fMRI associated with parameters that allow for whole-brain, sub-second data collection with both block and event-related design paradigms. tstats were significantly higher in the block-design short-TR scans, due to the much larger number of time-points acquired. However, a significant increase is still observed after auto-regressive modeling correction, indicating benefits to block-design experiments from high temporal resolution acquisitions. The observation of higher betas in the event-related design experiment with the short-TR scans is unexpected, and perhaps indicates the hemodynamic response is captured more accurately with the higher temporal resolution.

Sascha Brunheim^1,2, Helen C. Lückmann¹, Prantik Kundu³, Rainer Goebel^1,2, and Benedikt A. Poser^1
1Faculty of Psychology and Neuroscience, Department of Cognitive Neuroscience, Maastricht University, Maastricht, Netherlands, ²Brain Innovation B.V., Maastricht, Netherlands, ³Section on Functional Imaging Methods, Laboratory of Brain and Cognition, National Institutes of Health, Bethesda, MD, United States

Multi-echo and multi-band techniques have independently and rapidly established themselves as new standards in fMRI acquisition. Bringing them together offers tremendous potential to reap the benefit of both: faster TR to better resolve confounding physiological noise such as heart beats, and the multi-echo information to then separate out these signals. We evaluate different combinations of slice- and in-plane acceleration with different numbers of echoes in conjunction with multi-echo denoising using ME-ICA.

Ottavia Dipasquale^1,2, Arjun Sethi³, Maria Marcella Laganà², Francesca Baglio², Prantik Kundu⁴, Giuseppe Baselli¹, Neil A Harrison³, and Mara Cercignani^3
1Politecnico di Milano, Milan, MI, Italy, ²IRCCS, Don Gnocchi Foundation, Milan, MI, Italy, ³Clinical Imaging Sciences Centre, Brighton and Sussex Medical School, Brighton, United Kingdom, ⁴Section on Advanced Functional Neuroimaging, Brain Imaging Center, Icahn School of Medicine at Mount Sinai, New York, NY, United States

Artifact removal is an intrinsic challenge of resting state fMRI data, since images are acquired without experimental modulation of brain function, thus with no a-priori knowledge about the signal of interest. The aim of this study was to compare some data-driven cleaning procedures based on single- and multi-echo acquisitions. We qualitatively show the ability of the different methods to remove artifacts from a dataset that includes both typical subjects and subjects with a neuropsychiatric disorder, characterised by restlessness and high degree of head movements.

Wietske van der Zwaag¹, Mayur Narsude², Marzia Restuccia², Olivier Reynaud^1,3, Daniel Gallichan¹, and Jose P. Marques^1
1CIBM, EPFL, Lausanne, VD, Switzerland, ²LIFMET, EPFL, Lausanne, VD, Switzerland, ³Department of Radiology, Bernard and Irene Schwartz Center for Biomedical Imaging, NYU School of Medicine, New York, NY, United States

The 3D-EPI-CAIPI sequence was combined with a 2D-RF pulse to selectively image a cylindrical region of interest (such as the brainstem) with high spatial (1.5mm) and high temporal resolution (1s/volume). Use of the cylindrical excitation pulse did not reduce tSNR or BOLD sensitivity within the region of interest. This sequence allows a significant reduction of the FOV compared to full slab acquisitions, which can be used to increase spatial and/or temporal resolution and to reduce susceptibility-induced distortions, finally resulting in functional data with high spatio-temporal resolution, high anatomical fidelity and high BOLD sensitivity.

Nick Todd¹, Steen Moeller², Edward J. Auerbach², Essa Yacoub², Guillaume Flandin¹, and Nikolaus Weiskopf^1
1Wellcome Trust Centre for Neuroimaging, University College London, London, United Kingdom, ²Center for Magnetic Resonance Research, University of Minnesota, Minnesota, United States

This study evaluates a 2D multiband EPI sequence for 1.5mm high resolution, whole brain fMRI applications at 3T in terms of BOLD sensitivity and possible false activation due to signal leakage between the simultaneously excited slices. Ten healthy volunteers were imaged performing a visual/motor task under a 4 x 2 within-subject factorial design that considered multiband factors 1, 2, 4, and 6 (all with factor 2 in-plane GRAPPA), and reconstruction methods Slice-GRAPPA and Split Slice-GRAPPA. Results show clear gains in BOLD sensitivity for multiband factors 2, 4, and 6, manifested as a greater number of total voxels activated and higher maximum t-score values. However, signal from highly activated clusters was leaking into adjacent simultaneously excited slices and causing false positive activation at known alias locations. The effect was strongest for multiband factors 4 and 6 reconstructed with Slice-GRAPPA, but was greatly reduced using the Split Slice-GRAPPA approach.

caiyun shi¹, xiaoyong zhang^1,2, guoxi xie¹, lijuan zhang¹, chunxiang jiang¹, and xin liu^1
1Shenzhen Key Lab for MRI, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, shenzhen, guangdong, China, ²Centers for Biomedical Engineering, College of Information Science and Technology, University of Science and Technology of China, Hefei, China

Function magnetic resonance imaging(fMRI), which is a technique that measures the hemodynamic response realted to neural activity in the brain, requires a high spatial and temporal resolution. In this paper,we proposed a new method for accelerating fMRI experiments using sparse sampling of (k, t)-space which can greatly improve the temporal resolution than EPI. The performance of the proposed method has been evaluated using retrospective undersampling of EPI-based fMRI data and in vivo experiment. The proposed method produced accurate reconstruction of both the gray-scale images and the activation maps, and captured the BOLD signal with reduced imaging time for fMRI study.

Valur Olafsson¹, Prantik Kundu², and Thomas Liu^3
1Neuroscience Imaging Center, University of Pittsburgh, Pittsburgh, PA, United States, ²Dept. of Radiology, Icahn School of Medicine at Mount Sinai, New York, NY, United States, ³Center for functional MRI, UCSD, La Jolla, CA, United States

The recent emergence of fast simultaneous multi-slice functional MRI acquisitions has increased interest in exploring high frequency resting-state networks for functional connectivity MRI. Although studies have reported detecting high frequency networks, little has been done to investigate if the underlying source is truly BOLD based. Here, we propose to investigate the occurrence of whole brain high frequency BOLD resting-state networks, using multi-echo independent component analysis (ME-ICA) of high-pass filtered multi-echo simultaneous multi-slice (MESMS) data, which allows for automatic identification of high frequency BOLD and non-BOLD networks. We find that BOLD networks at frequencies higher than 0.2Hz are largely nonexistent.

Andrew S Nencka¹, Andrew M Huettner², L. Tugan Muftuler³, Kevin M. Koch¹, and Rasmus Birn^4
1Departments of Biophysics and Radiology, Medical College of Wisconsin, Milwaukee, WI, United States, ²Department of Biophysics, Medical College of Wisconsin, Milwaukee, WI, United States, ³Department of Neurosurgery, Medical College of Wisconsin, Milwaukee, WI, United States, ⁴Department of Psychiatry, University of Wisconsin, Madison, WI, United States

Simultaneous multislice (SMS) imaging is becoming wide spread in resting state acquisitions following developments from the Human Connectome Project. While most SMS studies are focused upon increasing both spatial and temporal resolution, this study is focused primarily upon increasing volume acquisition time to effectively freeze subject motion. In cases of challenging patient populations, including unsedated children, subject motion can destroy functional connectivity measures. The goal of this work is to acquire the full brain volume as quickly as possible in a resting state acquisition that is expected to be corrupted by motion. In this case, intra-volume motion artifacts can be eliminated. Although full volume coverage with 8mm isotropic voxels can be achieved in 150 ms, data acquired over 390 ms with spatial resolution of 2.6 x 2.6 x 4 mm³ was found to sufficiently avoid intra-acquisition motion artifacts in the presence of substantial motion. This work offers a new perspective for the application of SMS acquisitions in patient populations.

Bruno Riemenschneider¹, Pierre Levan¹, Marco Reisert¹, and Jürgen Hennig^1
1University Medical Center Freiburg, Freiburg, Germany

It is shown that when restricting to a number of predefined volumes, nonlinear trajectories can be used for real-time fMRI. The average time series of the the volumes can be computed efficiently after performing pre-scan calculations. A comparison of two methods is conducted using simulated data, and exemplary time series is reconstructed using real data.

Laurentius Huber¹, Maria Guidi¹, Jozien Goense², Toralf Mildner¹, Robert Trampel¹, Jessika Schulz¹, Cornelius Eichner¹, Robert Turner¹, and Harald E Möller^1
1Max Planck Institute for Human Cognitive & Brain Sciences, Leipzig, Germany, ²University of Glasgow, United Kingdom

We investigated here how the T₂^*-decay during k-space acquisition in EPI affects the blurring of the resulting image. We show that the corresponding PSF broadening is an inadequate measure of describing the associated T₂^*-blurring. Based on theoretical and experimental results shown, we conclude that full k-space acquisition GE-EPI has higher effective resolution than half k-space acquisition GE-EPI or than SE-EPI. This helps to disentangle layer-dependent activation with higher effective resolution.

Eric Wong^1
1Radiology/Psychiatry, UC San Diego, La Jolla, CA, United States

In fMRI, image acquisition and processing are typically treated as separate processes. We show here that when the goal is to obtain the functional time course of brain networks or nodes, the SNR per unit time can be dramatically increased by collecting highly under sampled data and estimating the time courses directly from raw k-space data, bypassing the formation of conventional images.

Ikuhiro Kida^1,2, Takashi Ueguchi^1,2, Yuichiro Matsuoka^1,2, and Maxim Zaitsev^3
1Center for Information and Neural Networks, National Institute of Information and Communications Technology, Suita, Osaka, Japan, ²Graduate School of Frontier Biosciences, Osaka University, Suita, Osaka, Japan, ³University Medical Centre Freiburg, Freiburg, Germany

Head motion may have a negative impact on fMRI results with high spatial resolution at ultra-high magnetic field. Prospective motion correction (PMC) using external position information based on optical tracking is one of the solutions for mitigating motion-related false positive activation (FPA). We evaluated the PMC with 1 mm isotropic resolution at 7T. A great improvement of FPA for phantom study suggest that PMC using optical tracking system promises an improvement of motion-induced artifact at 1 mm isotropic resolution. In a human study only a slight improvement was observed, which suggests improvements are needed for both the present study design and possibly the PMC set-up.

Dimo Ivanov¹, Markus Barth², Kâmil Uludağ¹, and Benedikt A Poser^1
1Department of Cognitive Neuroscience, Maastricht University, Maastricht, Netherlands, ²University of Queensland, Brisbane, Australia

3D echo planar imaging (3D EPI) has become a popular fast imaging sequence with numerous applications especially at ultra-high field. Here, we investigate different ACS acquisition schemes for 3D EPI GRAPPA reconstructions and propose a robust segmentation scheme that considerably reduces residual aliasing and provides high temporal stability.

Lars Kasper^1,2, Maximilian Haeberlin¹, Saskia Bollmann¹, S. Johanna Vannesjo¹, Bertram J. Wilm¹, Benjamin E. Dietrich¹, Simon Gross¹, Klaas E. Stephan², and Klaas P. Pruessmann^1
1Institute for Biomedical Engineering, University of Zurich and ETH Zurich, Zurich, Switzerland, ²Translational Neuromodeling Unit, Institute for Biomedical Engineering, University of Zurich and ETH Zurich, Zurich, Switzerland

High-resolution single-shot variable density spirals with acquisition densities matched to a Gaussian smoothing kernel are presented. Concurrent field monitoring and advanced image reconstruction (iterative SENSE and static B0-correction) proved to be instrumental to arrive at a high image quality. Matched-filter spirals increased tSNR of time series after fMRI-typical post-processing compared to Archimedean spirals and in summary, realize the best match to Gaussian smoothing kernel to date.

R. Allen Waggoner¹, Zhentao Zuo², Yan Zhuo², Topi Tanskanen¹, Kenichi Ueno³, Keiji Tanaka¹, and Kang Cheng^1,3
1Laboratory for Cognitive Brain Mapping, RIKEN - Brain Science Institute, Wako-shi, Saitama, Japan, ²State Key Laboratory of Brain and Cognitive Science,Beijing MRI Center for Brain Research, Institute of Biophysics, Chinese Academy of Sciences, Beijing, Beijing, China, ³RRC, RIKEN - Brain Science Institute, Wako-shi, Saitama, Japan

Physiological noise correction (PNC) is now commonly used to minimize the variance in fMRI data. The correction is most often performed in image space, however this methodology was originally developed as a k-space based correction. GRAPPA is widely used to accelerate data acquisition. The ACS lines are usually acquired separately, however this temporal mismatch could allow physiological noise to degrade the GRAPPA reconstruction. To explore the influence of physiological noise on GRAPPA reconstruction, we have applied PNC at various points in the processing pipeline of both normal GRAPPA reconstructed EPI data and data with temporally updated GRAPPA weights (tGRAPPA).

Alan Chu¹ and Douglas Noll^1
1Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States

Multiple receive coils are needed for simultaneous multislice (SMS) imaging, resulting in large raw data storage requirements and long computation times for a typical SMS fMRI scan. In this abstract, we compare and evaluate the activation performance of varying levels of coil compression on both SMS and conventional fMRI scans. We use an efficient concentric ring trajectory that has the susceptibility signal recovery benefit of spiral-in imaging, and also facilitates the use of an SMS GRAPPA reconstruction. GRAPPA and SENSE SMS reconstruction results are compared with each other and with those of a conventional scan.

Ana Beatriz Solana Sánchez¹, Anne Menini¹, Laura Sacolick¹, Nicolas Hehn¹, and Florian Wiesinger^1
1GE Global Research, Garching bei Muenchen, Bayern, Germany

This work presents a novel T2-prepared RUFIS method that tackles two eminent technical challenges of current BOLD fMRI: acoustic noise and geometric distortions. The method was tested and compared with GE-EPI and SE-EPI in healthy volunteers scanned at 3T using a two-segment T2 prepared RUFIS approach in motor and auditory tasks. Additionally, an improved approach using single-shot and further reduced acoustic noise was tested. The method showed a reduction of 40 dB(A) in LAeq acoustic noise measurement with respect to EPI-based methods, lack of geometrical distortions and consistent BOLD response, less sensitive (~1.3% BOLD change) but apparently spatially more specific.

Jingyuan Chen¹ and Gary Glover^1
1Electrical Engineering, Stanford University, Stanford, CA, United States

In the present study, we demonstrated that (1) the scale-free dynamics of BOLD time series can be simulated by the convolution of spontaneous event trains and hemodynamic response functions; (2) the scale-free BOLD time series driven by the spontaneous event train may explain (at least partially) the scale-free property reported by previous fMRI literature; (3) Scale-free dynamics of fMRI time series may carry non-neural information, e.g. local hemodynamic fluctuations, suggesting caution in studies attempting to employ metrics such as Hurst exponent as biomarkers for neuroimaging investigations.

Siyuan Liu¹, Lalith Talagala², Souheil Inati³, Yisheng Xu¹, Ho Ming Chow¹, Gang Chen⁴, and Allen Braun^1
1NIDCD, National Institutes of Health, Bethesda, Maryland, United States, ²NMRF/NINDS, National Institutes of Health, Bethesda, MD, United States, ³FMRIF/NIMH, National Institutes of Health, Bethesda, MD, United States, ⁴SSCC/NIMH, National Institutes of Health, Bethesda, MD, United States

GRAPPA parallel acquisition technique (PAT) is frequently used in current fMRI EPI protocols to reduce EPI distortions and to increase the time resolution. Here, we demonstrated that when FLASH rather than EPI based reference scans were used in GRAPPA accelerated EPI, the sensitivity to detect brain activations increased in task-related fMRI and such an increase could be critical in detecting activated regions with small effect sizes at the group level. Using FLASH reference scans also improved the group-level resting-state functional connectivity results. All improvements are attributed to the enhancement of temporal SNR when using FLASH based reference scans.

Lalit Gupta¹, Mariëlle Vlooswijk², Rob P. W. Rouhl², Rick Janssens², Anton de Louw³, Bert Aldenkamp³, Shrutin Ulman¹, René M.H. Besseling⁴, Paul A.M. Hofman², Jacobus F. A. Jansen⁴, and Walter H Backes^4
1Philips India Ltd., Bangalore, Karnataka, India, ²Dept of Neurology, Maastricht University Medical Center, Maastricht, Netherlands, ³Epilepsy Center Kempenhaeghe, Heeze, Netherlands, ⁴Department of Radiology, Maastricht University Medical Center, Maastricht, Netherlands

Temporal heterogeneity in BOLD time-series in patients with a suspect of a single epileptic seizure (first-fit) has been compared to patients with new-onset epilepsy and healthy controls. Known temporal heterogeneity measures, including fractional amplitude of low frequency fluctuations and regional homogeneity, were used for time-series analysis. Results suggest that particularly for the low frequency range, 0.02-0.09 Hz, first-fit patients show decreased amplitude of the BOLD fluctuations. These effects show that the analysis of frequency fluctuations has potential to predict which patients are at increased risk to develop epilepsy after a single seizure.

Samira M Kazan¹, Siawoosh Mohammadi¹, Martina F Callaghan¹, Guillaume Flandin¹, Robert Leech², Aneurin Kennerley³, Christian Windischberger⁴, and Nikolaus Weiskopf^1
1Wellcome Trust Centre for Neuroimaging, UCL Institute of Neurology, London, United Kingdom, ²Cognitive, Clinical and Computational Neuroimaging Lab, University of London, Imperial College, London, United Kingdom, ³Department of Psychology, University of Sheffield, Sheffield, United Kingdom, ⁴MR Centre of Excellence, Centre for Medical Physics and Biomedical Engineering, Medical University of Vienna, Vienna, Austria

Functional MRI group studies play a central role in basic and clinical neuroscience, but without large group sizes their statistical powers are rather low, mainly due to high inter-subject variance. We present a novel auto-calibration method, termed VasA fMRI, which significantly reduces variation and increases statistical power by accounting for vascularization differences between subjects. Sensitivity increases exceeded 20% in different brain areas, which is comparable to the gain achieved by a 40% increase in group size or by going from 1.5T to 3T. VasA fMRI does not require additional reference scans and can be applied to any task-related fMRI dataset. It facilitates investigation of subtle effects, the use of smaller groups and revisiting previous studies that were statistically underpowered.

Tim M Tierney¹, Louise J Croft², Maria Centeno¹, Elhum A Shamshiri¹, Suejen Perani^1,3, Torsten Baldeweg², Christopher A Clark¹, and David W Carmichael^1
1Developmental Imaging and Biophysics, UCL Institute of Child Health, London, United Kingdom, ²Cognitive Neuroscience and Neuropsychiatry, UCL Institute of Child Health, London, United Kingdom, ³Department of Basic and Clinical Neuroscienc, KCL Institute of Psychiatry, Psychology & Neuroscience, London, United Kingdom

Subject head motion negatively affects task based fMRI and connectivity based studies. We have developed a biophysical model named FIACH to address these issues. FIACH is a biophysical model of motion correction that employs a two-step procedure for motion correction. We have demonstrated it performs better than four other methods of retrospective motion correction. This was demonstrated in a paediatric population during overt speech where subject motion is common. FIACH is also capable of correcting data near inferior temporal areas. These areas have proven problematic for fMRI and FIACH provides the opportunity to improve knowledge of these areas function.

Muhammad Kaleem¹ and Dietmar Cordes^2
1Ryerson University, Toronto, Ontario, Canada, ²Lou Ruvo Center for Brain Health, Las Vegas, Nevada, United States

The abstracts presents a study of the Hurst exponent of fMRI BOLD time-series acquired at different sampling rates (TR). The Hurst exponent is related to the spectral characteristics of a time-series, and the change in the Hurst exponent value at different TRs due to frequency aliasing is demonstrated.

Mary C. Kociuba¹ and Daniel B. Rowe^1,2
1Department of Mathematics, Statistics, and Computer Science, Marquette University, Milwaukee, Wisconsin, United States, ²Department of Biophysics, Medical College of Wisconsin, Milwaukee, Wisconsin, United States

Given a dual-task experimental design with different task periods, two distinct task-activated peaks are expected. Signal processing will alter the activated voxel’s temporal frequency spectrum, by spreading voxel task activated peaks temporally and spatially into neighboring frequencies and voxels. If the period of both tasks falls within a close range, without knowledge of the expected task-activated regions, it is difficult to resolve the task to the associated region after processing. Modeling the impact of signal processing, with respect to the task-activated frequency, is critical to developing methods to account for artificial correlations, while preserving the signal of interest.

Tim van Mourik¹, Jan PJM van der Eerden¹, and David G Norris^1
1Donders Centre for Cognitive Neuroimaging, Radboud University Nijmegen, Nijmegen, Netherlands

We propose a new method to reliably and accurately extract laminar time courses from fMRI data. We use a spatial General Linear Model to disentangle laminar signals instead of interpolating through and integrating over a cortical area. We show that the Point Spread Function of the method is much sharper, proving better signal extraction. This may allow for larger voxel sizes to get around problems such as low SNR and severe distortions in EPI with which laminar analysis has dealt.

Klaus Scheffler¹, Gabriele Lohmann¹, Christian Mirkes¹, Shajan Gunamony¹, and Philipp Ehses^1
1MRC department, Max Planck Institute for Biological Cybernetics, Tübingen, Germany, Tübingen, Germany

Balanced OVOC is a variant of the FLASH-based One-Voxel-One-Coil (OVOC) method introduced by Jürgen in 2006. It aims to measure functional responses at a very high temporal resolution as no or just one spatial encoding direction is applied. Measurements of functional responses to visual checkerboard stimulation at a temporal resolution of 5 ms and one-dimensional resolution of 1 mm using a small surface coil of 3 cm are presented. With an echo time of 2.5 ms for bOVOC signal changes are mostly related to T2 and diffusion changes.

Jonas Bause^1,2, Myung-Ho In³, Philipp Ehses^1,4, G. Shajan¹, Oliver Speck³, Rolf Pohmann¹, and Klaus Scheffler^1,4
1High-Field Magnetic Resonance Center, Max Planck Insitute for Biological Cybernetics, Tuebingen, Germany, ²Graduate Training Centre for Neuroscience, University of Tuebingen, Tuebingen, Germany, ³Department for Biomedical Magnetic Resonance, University of Magdeburg, Magdeburg, Germany, ⁴Department of Biomedical Magnetic Resonance, University of Tuebingen, Tuebingen, Germany

High-resolution EPI at ultra-high field often suffers from distortions and requires a short time to k-space center in order to achieve the optimal BOLD contrast. In this work, zoomed functional imaging was combined with the point spread function mapping technique in order to obtain images with re-duced distortions acquired with the optimal TE for fMRI at 9.4T. The approach was tested in a functional experiment with 0.8 mm and 0.65 mm iso-tropic resolution. Activation maps obtained from the distortion corrected images showed a higher number of activated voxels then the maps from the undistorted images and match well with anatomy.

Pradyumna Sepúlveda^1,2, Ranganatha Sitaram^3,4, Mohit Rana^3,5, Tomás Ossandón⁶, Marcelo Andía^1,7, Cristián Montalba¹, Sergio Uribe^1,7, Pablo Irarrázaval^1,2, Sergio Ruiz^4,6, and Cristián Tejos^1,2
1Biomedical Imaging Center, Pontificia Universidad Católica de Chile, Santiago, Chile, ²Department of Electrical Engineering, Pontificia Universidad Católica de Chile, Santiago, Chile, ³Department of Biomedical Engineering, University of Florida, Gainesville, Florida, United States, ⁴Institute of Medical Psychology and Behavioral Neurobiology, University of Tübingen, Tübingen, Germany, ⁵Graduate School of Neural & Behavioural Sciences, International Max Planck Research School, University of Tübingen, Tübingen, Germany,⁶Department of Psychiatry, Faculty of Medicine, Interdisciplinary Center for Neuroscience, Pontificia Universidad Católica de Chile, Santiago, Chile, ⁷Radiology Department, Faculty of Medicine, Pontificia Universidad Católica de Chile, Santiago, Chile

Healthy subjects and patients can self-regulate brain activity locally or globally using real time fMRI neurofeedback (rtfMRI NF), often accompanied by observable behavioral changes. Furthermore, new therapeutic and research applications have encouraged a growing number of studies using rtfMRI NF. Despite the interest, the relevance of different factors in the self-regulation learning process is yet to be known. In this study we compared the influence of using a conscious mental strategy during neurofeedback versus pure contingent feedback. Results indicate that the use of a conscious strategy does not necessarily improve learning with respect to only NF.

Mitsunobu Kunimi¹, Sachiko Kiyama¹, and Toshiharu Nakai^1
1National Center for Geriatrics and Gerontology, Obu, Aichi, Japan

This study attempted to demonstrate the extent of the task load-dependent augmentation of brain activation and the change of functional connectivity during the TSP. Twenty healthy young adults (mean age: 22.80 �}3.44, 9 males) participated in the study. The significant activation was observed in the frontal and parietal region and it was augmented as task load increased. In addition, the functional connectivity was extended depending on task load. These results were considered that the network of these regions works as a central executive, and it controls switch of attention, task-set reconfiguration and inhibition during TSP.

Klaus Scheffler^1,2 and Philipp Ehses^1,2
1Dept. of Biomedical Magnetic Resonance, University of Tübingen, Tübingen, Germany, ²High-Field MR Center, Max Planck Institute for Biological Cybernetics, Tübingen, Germany

Passband bSSFP at 9.4T provides highly reproducible and stable functional activation maps. Results were compared to GE-EPI and SE-EPI. Based on the very short TE of 2.1 ms for bSSFP, T2*-related signal changes are expected to be very small, which is supported by the similarity of bSSFP and SE-EPI activation maps. As bSSFP is distortion-free, direct overlay to anatomical images is possible. Furthermore, the spatial resolution is not smoothed along the PE-direction due to T2*-related blurring as in EPI-based methods.

Tynan Stevens¹, Tim Bardouille^2,3, Gerhard Stroink¹, David Clarke^1,4, Ryan D'Arcy⁵, and Steven Beyea^1,2
1Dalhousie University, Halifax, Nova Scotia, Canada, ²BIOTIC, Halifax, Nova Scotia, Canada, ³IWK Hospital, Halifax, Nova Scotia, Canada, ⁴QEII Health Sciences Centre, Halifax, Nova Scotia, Canada, ⁵Simon Fraser University, Burnaby, British Columbia, Canada

Functional MRI for pre-surgical mapping is hampered by a high degree of individual variability, requiring experts to make appropriate data processing choices. We have developed a method for automatic and individualized processing of pre-surgical mapping using the ROC-r framework. This method optimizes both pre-processing pipelines and activation thresholds on a case-by-case basis. We compare localization obtained for fMRI with more direct measures of neural activity (MEG, cortical stimulation, and sensory evoked potentials). We find highly concordant spatial localization between modalities, showing the potential of automated optimization of fMRI processing for pre-surgical mapping.

Functional Connectivity Method & Applications

Wednesday 3 June 2015

Vincent Kyu Lee^1,2, Mark DeBrunner³, Jennifer A. Johnson³, Jodie Votava-Smith⁴, Vidya Rajagopalan⁵, Rafael Ceschin^1,2, Michelle Gruss², Frederick S. Sherman³, and Ashok Panigrahy^1,2
1Radiology, University of Pittsburgh, Pittsburgh, Pennsylvania, United States, ²Radiology, Children's Hospital of Pittsburgh, Pittsburgh, Pennsylvania, United States, ³Cardiology, Children's Hospital of Pittsburgh, Pennsylvania, United States, ⁴Cardiology, Children's Hospital of Los Angeles, California, United States, ⁵Children's Hospital of Los Angeles, Pennsylvania, United States

This is an initial presentation of 3T parallel transmission study on fetal complex congenital heart disease (CHD). ICA analysis was used with resting BOLD of fetal CHD patients to test evidence of disruption in fetal resting state networks (RSN) compared to healthy controls, and whether the degree of disruption of fetal RSN within the complex CHD group is related to the degree of presumed fetal cerebral oxygenation saturation. Our preliminary results provide evidence of disruption of the development of normal RSN in the frontal and posterior regions of the brain in complex CHD fetal cases compared to healthy fetal controls.

Maryam Abaei¹, Eugene P Duff², Tomoki Arichi^1,3, Jonathan O'Muircheartaigh^1,4, Emer Hughes¹, Giulio Ferrazzi¹, Steve M Smith², Serena Counsell¹, A David Edwards^1,5, Daniel Rueckert⁶, and Joseph V Hajnal^1,5
1Centre for the Developing Brain, King's College London, London, London, United Kingdom, ²Department of Clinical Neurosciences, Oxford University, Oxford, United Kingdom,³Department of Bioengineering, Imperial College, London, United Kingdom, ⁴Institute of Psychiatry, King's College London, London, United Kingdom, ⁵Division of Imaging Sciences and Biomedical, King's College London, London, United Kingdom, ⁶Biomedical Image Analysis Group, Department of Computing, Imperial College, London, United Kingdom

blood oxygen level dependen contrast-to-noise ratio is known to be maximal when the acquisition sequence echo time is matched to the T2* of the tissue of interest.We aimed to identify the optimal echo-time for performing resting-state fMRI studies in a test subject group.

Maryam Falahpour¹, Lori Haase², Martin P. Paulus², and Thomas T. Liu^1
1Center for Functional MRI, University of California San Diego, La Jolla, CA, United States, ²Department of Psychiatry, University of California San Diego, La Jolla, CA, United States

Mindfulness is a complex construct consisting of regulated attention, knowledge that events are momentary, and absence of emotion or cognitive appraisal of events. The primary aim of the present study was to investigate the relation between the functional connectivity and the changes in self-assessment scores after training in elite athletes. We found that measures of functional connectivity obtained before training can predict changes in subjects’ self-assessment of interoceptive awareness.

Christina Hamilton¹, Yuncong Ma¹, and Pablo Perez^1
1Pennsylvania State University, State College, PA, United States

Anesthesia is an altered state of consciousness that can be characterized by a complex interaction between an individual’s level of arousal and awareness. Resting-state functional magnetic resonance imaging (rsfMRI) has been shown to be a useful tool for investigating systems-level changes in functional brain connectivity during anesthesia-induced unconsciousness (AIU). Here we use rsfMRI and behavioral measures of anesthetic depth to investigate the dose-dependent effect of isoflurane on functional brain connectivity. These findings provide insight into the relationship between neuroimaging data and behavioral measures of anesthetic depth, which may provide us with a better understanding of the neural basis of AIU.

Xiaopeng Zhou¹, Katherine A Koenig¹, Erik B Beall¹, Lael Stone¹, Robert Bermel¹, Michael D Phillips¹, and Mark J Lowe^1
1The Cleveland Clinic, Cleveland, Ohio, United States

We investigated the modulation of motor connectivity in patients with multiple sclerosis (MS) and controls by acquiring fc in 3 conditions (rest, 2 Hz paced finger tapping and self-paced complex finger tapping).Significant fc changes in some brain regions during 2 motor tasks from rs were exhibited, showing fc network can be modulated under different states in controls and patients with MS. There was decreased fc in certain regions of patient compared with controls. The observation that more group differences were observed in the resting state suggests that this may be the most effective way of studying functional impairment in MS.

Guangyu Chen¹, Xiaolin Liu¹, Anthony G Hudetz², and Shi-Jiang Li^1
1Biophysics, Medical College of Wisconsin, milwaukee, WI - Wisconsin, United States, ²Department of Anesthesiology, Medical College of Wisconsin, milwaukee, WI - Wisconsin, United States

Time delayed network mutual information index (TD-NMI) was used to measure the dynamics of functional network modules in wakeful baseline, deep sedation with propofol, and recovery. Compared with wakeful baseline, all identified brain modules showed significantly reduced dynamics in deep sedation. During recovery, the frontal-posterior control module showed fully recovered dynamics, visual module showed partial recovery, while the default mode network and motor-sensory modules showed no sign of recovery. This study shed light, from a network dynamics perspective, on the relative importance of brain network modules and their association with the different states of propofol sedation.

Maryam Falahpour¹, Wesley K. Thompson², Angela E. Abbott³, Mark E. Mulvey³, Michael Datko³, Ralph-Axel Müller³, and Thomas T. Liu^1
1Center for Functional MRI, University of California San Diego, La Jolla, CA, United States, ²Department of Psychiatry, University of California San Diego, La Jolla, CA, United States,³Brain Development Imaging Lab, Department of Psychology, San Diego State University, CA, United States

fMRI resting state studies of subjects with autism spectrum disorder (ASD) have revealed altered functional connectivity (FC) in the default mode network (DMN), as compared to typically developed subjects. We hypothesized that the group differences in FC might reflect differences in the temporal variability of the FC measures. We used mediation analysis to examine the relation between FC and the temporal variability in FC between the two major nodes of the DMN, the posterior cingulate cortex and the medial prefrontal cortex. Our results suggest that under-connectivity in ASD population may be in part due to higher temporal variability in FC.

Marta Bianciardi¹, Nicola Toschi^1,2, Cornelius Eichner¹, Kawin Setsompop¹, Jonathan R. Polimeni¹, Bruce R. Rosen¹, and Lawrence L. Wald^1
1Department of Radiology, A.A. Martinos Center for Biomedical Imaging, MGH and Harvard Medical School, Boston, MA, United States, ²Medical Physics Section, Department of Biomedicine and Prevention, Faculty of Medicine, University of Rome “Tor Vergata”, Rome, Italy

The aim of this work was to validate our recently developed in-vivo structural template of human brainstem-nuclei (Bn) of the ascending-arousal, autonomic and motor systems by 1.1mm-isotropic fMRI at 7Tesla. For each of the 18 Bn-labels of the structural template, a spatially matched functional parcel in the brainstem was identified by ICA. Our results also showed that structural Bn-labels were functionally connected to specific brain regions as expected from previous work. This work provides a preliminary validation of the in-vivo structural Bn template, which is a key step towards its use in future studies of human brainstem function and pathology.

Brice Fernandez¹, Victor Spoormaker², Philipp Sämann², and Michael Czisch^2
1Applications & Workflow, GE Healthcare, Munich, Germany, ²Neuroimaging Unit, Max Planck Institute of Psychiatry, Munich, Germany

Resting state functional connectivity is an attractive tool to study affective and mood disorders. Gradient-echo EPI (GRE-EPI) is affected by strong signal drops in brain regions critical for emotional processing. Spin-echo EPI (SE-EPI) is known to not suffer from this problem. Here, we use a Hybrid EPI pulse sequence to simultaneously acquire both the GRE-EPI and the SE-EPI. The comparison of both acquisition methods shows that meaningful signal fluctuations can be recovered in susceptibility affected brain regions using SE-EPI, while minor differences in functional connectivity were observed for unaffected brain areas. Furthermore, GRE-EPI is more prone to global signal artefacts.

Christine Law¹, Haisam Islam¹, Gary Glover¹, and Sean Mackey^1
1Stanford University, Stanford, CA, United States

Simultaneous brain/spine fMRI has not been adopted due to the challenge of shimming multiple regions of high magnetic field inhomogeneity. A dynamic shim update technique was recently proposed to address this issue that used a separate set of the linear gradients and a frequency offset for the brain and spinal cord. We propose a technique to advance this concept by applying optimal linear shim values on a per-slice basis. Our slice-based technique, which calculates optimal shimming from a field map, would not be possible using the shim procedure provided by a scanner manufacturer.

Jürgen Finsterbusch^1,2
1Department of Systems Neuroscience, University Medical Center Hamburg-Eppendorf, Hamburg, Germany, ²Neuroimage Nord, University Medical Centers Hamburg-Kiel-Lübeck, Hamburg-Kiel-Lübeck, Germany

The feasibility of inner-field-of-view EPI based on 2D-selective RF excitations to acquire T2*-weighted images of the human spinal cord in vivo is demonstrated. Compared to conventional, slice-selective EPI, the readout echo train is shortened reducing geometric distortions and lower sections of the spinal cord can be covered without the need for excessive FOVs to avoid aliasing artifacts. Like for conventional EPI, slice-specific z-shimming can be applied to minimize signal dropouts related to through-slice dephasing. Thus, BOLD-based functional imaging may become feasible throughout the spinal cord.

Jennifer Evans¹, Prantik Kundu², and Peter Bandettini^1
1NIH, Bethesda, Maryland, United States, ²Mount Sinai, New York, United States

Physiological noise removal is an important part of the fMRI preprocessing pipeline but the collection of good physiological data adds an extra complication to data acquisition. Here we look at ICA denoising techniques that are currently available that attempt to remove “non-signal” components either by prior identification (FIX) or based on the amount of BOLD scaling (MEICA) using multi-echo acquisition and the effect of the use of physiological noise regressors on the denoised data.

Catie Chang¹, David A Leopold², Hendrik Mandelkow¹, Marieke L Schölvinck³, and Jeff H Duyn^1
1Advanced MRI Section, Laboratory of Functional and Molecular Imaging, NINDS, NIH, Bethesda, MD, United States, ²Section on Cognitive Neurophysiology and Imaging, Laboratory of Neuropsychology, NIMH, NIH, Bethesda, MD, United States, ³Ernst Strüngmann Institute (ESI) for Neuroscience in Cooperation with Max Planck Society, Frankfurt am Main, Germany

Here, we investigated the spatial and temporal relationship between neural activity and a behavioral index of vigilance state (spontaneous eye open/closure in darkness) using simultaneous fMRI and electrophysiological data from unanesthetized macaques in the resting state. We observed a robust relationship between resting-state electrophysiological and fMRI signals and our measure of eye open/closure, with eye opening correlated with decreases in fMRI signal across most of the cortex, and with increases in the thalamus and cerebellum. Temporal analysis indicated that thalamic and cerebellar regions exhibited increases in signal that preceded the signal decreases in cortex.

Xiaolin Liu¹, Kathryn K. Lauer², B. Douglas Ward¹, Jeffrey R. Binder³, Shi-Jiang Li¹, and Anthony G. Hudetz^2
1Biophysics, Medical College of Wisconsin, Milwaukee, Wisconsin, United States, ²Anesthesiology, Medical College of Wisconsin, Milwaukee, Wisconsin, United States, ³Neurology, Medical College of Wisconsin, Milwaukee, Wisconsin, United States

We proposed a neuroimaging-based anatomical-functional parcellation algorithm for defining brain networks at arbitrary spatial scales and extended this approach to determine global changes in brain functional connectivity during propofol sedation at two sedation levels as compared to wakeful baseline and recovery. Propofol sedation and its deepening are associated with a global reduction of brain functional connectivity. Reduced node connectivity in deep sedation showed different degrees of involvement of cortical and subcortical systems. The study represents a way of comprehensively characterizing complex functional and behavioral changes in sensory, attentional, cognitive and motor systems of the brain during anesthetic sedation.

Ann Sunah Choe^1,2, Craig K Jones^3,4, Suresh E Joel^3,4, John Muschelli⁵, Visar Belegu^6,7, Martin A Lindquist⁵, Brian S Caffo⁵, Peter CM van Zijl^3,4, and James J Pekar^3,4
1Radiology and radiological sciences, Johns Hopkins University School of Medicine, Baltimore, MD, United States, ²F. M. Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, MD, United States, ³Radiology and radiological sciences, Johns Hopkins School of Medicine, MD, United States, ⁴F. M. Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, MD, United States, ⁵Biostatistics, Bloomberg School of Public Health, Johns Hopkins University, MD, United States, ⁶Neurology, Johns Hopkins School of Medicine, MD, United States, ⁷International Center for Spinal Cord Injury, Kennedy Krieger Institute, MD, United States

Despite strong interest in using resting state fMRI (rsfMRI) outcome measures as imaging biomarkers for clinical studies, the temporal structure (e.g., seasonality) of such measures is poorly understood. This study aimed to assess the existence of temporal structure in three commonly used rsfMRI outcomes measures; namely spatial map similarity, temporal fluctuation magnitude, and between-network connectivity. A unique longitudinal dataset reporting on one healthy adult subject scanned on a weekly basis over 185 weeks enabled timeseries analysis on the measures of interest. Results revealed significant linear trend, annual periodicity, and persistence in many resting state networks, for all outcome measures.

Rainer Boegle^1,2, Thomas Stephan^1,3, Matthias Ertl^2,3, Stefan Glasauer^1,4, and Marianne Dieterich^1,3
1German Center for Vertigo and Balance Disorders, DSGZ IFB-LMU, Munich, Bavaria, Germany, ²Graduate School of Systemic Neurosciences, LMU, Munich, Bavaria, Germany,³Department of Neurology, LMU, Munich, Bavaria, Germany, ⁴Center for Sensorimotor Research, LMU, Munich, Bavaria, Germany

Recently it was demonstrated that healthy subjects inside MR machines develop a persistent nystagmus in darkness, while patients who are lacking bilateral peripheral vestibular function do not. We studied the influence of this magnetic vestibular stimulation (MVS) on fMRI resting-state fluctuations in healthy subjects, at 1.5Tesla and at 3Tesla. We found that significant modulation of the default mode network occurs, mainly in areas associated with vestibular function. As proposed for MVS, the modulation-scaling is significantly higher than the expected BOLD signal increase due to B0 without an additional modulation effect. We conclude that MVS does significantly modulate fMRI resting-state networks.

Peter S LaViolette¹, Sean Lew², Scott D Rand¹, Manoj Raghavan³, Kurt Hecox³, and Mohit Maheshwari^1
1Radiology, Medical College of Wisconsin, Milwaukee, WI, United States, ²Neurosurgery, Medical College of Wisconsin, Milwaukee, WI, United States, ³Neurology, Medical College of Wisconsin, Milwaukee, WI, United States

Interhemispheric functional connectivity can be modeled with a new technique called functional homotopy. We acquired resting-state fMRI (rs-fMRI) in two pediatric patients following corpus callosotomy (severing the corpus callosum) for treatment of refractory epilepsy. We calculated homotopy and compared their data to a control dataset to determine differences. We find global changes in interhemispheris connectivity due to the severing of the corpus callosum.

Jun Young Jeong¹, Julia Druzbicki², Kun-Han Lu¹, Haiguang Wen¹, and Zhongming Liu^1,3
1Electrical and computer engineering, Purdue University, West Lafayette, IN, United States, ²Department of statistics, Purdue University, IN, United States, ³Weldon school of biomedical engineering, Purdue University, IN, United States

To facilitate the functional interpretation of the spatial patterns of resting state fMRI extracted by using either the K-means spatial clustering or spatial ICA, we compared them with the patterns obtained by applying these two methods to a large set of task activation maps. We confirmed the close correspondence between the resting state and task activation patterns extracted by using spatial ICA. However, we failed to establish a similar level of correspondence for the spatial k-means clustering. This distinction is speculated to have a methodological basis.

Cameron William Trapp¹, Kishore Vakamundi², and Stefan Posse^3
1Physics, UNM, Coralles, NM, United States, ²DEPARTMENT OF PHYSICS AND ASTRONOMY, UNIVERSITY OF NEW MEXICO, ALBUQUERQUE, NEW MEXICO, United States,³DEPARTMENT OF NEUROLOGY, UNIVERSITY OF NEW MEXICO, NM, United States

Mapping of intrinsic high frequency (> 0.5 Hz) resting state connectivity has been demonstrated in recent studies in motor and visual cortex. In this study we use multi-slab echo volumar imaging with 136 ms temporal resolution to compare resting state connectivity in 6 major resting state networks at frequencies below 0.3 Hz and at frequencies between 0.5 and 3.6 Hz. The high frequency connectivity in the brain corresponds strongly to connectivity observed within the low frequency RSNs. This preliminary study establishes resting state connectivity at frequencies > 0.5 Hz in multiple major RSNs.

Sadia Shakil¹, Chin-Hui Lee¹, and Shella Keilholz^1,2
1Georgia Institute of Technology, Atlanta, Georgia, United States, ²Emory University, Atlanta, Georgia, United States

Window size and offset in sliding window correlation may have an effect on correct identification of functional connectivity states. This effect was examined on controlled simulated networks formed with data taken from resting-state functional MRI scans.

Michaël Bernier¹, Maxime Chamberland¹, Stephen Cunnane², and Kevin Whittingstall^3
1Nuclear medecine and radiobiology, Université de Sherbrooke, Sherbrooke, QC, Canada, ²Institut universitaire de gériatrie de Sherbrooke, Université de Sherbrooke, Sherbrooke, QC, Canada, ³Diagnostic radiology, Université de Sherbrooke, Sherbrooke, QC, Canada

Using non-local mean denoising, we present a reliable method to efficiently uncover subcortical correlated areas in task-based functional network. The regions of interest (ROIs) obtained in both subcortical areas and key structures involving task-based activations with our denoising pipeline allow a better representation of known functional networks involving subcortical structures in resting state fMRI in typical clinical MRI (1.5 T).

Jingwei Sheng¹, Yuhui Chai¹, Bing Wu², Weinan Tang¹, and Jia-Hong Gao^1
1Center for MRI Research, Peking University, Beijing, Beijing, China, ²GE Healthcare MR Research China, Beijing, China

We demonstrated the superiority of the spin-locked oscillatory excitation (SLOE) method proposed by our group in detecting sub-nanotelsa oscillatory fields in a phantom study. Furthermore, this sensitivity was verified in anesthetic rats with similar applied oscillatory fields. Our proposed SLOE method showed extremely high detection sensitivity that enables 3T MR scanners for the first time to direct detect applied sub-nT oscillatory magnetic fields both in phantom and rats. The milestone achieved in this work paves the way towards direct mapping neuronal currents by MRI in the study of human brain function.

Venkatagiri Krishnamurthy¹, Romeo S. Cabanban², and Kaundinya S. Gopinath^1
1Dept. of Radiology and Imaging Sciences, Emory University, Atlanta, GA, United States, ²Center for Systems Imaging, Emory University, Atlanta, GA, United States

Blood oxygen level dependent (BOLD) resting state fMRI (rsFMRI) has been the most widely employed method to explore resting-state functional connectivity (rsFC) in brain networks. Signal enhancement by extravascular protons (SEEP) has been advanced as an alternate mechanism to assess brain activation. In this study we compared rsFC networks highlighted by BOLD- and SEEP-contrast rsFMRI. The results of this study show that rsFMRI obtained with SEEP contrast exhibits differential sensitivity to functional connectivity networks compared to BOLD rsFMRI. While BOLD rsFC networks exhibit higher sensitivity to state-dependent brain networks (e.g. DMN), the SEEP contrast exhibits higher sensitivity to anatomic connections.

KISHORE VAKAMUDI^1,2, ELENA ACKLEY², and STEFAN POSSE^1,2
1DEPARTMENT OF PHYSICS AND ASTRONOMY, UNIVERSITY OF NEW MEXICO, ALBUQUERQUE, NEW MEXICO, United States, ²DEPARTMENT OF NEUROLOGY, UNIVERSITY OF NEW MEXICO, ALBUQUERQUE, NEW MEXICO, United States

Seed-based connectivity analysis (SCA) is widely used to study functional connectivity. However, it suffers from variability inherent in investigator-specific and subject-specific seed selection dependencies. The current Automated Subject-Specific Seed Optimization (ASSSO) method uses an iterative brain atlas based approach identifies the optimal seed locations to maximize the detected RSN connectivity in individual subjects. This seed-selection will maximize the sensitivity for detecting RSN dynamics at short and long time scales across the entire brain in real-time. This methodology is expected to have important applications in presurgical mapping.

Shiyang Chen^1,2, Hasan Ertan Cetingul², Xiaoping Hu^1,3, and Mariappan S. Nadar^2
1The Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, Atlanta, GA, United States, ²Imaging and Computer Vision, Siemens Corporation, Corporate Technology, Princeton, NJ, United States, ³Biomedical Imaging Technology Center, Emory University, Atlanta, GA, United States

Resting state networks (RSNs) are important biomarkers for disease diagnosis, (e.g. Alzheimer’s disease). The conventional methods to detect RSNs require a spatial smoothing step to compensate the low SNR of fMRI, and they assume the brain connectivity is voxel-to-voxel. These steps and assumptions are still controversial. We propose a searchlight plus multivariate regression method to detect the RSNs, which is able to detect region-to-region brain connection and can be performed on unsmoothed or slightly smoothed fMRI data. We validate this method by detecting the conventional RSNs, and we also found additional brain regions that are connected in the RSNs.

Md Taufiq Nasseef^1,2, Adam Liska^1,2, Stefano Panzeri¹, and Alessandro Gozzi^1
1Center for Neuroscience and Cognitive Systems @UniTn, Istituto Italiano di Tecnologia, Rovereto, TN, Italy, ²Center for Mind/Brain Sciences, University of Trento, Rovereto, TN, Italy

We have recently shown that the mouse brain contains distributed resting-state fMRI (rsfMRI) networks including a putative “default-mode” like network (DMN). To map the direction of information flow within this network, we have used Granger Causality (GC), a directed measure of functional connectivity, to describe the dominant direction of causation among four key nodes of the mouse DMN. We provide evidence of directional information transfer between nodes of this network that replicate patterns recently described for the human DMN, thus corroborating the presence of an evolutionary conserved DMN precursor in the rodent brain.

Erik Beall¹ and Mark Lowe^1
1Imaging Institute, Cleveland Clinic, Cleveland, OH, United States

Head motion in BOLD functional imaging is a serious problem. Current methods all assume the artifact is linearly related to head motion. We show this assumption is incorrect and demonstrate a fully retrospective spin history modeling and correction tool.

Erik Beall¹ and Mark Lowe^1
1Imaging Institute, Cleveland Clinic, Cleveland, OH, United States

Head motion is a major confound in BOLD imaging. Until very recently, it was only possible to use volumetric motion parameters. Hence, the distribution of typical head motion and the frequency spectrum at the slice level are unknown. We investigate these and find that while these are different, its not unexpected given the different frequency with which each measure is operating. However, we also find clear physiologic noise signatures and demonstrate that SLOMOCO can be used to obtain pulse and respiration traces.

fMRI:Bold Physiology & Multimodal Imaging

Wednesday 3 June 2015

Vadim Zotev¹, Ahmad Mayeli^1,2, and Jerzy Bodurka^1,3
1Laureate Institute for Brain Research, Tulsa, OK, United States, ²Electrical and Computer Engineering, University of Oklahoma, Tulsa, OK, United States, ³College of Engineering, University of Oklahoma, Tulsa, OK, United States

Implementation of EEG neurofeedback (EEG-nf) with simultaneous fMRI requires an accurate removal of EEG-fMRI artifacts in real time. We describe a simple and safe modification of a standard MR-compatible EEG cap that substantially improves signal quality for EEG-nf based on frontal EEG asymmetry. The modified cap includes two reference contours for two frontal EEG channels used to provide EEG-asymmetry-nf. Subtraction of fitted reference signals provides random-motion and cardioballistic artifact power suppression by 6...8 dB. The proposed EEG signal correction should be performed in real time after the MR artifact removal and prior to the average cardioballistic artifact subtraction.

Xiao Liu¹, Toru Yanagawa², David A Leopold³, Marieke Schölvinck⁴, Catie Chang¹, Hiroaki Ishida⁵, Naotaka Fujji², and Jeff H. Duyn^1
1AMRI, LFMI, NINDS, NIH, Bethesda, MD, United States, ²BSI, RIKEN, Saitama, Japan, ³Laboratory of Neuropsychology, NIMH, NIH, Bethesda, MD, United States, ⁴Ernst Strüngmann Institute for Neuroscience, Frankfurt, Hessen, Germany, ⁵Tokyo Metropolitan Institute of Medical Science, Tokyo, Japan

To investigate the neural correlate of large-scale signal fluctuation in resting state fMRI (rsfMRI) and understand its relationship to brain state, we examined the ECoG power recorded from large electrode grids in macaques during eyes-open, eyes-closed wakefulness, and sleep conditions. We observed a brain-state-sensitive, stereotypical event whose spatiotemporal characteristics resembled the rsfMRI characteristics. With an independent dataset of concurrent fMRI-electrophysiology, we further confirmed that the identified events were directly associated with global changes in resting-state fMRI signal.

Eleonora Maggioni^1,2, Claudio Zucca¹, Gianluigi Reni¹, Fabio Maria Triulzi³, Anna Maria Bianchi², and Filippo Arrigoni^1
1Scientific Institute IRCCS E.Medea, Bosisio Parini, LC, Italy, ²Department of Electronics Information and Bioengineering, Politecnico di Milano, Milano, MI, Italy, ³Neuroradiology Unit, Fondazione IRCCS Cà Granda, Ospedale Maggiore Policlinico, Milano, MI, Italy

This contribution is focused on the analysis of physiological response to intermittent photic stimulation using simultaneous EEG-fMRI data. The EEG-fMRI integration allowed to investigate the phenomena of neurovascular coupling, in particular it gave insight into the neural mechanisms underlying negative BOLD responses to visual stimulations.

Garth John Thompson¹, Valentin Riedl^2,3, Timo Grimmer^3,4, Alexander Drzezga⁵, Peter Herman¹, and Fahmeed Hyder^1,6
1Diagnostic Radiology, Magnetic Resonance Research Center, Yale University, New Haven, CT, United States, ²Neuroradiology, Nuclear Medicine, Universität München, München, Germany, ³Technische, Universität München - Neuroimaging Center, München, Germany, ⁴Psychiatry, Universität München, München, Germany, ⁵Nuclear Medicine, Uniklinikum, Koeln, Germany, ⁶Biomedical Engineering, Yale University, New Haven, CT, United States

While resting-state fMRI (R-fMRI) is powerful, specific "rest" condition parameters (e.g., eyes closed vs. open) vary study to study. Subjects underwent simultaneous R-fMRI and PET scans with eyes open and closed. R-fMRI data were used to generate networks, whereas PET data mapped glucose consumption (CMR_glc). While globally higher CMR_glc was observed with eyes open, there were insignificant CMR_glc differences across networks for either state. Since all R-fMRI networks were detected regardless of the state, CMR_glcconstancy across networks for a state suggests that the brain easily transitions between these networks and either "rest" condition can be used for R-fMRI.

Mwape Mofya¹, Alkatafi Ali Alhamud¹, Paul A. Taylor^1,2, André J. W. van der Kouwe³, and Ernesta M. Meintjes^1
1MRC/UCT Medical Imaging Research Unit, Department of Human Biology, University of Cape Town, Cape Town, South Africa, ²African Institute for Mathematical Sciences (AIMS), South Africa, ³Massachusetts General Hospital, Boston, Massachusetts, United States

Simultaneous DTI-fMRI data collection has previously been unsuccessful, due to mutual interference of the acquired DTI and fMRI signals. In this work, a single 3D EPI navigator, which has been used for motion correction and shown not to interfere with DWI collection, has been inserted after each DWI volume for BOLD measurement. For faster temporal sampling, we also test the inclusion of a second 3D EPI navigator in the middle of each DTI volume acquisition. Resting state networks produced in the standard BOLD and double navigated DTI-fMRI showed quite similar spatial extent and connectivity. Diffusion data produced similar whole brain distributions of tensor parameters, as well as strikingly similar tractographic reconstructions.

Katarzyna Bienkowska¹ and Valentin Riedl^1
1Neuroradiology, Technische Universitat Munchen, Munich, Germany

In presented study, a combined fMRI-MRS study has been performed to examine interactions between GABA/Glx levels and local intrinsic brain activity (iFC). Recently there has been considerable progress done in investigating interactions between GABA level and BOLD signal change during visual stimulation. What has not been carefully studied yet is the relationship between resting state coherent BOLD signal change (functional connectivity) and excitatory/inhibitory neurotransmitter balance. Therefore, GABA and Glx concentration in the occipital cortex were correlated with local functional connectivity (lFC) and amplitude of low frequency fluctuations (ALFF) of the resting brain in eyes-open and eyes-closed condition.

Frédéric Grouiller¹, João Jorge^2,3, Francesca Pittau⁴, Pascal Martelli¹, Wietske van der Zwaag⁵, Christoph M. Michel⁶, Serge Vulliémoz⁴, Maria Isabel Vargas¹, and François Lazeyras^1
1Department of Radiology and Medical Informatics, Geneva University Hospital, Geneva, Switzerland, ²Department of Bioengineering, Institute for Systems and Robotics, University of Lisbon, Lisbon, Portugal, ³Laboratory for Functional and Metabolic Imaging, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, ⁴EEG and Epilepsy Unit, Department of Neurology, Geneva University Hospital, Geneva, Switzerland, ⁵Biomedical Imaging Research Center (CIBM), Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, ⁶Functional Brain Mapping Laboratory, University of Geneva, Geneva, Switzerland

We report the feasibility to record simultaneous EEG-fMRI at 7T in epileptic patients. We successfully located epileptic networks using an optimized setup to reduce environmental noise and appropriate artefacts removal algorithms. The quality of intra-MR EEG allows detecting interictal epileptiform discharges and performing noise-sensitive topography-related analyses to detect epilepsy-specific haemodynamic changes. These results open new perspectives to better characterize epileptic networks at higher field.

Florian Schmid¹, Miriam Schwalm², Lydia Wachsmuth¹, Cornelius Faber¹, and Albrecht Stroh^2
1Department of Clinical Radiology, University of Münster, Münster, Germany, ²Institute of Microscopic Anatomy and Neurobiology, Johannes Gutenberg-University Mainz, Mainz, Germany

We investigated the BOLD correlate of locally detected Ca²⁺-recordings of spontaneous cortical and thalamic slow oscillations in rats and found BOLD activation of the cortex correlated to both cortical and thalamic up-states.

Laurentius Huber¹, Jozien Goense², Aneurin Kennerley³, Maria Guidi¹, Robert Trampel¹, Robert Turner¹, and Harald E Möller^1
1Max Planck Institute for Human Cognitive & Brain Sciences, Leipzig, Germany, ²University of Glasgow, United Kingdom, ³University of Sheffield, United Kingdom

This study investigates cortical layer-dependent functional changes of cerebral blood volume (CBV) in humans, monkeys and rats at 7 T. Using multiple imaging modalities including vascular space occupancy (VASO), iron-oxide contrast agent based fMRI, and optical imaging spectroscopy, we investigated the contributions of pial macrovasculature and layer-specific microvasculature to CBV based fMRI signals. Based on the results presented here, we suggest that the pial macrovasculature can introduce amplifying or attenuating signal contributions to layer-dependent CBV responses. These contaminations manifest themselves differently, depending on the units (ml or %) used in the evaluation procedure.

Qingfei Luo¹, Chung-Ki Wong¹, Han Yuan¹, Vadim Zotev¹, Wen-Ming Luh², and Jerzy Bodurka^1,3
1Laureate Institute for Brain Research, Tulsa, OK, United States, ²Cornell MRI Facility, Cornell University, Ithaca, NY, United States, ³College of Engineering, Center for Biomedical Engineering, University of Oklahoma, Norman, OK, United States

Pseudo-continuous arterial spin labeling (PCASL) is a quantitative CBF-fMRI pulse sequence. We investigate feasibility of simultaneous electroencephalography (EEG) and PCASL multi-modal measurement. Specific absorption rate (SAR) of PCASL, may increase the temperatures of EEG electrodes and thus have potential safety risks. To evaluate the feasibility and safety we performed real-time measurements of EEG electrode temperatures during concurrent EEG&PCASL (acquired with standard imaging hardware) at 3 Tesla. The results showed that the electrode temperature after a long PCASL scan will reach about 34°C - well below the safety limit (41°C). Results show feasibility of multi-modal EEG&PCASL experiments.

Chi Wah Wong¹ and Thomas Liu^1
1Center for Functional MRI, University of California San Diego, La Jolla, CA, United States

The amplitude of the resting-state fMRI global signal was compared between eyes open (EO) and eyes closed (EC) states. The amplitude of the global signal was significantly lower in the EO state and the magnitude of the decrease was correlated with increases in EEG vigilance. The relation between the EO-EC changes in global signal amplitude and EEG vigilance was found to be similar to that previously observed for caffeine-related changes in global signal amplitude and vigilance.

Maosen Wang¹, Yi He¹, Yaohui Tang¹, Dávid Zsolt Balla², Chunqi Qian³, and Xin Yu^1
1Research Group of Translational Neuroimaging and Neural Conteol,High Field Magnetic Resonance, Max Planck Institute for Biological Cybernetics, Tuebingen, Baden-Wuerttemberg, Germany, ²Department of Physiology of Cognitive Processes, Max Planck Institute for Biological Cybernetics, Tuebingen, Baden-Wuerttemberg, Germany,³Laboratory of Functional and Molecular Imaging, National Institute of Neurological Disorders and Str, National Institutes of Health, Bethesda, MD, United States

It remains ambiguous how the direct fiber optic insertion affects the local fMRI signal by optical stimulation. The fiber optic was inserted to target the deep layer cortex expressing Channelrhodopsin 2(ChR2). Robust fMRI signal was detected in the cortical regions close to the fiber tip with varied light pulse parameters on frequency, pulse duration and power level. The light evoked local field potential was also recorded by electrodes inserted into the cortex expressing ChR2. This work provides us a robust light-driven fMRI platform in combination with in vivo recording, which will facilitate the study to decipher cellular contribution to fMRI signal from the local neurovascular network.

Elhum A Shamshiri¹, Maria Centeno¹, Tim Tierney¹, Kelly St Pier², Ronit Pressler², Suejen Perani^1,3, Helen J Cross⁴, and David W Carmichael^1
1Developmental Imaging and Biophysics Section, UCL Institute of Child Health, London, United Kingdom, ²Epilepsy Unit, Great Ormond Street Hospital, London, United Kingdom, ³Department of Basic and Clinical Neurosciences, Institute of Psychiatry, Psychology, and neuroscience, London, United Kingdom, ⁴Neurosciences Unit, University College London, London, United Kingdom

Previously, cognitive network aberrations have been shown in epilepsy patients with regards to resting state networks such as the DMN (default mode network). However, the influence of interictal activity on these networks has yet to be properly explored. This study uses EEG-fMRI to explore the effect of spikes as a regressor of no interest in cognitive networks using a general linear model (GLM) and functional connectivity.

Elise Bannier^1,2, Marsel Mano^2,3, Robert Stroemer⁴, Isabelle Corouge², Lorraine Perronnet^2,3, Jussi T. Lindgren³, Anatole Lecuyer³, and Christian Barillot^2
1Radiology, University Hospital of Rennes, Rennes, France, ²Unité VISAGES U746 INSERM-INRIA, IRISA UMR CNRS 6074, University of Rennes, Rennes, France, ³Unité HYBRID INRIA, IRISA UMR CNRS 6074, Rennes, France, ⁴Brainproducts GmbH, Gilching, Germany

The aim of the study was to assess the feasibility and specificity of simultaneous ASL MRI and EEG recording at 3T in terms of safety, EEG and MR signal quality.

M Ethan MacDonald^1,2, Avery Berman^2,3, Rebecca J Williams^1,2, Erin L Mazerolle^1,2, and G Bruce Pike^1,2
1Radiology and Clinical Neurosciences, University of Calgary, Calgary, Alberta, Canada, ²Hotchkiss Brain Institute, Foothills Medical Centre, Alberta Health Services, Calgary, Alberta, Canada, ³Biomedical Engineering, McGill University, Montreal, Quebec, Canada

In this work, we use a quantitative susceptibility technique calculated from the phase data from BOLD-fMRI. Measurements of the susceptibility time course (BOLD-QSM) are compared to BOLD-fMRI in visual and motor regions. Subject’s heads are moved and imaging is repeated. Contrast to noise ratio is calculated in the signals from both techniques and found to be comparable. Five of eight measured signals showed higher CNR with the BOLD-QSM method. A good level of correlation is obtained between the two methods.

Toshihiko Aso¹ and Hidenao Fukuyama^1
1Human Brain Research Center, Kyoto University Graduate School of Medicine, Kyoto, Kyoto, Japan

Mapping of temporal delay in rs-fMRI signal is potentially useful as a contrast agent-free hemodynamic imaging. Our test-retest reliability study revealed that time-lag, magnitude and maximum-Z maps all have higher consistency than a functional connectivity map. Particularly the striking day-to-day stability encourages use of this technique for longitudinal observation in patients.

Raphael T Gerraty¹, David B Parker², Alayar Kangarlu³, and Qolamreza R Razlighi^2,4
1Psychology, Columbia University, New York, NY, United States, ²Biomedical Engineering, Columbia University, NY, United States, ³Psychiatry, Columbia University, NY, United States, ⁴Neurology, Columbia University, New York, NY, United States

Hemodynamic Impulse Response Function varies across life span; however we showed that brain atrophy and inaccurate co-registration accounts for most of these age related variabilities.

Saeedeh Amirmohseni¹, Daniel Segelcke², Esther Pogatzki-Zahn², and Cornelius Faber^1
1Department of Clinical Radiology, University Hospital Muenster, Muenster, Germany, ²Department of Anaesthesiology, Intensive Care and Pain Medicine, University Hospital Muenster, Muenster, Germany

In human fMRI studies, BOLD response to constant painful electrical and thermal stimulation has shown deviations from the canonical hemodynamic response function, leading to possible false negatives in activation maps when using a boxcar analysis model. In this work, we have considered possible signal variations during the BOLD response in the analysis, to process the data of noxious electrical stimulation of the hindpaw in rat models of incisional and inflammatory pain, and demonstrated the importance of accounting for these variations in order to fit the actual shape of the BOLD response.

Tzu-chen Yeh^1,2, Chou-ming Cheng³, Jin-jie Hong², Sheng-che Hung¹, Muh-Lii Liang⁴, and Jen-chuen Hsieh^2,3
1Department of Radiology, Taipei Veterans General Hospital, Taipei, Taiwan, Taiwan, ²Institute of Brain Science, National Yang-Ming University, Taipei, Taiwan, Taiwan,³Department of Medical Research, Taipei Veterans General Hospital, Taipei, Taiwan, Taiwan, ⁴Neurosurgery Neurological Institute, Taipei Veterans General Hospital, Taipei, Taiwan, Taiwan

Impaired cerebrovascular reserve (CVR) was supposed to cause dissociation between functional integrity and functional connectivity due to regional variation of CVR. And reversed linearity between functional connectivity and carbogen-based CVR suggested enhanced or neuroplasticity in patients of moyamoya disease. Underestimated or aberrant FC should be considered with impaired CVR, and studies of FC in diseased states need evidence of preserved CVR.

Hannah Hare¹ and Daniel Bulte^1
1FMRIB, University of Oxford, Oxford, Oxon, United Kingdom

A calibrated BOLD experiment was performed with a multi-echo sequence to determine how BOLD response to hypercapnia and the calibration parameter M vary with echo time. Theory has predicted these relationships to be directly proportional, but is based on the assumption of negligible intravascular signal contribution. We performed experiments both with and without the use of intravascular crushers, and show that although M is a linear function of echo time, the intercept is not zero. This result has implications for the commonly employed scaling method used to better compare M values from experiments with differing echo times.

Isabelle Lajoie¹, Kenneth S Dyson², Scott Nugent², Felipe D Tancredi^1,3, and Richard D Hoge^2
1Centre de recherche de l'IUGM, Université de Montréal, Montreal, Quebec, Canada, ²McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada, ³Albert Einstein Jewish Hospital, Brazil

It is believed that the delivery and consumption of oxygen have an important role in the pathogenesis of conditions such as Alzheimer’s Disease (AD), rising the pertinence of developing technologies that provide such information. We demonstrate here the potential of the MRI method introduced in our team, dubbed QUO2, for characterizing age-related differences in cerebral physiology, more precisely in resting CBF, CVR, M, OEF and CMRO2. In the next step, we will use our protocol to evaluate the level of discrepancy between individuals diagnosed with AD and age-matched controls.

Alex A Bhogal¹, Jeroen C.W. Siero², Marielle E. Philippens², Esben T. Petersen², Martijn Froeling², Jeroen Hendrikse², Manus J. Donahue³, and Hans Hoogduin^2
1University Medical Center, Utrecht, Utrecht, Netherlands, ²University Medical Center, Utrecht, Netherlands, ³Vanderbilt University School of Medicine, Tennessee, United States

Hypercapnia and hyperoxia, induced using targeted gas challenges, may be used to probe physiological effects related to the Blood Oxygen Level Dependent (BOLD) signal. Multi-echo BOLD data is acquired allowing changes in T2* and S0 parameters in response to breathing challenges to be examined. Results are explained in the context of physiological phenomenon.

Powell Pui Wai Chu^1,2, Ali M. Golestani¹, Jonathan B. Kwinta^1,2, Yasha B. Khatamian¹, and Jean J. Chen^1,2
1Rotman Research Institute, Baycrest, Toronto, Ontario, Canada, ²Medical Biophysics, University of Toronto, Toronto, Ontario, Canada

Resting-state fMRI functional connectivity (rs-fcMRI) measurements are based on the BOLD fMRI signal, which is influenced by systemic physiological variability. Thus, fcMRI data interpretation may be difficult across different subjects without the knowledge of these physiological influences. In this work, in order to improve the reliability of rs-fcMRI measurements, we assess the influence of physiological factors including baseline cerebral blood flow and venous blood oxygenation on functional connectivity estimates. We found a significance relationship between rs-fcMRI values and the above two physiological variables.

Bryon A. Mueller¹, Nicholas Evanoff², Kara L. Marlatt², Justin R. Geijer³, Kelvin O. Lim¹, and Donald R. Dengel^2
1Department of Psychiatry, University of Minnesota, Minneapolis, MN, United States, ²School of Kinesiology, University of Minnesota, Minneapolis, MN, United States,³Department of Health, Exercise and Rehabilitative Sciences, Winona State University, Winona, MN, United States

BOLD imaging during a hypercapnic challenge is often used to measure cerebral vascular reactivity. Determining the within and between day reproducibility of the BOLD signal change in response to a reproducible hypercapnic challenge will help determine if this method is precise enough to be used as a research tool to compare specific groups to control populations. We have measured the intraclass correlation coefficient and coefficient of within-subject variation for BOLD imaging during an iso-oxic, hypercapnic square stimulus. Results show excellent within day reproducibility and good between day reproducibility within the cerebral gray matter.

Avery J.L. Berman^1,2, Yuhan Ma¹, Richard D. Hoge^1,3, and G. Bruce Pike^1,2
1Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada, ²Department of Radiology/Hotchkiss Brain Institute, University of Calgary, Calgary, Alberta, Canada, ³Unité de neuroimagerie fonctionelle, Centre de recherche de l'institut de gériatrie de Montréal, Montreal, Quebec, Canada

The susceptibility of dissolved O₂ in blood was studied, focusing on the implications for BOLD imaging using hyperoxia. We proposed a model for the susceptibility of blood that incorporates dissolved O₂ and compared it with another previously proposed model. Distilled water and samples of bovine plasma were oxygenated and their susceptibilities were determined using gradient echo phase imaging. Measured changes in susceptibility vs. O₂ concentration were linear, with slopes that were dramatically less than previously predicted and in good agreement with our model. Under most physiological conditions, the susceptibility of dissolved O₂ can be disregarded in MRI studies employing hyperoxia.

Michael Germuska¹, Alberto Merola¹, Alan Stone², Kevin Murphy¹, and Richard Wise^1
1Cardiff University, Cardiff, Wales, United Kingdom, ²Oxford University, Oxfordshire, United Kingdom

Recently, calibrated MRI methods of estimating resting oxygen extraction fraction (OEF) have been developed. These methods rely on the independent quantification of BOLD and blood flow responses to respiratory challenges. The resulting estimates are then fed into a physiological model to solve for OEF. Data analysed in such a step-wise manner are susceptible to the propagation of errors along the pipeline, producing unstable estimates of OEF. Here we re-pose the analysis in a Bayesian framework to solve for the underlying physiological parameters in a one-step solution. In-vivo data demonstrates stable estimates of OEF within the expected range for healthy tissue.

Ian D Driver¹, Joseph Whittaker¹, Molly G Bright¹, Suresh D Muthukumaraswamy^1,2, and Kevin Murphy^1
1CUBRIC, School of Psychology, Cardiff University, Cardiff, United Kingdom, ²Schools of Pharmacy and Psychology, Auckland University, Auckland, New Zealand

Understanding the effect of hypercapnia on neuronal oscillations and cerebral metabolism is an important consideration for physiological and calibrated fMRI studies. Here we use MEG measurements to study broadband neuronal desynchronisation during graded levels of hypercapnia. Neuronal desynchronisation persisted even in mild levels of hypercapnia, below the sensory threshold (subjects performed a breathlessness rating task). There appears to be a linear dependence of the neuronal desynchronisation on the level of hypercapnia, presenting a possible solution for hypercapnia calibrated fMRI to overcome the violation of the isometabolic assumption with hypercapnia.

Alberto Merola¹, Kevin Murphy¹, Alan J Stone¹, Michael A Germuska¹, Valerie E M Griffeth², Nicholas P Blockley³, Richard B Buxton^3,4, and Richard G Wise^1
1CUBRIC, School of Psychology, Cardiff University, Cardiff, United Kingdom, ²Department of Bioengineering and Medical Scientist Training Program, University of California San Diego, La Jolla, California, United States, ³Center for Functional Magnetic Resonance Imaging, Department of Radiology, University of California San Diego, La Jolla, California, United States, ⁴Kavli Institute for Brain and Mind, University of California San Diego, La Jolla, California, United States

In this simulation study we analyze an existing mathematical model for BOLD calibration and assessment of oxygen extraction fraction (OEF). We have generated datasets of synthetic BOLD signals arising from a wide range of simulated physiological conditions and a variety of hypercapnic and hyperoxic respiratory tasks. OEF estimates demonstrate the inaccuracy of the current model and let us proposing a new approach for optimizing it. We were then able to optimize the current model and propose a new simplified model, achieving greatly improved performances. This represents a significant step forward towards an accurate and reliable quantification of oxygen metabolism in brain.

Joseph Whittaker¹, Ian Driver¹, Molly Bright¹, and Kevin Murphy^1
1CUBRIC, School of Psychology, Cardiff University, Cardiff, United Kingdom

The BOLD fMRI response depends on multiple physiological factors, in particular, the ratio of fractional changes in CBF and CMRO2, defined as n. Meaningful comparisons of BOLD responses between stimulus conditions rely upon a constant n, which may not be the case. The BOLD and CBF responses to 4 visual contrast levels were measured in 9 subjects during normocapnia and hypercapnia. We found larger BOLD responses during hypercapnia, suggesting altered CBF/CMRO2 coupling dynamics. These data show BOLD changes that are consistent with varying n values across contrast, and which are modulated further by changes in baseline physiology.

Michael Germuska¹, Alberto Merola¹, Kevin Murphy¹, and Richard Wise^1
1Cardiff University, Cardiff, Wales, United Kingdom

The estimation of the resting oxygen extraction fraction (OEF) has been demonstrated by calibrated measurement of the BOLD signal with serial hypercapnic and hyperoxic stimuli. However, the optimal respiratory paradigm (number, timing and duration of stimuli) has yet to be investigated. Here we utilise a detailed signal model and a realistic physiological noise model to optimise such a paradigm for the estimation of OEF. Our results suggest that the derived paradigm offers a significant improvement in the accuracy of OEF estimates compared to previously implemented designs of the same (18 minute) duration.

Wen-Ju Pan¹, Jacob Billings¹, and Shella Keilholz^1
1Biomedical Engineering, Emory University/ Georgia Institute of Technology, Atlanta, Georgia, United States

Intrinsic optical spectroscopy (IOS) may provide multiple hemodynamic parameters in high time resolution in functional optical imaging but rarely used within brain study. We developed a combination of optic fiber-based IOS, ~100um in diameter, and microelectrode neuronal LFP signal recording within brain at similar spatial range, a few hundred micros. This simultaneous recording design may be a promising approach for examining target sites of neurovascular coupling relations in brain.

Yongxia Zhou¹, Zachary B Rodgers¹, and Felix W Wehrli^1
1Radiology, University of Pennsylvania, Philadelphia, Pennsylvania, United States

An integrated MRI protocol was developed for simultaneous mapping of CBF with acceleration-based ASL and OEF with GESFIDE. The GESFIDE module was integrated into the post-label delay control module of the ASL sequence and total scan time was 405sec. Preliminary results in eight subjects, after averaging of data across the brain and subjects yielded values of 47.7±4.8mL/100g/min, 32.7±2.6%, and 123.8±11.1µmol/100g/min for CBF, OEF and CMRO2, respectively, in good agreement with the literature. Positive correlations between CBF and CMRO2 and negative trend between CBF and OEF were found. Global CMRO2 value correlated significantly with other established methods for the same subjects.

Kay Jann¹, Edgar A Rios Piedra¹, Robert X Smith¹, and Danny JJ Wang^1
1Department of Neurology, University of California Los Angeles, Los Angeles, California, United States

Noise-regression in BOLD-fMRI is necessary to minimize influences of motion, heart rate and respiration on functional connectivity (FC). Recently, modern pseudo-continuous ASL (pCASL) sequences have been shown to provide comparable FC networks as BOLD. However, the influence of head motion and physiological noise on ASL-FC has not yet been investigated. Here we report effects of different noise-regression strategies on FC using 2D-EPI and 3D-BS-GRASE pCASL. We found that ASL-FC analyses are feasible especially for 3D-BS-GRASE pCASL, but proper noise regression is still mandatory to prevent erroneous findings related to motion or physiological noise.

Tianyi Qian¹, Yinyan Wang^2,3, and Tao Jiang^2,3
1MR Collaborations NE Asia, Siemens Healthcare, Beijing, Beijing, China, ²Beijing Neurosurgical Institute, Beijing, China, ³Beijing Tiantan Hospital, Capital Medical University, Beijing, China

In previous resting-state functional magnetic resonance imaging (rs-fMRI) studies, the global signal (the averaged time series of BOLD signal across the whole brain) has always been treated as the background noise and is typically regressed out. Here, an iterative algorithm was used to extract the time-shift-corrected global signal. By calculating the time delay and weight with respect to the corrected global signal in each voxel, we were able to measure the transient time of the whole brain. Simultaneous multislice acceleration acquisition combined with rs-fMRI was used to acquire the dataset used in this study to further improve the temporal and spatial resolution of this method.