Remi Patriat¹, Erin Molloy², and Rasmus M Birn^3,4
1Medical Physics, University of Wisconsin Madison, Madison, Wisconsin, United States, ²University of Illinois Urbana-Champaign, Illinois, United States,³Psychiatry, University of Wisconsin Madison, Madison, Wisconsin, United States, ⁴Medical Phtysics, University of Wisconsin Madison, Madison, Wisconsin, United States

We developed a new motion correction method for resting state fMRI analysis that makes use of information contained at the edge of the brain to create a set of regressors that explain more variance and improve image quality compared to the current standard methods.

Jian Weng¹, Shanshan Dong¹, Feiyan Chen¹, and Hongjian He^2
1Physics Department, Zhejiang University, Hangzhou, Zhejiang, China, ²Biomedical Engineering Department, Zhejiang University, Hangzhou, Zhejiang, China

Spatial normalization is essential for most functional MRI studies. such a transformation could introduce unexpected registration error in practice and increase individual variations. We compared a usage of three common templates for normalization, and proposed a correction method.

Francisca Marie Tan^1,2, Karen Mullinger¹, César Caballero Gaudes³, Yaping Zhang², David Siu-Yeung Cho², Yihui Liu⁴, Susan Francis¹, and Penny Gowland^1
1Sir Peter Mansfield Magnetic Resonance Centre, University of Nottingham, Nottingham, Nottinghamshire, United Kingdom, ²Department of Electrical and Electronic Engineering, University of Nottingham Ningbo China, Ningbo, Zhejiang, China, ³Basque Center on Cognition, Brain and Language, Donostia, Spain, ⁴School of Information Science, Qilu University of Technology, Jinan, Shandong, China

Decoding mental activity at rest is a challenge because spontaneous events occur in the brain without any attributed task or prior stimulus timing. In this study, we validate the use of Sparse Paradigm Free Mapping prior to Temporal Independent Component Analysis (tICA) on a movement task to detect discrete motor events. The tICA components are assessed against EMG and classified using a meta-analysis, with 78 % of task-driven events identified by tICA. Results suggest that this method can be used in future studies of resting data to detect events and map these to functional areas using a meta-analysis.

Feng Xu^1,2, Joseph S. Gillen^1,2, Hongjun Liu³, Ann Choe^1,2, Hua Jun^1,2, Craig K. Jones^1,2, Suresh E. Joel¹, Brian S. Caffo⁴, Martin A. Lindquist⁴, Ciprian M. Crainiceanu⁴, Peter C. van Zijl^1,2, and James J. Pekar^1,2
1Russell H. Morgan Department of Radiology, Johns Hopkins University, Baltimore, MD, United States, ²F.M. Kirby Research Center, Kennedy Krieger Institute, Baltimore, MD, United States, ³Department of Radiology, Guangdong General Hospital, Guangdong Academy of Medical Sciences, Guangzhou, Guangdong, China, ⁴Biostatistics, School of Public Health, Johns Hopkins University, Baltimore, MD, United States

Spatial smoothing is the most popular way to enhance sensitivity in fMRI analysis, at a cost of coarsened spatial specificity. Multi-echo acquisitions can enhance specificity in fMRI by allowing analysis of effective transverse relaxation rate (R2*) via least-squares (LS) fitting to each voxel’s echo decay. Bayesian shrinkage improves parallel simultaneous estimation of many similar parameters by “borrowing strength” from parallel measurements. Here, we “shrink over grey matter” by applying Bayesian shrinkage to estimation of R2* in grey matter voxels, and show that shrinkage increases fMRI sensitivity (with respect to LS fitting) without the “blurring” caused by spatial smoothing.

Huanjie Li¹, Lisa D. Nickerson², Jinhu Xiong³, and Jia-Hong Gao^1
1Peking University, Beijing, Beijing, China, ²Harvard Medical School, Massachusetts, United States, ³University of Iowa, Iowa, United States

Most existing cluster-size tests used in fMRI data analysis to detect brain activation were formulated and validated under sufficiently smooth image conditions. Unfortunately, spatial smoothing degrades spatial specificity and increases false positives. Recently, a threshold-free cluster enhancement (TFCE) technique was proposed which does not require spatial smoothing, but this method can only be used for group level analysis. We propose a more reliable and effective 3D cluster-based method which can keep a higher sensitivity for localizing activation regions for both single-subject and group level analysis without the requirement of spatial smoothness.

Dietmar Cordes^1,2, Rajesh R. Nandy³, Scott Schafer², and Tor D. Wager^2
1Ryerson University, Toronto, Ontario, Canada, ²University of Colorado, Boulder, Colorado, United States, ³University of North Texas Health Science Center, Texas, United States

The physiological noise in the lower brain areas (brain stem and nearby regions) are investigated in resting-state data and a novel method is presented for computing both low-frequency and high-frequency physiological regressors. In particular, using a novel optimization algorithm that penalizes curvature, the cardiac -and respiratory-related low-frequency response functions are computed. Also, the frequency-aliasing property of the high-frequency cardiac waveform as a function of the repetition time (TR) is investigated. It is shown that for brain areas associated with large pulsations of the cardiac rate, the temporal SNR associated with the BOLD response has maxima at subject-specific TRs.

Hui Shen^1,2, Longchuan Li^1,3, Kaiming Li¹, Bing Ji¹, and Xiaoping Hu^1
1Department of Biomedical Engineering, Biomedical Imaging Technology Center, Emory University, Atlanta, GA, United States, ²College of Mechatronics and Automation, National University of Defense Technology, Changsha, Hunan, China, ³Marcus Autism Center, Children¡¯s Healthcare of Atlanta, Emory University School of Medicine, Atlanta, GA, United States

The aim of this work is to characterize oscillation variability in dynamic local and distant task-free functional connectivity with a sliding windows approach. Compared with local connectivity, distant connectivity exhibited significantly more intensive fluctuation, suggesting task-free functional connectivity dynamics may be mainly accounted for by long-distance functional interaction across distributed regions. Furthermore, the most stable and most instable areas were localized at the sensorimotor cortices and the default mode network (DMN) extending to the adjacent frontoparietal network, respectively. These findings shed new light on cortical organization in dynamic functional connectivity, and also highlight the importance of long-range dynamic functional interaction.

Li-Ming Hsu¹, Jennifer A. Stark¹, Julia K. Brynildsen¹, Hong Gu¹, Hanbing Lu¹, Elliot A. Stein¹, and Yihong Yang^1
1Neuroimaging Research Branch, National Institute on Drug Abuse, Baltimore, Maryland, United States

Resting-state fMRI of animal models has the advantage to assess the trajectory of a disease using longitudinal paradigms, but the reproducibility of the observations is critical and less known. Here, we investigate the reliability of resting fMRI signal of rats in short (10 min), middle (2 weeks) and long (3 months) terms. Our data showed that the mean ICC across brain networks during short- or mid-term scans (0.63 and 0.57 respectively) was significantly higher than that of long-term scans (0.22), suggesting that longitudinal experiments within weeks would have good reproducibility, but studies across months should be practiced with caution under the current conditions.

Muhammad Asad Lodhi¹, Matthew E Magnuson², Shella D Keilholz², and Waqas Majeed^1
1Department of Electrical Engineering, Lahore University of Management Sciences, Lahore, Punjab, Pakistan, ²Biomedical Engineering, Georgia Institute of Technology/Emory University, GA, United States

Previous studies have demonstrated the presence of repeated propagating spatiotemporal patterns in resting state fMRI. This paper describes an ICA-based method to automatically detect propagating spatiotemporal patterns in resting state BOLD fMRI. The proposed method does not require specification of a region of interest (ROI) and is more time-efficient compared with the available alternative.

Mauro DiNuzzo^1,2, Daniele Mascali^1,2, Marta Moraschi^1,3, Michela Fratini^1,3, Tommaso Gili³, Girolamo Garreffa^1,3, Bruno Maraviglia^1,3, and Federico Giove^1,2
1MARBILab, Enrico Fermi Center, Rome, Rome, Italy, ²Department of Physics, U Sapienza, Rome, Rome, Italy, ³Santa Lucia Foundation IRCCS, Rome, Italy

We applied Shannon entropy method to fMRI time series in order to examine whether and how information can be extracted from different experimental paradigms, namely evoked or resting brain (RS) activity. Shannon entropy measures information content of the signal without making a priori assumptions. We found a striking match between the high-entropy voxels and “activated” voxels, while RS data did not reveal any cluster of high-entropy values. This finding indicates that RS activity cannot be extracted using a code (i.e., probability distribution) determined at the voxel-level, paving the way for different approaches to determine the code underlying RS activity.

Christine Law^1,2
1Stanford University, Stanford, CA, United States, ²Nuffield Department of Clinical Neurosciences, University of Oxford, Oxford, Oxfordshire, United Kingdom

A novel technique to concurrently resolve HRF and quantify fMRI activation response is presented. The problem formulation involves a rank 1 constraint (nonconvex), which makes it very difficult to solve. We were able to compute rank minimization by reformulating the nonconvex problem into a sequence of convex problems. We also developed an acceleration method such that computation time (global convergence) is significantly improved. We achieved over 99.9% accuracy on both HRF and stimulation response in simulations using parameters typical to fMRI studies. Our technique can benefit research involving a population having HRF that deviates from the canonical assumption.

Robert X. Smith¹, Kay Jann², Beau Ances³, and Danny J.J. Wang^4,5
1Neurology, UCLA, Los Angeles, CA, United States, ²UCLA, CA, United States, ³Neurology, Washington University School of Medicine, St. Louis, MO, United States, ⁴Neurology, UCLA, CA, United States, ⁵Radiology, UCLA, CA, United States

Our aim is the quantification of the complex neural fluctuations seen in resting state fMRI to provide a measure of mental health and cognitive function. We present here a wavelet based multiresolution entropy calculation that employs noise estimation measures to determine the complexity of the underlying neural behavior. In the presence of nonstationary data, wavelet analysis holds a significant advantage over Fourier analysis. We develop a pseudo-complexity measure using the stationary wavelet transform (SWT) of the original rs-fMRI time series to investigate the intrinsic irregularity of the energy density fluctuations at multiple temporal scales. We apply our measure to a cohort of 26 cognitively normal (clinical dementia rating scale (CDR) = 0) and 26 mild cognitively impaired (CDR = 0.5) individuals from the Healthy Aging and Senile Dementia program project. We report a reduced entropy seen in various resting state networks including default mode regions for CDR=0.5 individuals.

Pedro Cardoso¹, Florian Fischmeister^1,2, Alexander Geissler^1,2, Moritz Wurnig^1,2, Siegfried Trattnig¹, Roland Beisteiner^1,2, and Simon Daniel Robinson^1
1Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria, ²Department of Neurology, Medical University of Vienna, Vienna, Austria

FMRI is increasingly being used in presurgical planning in patients with brain tumors and epilepsy to facilitate resection of affected tissue without harming essential function. In clinical populations, the HRF may be modified in regions of pathology. Being model-free, reproducible and able to automatically identify and remove unreliable runs from the analysis, uNBIASED may aid identifying neuronal activation when poor performance or artifact contamination is present, or the response does not agree with the prediction due to compromised performance or modified hemodynamic coupling.

Oscar San Emeterio Nateras^1,2, Fang Yu¹, Carlos Bazan III¹, Anderson Houyun Kuo¹, Jinqi Li², and Timothy Q Duong^1,2
1Radiology, University of Texas Health Science Center, San Antonio, TX, United States, ²Research Imaging Institute, San Antonio, TX, United States

This study demonstrates a novel rsfMRI application in the human spinal cord. The challenges and solutions were detailed. Reproducibility within and across subject was demonstrated. A major finding is that there were multiple prominent rsfMRI patterns in the spinal cord, showing lateral connectivity, unilateral connectivity and top-down connectivity. Future studies will improve spatial resolution, image the entire spinal cord, and map spinal cord connectivity to the brain, as well as explore clinical applications.

Daniel Lee Shefchik¹, Andrew Scott Nencka¹, and James S. Hyde^1
1Biophysics, Medical College of Wisconsin, Milwaukee, WI, United States

This abstract aims to help address multiple comparison issues in functional MRI experiments. It does this through the utilization of directly registered tissue segmentation maps obtained through a multiple-echo, echo planar imaging (EPI) acquisition. The tissue maps are used to mask out the gray mater, which reduces the amount of voxels being analyzed, and provide a functional map with fewer false positives and negatives.

Valur Olafsson¹, Prantik Kundu², Chi Wah Wong³, Jia Guo³, Kun Lu³, Eman Ghobrial³, Peter Bandettini^2,4, Eric Wong³, and Thomas Liu^3
1Neuroscience Imaging Center, University of Pittsburgh, Pittsburgh, PA, United States, ²Section on Functional Imaging Methods, NIMH, Bethesda, MD, United States, ³Center for Functional MRI, UCSD, La Jolla, CA, United States, ⁴Functional MRI Facility, NIMH, Bethesda, MD, United States

The increased sampling rate of simultaneous multi-slice (SMS) fMRI acquisitions can increase tSNR and statistical power for resting state functional connectivity (fc) MRI. Multi-echo acquisitions with ICA analysis have also been proposed as a method to improve the detection of resting-state networks. Here we investigate the benefits of combining the two approaches and compare the performance of multi-echo SMS (MESMS) and multi-echo single-slice (MESS) acquisitions. We find that the higher sampling rate of the MESMS acquisition enables the identification of BOLD-like independent components that contain high frequency energy.

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

With seed-based correlation analysis, literatures on brain spontaneous activity have demonstrated that the default-mode network (DMN) is negatively correlated with a set of brain regions, referred to as the task-positive network (TPN) at rest[1]. However, regions compromised in the TPN and the extent of anti-correlations are inconsistent across different studies. It’s widely acknowledged that the reported inconsistency derives from specific MR acquisitions and distinct preprocessing steps: studies without correcting for physiological noise may fail to unveil anti-correlations buried in the physiological noises; while those conducting global signal regression (GSR) may demonstrate spurious anti-correlations due to the improper removal of informative neural information. Recently, it has been shown that, posterior cingulate cortex (PCC), the typical seed adopted by conventional analysis to study functional connectivity with respect to the DMN, has heterogeneous functions within its subparts. It’s likely that seeds residing in different functional units may lead to discrepant positive/negative correlation patterns, which has never been addressed in prior studies. Here, we first obtained different PCC seeds via parcellation, then employed conventional correlation analysis and recently proposed point-process analysis[6] to study such seed dependences of the observed anti-correlations between the DMN and TPN.

Yue Zhang¹, Xiaoying Wang², Jue Zhang¹, Xiaoping Hu³, and Jing Fang^1
1College of Engineering, Peking University, Beijing, Beijing, China, ²Department of Radiology, Peking University First Hospital, Beijing, Beijing, China,³Department of Biomedical Engineering, Georgia Institute of Technology / Emory University, Atlanta, Georgia, United States

The reliability is important for functional magnetic resonance imaging (fMRI) data. Previous study has developed a statistical model (independence model) to quantify the test-retest reliability with only two states (active or inactive). More and more fMRI experiments have detected three state regions in the brain, including active, deactive and non-significant regions. In order to quantify the test-retest reliability of the fMRI data with three state regions, this study developed the three-state independence model. The model was applied to acupuncture fMRI data, the results indicated that the reliability of BOLD was higher than that of CBF.

Yu-Ting Ko¹, Pai-Chuan Hung¹, Geng-Hong Lin¹, Pei-Jung Tsai², Ching-Po Lin³, and Changwei W Wu^1
1Graduate Institute of Biomedical Engineering, National Central Unversity, Taoyuan, Taiwan, ²Department of Biomedical Imaging and Radiological Sciences, National Yang-Ming University, Taipei, Taiwan, ³Institute of Neuroscience, Nation Yang-Ming University, Taipei, Taiwan

The resting-state fMRI signal is the baseline fluctuations with non-linear and non-stationary properties; thus its underlying mechanism was quite fussy. From the frequency viewpoints, we attempted to retrieve its physiological implications by observing the frequency correspondence between fMRI and EEG. Therefore, we simultaneously recorded fMRI and EEG signals in 2 conditions (before and after sleep) and used the Hilbert-Huang Transform to extract their spectral correspondence. In our results, we found the opposite frequency correspondences between EEG and fMRI in the resting state, especially in the low frequency range, resembling previous studies. Furthermore, such correspondence deviates in different physiological conditions.

Martin Miguel Merener¹, Richard Byrd², Rajesh R. Nandy³, and Dietmar Cordes^1,4
1Physics, Ryerson University, Toronto, Ontario, Canada, ²Computer Science, University of Colorado Boulder, Boulder, CO, United States, ³School of Public Health, University of North Texas, Fort Worth, TX, United States, ⁴Department of Psychology and Neuroscience, University of Colorado Boulder, CO, United States

This study provides important improvements in fMRI data analysis techniques for the detection of active brain areas. We propose and study a family of constraints for CCA, which naturally generalizes two interesting previously studied models. The solutions for these models can be found numerically and efficiently. For several choices of these constraints, the performance of the method in determining active voxels is excellent as measured via ROC simulations, and provide a significant improvement compared to previously published models in constrained CCA.

M. Muge Karaman¹, Daniel B. Rowe^1,2, and Andrew S. Nencka^2
1Department of Mathematics, Statistics, and Computer Science, Marquette University, Milwaukee, WI, United States, ²Department of Biophysics, Medical College of Wisconsin, Milwaukee, WI, United States

In fMRI and fcMRI, many studies have aimed to alleviate the data through spatial and temporal processing. While such processing alleviates the noise, it alters the statistical properties of the data by inducing correlations of no biological origin. We propose a linear model to precisely quantify the correlations induced by spatiotemporal processing, and expand the current complex-valued fMRI model to incorporate the effects of processing into the final analysis. The proposed model provides a true interpretation of the acquired data and in turn contributes to producing more accurate functional activation and connectivity statistics by decreasing false negative and positives.

Steffen Bollmann¹, Lars Kasper², Carmen Ghisleni¹, Simon-Shlomo Poil¹, Peter Klaver³, Lars Michels⁴, Dominique Eich-Höchli⁵, Daniel Brandeis^6,7, and Ruth L. O'Gorman^1
1Center for MR-Research, University Children's Hospital, Zurich, Zurich, Switzerland, ²Institute for Biomedical Engineering, University and ETH Zurich, Zurich, Switzerland, ³Institute of Psychology, University of Zurich, Zürich, Switzerland, ⁴Institute of Neuroradiology, University Hospital of Zurich, Zürich, Switzerland, ⁵Psychiatric University Hospital, Zürich, Switzerland, ⁶Department of Child & Adolescent Psychiatry, University of Zurich, Zürich, Switzerland, ⁷Central Institute of Mental Health Mannheim / Heidelberg University, Germany

Although physiological noise correction is considered to be important, there is little known about the impact of physiological noise correction on task based fMRI group studies. We therefore investigated the effect of RETROICOR regressors on a working memory paradigm comparing healthy adults to patients with ADHD. By including physiological noise regressors into a task-based fMRI analysis, we observed an increase in power in task-relevant regions. At the same time, presumably spurious activation in areas previously associated with physiological noise was diminished. Physiological noise correction for fMRI therefore appears to reduce the risk of interpreting group differences caused by physiological artifacts.

Xiaojie Zhao¹, Xiaofei Li¹, and Li Yao^1
1College of Information Science and Technology, Beijing Normal University, Beijing, Beijing, China

As a common data preprocessing procedure for fMRI data, spatial normalization can provide abundant referential information for the brain region recognition. However, for real-time fMRI (rtfMRI), which requires the entire data processing within a single TR, spatial normalization is too time-consuming to include in the data preprocessing in rtfMRI. In this paper, we discussed the cutoff frequency and iteration number using bisection method in nonlinear registration of spatial normalization, proposed an adjusted nonlinear registration method to meet the real-time requirement of rtfMRI.

Mariela Hidalgo¹, Juan Vielma², Rodrigo Salas¹, Alejandro Veloz^1,3, and Steren Chabert^1
1Biomedical Engineering Department, Universidad de Valparaíso, Valparaíso, Quinta Region, Chile, ²Medicine Faculty, Universidad de Valparaíso, Valparaíso, Quinta Región, Chile, ³Informatics Department, Universidad Técnica Federico Santa María, Valparaíso, Quinta Región, Chile

In many studies of functional MRI the existence of inter individual variability in the hemodynamic response function has been demonstrated and influences the cortical activation detection. This study consists in evaluating this effect among seven healthy volunteers when an individual HRF is used in the detection of brain activation in the general lineal model analysis. In about half of the cases, similar activation volumes are detected in both cases. Our data show low variability in Time-To-Peak, with a mean TTP longer than the canonical TTP. A robust methodology is still to be defined to reduce possible the inter-individual volume variability.

Daniel F. Arteaga¹, Megan K. Strother¹, Carlos C. Faraco¹, Lori C. Jordan², Travis R. Ladner¹, Lindsey M. Dethrage¹, Robert J. Singer³, J Mocco⁴, Paul F. Clemmons⁵, Michael J. Ayad⁶, and Manus J. Donahue^1,2
1Department of Radiology, Vanderbilt University, Nashville, Tennessee, United States, ²Department of Neurology, Vanderbilt University, Nashville, Tennessee, United States, ³Section of Neurosurgery, Dartmouth Hitchcock Medical Center, Lebanon, New Hampshire, United States, ⁴Department of Neurosurgery, Vanderbilt University, Nashville, Tennessee, United States, ⁵Department of Radiology Nursing, Vanderbilt University, Nashville, Tennessee, United States, ⁶Department of Neurosurgery, New York Methodist Hospital, Brooklyn, New York, United States

Vascular steal has been proposed as a compensatory mechanism in hemodynamically-compromised ischemic parenchyma. Here, independent measures of changes in CBF and BOLD MRI contrast in response to a vascular stimulus in patients (n=40) with ischemic cerebrovascular disease are recorded. 15/40 participants exhibited negative BOLD reactivity. Of these, three participants exhibited significant (P<0.01) reductions in CBF with hypercarbia; eight exhibited increases (P<0.01) in CBF and the remaining four participants exhibited no statistical change in CBF. These findings suggest that the origins of negative BOLD responses in stroke patients are most frequently not due to vascular steal.

Luke Norman¹, Andrew Iles², Natalia Lawrence², Abdelmalek Benattayallah³, and Anke Karl^2
1Institute of Psychiatry, Department of Child and Adolescent Psychiatry, King's College, London, London, United Kingdom, ²Mood Disorders Centre, School of Psychology, University of Exeter, Exeter, United Kingdom, ³Exeter MR Research Centre, University of Exeter, Exeter, United Kingdom

Expressive-suppression is a maladaptive emotion-regulation method, which is associated with increased post-traumatic symptoms following trauma. Thirty-five individuals took part in the first neuroimaging study of expressive-suppression in a trauma-exposed population. We found that self-reported use of expressive-suppression was associated with decreased activation in the mPFC, and increased activation in the insula, in an emotional-faces task. Our findings suggest that deficits in mPFC –limbic circuitry may prompt compensatory use of expressive-suppression in trauma exposed individuals. Furthermore, they suggest that insula hyperactivation in PTSD may partially result from increased habitual expressive-suppression suppression to emotional material in this population.

Xiaolin Liu¹, 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

In this study, we investigated the effects of propofol sedation on the functional connectivity of the precuneus, an important neural structure to human consciousness in the posteromedial parietal cortex. We found that, contrary to our hypothesis, compared with wakefulness deep sedation at the point of unresponsiveness (with auditory stimuli continuously supplied) is marked by an increase of precuneus connectivity, particularly in the dorsal medial prefrontal and visual cortices. The enhanced cortical connectivity of the precuneus may reflect disconnected endogenous mentation or dreaming that continues at a lower rate of energy consumption during the unresponsive state of propofol sedation.

Shiliang Huang¹, Lora Talley Watts¹, Justin Long¹, Qiang Shen¹, and Timothy Duong^1
1Research Imaging Institute, University of Texas Health Science Center at San Antonio, San Antonio, TX, United States

This study examined longitudinally the rsfMRI changes in mild TBI of the (unilateral) primary somatosensory cortex in rats during hyperacute and chronic phase up to 14 days. Quantitative correlations were made with T2, DTI, fractional anisotropy (FA), and functional outcomes (forelimb placement asymmetry and foot fault scores). rsfMRI z-scores reduced after mild TBI but improved with time. The trend of improvement parallels those of behavioral scores. rsfMRI did not significantly correlate with ADC and FA changes, suggesting they provide complementary information. This study demonstrated that rsfMRI offers novel insights into functional connectivity in mild TBI.

Shun Qi¹, Hong Yin², and Yunfeng Mu^3
1Xijing Hospital of The Fourth Military Medical University, Xi¡¯an, shaanxi, China, ²Xijing Hospital of The Fourth Military Medical University, shaanxi, China, ³Xijing Hospital of The Fourth Military Medical University, China

First, the PTSD which was caused by a very rare accident, happened around 8:40 a.m. on July 29th, 2007. A severe coalmine-flood disaster occurred at Zhijian Coalmine in Shanxian County, about 200 km west of Zhengzhou, the capital of Henan Province in central China (USA Today News, 2007). Sixty-nine male miners were trapped in a nearly 1400 m underground coal pit. Fortunately, all of them were rescued after 75 hours of the ordeal in the darkness, with no deaths and severe body injuries. This study is the first evidence to find the cortex thickness reduction by surface-based morphometry based on such serious, sustained, direct, high-intensity and acute trauma.

Kapil Chaudhary¹, S SENTHIL KUMARAN², Poodipedi Sarat Chandra³, and Manjari Tripathi^1
1Neurology, All India Institute of Medical Science, New Delhi, Delhi, India, ²Department of NMR, ALL INDIA INSTITUTE OF MEDICAL SCIENCES, New Delhi, Delhi, India, ³Neurosurgery, All India Institute of Medical Science, New Delhi, Delhi, India

The localizing language and memory in intractable epilepsy has become increasingly elucidated. We describe semantic noun naming and complex scene working memory dysfunction in a right-handed patient using functional MRI and define clinical correlation with functional imaging. A 10-year-old right-handed girl with Rasmussen’s syndrome was referred to our institute. Intractable seizures may need to be treated with surgery if the seizure control in the patient is not achieved with medications. Although surgical resection is successful in curing up to 70% of patients, a significant risk of post-operative cognitive decline continues to limit its application. Functional MRI was used to observe the role of language and memory function in this patient.

Fatemeh Mokhtari¹, Elizabeth Davenport¹, Jillan Urban¹, Naeim Bahrami¹, Christopher Whitlow², Alex Powers³, Joel Stitzel¹, and Joseph Maldjian^2
1Biomedical Engineering, Wake Forest University School of Medicine, Winston Salem, NC, United States, ²Radiology, Wake Forest University School of Medicine, Winston Salem, NC, United States, ³Neurosurgery, Wake Forest University School of Medicine, Winston Salem, NC, United States

The primary goal of this study is to determine if cumulative head impacts over a season of high school football has an effect on frontal lobe interhemispheric connectivity. In order to explore this relationship a multiple regression analysis was performed using a linear model of a score of head impacts vs. pre-post difference in fMRI connectivity. Results indicate changes in connectivity of frontal structures in non-concussed subjects during a season of football. These findings add to a growing body of literature that cumulative subconcussive sports-related impacts may have an effect on the brain function.

Gopikrishna Deshpande^1,2, Hao Jia¹, Xiaoping Hu³, Changfeng Jin⁴, Lingjiang Li⁴, and Tianming Liu^5
1MRI Research Center, Department of Electrical and Computer Engineering, Auburn University, Auburn, Alabama, United States, ²Department of Psychology, Auburn University, Auburn, Alabama, United States, ³Biomedical Imaging Technology Center, Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, Atlanta, Georgia, United States, ⁴The Mental Health Institute, The Second Xiangya Hospital, Central South University, Changsha, China, ⁵Department of Computer Science, University of Georgia, Athens, Georgia, United States

It is acknowledged that functional connectivity (FC) in the brain obtained from resting state fMRI dynamically changes with time. Further, it has been shown that dynamic changes in FC and effective connectivity (EC) are relevant to disease processes. However, an outstanding question that remains is whether dynamic information from FC and EC provide increased sensitivity for identifying brain pathologies in addition to that obtained by static connectivity metrics? Here, we provide answers to these questions by demonstrating that information from temporal variations in FC and EC provides better accuracy for classifying subjects with PTSD (post-traumatic stress disorder) from healthy controls.

Benito de Celis Alonso¹, Silvia Hidalgo Tobón², Pilar Dies Suarez², and Eduardo Barragán Pérez^2
1BUAP, Puebla, Puebla, Mexico, ²Hospital Infantil de México Federico Gómez, Mexico DF, Mexico

The Posterior Cingulate Cortex is a key node of the default mode network of resting states and has been shown to be affected by ADHD. Several studies exist on the resting states of this disorder but little to no amount of work to our knowledge exists studying the specific ROI to ROI connectivity. In this project we compared ADHD children patients with healthy ones. We assessed the differences in connectivity in their resting states with a special focus in the role of the posterior cingulate and retrosplenial cortex.

Ludovica Griffanti^1,2, Ottavia Dipasquale^1,2, Francesca Baglio¹, Raffaello Nemni^1,3, Mario Clerici^1,3, and Giuseppe Baselli^2
1IRCCS, Fondazione don Carlo Gnocchi, Milano, Milan, Italy, ²Department of Electronics, Information and Bioengineering, Politecnico di Milano, Milan, Italy, ³Physiopatholgy Department, Università degli Studi di Milano, Milan, Italy

With high-dimensional independent component analysis (ICA) the resting state (RS) networks typically found with low-dimensional ICA are decomposed in sub-networks, giving further insight into functional connectivity changes in pathological conditions, e.g. in Alzheimer's disease (AD). We performed temporal analyses of RS-fMRI data in healthy subjects and AD patients, focusing on the primarily altered default mode network (DMN) and exploring the sensory motor network. Low-dimensional results confirmed literature, while high-dimensional decomposition in sub-networks was essential to better localize functional connectivity alterations in AD, suggesting that the connectivity damage is not confined to the DMN.

Sandrine Saillet^1,2, Olivier David^1,2, and Jan M. Warnking^1,2
1U836, Inserm, Grenoble, France, ²Grenoble Institut des Neurosciences, Université Joseph Fourier, Grenoble, France

Recording EEG together with fMRI has permitted to unravel the neuronal correlates of spontaneous brain activity. However the mechanisms involved in the coupling between epileptic discharges (EDs) and the hemodynamic response are partially known. Actually, the interpretation of fMRI results (“activated” or “deactivated”), in terms of underlying physiological processes crucially depend on the understanding of these mechanisms. In this work, we measured local field potentials simultaneously with changes in BOLD signal, cerebral blood flow (via ASL) and cerebral blood volume (via Mion fMRI) in the rat somatosensory cortex following intracortical bicuculline injection eliciting interictal-like discharges.

Svyatoslav Vergun¹, Wolfgang Gaggl², Veena A Nair², Rasmus M Birn³, M. Elizabeth Meyerand³, James Reuss⁴, Edgar A DeYoe⁵, and Vivek Prabhakaran^2
1Medical Physics, UW-Madison, Madison, WI, United States, ²Radiology, UW-Madison, WI, United States, ³Medical Physics, UW-Madison, WI, United States, ⁴Prism Clinical Imaging, Inc, WI, United States, ⁵Radiology, Medical College of Wisconsin, WI, United States

We present a clinically motivated, automated component decomposition and classification method using resting state functional MRI data of epilepsy and vascular/tumor patients. Preprocessed resting state scans are decomposed, with respect to their functional time series signal, using spatial independent component analysis. The resultant components are used in the classification step in which they are spatially correlated with a template compiled by a previous study. The automated classifier achieved promising performance for the visual, sensorimotor, default-mode and auditory networks.

Tynan Stevens^1,2, David Clarke^3,4, Ryan D'Arcy^5,6, Gerhard Stroink¹, and Steven Beyea^2,7
1Physics, Dalhousie University, Halifax, NS, Canada, ²Neuroimaging Research Lab, BIOTIC, Halifax, NS, Canada, ³Surgery, Dalhousie University, Halifax, NS, Canada, ⁴Neurosurgery, QEII Health Sciences Centre, Halifax, NS, Canada, ⁵Applied Science, Simon Frasier University, Burnaby, BC, Canada,⁶Surrey Memorial Hospital, Surrey, BC, Canada, ⁷Radiology, Dalhousie University, Halifax, NS, Canada

We demonstrate a novel data-driven method for selecting thresholds for pre-surgical fMRI data, based on reliability of the activation patterns in just a single fMRI run. Our new method incorporates spatial information not present in histogram based thresholding methods, and alleviates the need for test-retest imaging of existing reliability optimization methods. The new method produces significantly higher test-retest overlap when compared to established threshold optimization methods, particularly for low CNR situations like language mapping in patient populations. This analysis therefore provides the most robust automated thresholds, and unlike other techniques can be applied to any existing fMRI paradigm without modification.

Chuangjian Cai¹, Xue Xiao¹, Yan Zhu², and Kui Ying^3
1Department of Biomedical Engineering, Tsinghua University, Beijing, Beijing, China, ²Yuquan Hospital, Tsinghua University, Beijing, China, ³Department of Engineering Physics, Tsinghua University, Beijing, China

Machine learning techniques for fMRI help to identify the features of some brain disease, such as depression. We utilized dynamic functional connectivity analysis of resting state fMRI with BP ANNs to investigate the dynamic differences and differentiate between the depression and the control group, with cross validation and permutation test to test its feasibility. A general rate of 95.45% was achieved, better than the traditional method that combines support Vector Machine and static analysis only. The new method certificated the internal regularity of dynamic functional connectivity, found brain regions with highly discriminative power and supplied an effective model for dynamic functional connectivity investigation.

Kai Ai¹, Gang Yu¹, Pan Lin², and YanHua Gao^3
1School of Geosciences and Info-Physics, Central South University, Changsha, China, ²Institute of Biomedical Engineering, Xi'an Jiaotong University, Xi'an, China, ³Department of B Ultrasound, Shaanxi Provincial People's hospital, Xi'an, China

Emotional memory is an important brain cognitive function and shows a positive affective bias in healthy aging. However, previous studies considered the extracted time series as stationary processes. Recently, studies show that dynamic coupling or functional connectivity (FC) patterns between brain regions are distinct from underlying anatomical links. Emerging evidence suggests that dynamic FC may index changes in brain function or clinical biomarker. Understanding of dynamic emotion networks is important to characterize aging brain development and the related brain disorder. This study investigates the time-frequency features of dynamic emotional memory network FC in aging brain by using wavelet transform coherence.

Maryam Abaei^1,2, Devi Sagar^2,3, Clair Spicer³, Elizabeth G Stockley², Malcolm J.W Prior⁴, Henryk Fass¹, David A Walsh^2,5, Victoria Chapman^2,3, and Dorothee P Auer^1,2
1Radiological Sciences, Division of Clinical Neurosciences, Nottingham University, Nottingham, Nottinghamshire, United Kingdom, ²Arthritis Research UK Pain Centre, Nottingham University, Nottingham, Nottinghamshire, United Kingdom, ³School of Life Sciences, Nottingham University, Nottingham, Nottinghamshire, United Kingdom, ⁴School of Medicine, Nottingham University, Nottingham, Nottinghamshire, United Kingdom, ⁵Academic Rheumatology, Nottingham University, Nottingham, Nottinghamshire, United Kingdom

Ph-fMRI, Brain, Capsaicin, MIA, Osteoarthritis, Animal, Neuronal, Sensitization

Giusy Olivito^1,2, Marco Bozzali³, Marco Molinari², Maria Leggio^1,2, and Mara Cercignani^3,4
1Department of Psychology, Sapienza University of Rome, Rome, Italy, ²Ataxia Laboratory, IRCCS Santa Lucia, Rome, Italy, ³Neuroimaging Laboratory, IRCCS Santa Lucia, Rome, Italy, ⁴Clinical Imaging Science Center, Brighton and Sussex Medical School, Brighton, United Kingdom

The cerebellar output channels have been demostrated to be spatially segregated and to focus on functionally distinct cortical systems, supporting the cerebellar role in cognition. 3 cases of patients with left cerebellar lesions were used to demonstrate the sensitivity of resting-state (RS) fMRI to changes in functional connectivity caused by the presence of the lesion. Using the resting-state approach, we were able to define the pattern of connectivity between cerebral cortex and specific cerebellar lobuli. When comparing each patient to the healthy participants, we found different patterns of altered connectivity, involving both contralateral and ipsilateral cortical areas.

Jinze Li¹, Gang Yu¹, Pan Lin², Yanhua Gao³, and Kai Ai^1
1School of Geosciences and Info-Physics, Central South University, Changsha, China, ²Institute of Biomedical Engineering, Xi'an Jiaotong University, Xi'an, China, ³Department of B Ultrasound, Shaanxi Provincial People's hospital, Xi'an, China

Psychiatric disorders are harmful to children and adolescents. And it¡¯s a hard work to distinguish the corresponding patients from the healthy in early diagnosis. Previous studies have proved that the brain functional networks show abnormal pattern in children and adolescents who suffer from mental diseases such as attention deficit hyperactivity disorder (ADHD) and autism disorder. This paper presents a combined method based on independent components analysis (ICA) and support vector machine to classify ADHD, Autism and control group automatically. Based on the combined method, more psychiatric disorders of children and adolescents are expected to be automatically distinguished in the future.

Ravjit Singh Sagoo¹, Habib Ganjgahi², Eddie Ng'andwe¹, Mahmud Saedon³, Sarah Wayte⁴, Alex Wright^5,6, Arthur Bradwell^5,6, Charles Hutchinson^1,7, Christopher Imray^3,6, and Thomas Nichols^2
1Department of Imaging, University Hospitals Coventry and Warwickshire NHS Trust, Coventry, United Kingdom, ²Department of Statistics, University of Warwick, Coventry, United Kingdom, ³Department of Surgery, University Hospitals Coventry and Warwickshire NHS Trust, Coventry, United Kingdom,⁴Department of Medical Physics, University Hospitals Coventry and Warwickshire NHS Trust, Coventry, United Kingdom, ⁵University of Birmingham, Birmingham, United Kingdom, ⁶Birmingham Medical Research Expeditionary Society, Birmingham, United Kingdom, ⁷University of Warwick, Coventry, United Kingdom

Rapid ascent to high altitude results in arterial hypoxaemia, frequently leading to acute mountain sickness (AMS). The precise mechanisms remain poorly understood. AMS symptoms can include cognitive and behavioural changes. We postulate that changes in cerebral blood flow in response to hypoxia may result in changes in the brain’s resting-state functional connectivity and, therefore, could be studied using resting-state functional MRI. 12 subjects underwent serial scans over a 22-hour period of normobaric hypoxia. The results did not show a global increase in connectivity to account for the cognitive/behavioural symptoms. This suggests that alternative pathophysiological processes may contribute to these symptoms.

Joanna L Varley¹, Jonathan Fulford², and Craig A Williams^1
1Children's Health & Exercise Research Centre, University of Exeter, Exeter, Devon, United Kingdom, ²Exeter NIHR Clinical Research Facility, University of Exeter, Exeter, Devon, United Kingdom

Previous studies have indicated the detrimental effect of missing breakfast on cognitive performance in school. The aim was to investigate the feasibility of utilizing functional magnetic resonance (fMRI) techniques with children and to examine changes in brain activity when undertaking cognitive tasks between a breakfast fasted and satiated state. Significant positive activations were found in Broadmann areas 6, 17 and 45 when comparing the satiated state to the fasted. The findings show that the impact of breakfast consumption can be observed through fMRI activated areas of the brain when completing cognitive tasks, compared to a fasted state in children.

Feng Xu^1,2, Peiying Liu³, James J. Pekar^1,2, and Hanzhang Lu^3
1Russell H. Morgan Department of Radiology, Johns Hopkins University, Baltimore, MD, United States, ²F.M. Kirby Research Center, Kennedy Krieger Institute, Baltimore, MD, United States, ³Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, TX, United States

Caffeine has a vasoconstriction effect on vasculature. However, its exact neuro-metabolic effect has not been examined. We used TRUST MRI, phase-contrast MRI and PCASL MRI to examine the dynamic changes of whole-brain cerebral metabolic rate of oxygen (CMRO2), venous oxygenation (Yv), cerebral blood flow (CBF) after caffeine ingestion. Significant decreases of whole-brain CBF and Yv were found, while no changes were present in whole-brain CMRO2. Regional CBF revealed various decline rates from global CBF, suggesting that neural metabolism might paly a role of enhancing and suppressing CBF in addition to vasoconstriction. Co-existing neural effects lead to whole-brain CMRO2 unchanged.

Maria A. Rocca¹, Paola Valsasina¹, Alvino Bisecco¹, Alessandro Meani¹, Laura Parisi¹, Maria Josè Messina², Bruno Colombo², Andrea Falini³, Giancarlo Comi², and Massimo Filippi^1
1Neuroimaging Research Unit, Institute of Experimental Neurology, San Raffaele Scientific Institute, Vita-Salute San Raffaele University, Milan, MI, Italy,²Department of Neurology, San Raffaele Scientific Institute, Vita-Salute San Raffaele University, Milan, MI, Italy, ³Department of Neuroradiology, San Raffaele Scientific Institute, Vita-Salute San Raffaele University, Milan, MI, Italy

Resting state functional MRI (RS fMRI) and graph theory were applied to explore abnormalities of large-scale brain networks (connectome) in 64 patients with multiple sclerosis (MS) and fatigue (F). As control groups, 60 MS patients without fatigue (NF) and 59 healthy controls (HC) were included. F-MS patients, unlike HC and NF-MS patients, lost hubs in the thalami and middle cingulate cortex. Compared to HC and NF-MS patients, F-MS patients experienced a decreased degree in the bilateral thalamus. Fatigue in MS is related to a functional disruption of the thalamic connector, which should be the target of potential therapeutic interventions.

Marta Maieron¹, Barbara Tomasino², Serena D'Agostini³, Miran Skrap⁴, and Ferdinando Calzolari^5
1SOC Medical Physics, AOU - S. Maria della Misericordia di Udine, Udine, UDINE, Italy, ²IRCCS Medea: Nostra Famiglia, Pasian di Prato, Udine, Italy, ³SOC of Neuroradiology, AOU S. Maria della Misericordia di Udine, Udine, Italy, ⁴SOC of Neurosurgery, AOU S. Maria della Misericordia di Udine, Udine, Udine, Italy, ⁵SOC of Neuroradiology, AOU S. Maria della Misericordia di Udine, Udine, Udine, Italy

In a routine clinical practice, fMRI plays an important role ad non-invasive tool for pre-surgical functional mapping of eloquent cortex. The standard approach is to acquire fMRI data while the patient perform a task designed to target a specific function. However there are some limitations: the patients can have difficulty in performing required tasks and the task-based approaches may be unreliable. Functional mapping based on spontaneous intrinsic activity, referred to as resting state fMRI, offers a different option for presurgical mapping. The aim of our study was to confirm the important role that this methods could have in clinical field.

Hesamoddin Jahanian¹, Wendy W Ni¹, Thomas Christen¹, Michael E Moseley¹, Manjula K Tamura², and Greg Zaharchuk^1
1Stanford University, Department of Radiology, Stanford, California, United States, ²Stanford University, Division of Nephrology, Stanford, California, United States

In this work, we hypothesized that the spontaneous resting state BOLD signal fluctuations can be viewed as the response of the brain to the internal challenges to the cerebrovascular system, including heartbeat, inhalation, and baseline neuronal activity and may provide information about cerebrovascular reactivity. To test this hypothesis we compared the magnitude of rsBOLD signal fluctuations to the cerebrovascular reactivity measured as percentage signal change during a breath-holding challenge in a population of older adults (N=30). Our results indicate a strong linear relation between the magnitude of rsBOLD signal fluctuations and breath-holding percentage signal change across subjects.

Wendy W Ni^1,2, Catie Chang³, Hesamoddin Jahanian², Gary H Glover², and Greg Zaharchuk^2
1Department of Electrical Engineering, Stanford University, Stanford, CA, United States, ²Department of Radiology, Stanford University, Stanford, CA, United States, ³Advanced MRI Section, LFMI, NINDS, NIH, Bethesda, MD, United States

In this study, we analyze the percentage of resting state BOLD fMRI signal variance attributable to cardiac and respiratory fluctuations, as quantified using the RVHR model, in young normal subjects, elderly subjects with hypertension and chronic kidney disease, and elderly normal subjects. We found a statistically significant difference (p<0.05 with Bonferroni correction) between the young group and each elderly group, but not between the elderly groups. This finding supplements previous work to indicate an association of aging with increased non-physiological fluctuations and/or reduction or change in neurovascular response to physiological stimulation in resting state.

Gustavo Pamplona¹, Gérson Santos Neto², Sara Rosset², and Carlos Ernesto Garrido Salmon^1
1Department of Physics, FFCLRP - USP, Ribeirão Preto, São Paulo, Brazil, ²FMRP - USP, Ribeirão Preto, São Paulo, Brazil

It is known that multiple brain areas, even to an individual at rest, work synchronously even if they are anatomically separated, suggesting functional and structural connections. In this way, our brain can be considered a complex network, in which nodes can be the different areas and edges can be the measurements of functional connectivity between time series of the magnetic resonance signal of each area. In this study, we purpose to analyze the relationship between network topological parameters and intellectual performance, using magnetic resonance images and considering weighted and binary functional connectivity networks.

Shih-Yen Lin^1,2, Hengtai Jan¹, Tsang-Chu Yu³, Yi-Ping Chao³, Kuan-Hung Cho⁴, and Li-Wei Kuo^1
1Institute of Biomedical Engineering and Nanomedicine, National Health Research Institutes, Miaoli, Taiwan, ²Department of Computer Science, National Chiao Tung University, Hsinchu, Taiwan, ³Department of Computer Science and Information Engineering, Chang Gung University, Taoyuan, Taiwan,⁴Institute of Brain Science, National Yang-Ming University, Taipei, Taiwan

To investigate the brain networks at multiple scales, recent studies have attempted to divide the cortical regions into smaller parcels at multiple levels of granularity. In this study, we proposed a parcellation method based on the weighted k-means algorithm with the following desirable features, including similar subdivision volume size over the whole brain, not fragmented, fully deterministic and highly reproducible. A quantitative evaluation with calculating the coefficient of variance among all parcels was performed. Our results show the variances significantly drop between intermediate to finest levels, suggesting that the clustering sizes become more uniformly. Future works include developing more quantitative evaluation parameters, demonstration on other brain atlases and application on brain network analysis at multiple scales.

Jiancheng Zhuang¹ and Bosco Tjan^1
1University of Southern California, Los Angeles, CA, United States

We describe an approach of using dynamic Structural Equation Modeling (SEM) analysis to estimate the effective connectivity networks from resting-state fMRI data measured by a multiband EPI sequence. Two structural equation models were estimated at each voxel with respect to the sensory-motor network and default-mode network. The resulting connectivity maps indicate that supplementary motor area has significant connections to left/right primary motor areas, and medial prefrontal cortex link significantly with posterior cingulate cortex and inferior parietal lobules. The results imply that high temporal resolution images obtained with multiband fMRI data can provide dynamic and directional information on effective connectivity.

Chiara Mastropasqua^1,2, Marco Bozzali¹, Giovanni Giulietti¹, Giacomo Koch³, and Mara Cercignani^1,4
1Neuroimaging Laboratory, IRCCS Santa Lucia, Rome, Italy, ²Trieste University, Rome, Italy, ³Department of Clinical and Behavioural Neurology, IRCCS Santa Lucia, Rome, Italy, ⁴Clinical Imaging Sciences Centre - University of Sussex, Brighton and Sussex Medical School, Brighton, United Kingdom

With the aim of comparing thalamo-cortical functional and structural connections, we used independent component analysis of resting state (RS) fMRI data to detect thalamic functional components at rest. Next, we used these thalamic components as seeds for RS seed based analysis and probabilistic tractography to identify cortical regions structurally and functionally connected with each thalamic components. The results were visually compared to cross-validate these two commonly used approaches. These two methods yield partially consistent results: the partial overall correspondence between structural and functional connections suggests that they provide complementary information.

Chiara Mastropasqua^1,2, Marco Bozzali¹, Mara Cercignani^1,3, Viviana Ponzo⁴, and Giacomo Koch^4
1Neuroimaging Laboratory, IRCCS Santa Lucia, Rome, Italy, ²Trieste University, Trieste, Italy, Italy, ³Clinical Imaging Sciences Centre - University of Sussex, Brighton and Sussex Medical School, Brighton, United Kingdom, ⁴Department of Clinical and Behavioural Neurology, IRCCS Santa Lucia, Rome, Italy

Our aim was to combine transcranial-magnetic stimulation and resting-state (RS) fMRI to investigate changes in functional connectivity at rest induced by prefrontal continuous theta-burst stimulation (cTBS). Seed-based analysis was used to identify the main connections to the prefrontal cortex. The adjacency matrix summarising the correlation between the resulting 29 regions was computed before and after cTBS, and compared using the Network Based Statistics toolbox. After cTBS, thecorrelation between the right prefrontal cortex and right parietal cortex was decreased, demonstrating for the firts time the possibility to induce selective changes in a specific region without interfering with functionally correlated area.

Anand Narasimhamurthy¹, Ashish Anil Rao¹, Ek Tsoon Tan², Rakesh Mullick¹, and Suresh Emmanuel Joel^1
1General Electric Global Research, Bangalore, Karnataka, India, ²General Electric Global Research, Niskayuna, New York, United States

In this work we investigate the short term dynamics of brain connectivity using a graph theoretic representation. While there are known situations where brain connectivity changes, for instance following a traumatic brain injury; these changes take place over a length of time, brain connectivity could be quite dynamic even in a short time scale. While graph theoretical methods have been extensively used in neuroimaging, the dynamics aspect of brain connectivity is relatively less explored. We report the dynamics at the local level by quantifying changes in the neighborhood connectivity of nodes across time.

Hao Jia¹, Xiaoping Hu², and Gopikrishna Deshpande^1,3
1MRI Research Center, Department of Electrical and Computer Engineering, Auburn University, Auburn, Alabama, United States, ²Biomedical Imaging Technology Center, Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, Atlanta, Georgia, United States,³Department of Psychology, Auburn University, Auburn, Alabama, United States

Dynamic functional connectivity (FC) analysis has received increasing attention since it is hoped that dynamics of FC could provide potentially more information than its static counterpart, benefiting neuroscientific and clinical research. However, the key question that is yet to be answered is whether connectivity dynamics has any behavioral relevance over and above that obtained from static FC. In this work, we describe a principled framework for calculating dynamic FC metrics from the whole brain using data-driven adaptive windows and evolutionary clustering. In addition, we demonstrate that dynamic FC explains more variance in behavior as compared to static FC metrics.

Ashish Anil Rao¹, Hima Patel¹, Ek Tsoon Tan², Rakesh Mullick¹, and Suresh Emmanuel Joel^1
1General Electric Global Research, Bangalore, Karnataka, India, ²General Electric Global Research, New York, United States

Automated classification of ICA derived components in to components of neuronal origin and components of noise origin will be very useful. Several attempts with modest results have been reported previously. Recently a method for classifcation of ICA derived from high resolution, long duration multiband scans has been reported. Here we present accurate automated classifier at a single subject single run level for the widely used conventional resting state fMRI.

Olga Tymofiyeva¹, Etay Ziv¹, Donna M Ferriero², A James Barkovich¹, Christopher P Hess¹, and Duan Xu^1
1Department of Radiology & Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States, ²Department of Pediatrics, University of California, San Francisco, San Francisco, CA, United States

Through the abstraction from the anatomy, the developed framework allows for unbiased construction and connection-wise comparison of diffusion MRI-based brain networks. Brain alignment is performed in the network domain and, therefore, can be applied to subjects at any stage of development with any potential anatomical abnormalities.

Ehsan Eqlimi¹, Nader Riyahi Alam¹, MA Sahraian², A Eshaghi², and Hamidreza Saligheh Rad^1,3
1Medical Physics and Biomedical Engineering Department, Tehran University of Medical Sciences, Tehran, Tehran, Iran, ²Sina MS Research Center, Sina Hospital, Tehran, Tehran, Iran, ³Quantitative MR Imaging and Spectroscopy Group, Research Center for Molecular and Cellular Imaging, Tehran, Tehran, Iran

Functional magnetic resonance imaging (fMRI) can be applied to investigate resting state functional connectivity in brain without any stimulation paradigm. Resting state communication patterns between brain areas is a key to understand how brain functions. Furthermore, abnormal functional connectivity within brain networks is thought to be responsible for some pathologies. In this work, we proposed mutual information weighted graphs instead of classic correlation graphs to model brain functional networks, and extracted clustering coefficient, degree and eigenvector centrality as principal graph theoretical features for each node of graphs to demonstrate alterations in functional connectivity patterns of patients with multiple sclerosis.

Burak Akin¹, Hsu-Lei Lee¹, Nadine Beck¹, Jürgen Hennig¹, and Pierre Levan^1
1Medical Physics, University Medical Center, Freiburg, Germany

Resting-state networks (RSN) are becoming an important tool for the study of brain function. The advent of novel fast fMRI sequences has led to improved sensitivity in the statistical analysis of fMRI data. In this study an ultra-fast imaging technique called MR-encephalography (MREG) is compared with standard EPI. RSN analysis is assessed for both datasets using ICA. The 25-fold increase in sampling rate of MREG relative to conventional fMRI resulted in improved sensitivity and a higher number of components associated with standard RSN in individual subjects. Compared to EPI, MREG might thus greatly improve analyses of intra- and inter-network connectivity.

Hsu-Lei Lee¹, Jakob Assländer¹, Pierre LeVan¹, and Jürgen Hennig^1
1University Medical Center Freiburg, Freiburg, Germany

An fMRI acquisition of 10 Hz sampling rate was achieved by using MR-Encephalography with a three-dimensional single-shot stack of spirals trajectory which provides the possibility to inspect brain network structures in spatial, temporal and frequency domains. In this study we used ICA and partial correlation analysis to construct the brain networks and found hub regions and compare network characteristics in the resting-state functional structure at frequencies as high as 5 Hz.

Shantanu Majumdar^1,2, Feng Wang^1,2, Li Min Chen^1,2, and John C Gore^1,2
1Vanderbilt University Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States, ²Department of Radiology and Radiological Sciences, Vanderbilt University, Nashville, TN, United States

The study of inter-laminar functional connectivity within the primary somatosensory (S1) cortex provides great depth of knowledge about input and output information process in S1 cortex for somatic sensation, but few studies have resolved the contributions of different cortical layers to measured BOLD signals. In this work, we performed sub-millimeter resolution resting-state functional MRI in the S1 cortex in anesthetized squirrel monkeys and examined the laminar specific functional connectivity between sub-regions (area 3a, area 3b and area 1) of the S1 cortex.

Eleonora Fornari¹, Romana Rytsar², and Maria G Knyazeva^3,4
1CIBM, Dept. of Radiology, CHUV, Lausanne, Switzerland, ²Department of Clinical Neuroscience, CHUV, Switzerland, ³Department of Clinical Neuroscience, CHUV, Lausanne, Switzerland, ⁴Department of Radiology, CHUV, Lausanne, Switzerland

In humans, spatial integration develops slowly, continuing through childhood into adolescence. On the assumption that this protracted course depends on the formation of networks with slowly developing top-down connections, we compared effective connectivity in the visual cortex between 13 children (age 7–13) and 14 adults (age 21-42) using a passive perceptual task. The subjects were scanned while viewing bilateral gratings, which either obeyed Gestalt grouping rules (colinear gratings, CG) or violated them (non-colinear gratings, NG). An analysis of effective connectivity showed that top-down modulatory effects generated at an extrastriate level and interhemispheric modulatory effects between primary visual areas (all inhibitory) are significantly weaker in children than in adults, suggesting that the formation of feedback and interhemispheric effective connections continues into adolescence.

Russell W. Chan^1,2, Shu-Juan J. Fan^1,2, Patrick P. Gao^1,2, Iris Y. Zhou^1,2, Adrian Tsang^1,2, and Ed X. Wu^1,2
1Laboratory of Biomedical Imaging and Signal Processing, The University of Hong Kong, Hong Kong, China, ²Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, China

This study investigates the laminar interconnections in the rat primary visual cortex (V1) and visualizes layer specific neuroanatomy using resting-state fMRI (rsfMRI) and manganese-enhanced MRI (MEMRI), respectively. The rsfMRI results indicated that the V1 layers II/III and layers V/VI are more functionally connected during resting-state using seed-based and independent component analyses, which could be associated with the intracortical processing in V1 layers II, III, V and VI during visual stimulation. We also demonstrated layer specific manganese enhancement in the rat V1, revealing the neuroanatomical structure. The laminar rsfMRI connectivity may provide further insights in intracortical and intrahemispheric neural communication.

Madalina E Tivarus¹, Britta Pester², Tong Zhu³, Christoph Schmidt², Thomas Lehmann², Jianhui Zhong¹, Lutz Leistritz², and Giovanni Schifitto^1
1University of Rochester, Rochester, NY, United States, ²Institute of Medical Statistics, Computer Sciences and Documentation, Jena University Hospital, Friedrich Schiller University, Jena, Germany, ³University of Michigan, Ann Arbor, MI, United States

Generalized Partial Directed Coherence (gPDC) was applied on fMRI data to evaluate changes in functional networks and their connectivity in HIV infected individuals treated with lithium for HIV-associated cognitive impairment. GPDC analysis shows that lithium affected functional connectivity and it provides a paradigm to investigate functional and anatomical interrelationships in the context of clinical changes. Therefore, applying gPDC on functional MRI data provides an opportunity to further dissect the functional changes observed in relevant networks affected by the intervention.

Daniele Procissi¹, Kathleen Anne Williams¹, Lakshmi Rajagopal², Yoshihiro Oyamada^2,3, and Herbert Meltzer^2
1Radiology, Northwestern University, Chicago, IL, United States, ²Psychiatry and Behavioral Sciences, Northwestern University, Chicago, Illinois, United States, ³Dainippon Sumitomo Pharma Co., Ltd, Japan

The aim of this work is to use 1H-MRS and fMRI to detect changes in metabolism and connectivity that reflect the behavioral changes and ultimately to evaluate MRI as a potential predictor of successful pharmacological prevention of cognitive impairment in a PCP rat model of schizophrenia .

Akira Matsushita¹, Kousaku Saotome¹, Kei Nakai², Kiyoshi Eguchi², Yoshiyuki Sankai³, and Akira Matsumura^2
1Center for Cybernics Research, University of Tsukuba, Tsukuba, Ibaraki, Japan, ²Faculty of Medicine, University of Tsukuba, Tsukuba, Ibaraki, Japan,³Faculty of Engineering, Information and Systems, University of Tsukuba, Tsukuba, Ibaraki, Japan

Our facility has developed robot-assistive rehabilitation using robot suits, and tested them on people with gait impairment. In this study, we obtained gait performance and rsfMRI before and after rehabilitation to investigate the relationships between rsfMRI and rehabilitation. In the result, the rsfMRI findings in the supplementary motor area, premotor area, orbitofrontal cortex, and lateral prefrontal cortex were related to motor function recovery in rehabilitation. rsfMRI prior to rehabilitation may help to predict the recovery during rehabilitation.

Venkatagiri Krishnamurthy¹, Kaundinya S Gopinath¹, Gregory S Brown², and Benjamin M Hampstead^2,3
1Dept of Radiology and Imaging Sciences, Emory University, Atlanta, GA, United States, ²Dept of Rehabilitation Medicine, Emory University, Atlanta, GA, United States, ³Atlanta VAMC RR&D Center of Excellence in Visual and Neurocognitive Rehabilitation, Decatur, GA, United States

Spatial navigation ability declines in the elderly, and in Alzheimer’s disease. Enhancing navigation skills will result in functional improvement in these populations. Transcranial direct current stimulation (tDCS) can be used to modulate cortical excitability and brain cognition. In this preliminary study, we examined resting state functional connectivity (rsFC) in spatial navigation networks with functional MRI, after tDCS-based excitation of appropriate brain regions. RsFC among a number of areas involved in spatial navigation increased significantly after tDCS. The results can be employed to evolve a framework for evoking plastic reparatory changes in brain networks through tDCS and monitoring them with rsFMRI.

Chandler Sours^1,2, Gad Alon³, Steven Roys¹, and Rao P. Gullapalli^1,4
1Diagnostic Radiology and Nuclear Medicine, University of Maryland School of Medicine, Baltimore, Maryland, United States, ²Magnetic Resonance Research Center (MRRC), Baltimore, Maryland, United States, ³Physical Therapy and Rehabilitation Science, University of Maryland School of Medicine, Maryland, United States, ⁴Magnetic Resonance Research Center (MRRC), Maryland, United States

The effects of transcranial pulsed current stimulation (tPCS) on resting state functional connectivity (rs-FC) within the motor network were investigated. Four fMRI scans were acquired, one motor paradigm and three (PRESTIM, during STIM, POST-STIM) resting state fMRI scans. We demonstrated changes in connectivity patterns that are specific to tPCS including increased thalamo-cortical connectivity during STIM and reduced cerebellar-cortical involvement POST-STM. ROI analysis confirms reduced strength and diversity of the motor network during STIM, and reduced diversity POST-STIM. Our data confirm previous findings using tDCS and strengthen the evidence for the unique neuro-modulatory effects of tPCS.

Hui-Yu Wang¹, You-Yin Chen², Sheng-Huang Lin^3,4, and Jun-Cheng Weng^1,5
1School of Medical Imaging and Radiological Sciences, Chung Shan Medical University, Taichung, Taiwan, ²Department of Biomedical Engineering, National Yang-Ming University, Taipei, Taiwan, ³Department of Neurology, Tzu Chi General Hospital, Tzu Chi University, Hualien, Taiwan, ⁴Institute of Biomedical Engineering, National Taiwan University, Taipei, Taiwan, ⁵Department of Medical Imaging, Chung Shan Medical University Hospital, Taichung, Taiwan

Parkinson’s disease (PD) is a progressive neurodegenerative disorder that is characterized by dopamine depletion in the striatum, and it associated with predominantly motor, cognitive and affective symptoms. The clinical diagnosis of PD is extremely difficult, because the symptom is similar to other central nervous disorder, such as Alzheimer’s disease and Hydrocephalus. The most common diagnostic methods are neurologist’s inquiry and positron emission tomogram (PET) studies. One consistent pathophysiological hallmark of PD is the change in spontaneous oscillatory activity in the basal ganglia thalamocortical networks. Therefore, the goal of our study is to evaluate brain functional connectivity changes using frequency-specific resting-state functional MRI (rs-fMRI) in PD rat and baseline controls using three different seed regions analysis, motor cortex (M1), corpus striatum (CPu) and substantia nigra (SNr), and independent component analysis (ICA). Our results showed a PD-associated decrease in cortico-cortical and cortico-striatal functional connectivity and drops in the power content of cortical and striatal signals. Our results demonstrated that PD modulate cortical and striatal resting state BOLD signal oscillations and cortico-cortical as well as cortico-striatal network correlation.

Mihaela Onu¹, Liviu Badea², and Adina Roceanu^3
1Medical Imaging, Clinical Hospital "Prof. dr. Th. Burghele", Bucharest, Romania, ²National Institute for Research in Informatics, Bucharest, Romania,³University Emergency Hospital Bucharest, Bucharest, Romania, Romania

We hypothesized that, in Parkinson disease (PD), gray matter density and functional cerebral connectivity might develop compensatory behaviors in response to the damaged motor control loops. Using VBM and rsfmri analyses, we found a gray matter thalamic enlargement in parallel with increased connectivity at the basal ganglia/thalamic level which may reflect a compensatory effect in response to the damaged motor control loops.

Yeh-Hsiung Cheng¹, I-Jung Chen², Tzu-Cheng Chao^1,2, and Ming-Long Wu^1,2
1Institute of Medical Informatics, National Cheng Kung University, Tainan, Taiwan, ²Computer Science and Information Engineering, National Cheng Kung University, Tainan, Taiwan

Studies have shown that human brain activities are dynamic and could vary during fMRI experiment. Here, we propose a windowing-based Granger causality analysis for evaluating effective connectivity (EC) in fMRI data called dynamic Granger causality analysis (DGCA). Principal Granger causality patterns obtained from subjects reflect common brain states among subjects while processing the auditory-motor task. Results show that DGCA provides more EC information that could potentially provide more knowledge of dynamic changes in the brain.

Feng-Xian Yan¹, David D. Shin², Chi-Jen Chen¹, Thomas T. Liu², and Ho-Ling Liu^3
1Department of Radiology, Taipei Medical University - Shuang Ho Hospital, New Taipei City, Taiwan, ²Center for Functional MRI and Department of Radiology, University of California, San Diego, La Jolla, CA, United States, ³Department of Medical Imaging and Radiological Sciences, Chang Gung University, Taoyuan, Taiwan

This study aimed to investigate the perfusion-based modular architecture for the bilateral anterior, middle, and posterior cerebral artery (ACA, MCA, and PCA) territories, and default mode network (DMN). Whole-brain CBF maps from 116 healthy subjects, based on arterial spin labeling measurements, were included for the analysis. Brain regions exhibited significantly correlated CBF variations among subjects were identified by comparing each area in the AAL template with the seed region. The results demonstrated that significant short- and long-range connections were found in these territories and DMN. This study provides a preliminary result of intrinsic CBF modularity of the human brain.

Joao M. S. Pereira¹, João Duarte², Miguel Raimundo³, and Miguel Castelo-Branco^4
1Laboratory of Biostatistics, IBILI - Faculty of Medicine, University of Coimbra, Coimbra, Portugal, ²ICNAS - Faculty of Medicine, University of Coimbra, Coimbra, Portugal, ³Clinical School - Faculty of Medicine, University of Coimbra, Coimbra, Portugal, ⁴Visual Sciences Laboratory, IBILI - Faculty of Medicine, University of Coimbra, Coimbra, Portugal

For functional imaging purposes, Arterial Spin Labeling (ASL) has the advantage of measuring activity directly related to neuronal activation with greater localizing sensitivity than the more common BOLD signal analyses. The application of ASL in the fMRI context has, however, been blighted by subpar performances. This work addresses the variability of the latency in the hemodynamic response function (HRF) in both healthy controls and diabetes type II patients, known to have changes in brain vasculature. A method is presented to bypass this limitation in ASL fMRI.

Domenico Zacà¹, Uri Hasson^1,2, Ben Davis¹, Nicola De Pisapia², and Jorge Jovicich^1,2
1Center for Mind/Brain Sciences, University of Trento, Mattarello, TN, Italy, ²Department of Psychology and Cognitive Sciences, University of Trento, TN, Italy

The Amplitude Variance Asymmetry (AVA) of the BOLD resting state signal has been demonstrated to provide reproducible nonrandom patterns of high frequency resting state activity. To investigate its neural, vascular or hemodynamic sources we studied resting state AVA from perfusion fMRI data. Perfusion-derived BOLD AVA patterns replicated previous pure BOLD findings. A significant CBF AVA pattern was also detected with similar topological features but smaller spatial extent than the BOLD AVA. These results suggest a neuronal origin of these transients as CBF measurements are more coupled with metabolism than BOLD measurements.

Ali M Golestani^1,2 and J Jean Chen^1,2
1Rotman Research Institute, Baycrest, Toronto, ON, Canada, ²University of Toronto, Toronto, ON, Canada

End-tidal CO2 (PETCO2) drives the BOLD signal. This effect is not homogeneous across brain regions, and delay maps are estimated using breath holding or other respiratory tasks, which could be uncomfortable for subjects. We investigated regional variability of PETCO2 effect by estimating the response of the resting-state BOLD signal to PETCO2 fluctuations with high temporal resolution (TR = 0.323S). We estimated the PETCO2 response at each voxel and computed the time to peak (TTP) of estimated response for multiple brain region We show that it is possible to assess brain hemodynamics using resting-state BOLD TTP in response to PETCO2.

David C Zhu¹, Takashi Tarumi^2,3, Muhammad Ayaz Khan^2,3, and Rong Zhang^2,3
1Departments of Radiology and Psychology, Michigan State University, East Lansing, MI, United States, ²Institute for Exercise and Environmental Medicine, Texas Health Presbyterian Hospital Dallas, Dallas, TX, United States, ³Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, TX, United States

We used transcranial Doppler ultrasonography, near-infrared (NIR) spectroscopy and resting-state fMRI to demonstrate the presence of high-level coupling between the vascular signal fluctuations driven by cardiac activity and the fluctuations of resting-state fMRI and NIR BOLD signals. Findings from the present study raise a fundamental question of whether the BOLD signals used to assess brain functional connectivity are primarily due to the vascular effects produced from upstream changes in cerebral hemodynamics. The results demonstrate the importance and necessity to develop new methods to uncover the BOLD signal due to spontaneous neuronal activity from the strong vascular signal contamination.

Leon C. Ho^1,2, Peng Cao^1,2, Jevin W Zhang^1,2, Kevin C. Chan^3,4, and Ed X. Wu^1,2
1Laboratory of Biomedical Imaging and Signal Processing, The University of Hong Kong, Pokfulam, Hong Kong, China, ²Department of Electrical and Electronic Engineering, The University of Hong Kong, Pokfulam, Hong Kong, China, ³Neuroimaging Laboratory, Department of Radiology, University of Pittsburgh, Pittsburgh, Pennsylvania, United States, ⁴Department of Ophthalmology, University of Pittsburgh, Pittsburgh, Pennsylvania, United States

Under functional activation, both tissue and vasculature undergo physiological changes (such as cell swelling and vessel dilation). Non-diffusion-weighted BOLD and diffusion-weighted BOLD have been reported to reflect different tissue/vascular contribution when the strength of diffusion attenuation (b-value) is alternated. Different b-values signal fittings and kurtosis models are presented in this study to evaluate non-Gaussian distribution of diffusion displacement in brain tissue during activation. Kurtosis increased and was mainly contributed by the increase in blood flow during activation as indicated by the initial fast decay of diffusion signal.

Zhongwei Chen^1,2, Jing An³, Zhentao Zuo¹, Rong Xue¹, and Danny JJ Wang^4
1State Key Laboratory of Brain and Cognitive Science,Beijing MRI Center for Brain Research, Institute of Biophysics, Chinese Academy of Sciences, Beijing, Beijing, China, ²Graduate School, University of Chinese Academy of Sciences, Beijing, Beijing, China, ³Siemens Shenzhen Magnetic Resonance Ltd, Shenzhen, China, ⁴Department of Neurology, University of California Los Angeles, Los Angeles, United States

Balanced SSFP is a promising fMRI technique due to its reduced susceptibility artifacts and high spatial and temporal resolution compared to GE-EPI. In this study, a novel multi-phase passband cine SSFP technique was introduced for event-related (ER) fMRI with a temporal resolution of 50 ms and spatial resolution of a few mm3. ER-fMRI experiment at 7T demonstrated that this technique can reliably detect the initial dip and can differentiate a time delay of 200ms in stimulus presentation. Multiphase passband SSFP is a promising fMRI technique at ultra high magnetic fields (7 Tesla).

Richard Baheza¹, Nellie Byun², Adam Stark³, and John C. Gore^3,4
1RADIOLOGY AND RADIOLOGICAL SCIENCES, Vanderbilt University Medical Center, NASHVILLE, TN, United States, ²PHARMACOLOGY, Vanderbilt University Medical Center, NASHVILLE, TN, United States, ³VANDERBILT UNIVERSITY INSTITUTE OF IMAGING SCIENCE, Nashville, TN, United States, ⁴BIOMEDICAL ENGINEERING, Vanderbilt University, Nashville, TN, United States

We performed whole brain pharmacological MRI (phMRI) at 9.4T with the FDA approved T1 contrast agent Magnevist using an optimized 3D acquisition sequence to detect amphetamine-induced brain activity in rats. Increasing the psychostimulant dose from 2.0 to 4.0 mg/kg elevated the acquired signal increases from 1.87±0.55% to 4.71±0.33% in the caudate-putamen, demonstrating that the method is sensitive to drug concentration. These data validate the utility of using a T1 shortening agent with 3D acquisition for phMRI. This method has translational potential to clinical drug studies using FDA-approved contrast agents.

Joao Jorge^1,2, Patricia Figueiredo², Rolf Gruetter¹, and Wietske van der Zwaag^1
1Center for Biomedical Imaging, University of Lausanne/Ecole Polytechnique Federale de Lausanne, Lausanne, Switzerland, ²Institute for Systems and Robotics, Instituto Superior Tecnico, Universidade Tecnica de Lisboa, Lisbon, Portugal

During presentation of a stimulus, positive BOLD responses are generally attributed to local increases in neuronal activity. Sustained negative BOLD responses are also frequently observed, but their underlying neurovascular coupling mechanisms are less well understood. Here, we studied the negative BOLD response (both intra- and cross-modal) to contrast-varying visual stimuli. In grey matter, negative responses consistently decreased with increasing stimulus contrast in both visual and auditory regions. Although of larger amplitude, responses observed in draining veins tended to be less contrast-dependent.

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

Our objective was to compare the fMRI calibration parameter (M) quantified with two methods and to establish a mutual scaling factor. Deoxyhemoglobin based calibration was achieved by quantifying R2’ to yield MR2’ after magnetic field inhomogeneity correction. Calibration exploiting the fractional change in cerebral blood flow during hypercapnia was obtained by ASL, providing along with the measured change in BOLD Mhypercapnia via Davis’ model. Preliminary results show comparable calibration M values for the two methods. The quantitative scaling factor might provide potential support for the R2’ calibration method, which is more straightforward to implement.

Ali M Golestani^1,2 and J Jean Chen^1,2
1Rotman Research Institute, Baycrest, Toronto, ON, Canada, ²University of Toronto, Toronto, ON, Canada

End-tidal CO2 (PETCO2) fluctuations constitute a source of physiological noise in the BOLD fMRI signal. In this study we estimate the hemodynamic response of the resting-state BOLD signal to spontaneous PETCO2 fluctuations. Cardiac, respiratory and PETCO2 signals were recorded during resting-state fMRI scan. The response is estimated voxel-wise in 8 subjects. The estimated PETCO2 response has a peak around 9 seconds, which is in accordance with previously reported results. There is some overlap between the regions most affected by RVT and by PETCO2 fluctuations showing possible interaction between the two, but they each modulates the BOLD signal in unique ways.

Sungho Tak¹, Danny J. J. Wang², Jonathan R. Polimeni³, Lirong Yan², and J. Jean Chen^1
1Rotman Research Institute at Baycrest Centre, Toronto, ON, Canada, ²Laboratory of Functional MRI Technology (LOFT), University of California, Los Angeles, CA, United States, ³Athinoula A. Martinos Center for Biomedical Imaging, Harvard Medical School, Charlestown, MA, United States

In this study, we investigated regional associations of resting-state functional connectivity MRI (fcMRI) estimates with cerebral blood flow (CBF)-BOLD coupling, as well as the role of large vessel volume. Based on extensive analyses, we found that functional connectivity strength was significantly proportional to the regional strength in CBF-BOLD coupling, and inversely proportional to large-vessel volume fraction. Our work suggests that despite inherent ambiguity of fcMRI estimates, synchronized activities observed in the functional networks are not likely to be mediated by common vascular drainage linking distal cortical areas, but rather by tighter CBF-BOLD coupling, which might be associated with neuronal connection.

Daniil P Aksenov¹, Limin Li¹, Michael Miller¹, Gheorghe Iordanescu¹, Holden M Faber², Robert A Linsenmeier², and Alice M Wyrwicz^1
1NorthShore University HealthSystem, Evanston, IL, United States, ²Northwestern University, IL, United States

To evaluate the relationship among BOLD fMRI, single units (SU), local field potentials (LFP), and oxygen response we obtained fMRI data from rabbits with chronically implanted electrodes to measure both PO2 and neuronal responses to whisker stimulation in the cortex. Striking differences were observed in cortical BOLD, oxygen, and electrophysiological responses. The shape of oxygen response more closely resembles the shape of SU and LFP, whereas the PO2 response exhibits the post-stimulus offset and subsequent undershoot that characterize the BOLD response. These differences reflect the unique physiological characteristics of each signal.

Tommaso Gili^1,2 and Richard G Wise^1,2
1IRCCS Santa Lucia Foundation, Rome, Italy, ²Cardiff University Brain Research Imaging Centre, Cardiff, Wales, United Kingdom

We propose a method to estimate the level of variation of cerebral metabolic oxygen consumption associated with a visual resting state network. This focuses on extracting relative CBF changes associated with a BOLD-defined visual network. The linked BOLD and CBF variations were estimated using a Bayesian fitting of a non-linear model of dual-echo arterial spin labeling data. Although this relies on the BOLD signal to define the resting state network, the estimates of CMRO2 variation derived from BOLD and CBF estimates appear plausible at around 5%.

Hannah V Hare¹, Nicholas P Blockley¹, Alexander G Gardener¹, Michael A Germuska¹, Stuart Clare¹, and Daniel P Bulte^1
1FMRIB, University of Oxford, Oxford, United Kingdom

Calibrated functional MRI (fMRI) is most often performed at 3T, but there is increasing interest in implementing this method on 7T research and 1.5T clinical systems. It is currently unclear whether this affects the accuracy of the resultant measurements of oxygen metabolism. We investigated the robustness of such measurements by performing the same calibrated fMRI protocol at 1.5, 3 and 7T. The calibration parameter M was found to increase predictably with field strength, whilst the estimated change in oxygen metabolism in response to a motor task was independent of field strength.

Alan J Stone¹, Kevin Murphy¹, and Richard G Wise^1
1CUBRIC, School of Psychology, Cardiff University, Cardiff, United Kingdom

Dual calibrated FMRI (dcFMRI) is an extension of the calibrated BOLD methodology, capable of producing regional measurements of oxygen extraction fraction (OEF) across the brain. 6 normal healthy participants were scanned using a hypocapnic challenge to demonstrate the sensitivity of the dcFMRI technique to detect increases in OEF associated with reductions in CBF. The data acquired during normocapnia (baseline CBF) and hypocapnia (lowered CBF) shows a clear decrease in CBF and increase in OEF with hypocapnia. This demonstrates the sensitivity of dcFMRI suggesting the technique is appropriate for application to vascular dysfunction in which flow and metabolism may be impaired.

Alan J Stone¹, Kevin Murphy¹, Ashley D Harris^2,3, and Richard G Wise^1
1CUBRIC, School of Psychology, Cardiff University, Cardiff, United Kingdom, ²Russell H. Morgan Department of Radiology and Radiological Science, The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States, ³F.M. Kirby Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, Maryland, United States

To aid the implementation of dcFMRI in clinical and research scan protocols, it is necessary to produce reliable measures of absolute CMRO ₂ in a short scan time. However, reducing scan time must be done carefully and it is important to establish that the parameter measurements are not compromised. Here we investigate reducing scan times of an interleaved and simultaneous hypercapnic-hyperoxic dcFMRI acquisition using two analytical approaches. A “varying window duration” analysis is used to investigate if periods of respiratory challenges can be shortened and a “data-point resampling” analysis is used to investigate if fewer respiratory challenges can be performed.

Marta Bianciardi¹, Nicola Toschi^1,2, Jonathan R Polimeni¹, Himanshu Bhat³, Bruce R Rosen¹, David A Boas¹, and Lawrence L Wald^1
1Department of Radiology, A.A. Martinos Center for Biomedical Imaging, MGH, Harvard Medical School, Boston, MA, United States, ²Department of Medicine, University of Rome “Tor Vergata”, Rome, Italy, ³Siemens Medical Solutions, Boston, MA, United States

The aim of this work was to develop novel MRI-indicators that evaluate cerebrovascular compliance through direct measures of blood volume changes. We employed time-of-flight EPI-MRI, and constrained the sequence parameters to produce MRI-contrast dependent primarily on blood volume changes. A pulsatility-Volume-Index (pVI) was defined as the ratio of the signal during systole divided by the signal during diastole integrated within the vascular tree. pV1 was on average ~1.5 in larger cerebral arteries, and varied within each arterial segment and across arteries. pVI seems to be a promising MRI-indicator for the evaluation of cerebrovascular compliance in a highly localized and subject-specific-manner.

Richard Watts¹, Scott Hipko¹, Jay Gonyea¹, and Trevor Andrews^1,2
1Department of Radiology, University of Vermont College of Medicine, Burlington, VT, United States, ²Philips Healthcare, Cleveland, OH, United States

Functional brain imaging using T1ρ- and T2-weighted imaging provide qualitatively similar maps, but with some regions of significant difference, perhaps due to differences in the contrast mechanism.

Hye-Young Heo¹, Casey P Johnson¹, Daniel R Thedens¹, John A Wemmie^2,3, and Vincent A Magnotta^1,3
1Department of Radiology, University of Iowa, Iowa City, Iowa, United States, ²Department of Neurosurgery, University of Iowa, Iowa, United States,³Department of Psychiatry, University of Iowa, Iowa, United States

Recent experiments suggest that functional imaging of T1 relaxation in the rotating frame (T1ρ) can detect localized metabolic changes in the human visual cortex induced by a flashing checkerboard task. Possible sources of the functional T1ρ signal include changes in pH, glucose and glutamate concentrations, and cerebral blood volume. In this study we explored the relationship between the functional T1ρ signal and cerebral blood volume by employing an inferior saturation pulse. The results show that, although there is a contribution of cerebral blood volume to the functional T1ρ signal, a majority of the signal likely comes from the tissue compartment. Therefore, using spatial saturation pulses is an effective and efficient means to minimize blood volume contributions to the functional T1ρ signal.

Zahra Shams¹ and Abbas N Moghaddam^1,2
1BME, Tehran Polytechnic, Tehran, Tehran, Iran, ²School of Cognitive Sciences, Institute for Studies in Theoretical Physics and Mathematics, Tehran, Iran

In this work, we presented a new method for fMRI, termed Averaged BOSS (A-BOSS), in which the idea of phase transition is employed without the undesirable limited spatial coverage. The implementation of the proposed method involved compacting SSFP profile with a profile period at the order of pixel size, resulted in averaging the effect of frequency shift on the magnetization everywhere with no need for careful shimming. The analysis of experimental results highlighted the area of activity that was in accordance with BOLD activation map. In conclusion, A-BOSS overcomes BOSS limitations, creating considerably high signal level in comparison with BOLD.

Simon Daniel Robinson¹, Florian Fischmeister^1,2, Günther Grabner¹, Moritz Wurnig^1,2, Jakob Rath^1,2, Thomas Foki^1,2, Eva Matt^1,2, Siegfried Trattnig¹, Roland Beisteiner^1,2, and Alexander Geissler^1,2
1Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Vienna, Austria, ²Department of Neurology, Medical University of Vienna, Vienna, Austria

The specificity of the BOLD response to microvascular rather than draining vein signal is understood to increase with field strength. This understanding is based on image SNR and relaxation rates, however. In the light of recent studies showing less than linear increases of time-series SNR and activation statistics with field strength we re-examined the specificity of the BOLD response in high resolution fMRI at 3T and 7T with 12 subjects who performed a hand task. fMRI data were analysed with ICA and compared with 7T SWI. There was a significant increase in sensitivity but not specificity with field strength.

Jennifer Evans¹, Prantik Kundu¹, Silvina Horovitz², and Peter Bandettini^1
1SFIM/NIMH, NIH, Bethesda, Maryland, United States, ²NIH, Bethesda, United States

It’s known that fMRI time series tend to slowly drift, and removing drifts using linear regression also removes slow BOLD changes that might be occurring. Conventional single echo fMRI cannot separate non-BOLD based signal drifts from neurally related changes in BOLD. In this study, we demonstrate that multi-echo independent components analysis (MEICA) (Kundu, 2012) of multi-echo fMRI data can separate these two mixed low frequency signals and show very slow BOLD changes in the visual cortex from visual stimulation with slowly varying contrast and in resting state.

Cornelius Eichner^1,2, Berkin Bilgic¹, Marta Bianciardi¹, Jonathan Polimeni¹, Robert Turner², Jenni Schulz³, David G Norris³, Lawrence L Wald¹, and Kawin Setsompop^1
1Martinos Center for Biomedical Imaging, Charlestown, MA, United States, ²Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Saxony, Germany, ³Donders Centre for Cognitive Neuroimaging, Nijmegen, Netherlands

Acquisition of Multi-echo (ME) EPI has recently been shown to be a robust and reliable method for extracting valuable information from fMRI data. In this work, we combine a blipped CAIPI Simultaneous Multiple Slice (SMS) of Multi-echo EPI acquisition with a robust pipeline for reconstruction of time resolved phase and QSM data. Magnitude, Phase and QSM contrasts, which were acquired in one single shot, are shown for multiple echo times. In the future, this type of data can help to determine physiological underpinnings of BOLD activation.