ISMRM 23rd Annual Meeting & Exhibition • 30 May - 05 June 2015 • Toronto, Ontario, Canada

Scientific Session • fMRI: Resting-state Functional Connectivity

Monday 1 June 2015

Room 714 A/B

10:45 - 12:45


Mark J. Lowe, Ph.D., T.B.A.

10:45 0045.   
Comparison of BOLD and CBV-weighted resting state connectivity to an anatomical ‘gold standard’ in the motor network of the squirrel monkey brain
Yurui Gao1,2, Feng Wang2,3, Iwona Stepniewska4, Ann S Choe1,2, Kurt G Schilling1,2, Landman A Bennett2,5, Adam W Anderson1,2, Zhaohua Ding2,3, Limin Chen2,3, and John C Gore2,3
1Department of Biomedical Engeneering, Vanderbilt University, Nashville, Tennessee, United States, 2Institute of Imaging Science, Vanderbilt University, Nashville, Tennessee, United States, 3Department of Radiology and Radiological Science, Vanderbilt University, Nashville, Tennessee, United States, 4Department of Psychology, Vanderbilt University, Tennessee, United States, 5Department of Electrical Engeneering, Vanderbilt University, Nashville, Tennessee, United States

This study aims to compare functional connectivity acquired using resting state MRI against the anatomical connectivity revealed by neurotracers injected into primary motor cortex (M1). BOLD timecourse correlation, CBV-weighted timecourse correlation and histological fiber strength were calculated, spacial normalized and compared. The comparison indicates the relationship between function and anatomy in motor network of squirrel monkey during resting state.

10:57 0046.   
Remodeled resting state functional connectivity pattern in the default mode network and cortico – striatal circuitry of GPR88 knock-out mouse brain
Tanzil Mahmud Arefin1,2, Anna Mechling2,3, Thomas Bienert2, Hsu-Lei Lee2, Sami Ben Hamida4, Dominik V. Elverfeldt2, Jürgen Hennig2, Brigitte Kieffer5,6, and Laura-Adela Harsan2
1Computational Neuroscience, Bernstein Center Freiburg, University of Freiburg, Freiburg, Baden - Württemberg, Germany, 2Diagnostic Radiology, Medical Physics, University Hospital Freiburg, Freiburg, Baden - Württemberg, Germany, 3Faculty of Biology, University of Freiburg, Freiburg, Baden - Württemberg, Germany, 44Institut de Génétique et de Biologie Moléculaire et Cellulaire, Strasbourg, France, 5Douglas Research Center, McGill University, Montreal, Canada, 6Institut de Génétique et de Biologie Moléculaire et Cellulaire, Strasbourg, France

Functional communication between brain regions plays a key role in complex cognitive process. Emerging studies show that rsfMRI can reveal the modifications in functional brain connectivity due to psychiatric disorders or drug effects. This study was aimed to scrutinize the functional connectivity modifications in GPR88 Knock Out (KO) mice using rsfMRI technique, which has not been reported yet and might be interesting in the perspective of neurological or psychiatric disorders and drug research.

11:09 0047.   Voxel-scale mapping of the mouse brain functional connectome
Adam Liska1,2, Alberto Galbusera1, Adam J. Schwarz3, and Alessandro Gozzi1
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, 3Department of Psychological and Brain Sciences, Indiana University, Bloomington, IN, United States

Human brain imaging studies have revealed the presence of functionally specialized sub-systems interlinked by highly-connected “hub nodes”. However, the extent to which a similar functional architecture exists in the mouse brain remains to be fully elucidated. We have computed voxel-scale whole-brain functional connectivity maps to generate a comprehensive mouse brain “functional connectome”. Consistent with human studies, we show that mouse brain contains mutually-interconnected connector hubs in several sub-regions of a “default mode network”, and in well-known integrative cortical structures. Our findings suggest the presence of evolutionarily-conserved, mutually-interconnected functional modules and cortical hubs as a fundamental feature of the mammal brain.

11:21 0048.   
Mapping resting-state dynamics on spatio-temporal graphs: a combined functional and diffusion MRI approach
Alessandra Griffa1,2, Kirell Benzi3, Benjamin Ricaud3, Xavier Bresson3, Pierre Vandergheynst3, Patric Hagmann1,2, and Jean-Philippe Thiran1,2
1Signal Processing Laboratory 5 (LTS5), École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 2Department of Radiology, Lausanne University Hospital (CHUV) and University of Lausanne, Lausanne, Switzerland, 3Signal Processing Laboratory 2 (LTS2), École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland

Magnetic resonance imaging allows inferring overall brain structural and functional networks. A growing body of recent literature suggests that a static description of functional connectivity (e.g. with simple correlation measures) might by over simplistic. In the present work we propose a mathematically sound and flexible method for the mapping of dynamic spatio-temporal resting state patterns. Our framework is based on the representation of data on a spatio-temporal graph and exploits structural (diffusion-based) and functional information in a complementary manner. Nodes within isolated functional sub-networks are simultaneously close in space (the space of the anatomical connectivity substrate) and time (temporally co-active).

11:33 0049.   
Does vasomotion alter functional connectivity? A multi-modal study using Optical Imaging Spectroscopy and BOLD fMRI
Priya Patel1, Aneurin James Kennerley1, Luke Boorman1, Myles Jones1, and Jason Berwick1
1Psychology, University of Sheffield, Sheffield, South Yorks, United Kingdom

Slow cerebral oscillations, 0.1 Hz, termed as vasomotion, could confound neurovascular coupling within resting state fMRI BOLD signal, therefore inferring connectivity changes from BOLD fMRI signal in disease states problematic. We aim to utilize a systematic analysis of the BOLD fMRI signal and 2 - dimensional optical imaging spectroscopy (2D-OIS) data to examine the magnitude and spatial correlations of fluctuations in BOLD fMRI signals and hemodynamics following manipulation of the systemic blood pressure in anesthetized rodents; this will in part emulate physiological conditions such as in disease states. Changing the systemic blood pressure modulated the 0.1Hz vasomotion signal and we have seen a difference in the inferred connectivity maps before and after this change.

11:45 0050.   Can resting state fMRI be used to map cerebrovascular reactivity?
Peiying Liu1, Babu G Welch2, Darlene King2, Yang Li1, Marco Pinho1,3, and Hanzhang Lu1
1Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, Texas, United States, 2Neurological Surgery Clinic, University of Texas Southwestern Medical Center, Dallas, Texas, United States, 3Department of Radiology, University of Texas Southwestern Medical Center, Texas, United States

Hypercapnia inhalation is commonly used to map cerebrovascular reactivity(CVR) in clinical research studies. However, gas inhalation may not be feasible for acute patients (e.g., acute stoke, traumatic brain injury). In this work, we aim to explore an alternative approach without gas inhalation. We hypothesize that global BOLD signal can serve as a reliable regressor for CVR estimation from the resting state BOLD data. We showed that the resting CVR generated by this method is reproducible, and of similar physiological origin as CO2-inhalation-derived CVR. This method might be a reliable surrogate of CVR mapping when hypercapnia inhalation is not feasible.

11:57 0051.   Subject-specific modeling of physiological noise in resting-state fMRI at 7T
Sandro Nunes1, Marta Bianciardi2, Afonso Dias1, Rodolfo Abreu1, Juliana Rodrigues1, L. Miguel Silveira3, Lawrence L. Wald2, and Patricia Figueiredo1
1Institute for Systems and Robotics and Department of Bioengineering, Instituto Superior Técnico, Universidade de Lisboa, Lisbon, Lisbon, Portugal,2Department of Radiology, A.A. Martinos Center for Biomedical Imaging, MGH and Harvard Medical School, Boston, MA, United States, 3INESC-ID and Department of Electrical and Computer Engineering, Instituto Superior Técnico, Universidade de Lisboa, Lisbon, Lisbon, Portugal

A number of strategies have been proposed for modeling physiological noise in rs-fMRI, including different models of the respiratory volume per time (RVT) and heart rate (HR) contributions. We performed a systematic comparison of RVT and HR models within an extended RETROICOR model of physiological noise in rs-fMRI at 7T. We found that a dual-lag model with subject-specific lag optimization explained significantly more variance than single-lag or convolutions models, or group optimization. We conclude that taking into account the high inter-subject variability of RVT/HR responses significantly improves physiological noise modeling, and it should hence reduce inter-subject variability of rs-fMRI studies.

12:09 0052.   
Inter-Scanner Reliability of Graph-Theoretic Brain Network Metrics
Thomas Welton1, Dorothee P Auer1, and Robert A Dineen1
1Sir Peter Mansfield Imaging Centre, School of Medicine, University of Nottingham, Nottingham, Nottinghamshire, United Kingdom

We tested the between-scanner reliability of graph-theoretic brain network properties derived from resting-state fMRI data. This is a crucial prerequisite to the wider application of graph metrics in clinical neuroscience and, in particular, to future adoption in multicentre treatment trials. Using data from five healthy subjects scanned in 10 different scanners, we show that overall reliability was poor. Scanner field strength was not associated with reliability, but averaging over repeat scans did improve reliability. Graph metrics derived from human brain fMRI data are not yet reliable enough between scanners for use as surrogate outcomes in multicentre clinical trials.

12:21 0053.   
Anisotropy of local functional connectivity (LFC) in resting state fMRI time series: what does it say about the fmri signal?
Michael J. Tobia1, David Gallagher1, Rahul Dewal1, Prasanna Karunanayaka1, Sebastien Rupprecht1, and Qing X. Yang1
1Center for NMR Research, Penn State University, Hershey, PA, United States

This experiment investigated anisotropic local functional connectivity (LFC) in GE-EPI time series in phantom and human in vivo resting state fMRI data. LFC was computed in a neighborhood radius of 2 voxels using Pearson’s correlation. In vivo, LFC anisotropy varied across gray and white matter sites, resembling DTI in some aspects, while differing on others. Phantom experiments showed that fluctuating electric current is sufficient to generate anisotropic LFC proximal to the current-carrying filament. In conclusion, anisotropic correlations of fMRI time series may arise from an alternative non-BOLD contrast mechanism, potentially related to an electric current effect on Bo.

12:33 0054.   
fMRI-derived functional connectivity density mapping as a biomarker of state changes as reflected by glucose metabolism
Garth John Thompson1, Valentin Riedl2,3, Timo Grimmer3,4, Alexander Drzezga5, Peter Herman1, and Fahmeed Hyder1,6
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 a popular way to measure networks in the human brain, a lack of understanding in terms of glucose metabolism (CMRglc) has made translation to clinical settings difficult. Simultaneous fluorodeoxyglucose PET and R-fMRI data were collected from 22 subjects with eyes open or eyes closed. Various R-fMRI quantifications were tested to match the globally higher CMRglc observed with eyes open. Functional connectivity density (FCD) without any global signal regression reflected state change similar to that observed with CMRglc data. Thus FCD may be a viable biomarker for R-fMRI in clinical settings.