Anders Garpebring¹, Ronnie Wirestam², and Mikael Karlsson^1
1Radiation Sciences, Umeå University, Umeå, Sweden, ²Medical Radiation Physics, Lund University, Lund, Sweden

Phase sensitive MRI has shown great potential for quantification of the Arterial Input Function (AIF). Phase based AIFs are sensitive to motion and B₀ drift, and a background phase curve is thus normally subtracted from the AIF phase. However, this procedure can be complicated by accumulation of contrast agent (CA) in the background tissue region. The purpose of this study was to assess the importance of CA in the background ROI and to develop a method to compensate for this effect. The results showed that CA in the background was a significant problem and that the developed compensation was beneficial.

Kishor Karki¹, Ramesh Paudyal¹, Tavarekere N. Nagaraja², James R. Ewing^1,3, Joseph D. Fenstermacher², and Robert A. Knight^1,3
1Department of Neurology, Henry Ford Hospital, Detroit, Michigan, United States, ²Department of Anesthesiology, Henry Ford Hospital, Detroit, Michigan, ³Department of Physics, Oakland University, Rochester, Michigan, United States

The present study investigates the possibility of using an average arterial input function (AIF) obtained from direct blood sampling experiments to estimate blood–brain barrier (BBB) permeability or blood-to-brain transfer constant (K^trans) of Gd-DTPA for a step-down infusion (SDI) procedure instead of using the individual MRI derived AIF. Enhanced spatial resolution of areas with BBB opening was generally observed with the SDI technique. The K^trans values from both types of AIF’s agreed closely and were highly correlated, suggesting that an averaged AIF for a given contrast agent obtained from a standard technique can be used to assess vascular permeability.

Ross A Little¹, Marietta Scott², Anita Banerji¹, Yvonne Watson¹, Josephine Naish¹, and Geoff JM Parker^1
1Imaging Sciences & Biomedical Engineering, University of Manchester, Manchester, United Kingdom, ²AstraZeneca, Cheshire, United Kingdom

An arterial input function (AIF) model is presented that allows the effects of AIF dispersion to be characterised. This facilitates modelling of dispersion between the measurement point of an AIF and the tissue voxel of interest, providing an improved fit and a measure of AIF dispersion across the tissue. Simulations and application to clinical data show the applicability of the dispersed AIF method within the DCE-MRI analysis pipeline. They show that not accommodating dispersal leads to errors on the tissue microvascular parameters.

Claire Foottit¹, Greg O. Cron¹, Jean Francois Mercier¹, Viviane Thanh-Van Nguyen², Ian Cameron¹, Mark E Schweitzer¹, John Sinclair¹, John Woulfe¹, Matthew J Hogan³, and Thanh B Nguyen^1
1The Ottawa Hospital, Ottawa, Ontario, Canada, ²University of Montreal, ³Neuroradiology, The University of Ottawa, Ottawa, Ontario, Canada

In DCE-MRI studies of human brain tumors, a suitable artery for the arterial input function (AIF) is often not available or is insufficiently large. The venous output function (VOF) in the superior sagittal sinus is often used as a surrogate AIF. However, to our knowledge no DCE-MRI study has ever validated this practice for high-temporal-resolution data. In this work, we acquired phase-derived AIFs and VOFs simultaneously during DCE-MRI studies, with the hypothesis that tumor tracer kinetic parameters would be equivalent using either vascular function. The data support the use of a VOF in place of the AIF.

Shy-Chyi Chin¹, Yeng-Peng Liao², Ya-Ting Chuang¹, and Ho-Ling Liu^2,3
1Department of medical imaging and intervention, Chang-Gung Medical Center, Guei-Shan, Tao-Yuan, Taiwan, ²Chang Gung University, Department of Medical Imaging and Radiological Sciences, Guei-Shan, Tao-Yuan, Taiwan, ³Department of Medical Imaging and Intervention, Chang-Gung Medical Center

A new trend is to utilize MRI intra-operatively to assess the sufficiency of brain tumor resection, not just for better navigation. Both dynamic contrast-enhanced (DCE) and dynamic susceptibility-contrast (DSC) techniques are applied to patients after surgical resection. Our study, distinct from previous works biased toward DSC-MRI, has concluded that (1) no discrepancy between the normalized V_P from DCE-MRI and normalized CBV from DSC-MRI and (2) K^trans map from DCE-MRI seemingly more precise in denoting the permeability and extent of impaired brain-blood-barrier than K2 map from DSC-MRI.

Chih-Feng Chen¹, Lin-Wei Hsu², and Ho-Lin Liu^2
1Department of Radiology, Chang Gung Memorial Hospital, Chiayi, Taiwan, Taiwan, ²Department of Medical Imaging and Radiological Sciences Institute of Medical Physics and Imaging Sci, Chang Gung University, Taoyuan, Taiwan, Taiwan

The most common estimated parameters in DCE-MRI are Ktrans and Ve. The process of Ktrans estimation is not easy. but the IAUC is more feasible. In this study, we aim to further investigate the application of mIAUCktrans in clinical and find the correlation between mIAUCktrans and Ktrans. With the correlation coefficients of 0.95, mIAUCktrans was more correlated with Ktrans. Since the high correlation between Ktrans and mIAUCktrans was demonstrated in this study, it reveals that mIAUCktrans could be an alternative for physiological condition evaluation in DCE-MRI.

Sanjeev Chawla¹, Sungheon Kim^1,2, Larry Dougherty¹, Sumei Wang¹, Laurie A Loevner¹, Harry Quon³, and Harish Poptani^1
1Radiology, University of Pennsylvania, Philadelphia, PA, United States, ²Radiology, New York University, New York, NY, United States, ³Radiation Oncology, University of Pennsylvania, Philadelphia, PA, United States

Twenty-eight patients with squamous cell carcinomas of head and neck underwent DCE- MRI with a radial acquisition and dynamic K-space weighted image reconstruction contrast technique and diffusion weighted imaging. Motion correction algorithms were applied to further reduce motion artifacts. Strong positive correlations in Ktrans (r=0.684, p<0.001) and ADC (r=0.407, p=0.031) from primary tumors and metastatic nodes were observed. These findings indicate that both lesions have similar physiological characteristics. Good quality Ktrans and ADC maps from primary tumors, located in regions prone to physiological motion, is feasible and may help in assessing these tumors for therapeutic response using these imaging markers.

Bingsheng Huang¹, Pek Lan Khong¹, Chung-sing Wong¹, Dora Lai Wan Kwong², and Queenie Chan^3
1The University of Hong Kong, Hong Kong Island, Hong Kong SAR, Hong Kong, ²Clinical Oncology, The University of Hong Kong, ³Philips Healthcare

Dynamic Contrast-enhanced MRI (DCE-MRI) is useful for estimating the tumor malignancy because the contrast enhancement pattern reflects the perfusion and permeability of the tumor as well as the tumor angiogenesis. In this study we investigated the relationship between semi-quantitative contrast enhancement parameters in dynamic DCE-MRI and the diagnosing staging of nasopharyngeal carcinoma, and the correlation between these parameters and PET-CT scanning. Our results showed that some parameters from semi-quantitative analysis of DCE-MRI significantly correlate with the T stages of NPC tumors, but not with PET-CT parameter.

Bin Chen¹, Yi Dang¹, Xue Dong Yang², Jing Fang^1,3, Xiaoying Wang^1,2, and Jue Zhang^1,3
1Academy for Advanced Interdisciplinary Studies, Peking University, BEIJING, BEIJING, China, People's Republic of, ²Radiology, Peking University First Hospital, BEIJING, BEIJING, China, People's Republic of, ³College of Engineering, Peking University, BEIJING, BEIJING, China, People's Republic of

MR Renography was widely used to provide renal functional information in recent studies. However, conventional 3-D whole kidney scan takes more than ten minutes to obtain MRR curves in diffuse renal disease. The purpose in this study was to investigate a higher temporal resolution scan protocol to reflect renal functional information. The center thin slab scan employed, instead of the whole kidney scan protocol, was proved feasible for reflecting whole kidney perfusion information, and it also significantly improved the temporal resolution of MRR.

Liyong Chen^1,2, Ganesh Adluru¹, Matthias C. Schabel¹, Christopher J. McGann³, and Edward V.R. DiBella^1,2
1Utah Center for Advanced Imaging Research, Department of Radiology, University of Utah, Salt Lake City, Utah, United States, ²Department of Bioengineering, University of Utah, Salt Lake City, Utah, United States, ³Division of Cardiology & Radiology, University of Utah, Salt Lake City, Utah, United States

3D hybrid radial pulse sequence was applied dynamic contrast enhanced myocardial perfusion MR imaging.

Anthony Glenn Christodoulou¹, Cornelius Brinegar¹, Bo Zhao¹, Justin P Haldar¹, Haosen Zhang², Yi-Jen L Wu², T. Kevin Hitchens², Chien Ho², and Zhi-Pei Liang^1
1Department of Electrical and Computer Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ²Pittsburgh NMR Center for Biomedical Research, Department of Biological Sciences, Carnegie Mellon University, Pittsburgh, PA, United States

Myocardial perfusion imaging is an important and challenging application of cardiovascular MRI. This work demonstrates that sparse sampling of (k,t)-space with the joint use of partial and spatial-spectral sparsity constraints can significantly improve the spatiotemporal resolution of first-pass myocardial perfusion imaging experiments. Experimental results in rats show a 390 μm in-plane spatial resolution and 15 ms temporal resolution, representing an equivalent acceleration factor of 51.

Benjamin Leporq¹, Sorina Camarasu¹, Frank Pilleul^1,2, and Olivier Beuf^1
1CREATIS, CNRS UMR 5220, Inserm U1044, INSA-Lyon, Université Lyon 1, Villeurbanne, France, ²Département d'imagerie digestive, CHU Edouard Herriot, Hospices Civils de Lyon, Lyon, France

Liver fibrosis is an important cause of mortality and morbidity in patients with chronic liver diseases and can lead to cirrhosis which the complications involve 15,000 deaths per year in France. An early detection and a clinical follow-up of liver fibrosis are still needed. The histology after liver biopsy is the gold standard but has inherent risk, interobserver variability and sampling errors. It has been demonstrated that liver perfusion imaging has the potential to detect and assess vascular modifications associated with liver fibrosis. However, these methods are only ROI-based and regional variations often met in diffuses liver diseases couldn’t be observed. This work presents a post-processing method using EGI (European Grid Initiative) for parallel processing to allow 3D liver parametric mapping with a reasonable processing time. Acquisition is founded on MR-DCE imaging technique after injection of the MS-325 blood pool agent.

Si-Wa Chan¹, Yi-Jui Liu^2,3, Dah-Cherng Yeh⁴, Jeon-Hor Chen⁵, Fang-Yi Lee⁶, Huei-Jen Hsueh⁴, Kuo-Fang Shao³, and Hsiao-Wei Peng^1
1Department of Radiology, Taichung Veterans General Hospital, Taichung, Taiwan, Taiwan, ²Department of Automatic Control Engineering, Feng-Chia University, Taichung, Taiwan, Taiwan,³Master's Program in Biomedical Informatics and Biomedical Engineering, Feng-Chia University, Taichung, Taiwan, Taiwan, ⁴Division of General Surgery, Taichung Veterans General Hospital, Taichung, Taiwan, Taiwan, ⁵Center for Functional Onco Imaging, University of California, Irvine, CA, United States, ⁶Department of Physicain, Taichung Veterans General Hospital, Taichung, Taiwan, Taiwan

Dynamic contrast-enhanced MR (DCE-MRI) is able to distinguish malignancy from benign tissue by recognizing the differences in contrast enhancement that originates from their differences in microcirculation. DCE-MRI also evaluates some of the functional images, such as tissue perfusion and permeability of tumor vessels. Therefore, this ability can be used as a functional image to monitor the pathophysiological responsiveness to treatment. In this work, two-compartment model proposed by Buckley et al. was applied in DCE-MRI to monitor the sequential chemotherapy response of patient with LABC. We aim to find the perfusion parameters which are sensitive to chemotherapy response.

Jack T Skinner^1,2, Mary E Loveless^1,2, Todd E Peterson^2,3, Thomas E Yankeelov^2,3, and Mark D Does^1,2
1Biomedical Engineering, Vanderbilt University, Nashville, TN, United States, ²Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States, ³Radiology and Radiological Sciences, Vanderbilt University, Nashville, TN, United States

DCE-MRI characterizes tissue in terms of physical properties, such as the extracellular/extravascular volume fraction (v_e), by indirectly measuring the kinetics of a contrast agent into and out of the interstitial space. The fitted model parameters, including v_e, are potentially biased by confounding water dynamics, which are not accounted for in the model. In contrast to MRI, radiotracer imaging with SPECT is insensitive to water dynamics. Estimates of v_e were obtained using DCE-MRI and dual-isotope SPECT imaging in a rat glioma model and the results showed a large and consistent overestimation by DCE-MRI.

Martin Büchert¹, and Klaus Mross^2
1MRDAC Magnetic Resonance Development and Application Center, University Medical Center Freiburg, Freiburg, Germany, ²Klinik für Tumorbiologie

For assessing treatment response to cancer therapeutics dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) is a valuable tool. For DCE-MRI a 3D-Flash and an IR-TrueFISP protocol providing different spatial and temporal resolution were investigated in phantom and 12 tumor patients. Data of both methods were in good agreement with each other. IR-TrueFISP data showed a much lower variability compared to 3D-Flash results, leading to a three times higher average fit error in the model fit of the later. It was shown that the IR-TrueFISP protocol with its high temporal resolution and good accuracy is a suitable DCE-MRI acquisition method.

Abbas Babajani-Feremi¹, Rajan Jain^1,2, Jayant Narang¹, Ali Syed Arbab¹, Kourosh Jafari-Khouzani¹, Mohammad-Reza Nazem-Zadeh³, and Hamid Soltanian-Zadeh^1,4
1Department of Radiology, Henry Ford Hospital, Detroit, Michigan, United States, ²Department of Neurosurgery, Henry Ford Hospital, Detroit, Michigan, United States, ³Department of Neurology, Henry Ford Hospital, Detroit, Michigan, United States, ⁴CIPCE, Electrical and Computer Engineering Department, University of Tehran, Tehran, Tehran, Iran

We propose a method to find the best Patlak model that appropriately represents vascular permeability of tumor using dynamic contrast-enhanced T1-weighted magnetic resonance perfusion (DCET1MRP). Three Patlak models were developed . Model 1 estimates only vascular plasma volume (VP). Model 2 estimates VP and influx transfer constant (Ki). Model 3 estimates VP, Ki, and reverse transfer constant (Kb). Using least-square method, three models were fitted to dynamic data. F-statistic was used for model comparison. Based on analysis of 31 patients with brain tumors, we conclude that that F-statistic can be used to choose appropriate model for tumor and non-tumor regions.

Kenneth K Kwong¹, Ona Wu¹, Suk-tak Chan¹, Koen Nelissen¹, and David A Chesler^1
1Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, Massachusetts, United States

In MR perfusion analysis, the tissue mean transit time τ is normally estimated indirectly by using the ratio of regional cerebral blood volume (rCBV) and regional cerebral blood flow (rCBF). We propose two different methods to estimate τ directly using signals from the tissue contrast agent bolus concentration time course.

Dana Suciu Poole¹, Johannes Rolf Sikkema², Arnoldus M van den Maagdenberg³, and Louise van der Weerd^2,4
1Radiology, Leiden University Medical Centre, Leiden, Netherlands, ²Radiology, Leiden University Medical Centre, Netherlands, ³Human Genetics, Leiden University Medical Centre, Netherlands,⁴Anatomy and Embriology, Leiden University Medical Centre, Netherlands

Dynamic contrast-enhanced magnetic resonance imaging typically uses intravenous administration of a contrast agent. In mice, however, repeated intravenous injections often cause scarring, and insertion of permanent canulae in the tail or femoral vein require an invasive procedure and are prone to infection and damage by grooming. In this study, we have developed an alternative longitudinal DCE-MRI protocol at 9.4T using intraperitoneal administration of gadolinium-DOTA. We show that Patlak plots can be calculated following intraperitoneal administration of contrast agent on mice with unilateral blood-brain barrier impairment.

Ram Kishore Singh Rathore¹, PRATIVA sahoo², rishi awashti³, rakesh k gupta⁴, SANJAY verma⁵, and divya Rathore^5
1Mathematics and Statistics, IIT Kanpur, KANPUR, U.P., India, ²Mathematics and Statistics, IIT Kanpur, KANPUR, India, ³SGPGI, ⁴SGPGI, LUCKNOW, ⁵IIT Kanpur

A modified generalized tracer kinetic model incorporating a strong leakage compartment, presenting as a local sink to BBB, is proposed. The new model as such automatically takes care of the cerebral blood volume correction and avoids choosing a neighboring feeding artery. The additional term in the model in the absence of the leakage compartment would give insignificant values to the corresponding rate constant and therefore does not cause any significant change in the parameters in such cases. The results, however, show that such compartments occur in the case of high grade gliomas, as compared with the low grade ones.

Shy-Chyi Chin¹, Yeng-Peng Liao², Ya-Ting Chuang¹, and Ho-Ling Liu^2
1Department of medical imaging and intervention, Chang-Gung Medical Center, Guei-Shan, Tao-Yuan, Taiwan, ²Chang Gung University, Department of Medical Imaging and Radiological Sciences, Guei-Shan, Tao-Yuan, Taiwan

We concluded that the ideal application of the dynamic contrast-enhanced MRI (DCE-MRI) technique can be achieved in the clinical oncology even in patients with suboptimal scanning condition like oral cavity malignancy with inevitable head motion and image distortion. Most of the imaging distortion can be overcome by using the alternate DCE-MRI method, that is, reference region model (RRM) to substitute with the conventional general kinetic model (GKM). On the other hand, DCE-MRI derived parameters are not correlative to pathological grading, though. The more sophisticated analysis between these parameters deserves further investigation.

Philipp Krämer¹, Simon Konstandin¹, Melanie Heilmann¹, and Lothar Rudi Schad^1
1Computer Assisted Clinical Medicine, Heidelberg University, Mannheim, Germany

A new sequence is presented which uses the advantages of a Twisted Projection Imaging trajectory design to reduce measurement time in radial proton imaging. The SNR of this sequence is higher when compared to standard 3D radial imaging. It is shown that this sequence can be used for fast and accurate T1 measurement with a Variable Flip Angle approach. Especially in DCE-MRI this sequence has advantages for precontrast as well as dynamic measurement. It could be used where conventional radial approaches to DCE-MRI suffer from lower SNR and undersampling artifacts.

Ethan K. Brodsky^1,2, Kevin M. Johnson², Walter F. Block^2,3, and Scott B. Reeder^1,2
1Radiology, University of Wisconsin, Madison, WI, United States, ²Medical Physics, University of Wisconsin, Madison, WI, United States, ³Biomedical Engineering, University of Wisconsin, Madison, WI, United States

Detection, characterization and monitoring of hepatocellular carcinomas (HCC) in cirrhotic patients is challenging due to their variable and rapid arterial enhancement. Early efforts to supplant conventional multiphase CE-MRI with a time-resolved undersampled multi-echo 3DPR technique have shown promise, but require a wide temporal footprint to achieve adequate image quality, compromising temporal fidelity of small lesions enhancement and complicating quantitative modeling. In this work, we discuss improved k-space sampling trajectories and the incorporation of an iterative SENSE-based constrained reconstruction algorithm that lead to substantially improved temporal resolution and image quality. Consistency weighting allows for efficient use of data from multiple echo times.

Haoyu Wang¹, Da Wang¹, Shanglian Bao¹, and Jiani Hu^2
1Beijing Key Lab of Medical Physics and Engineering, Peking University, Beijing, China, People's Republic of, ²Department of Radiology, Wayne State University, Detroit, MI, United States

The purpose of this study is to investigate the feasibility of applying CS theory to breast DCE-MRI to improve the temporal resolution while faithfully reconstructing uptake curves. In the preliminary experiment, DCE images were reconstructed using Reference Image based Compressed Sensing technique (RICS). The uptake curve reconstruction experiment shows a high correlation between uptake curves reconstructed from fully sampled data by Fourier transform and from undersampled data by RICS, indicating high similarity between them. In conclusion, our preliminary results demonstrate the feasibility of RICS for improving temporal resolution and faithfully reconstructing uptake curves of DCE-MRI.

Lars Gerigk¹, Hendrik Laue², Lydia Schuster¹, Thomas Hauser¹, Ann-Kathrin Homagk³, Armin Nagel³, Marco Essig¹, Heinz-Peter Schlemmer¹, and Michael Bock^3
1Radiology, German Cancer Research Center, Heidelberg, Baden-Württemberg, Germany, ²Institute for Medical Image Computing, Fraunhofer MEVIS, Bremen, Bremen, Germany, ³Medical Physics in Radiology, German Cancer Research Center, Heidelberg, Baden-Württemberg, Germany

Since vascularization is essential in tumors, analyzing the amount and distribution of tissue perfusion is important to understand tumor growth and to monitor therapies. At lower field strengths, DCE-MRI has been used successfully to quantify permeability and flow, but has been limited in spatial resolution. To make use of the higher signal strength, we adapted this technique to a 7 T whole body system and applied it to patients with brain tumors. Due to the high spatial resolution at 7 T, tumor heterogeneity can be visualized.

Melanie Freed^1,2, Jacco A de Zwart³, Prasanna Hariharan⁴, Matthew R Myers⁴, and Aldo Badano^1
1CDRH/OSEL/DIAM, Food and Drug Administration, Silver Spring, MD, United States, ²Dept. Bioengineering, University of Maryland, College Park, MD, United States, ³Advanced MRI Section, LFMI, NINDS, National Institutes of Health, Bethesda, MD, United States, ⁴CDRH/OSEL/DSFM, Food and Drug Administration, Silver Spring, MD, United States

There is a need for realistic and well-characterized dynamic phantoms for dynamic contrast-enhanced MRI of the breast to aid in optimization and standardization of imaging protocols. We have developed a dynamic lesion model that mimics the border shape and washout curve behavior of benign and malignant breast lesions. Measurements of the true contrast agent concentration versus time were acquired with an x-ray system and compared with MRI measurements using different spatial and temporal resolutions.

Yanming Yu¹, Quan Jiang², Haoyu Wang³, Shanglian Bao³, E. Mark Haacke⁴, and Jiani Hu^4
1Logging technique research institute, great wall drilling company, China National Petroleum Corporat, Beijing, China, People's Republic of, ²Department of Neurology, Henry Ford Health Sciences Center, Detroit, MI, United States, ³Beijing Key Lab of Medical Physics and Engineering, Peking University, Beijing, China, People's Republic of, ⁴Department of Radiology, Wayne State University, Detroit, MI, United States

The purpose of this project is to investigate the effect of T2^* on permeability and effect of T1 on perfusion using a new pharmacokinetic model with simultaneously acquired T1 and T2^* DCE-MRI data. The preliminary results demonstrate that 1) T2^* has great effects on the estimation of permeability parameters; 2) T1 has great effects on the estimation of perfusion parameters; 3) both the T1 and T2^* effects are much greater in tumors when compared to healthy brain tissue. In conclusion, simultaneously acquired T1- and T2^*-weighted DCE-MRI data can be used to improve the estimation accuracy of permeability and perfusion parameters.

Ramesh Paudyal¹, Hassan Bagher-Ebadian¹, Robert A. Knight^1,2, Tavarekere N Nagaraja³, Joseph D Fenstermacher³, and James R Ewing^1,2
1Neurology, Henry Ford Hospital, Detroit, MI, United States, ²Physcis, Oakland University, Rochester, MI, United States, ³Anesthesiology, Henry Ford Hospital, Detroit, MI, United States

In this study, we evaluate the effects of equilibrium intercompartmental water exchange kinetics in a monoexponential estimate of tissue water protons relaxation rate R1 using an experimentally measured plasma concentration vs. time, an arterial input function (AIF), of a paramagnetic contrast agent like Gd-DTPA via a Look-Locker measurement in a three site two exchange [3S2X] model. The observable change in R1 was used in a Patlak graphical plot to asses the influence of protons of water exchange on the measured cerebrovascular parameters.

Aiming Lu¹, and Keith R Thulborn^1
1Center for MR Research, University of Illinois, Chicago, IL, United States

DCE-MRI has been applied to characterize tissue parameters such as cerebral blood volume (CBV) and permeability. Although whole brain imaging with high temporal and spatial resolution is desired, current acquisition techniques often compromise either resolution or coverage. Meanwhile, fitting the tracer models requires a fast and robust method for mapping tissue T1s prior to contrast injection. The multiple-echo 3D projection reconstruction sequence was adopted here to image the entire brain with high temporal and spatial resolution. Fast T1 mapping was achieved by linearly fitting the signals collected at different flip angles. The propose approach were demonstrated in brain tumor patients.

Magne Mørk Kleppestø^1,2, Oliver Marcel Geier¹, Christopher Larsson¹, Frederic Courivaud¹, Raimo Aleksi Salo¹, Petter Brandal³, Inge Andre Rasmussen¹, and Atle Bjornerud^1,4
1Interventional Centre, Oslo University Hospital, Oslo, Norway, ²Dept. of Physics, Univ. of Oslo, Oslo, Norway, ³Dept. of Oncology, Oslo University Hospital, Oslo, Norway, ⁴Dept. of Physics, Univ. of Oslo

T2*-effects in blood from a single bolus injection of gadolinium based contrast agents can be significant even when short TE heavily T1-weighted sequences are used. We investigated the effect of T2*-shortening on the measured permeability related parameters obtained from standard two-compartment modeling with vascular deconvolution by applying a double echo 3D-saturation recovery (SR) DCE sequence with a TR of 3.3 sec. Both the transfer constant (Ktrans) and plasma volume (Vp) were found to be significantly over-estimated when T2*-effects were not considered in the dose-response calculation and the error was largest for Vp

Aaron Oliver-Taylor¹, Roger J Ordidge¹, and David L Thomas^2
1Medical Physics and Bioengineering, University College London, London, England, United Kingdom, ²Institute of Neurology, University College London, London, England, United Kingdom

Vessel Selective Arterial Spin Labelling allows assessment and quantification of vessel specific perfusion territories. We make use of Parallel Transmission in order to create 'label' and 'no label' regions to selectively label arteries within the neck. An in-silico simulation was performed to assess the specificity of the method in comparison to using a single surface coil when attempting to only label the right side arteries in the human neck. Results show that for an increasing number of coils contralateral labelling of the left side arteries is reduced whilst maintaining high labelling efficiency on the right.

Daehyun Yoon¹, Hesamoddin Jahanian², Douglas C Noll², and Luis Hernandez-Garcia^2
1Electrical Engineering, University of Michigan, Ann Arbor, MI, United States, ²Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States

We present a novel approach with parallel excitation for spatially selective pseudo CASL to perform vessel selective arterial spin labeling (ASL). Previously, several methods have been proposed to perform vessel selective ASL, but they are either vulnerable to motion due to acquiring multiple images or limited in selectivity. We use multiple RF transmission coils to tag only the desired vessel and to leave the other vessels untagged. Our approach provides superior spatial selectivity and high inversion efficiency, making it an efficient tool for vessel selective ASL.

Michael Helle¹, Susanne Rüfer¹, Matthias van Osch², Olav Jansen¹, and David Gordon Norris^3,4
1Institute for Neuroradiology, Christian-Albrechts-Universität, UK-SH, Kiel, Germany, ²C.J. Gorter Center for high field MRI, Department of Radiology, Leiden University Medical Center, Leiden, Netherlands, ³Donders Institute for Brain, Cognition and Behaviour, Radboud University Nijmegen, Nijmegen, Netherlands, ⁴Erwin L. Hahn Institute for Magnetic Resonance Imaging, Essen, Germany

Superselective pseudo-continuous arterial spin labeling has been recently introduced for perfusion territory imaging. The method offers the possibility to change the size of the labeling spot, thus to adapt to individual arteries. However, this approach requires separate measurements for individual arteries. Here, we propose a new labeling scheme for superselective ASL that utilizes an elliptical labeling spot in order to label the blood in multiple vessels simultaneously. This results in decreased scan time, less artifacts and potentially better labeling efficiency.

Nolan S Hartkamp¹, M Helle², J Hendrikse¹, and M J P van Osch^3
1Radiology, UMC Utrecht, Utrecht, Netherlands, ²Institute of Neuroradiology, Christian-Albrechts-Universität, Kiel, Germany, ³C.J. Gorter Center, Leiden UMC, Leiden, Netherlands

A previous research group proposed there was little evidence of mixing of blood in the basilar artery as they observed separate perfusion territories for both vertebral arteries in the cerebellum and cerebrum. The proposition of blood not mixing in the basilar artery was demonstrated with regional perfusion arterial spin labeling (ASL) MR imaging. A method that depends on clustering voxels by using a cut-off value or by more advanced techniques such as k-means clustering. It may be argued these techniques do not accurately reflect mixed perfusion territories. Therefore we aim to investigate this observation with a more suitable method.

Cheng Ouyang^1,2, Keith Thulborne³, and Brad P. Sutton^1,2
1Bioengineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ²Beckman Institute, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ³3Center for Magnetic Resonance Research, University of Illinois at Chicago, Chicago, IL, United States

A novel vessel-encoded regional perfusion imaging technique is proposed by adding in-plane gradients into the pTILT sequence to provide differentiation of phase during flow encoding at different vessels. Simulations showed the labeling efficiency as the function of offset from encoded vessels follows an analytical sinusoid equation, which is useful to improve the separation of vascular perfusion regions. In vivo studies demonstrated the ability to provide robust delineation of vascular territories in regional perfusion imaging.

Alexander Graeme Gardener¹, Stuart Clare¹, and Peter Jezzard^1
1FMRIB, University of Oxford, Oxford, United Kingdom

Adaptive Sequence Design ASL optimises multi-TI timings in real-time whilst the subject is within the scanner. Previously, a voxel mask based on difference signal was used; this study uses prior-acquired and stored GM-weighted DIR images loaded into the ASL acquisition to generate perfusion-independent masks. It is shown in healthy adult subjects that both approaches give similar TI optimisation and perfusion quantification, with simple online regional segmentation showing later inflow from PCA. This technique will be of use when dealing with diseased subjects, such as stroke patients, where perfusion may be altered, for which the original approach would not be sensitive.

Johanna Kramme¹, Johannes Gregori¹, and Matthias Günther^1,2
1Fraunhofer MEVIS-Institute for Medical Image Computing, Bremen, Germany, ²Faculty of Physics and Electronics, University of Bremen, Germany

A challenge in Arterial Spin Labeling experiments is low Signal to Noise Ratio (SNR) and physiological noise, especially at inflow times above 2500ms. For time series over 3000ms a high number of averages was necessary to guaranty sufficient SNR, leading to long scan times. In this work a 3D-GRASE sequence is presented which is capable of adaptive averaging, allowing to average higher TIs more often than lower ones. Due to this scan times can be halved, by increasing the SNR by at least one halve compared to standard imaging where all inflow times are imaged equal number of times.

Weiying Dai¹, Ajit Shankaranarayanan², and David Alsop^1
1Radiology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, United States, ²Global Applied Science Laboratory, GE Healthcare, Menlo Park, CA, United States

Quantification of perfusion with arterial spin labeling (ASL) requires knowledge of the label delivery over time. Standard ASL perfusion quantification assumes simple plug flow, i.e., all labeled blood takes the same time to travel from the labeling plane to the imaging region. Simple plug flow may not well approximate the tracer delivery. Here, we show that approximating the transit time distribution with a gamma variate function produces a relatively simple analytic solution to the modified Bloch equations. Pixel-by pixel fitting to in-vivo data was used to measure transit time dispersion. Using a dispersion model provided better fits to the multiple delay data compared with the standard model. Dispersion varies with spatial locations: deep gray matter regions have less dispersion and posterior regions have more.

Fernando F Paiva¹, Bernd U Foerster², Rafael G Oliveira³, Fernanda Tovar-Moll¹, and Jorge Moll^1
1D'Or Institute for Research and Education, Rio de Janeiro, RJ, Brazil, ²Philips Medical Systems, ³InRad-Hospital das Clinicas, Magnetic Resonance Department, Faculty of Medicine of the University of São Paulo

Due to the intrinsic low SNR, EPI-based ASL requires several signal averages to achieve a reliable measurement. A possible approach to overcome this SNR issue is based on using single-shot 3D EPI. However, the ASL signal evolves during the data acquisition and this can result in image blurring if not taken into account. In the present study we present an optimization for a 3D gradient echo EPI acquisition scheme by exploring a modulation of the flip angle of the MR acquisition to keep the ASL contrast constant over the 3D image readout.

Huan Tan¹, W. Scott Hoge², and Robert A Kraft^1
1VT-WFU School of Biomedical Engineering and Sciences, Winston-Salem, NC, United States, ²Brigham and Women's Hospital and Harvard Medical School, Boston, MA, United States

Single-shot 3D GRASE is a popular acquisition technique for ASL imaging. The slice coverage is limited by T2 relaxation. Increasing the number of slices acquired aggravates through-plane blurring due to the longer acquisition window. A new 3D GRASE encoding scheme is developed which performs two kz encodes for each spin echo that doubles slice coverage without compromising image quality. The new encoding scheme is shown to achieve similar image quality as a standard 2-shot 3D GRASE for in-vivo ASL imaging.

Eidrees Ghariq¹, Wouter M. Teeuwisse¹, Andrew Webb¹, and Matthias J.P. van Osch^1
1C.J. Gorter Center for high field MRI, Radiology, Leiden University Medical Center, Leiden, Zuid-Holland, Netherlands

High field MRI holds promises for better SNR in ASL applications due to the prolonged blood T1. To fully reach this gain in SNR the label duration must be chosen optimally. In the present study we determined theoretically and experimentally the optimal label duration for pseudo-continuous ASL at 7Tesla. The optimal label duration was found in vivo to be approximately 1500ms, which is lower than the optimal theoretical value around 3000ms. This can probably be attributed to sub-optimal robustness in label efficiency, which in the end results in a shift of the optimal SNR to shorter label durations.

David Dongsuk Shin¹, Eric C Wong¹, Youngkyoo Jung¹, Ho-Ling Liu², and Thomas T Liu^1
1Center for Functional MRI, University of California, San Diego, La Jolla, CA, United States, ²Department of Medical Imaging and Radiological Sciences, Chang Gung University, Taiwan

Phase tracking errors in PCASL can lead to significant loss in inversion efficiency. This error can be significantly reduced by minimizing the duration of the RF-to-RF spacing of the PCASL pulse train. Based on a Bloch simulation study, we propose a set of tagging parameters that achieve a sub-millisecond RF-to-RF spacing while preserving high inversion efficiency over a wide range of flow velocities. These parameters were also chosen to be relatively insensitive to B1 and gradient imperfections. In-vivo measurements were used to compare the effectiveness of the new optimized tagging parameters with those from a previous study.

S. L. Talagala¹, W-M Luh², and H. Merkle^3
1NMRF/NINDS, National Institutes of Health, Bethesda, MD, United States, ²FMRIF/NIMH, National Institutes of Health, Bethesda, United States, ³LFMI/NINDS, National Institutes of Health, Bethesda, MD, United States

Velocity dependent labeling RF power modulation (VDLM) has been proposed as a possible method to achieve a high labeling efficiency throughout the cardiac cycle with minimum power deposition. Here, we investigate if real-time application of VDLM could be employed to increase the CASL signal when using a neck labeling coil. Current data indicate that the perfusion signal is dependent on the particular VDLM employed. When using the same average power, 3.5/1.5W VDLM was found to yield slightly higher perfusion signal compared to constant labeling. Further studies are needed to optimize VDLM in terms of perfusion signal. However, since VDLM can be used to reduce power deposition from the labeling pulse, it should prove useful for CASL perfusion MRI at higher magnetic field strengths.

David Dongsuk Shin¹, Ho-Ling Liu², Ajit Shankaranarayanan³, and Thomas T Liu^1
1Center for Functional MRI, University of California, San Diego, La Jolla, CA, United States, ²Department of Medical Imaging and Radiological Sciences, Chang Gung University, Taiwan, ³GE Healthcare, Waukesha, Wisconsin, United States

In PCASL, accurate and reliable CBF estimates can be complicated by variations in tagging efficiency (α) caused by phase tracking errors. We present a new technique named α-corrected PCASL, which compensates for tagging efficiency variations on a per voxel basis. The key component of this technique is the accurate estimation of the phase errors and the associated tagging efficiency map. In a healthy subject, CBF estimates obtained with OptPCASL and α-corrected PCASL were similar (87.7 vs. 86.0 ml/100ml-min), while conventional PCASL estimates were lower (78.4 ml/100ml-min). For phase tracking errors < 30 degrees, CBF quantification error is less than 2.85%.

Dennis Franciscus Ramon Heijtel¹, Matthias J P van Osch², Matthan W A Caan¹, Ed van Bavel³, and Aart J Nederveen^1
1Radiology, Academic Medical Center, Amsterdam, Netherlands, ²Radiology, Leiden University Medical Center, Leiden, Netherlands, ³Biomedical Engineering & Physics, University of Amsterdam, Amsterdam, Netherlands

The purpose of this study was to implement and investigate the performance of different available ASL sequences on a 1T open bore scanner. We demonstrate that ASL on a 1 Tesla open bore scanner is possible with a reasonable resolution and within a scanning period of approximately 5-6 min.

David Dongsuk Shin¹, Ho-Ling Liu², Eric C Wong¹, and Thomas T Liu^1
1Center for Functional MRI, University of California, San Diego, La Jolla, CA, United States, ²Department of Medical Imaging and Radiological Sciences, Chang Gung University, Taiwan

Background suppression (BGS) of static tissue can be effective for reducing structural and physiological noise and improving SNR in arterial spin labeling. However, BGS leads to a reduction in quantitative CBF estimates in a manner that has not yet been accurately predicted with theory. To address this issue, CBF estimates obtained with BGS are typically compensated using an empirically determined global scaling factor. We show that the BGS scaling factor varies across subjects and exhibits a dependence on baseline CBF. As a result, the use of a global scaling factor can lead to substantial errors in quantitative CBF estimates.

Cheng Ouyang^1,2, and Brad P. Sutton^1,2
1Bioengineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, ²Beckman Institute, University of Illinois at Urbana-Champaign, Urbana, IL, United States

Pseudo-continuous transfer insensitive labeling technique (pTILT) is a novel pseudo-continuous arterial spin labeling method, which employs non-adiabatic saturation RF pulses for tagging. However, magnetic field inhomogeneities, can compromise the labeling efficacy of pTILT, which induces loss in signal-to-noise ratio and perfusion quantification errors. We propose a method to restore the signal loss by correcting the field inhomogeneity effects through the use of variable phase offsets in the tagging RF pulse pairs. Simulation and in vivo results are shown to demonstrate the effectiveness of this approach. This will provide more robust perfusion measures than the conventional pTILT technique.

Marjorie Villien^1,2, Julien Bouvier³, Irène Tropres³, Matthias J. P. van Osch⁴, Christoph Segebarth^1,2, Jean-François Le Bas⁵, Alexandre Krainik^1,5, and Jan Martin Warnking^1,2
1Centre de Recherche Inserm, U836, Grenoble, France, ²Grenoble Institut des Neurosciences, Université Joseph Fourier, Grenoble, France, ³IFR 1, Université Joseph Fourier, Grenoble, France,⁴Department of Radiology, Leiden University Medical Center, Leiden, Netherlands, ⁵Service de Neuroradiologie, CHU Grenoble, Grenoble, France

Robust MRI methods to measure cerebral vasoreactivity in patients are increasingly sought. A challenge in such studies is to correctly model the signal during a capnic stimulus, potentially varying with subject compliance and response. This is especially critical in ASL with limited SNR, as any mismatch will further decrease sensitivity. Here we compare the performance of regressors derived from hypercapnia data collected during each scanning session for the analysis of ASL data to a standard block regressor. Capnia-derived regressors slightly but consistently outperformed block regressors, more than compensating any variability in the capnia measurement and increasing robustness to experimental variability.

Felipe Tancredi^1,2, Claudine Gauthier^1,2, Cécile Madjar², Joseph Fisher³, Danny JJ Wang⁴, and Richard Hoge^1,2
1Université de Montréal, Montreal, Quebec, Canada, ²Centre de recherche de l'institut universitaire de gériatrie de Montréal, Montreal, Quebec, Canada, ³University of Toronto, Toronto, Ontario, Canada, ⁴Neurology, UCLA, Los Angeles, California, United States

Hypercapnia is known to induce a global increase in cerebral blood flow and is an important component in MRI-based techniques measuring brain’s metabolic and vascular reactivity. While ASL methods have shown promise in these applications, there has been concern about possible underestimation of flow changes during global manipulations. We sought to compare pulsed and continuous labeling methods for detection and quantification of CBF changes induced by mild hypercapnia. Our results show that, although both offer comparable sensitivity to visually evoked CBF responses, continuous labeling yields higher flow change values for a given global manipulation and at higher SNR.

Love Erlandsson Nordin¹, Tie-Qiang Li^1,2, Jacob Brogren³, Niclas Sjögren³, Kristin Hannesdottir³, JiongJiong Wang⁴, and Per Julin^3,5
1Diagnostic Medical Physics, Karolinska University Hospital, Huddinge, Stockholm, Sweden, ²Clinical Science, Intervention and Technology, Division of Medical Imaging and Technology, Karolinska Institute, Stockholm, Sweden, ³AstraZeneca R&D Neuroscience, Södertälje, Sweden, ⁴Neurology, UCLA, Los Angeles, CA, United States, ⁵Section for Brain Injury Rehabilitation, Department of Rehabilitation Medicine, Danderyd University Hospital, Karolinska Institutet, Stockholm, Sweden

The aim of this study is to quantify and model neurological response pattern measured by regional CBF changes in relation to variations in plasma concentration of d-amphetamine after a single oral dose of d-amphetamine. Results from voxel- and region of interest-based analyses of arterial spin labeling (ASL) data have been compared. The study was carried out on a 3 T MRI unit using a 32 channel head coil. 12 randomized volunteers participated, 6 d-amphetamine (20mg)/6 placebo. The results show a significant reduction of global grey matter CBF. This shows that ASL technique has sufficient sensitivity to detect statistically significant changes of cerebral perfusion following administration of d-amphetamine.

Xiaoyun Liang¹, Jacques-Donald Tournier^1,2, Richard Masterton¹, Alan Connelly^1,2, and Fernando Calamante^1,2
1Brain Research Institute, Florey Neuroscience Institutes, Heidelberg West, VIC, Australia, ²Department of Medicine, University of Melbourne, Melbourne, VIC, Australia

3D GRASE has been shown to have some advantages over EPI, and ASL has the advantage of providing a more direct measure and better localization of activation than BOLD. However, the role of 3D GRASE ASL in resting-state functional connectivity has not been investigated. In this study, an improved 3D GRASE pCASL sequence was implemented with whole-brain coverage and then resting-state functional connectivity was investigated. BOLD data were acquired for comparison. In addition to the similarity of results of our method to that of BOLD, our method shows improved coverage in high susceptibility regions, which provides a viable means for studies that especially focus on those particular regions.

Ying Hao¹, Yin Jiang², Yue Zhang³, Cailian Cui², Xiaoying Wang^1,4, Jue Zhang^1,3, and Jing Fang^1,3
1Academy for Advanced Interdisciplinary Studies, Peking University, Beijing, Beijing, China, People's Republic of, ²Neuroscience Research Institute, Peking University, Beijing, Beijing, China, People's Republic of, ³College of Engineering, Peking University, Beijing, Beijing, China, People's Republic of, ⁴Dept. of Radiology, Peking University First Hospital, Beijing, Beijing, China, People's Republic of

This study determined the effect of a relatively long period (30min) of low-frequency electroacupuncture stimulation (EAS) on resting-state CBF utilizing arterial spin labeling (ASL) technique. We observed that blood flow significantly decreased within the ipsilateral IPL and dACC, as well as the contralateral IPFC and postcentral gyrus. The observation that the acupuncture did not significantly increase CBF within any brain region is contrary to previous reports using block-designed acupuncture assessed by functional MRI.

Iain D Wilkinson¹, Nyssa Craig¹, Elaine Cachia¹, Tim J B Hughes¹, Dan Warren¹, Solomon Tesfaye², Petersen T Esben³, Xavier Golay⁴, and Dinesh Selvarajah^2
1Academic Radiology, University of Sheffield, Sheffield, S Yorkshire, United Kingdom, ²Diabetes, Sheffield Teaching Hospitals, ³National University of Singapore, ⁴University College London

Cerebral ischemia and a reduced ability to recover from it are linked with type-2 diabetes mellitus (T2DM). This study uses QUASAR arterial spin labeling (ASL) to determine arterial cerebral blood flow (CBF) before and after a pharmacological vasodilatory challenge (acetazolamide), providing a measurement of cerebrovascular reserve (CVR) in groups of patients with T2DM and controls. Increases in gray matter CBF were seen following acetazolamide. Results suggest a subset of neurologically normal T2DM patients have limited CVR. This technique may provide clinically useful methods with which to identify such subgroups.

U C Anazodo^1,2, N Suskin³, J Wang⁴, J K Shoemaker⁵, and K St Lawrence^1,2
1Lawson Health Research Institute, St Joseph’s Health Care, London, Ontario, Canada, ²Department of Medical Biophysics, University of Western Ontario, London, Ontario, Canada, ³London Health Science Cardiology Rehabilitation Program, London, Ontario, Canada, ⁴Department of Neurology, UCLA, Almanson-Lovelace Brain Mapping Center, Los Angeles, California, United States,⁵Neurovascular Research Laboratory, School of Kinesiology, University of Western Ontario, London, Ontario, Canada

Recent evidence suggests that in cardiovascular disease patients, aerobic fitness can decrease the adverse effect of the disease on cognitive function and improve cerebrovascular health. A relative marker of cerebrovascular health is the measure of the brain’s response to change in arterial CO2 tension, known as cerebrovascular reactivity. In this study, the reproducibility of arterial spin labeling in assessment of cerebrovascular reactivity in cardiovascular disease patient is investigated. Our results demonstrate that the variation between repeated measures is minimal and, as such will not compromise further intra-subject comparisons on fitness effects.

Eric R Muir¹, René C Rentería^2,3, and Timothy Q Duong^1
1Research Imaging Institute, University of Texas Health Science Center, San Antonio, TX, United States, ²Department of Physiology, University of Texas Health Science Center, San Antonio, TX, United States, ³Center for Biomedical Neuroscience, University of Texas Health Science Center, San Antonio, TX, United States

Diabetic retinopathy (DR) is the leading cause of new blindness in working age adults. Vascular dysfunction is the most prominent feature of DR. The retina is nourished by two separate vasculatures, the retinal and choroidal vessels. Arterial spin labeling was used to image layer-specific, quantitative retinal and choroidal blood flow (RBF and ChBF) in the Akita mouse model of DR at 42x42x400 micron. RBF was significantly reduced in Akita mice, while ChBF was unchanged. Vision assessed by the optokinetic response was also worse in Akita mice. MRI provided a non-invasive method to monitor vascular changes in rodents in vivo.

Eric R Muir¹, Bryan H De La Garza¹, and Timothy Q Duong^1
1Research Imaging Institute, University of Texas Health Science Center, San Antonio, TX, United States

Retinitis pigmentosa (RP), which causes photoreceptor death and blindness, affects 1.5 million people worldwide. The vasculature and blood flow (BF) of the retina is attenuated in RP. The retina is nourished by two separate vasculatures, the retinal and choroidal vessels. Arterial spin labeling was used to image layer-specific, quantitative retinal and choroidal BF in the rd10 mouse model of RP at different stages of disease at 42x42x400 microns. Retinal BF was reduced as the retina degenerated, while choroidal BF was unchanged. MRI provided a non-invasive method to monitor vascular changes in rodents in vivo.

Xiang He¹, and Kyongtae Ty Bae^1
1Department of Radiology, University of Pittsburgh, Pittsburgh, Pennsylvania, United States

Tracking arterial spin labeled (ASL) water in the human brain provides key information on the dynamics of water exchange and brain perfusion measured by PET and MR imaging. In this study, we used the T1ρ characteristic of ASL signal to investigate the labeled water compartment structure. Our result is in agreement with PASL T2* studies and is consistent with the results from other PASL studies for flow quantification. We postulate that the discrepancy between the T1ρ PASL and T2 pCASL may be due to the difference between their techniques, and PASL technique provides more accurate information on water compartmental structure.

Lirong Yan¹, Cheng Li², Emily Kilroy¹, Felix Werner Wehrli², and Danny JJ Wang^1
1Department of Neurology, University of California Los Angeles, Los Angeles, CA, United States, ²Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States

A new method was introduced to measure arterial and microvascular cerebral blood volume (CBV) by combining multiphase trueFISP readout with arterial spin labeling. Both simulation and experimental results demonstrated well preserved longitudinal magnetization of flowing spins, which can be treated as an intravascular contrast agent for CBV measurement. The measured arterial and microvascular cerebral blood volume (CBV) was around 1.5%.

Anitha K Priya¹, Lirong Yan¹, and Danny JJ Wang^1
1Neurology, UCLA, Los Angeles, CA, United States

Perfusion Tensor Imaging has been introduced recently to demonstrate anisotropic microvascular flow. In the present study, a multi-directional diffusion-weighted arterial spin labeling (DW-ASL) sequence was performed with a post-labeling delay of 1.2s and a b value of 50s/mm2 to derive the directional information of capillary flow. Fractional anisotropy values were relatively high throughout the brain but were significantly greater in white matter than in gray matter. The primary eigenvectors showed random patterns in the brain. The largely isotropic microvascular flow or perfusion observed in this study implies that a single ASL scan may be enough to quantify cerebral perfusion.

Wouter M. Teeuwisse¹, Aart J. Nederveen^2,3, Eidrees Ghariq¹, Dennis F. Heijtel^2,3, and Matthias J.P. van Osch^1
1Radiology, C.J. Gorter Center for high field MRI, Leiden University Medical Center, Leiden, Netherlands, ²Radiology, Amsterdam Medical Center, Amsterdam, Netherlands, ³Spinoza Center, Amsterdam, Netherlands

Information on the dynamics of labeled spins is essential in ASL to enable absolute quantification and interpretation of the data. In this study multi-echo, multi-delay-time imaging is employed to study differences in spin dynamics of pseudo-CASL (PCASL) and velocity selective ASL (VSASL) and to study the influence of vascular crushing. It is shown that vascular crushing decreases the contribution of arterial signal in PCASL and both arterial and venous signal in VSASL. Delay times larger than 1800 ms or vascular crushing is needed to achieve sensitivity towards perfusion signal.

Jia Guo¹, and Eric C. Wong^2
1Bioengineering, University of California San Diego, La Jolla, California, United States, ²Department of Radiology and Psychiatry, University of California San Diego, La Jolla, California, United States

In Velocity Selective ASL (VSASL) and the QUIXOTIC method for measurement of venous T2, flow/diffusion weighting gradients are used to attenuate the signal of flowing spins. The flow weighting is strong enough that diffusion related attenuation in CSF, which has high diffusivity, can become significant, and the CSF signal can contaminate the desired vascular signal. We introduce here the use of a long TE image to estimate the CSF signal, and remove this contamination from VSASL and QUIXOTIC data. Results demonstrate CBF values in VSASL and T2 values in QUIXOTIC that are closer to literature values after CSF correction.

Feng Xu¹, Jinsoo Uh², and Hanzhang Lu^1
1University of Texas Southwestern Medical Center, Dallas, TX, United States, ²University of Texas Southwestern Medical Center

The T2-Relaxation-Under-Spin-Tagging (TRUST) MRI technique was developed to quantitatively estimate blood oxygenation via the measurement of pure blood R2. A limitation of the current TRUST is the long TR (=8s), as a shorter TR results in an over-estimation of R2. We applied a non-selective 90° RF pulse after the EPI acquisition, and achieved the unbiased R2 estimation in TR of 3s with the precision preserved. Furthermore, we also tested the impact of reducing the TE although previous TRUST already used a relatively short TE of 7ms. The TE of 3.6ms was found to reduce R2 estimation uncertainty by 50%.

Johannes Gregori¹, Norbert Schuff^2,3, and Matthias Günther^1,4
1Fraunhofer MEVIS, Bremen, Bremen, Germany, ²Radiology & Biomedical Imaging, University of California San Francisco, United States, ³Center for Imaging of Neurodegenerate Diseases (CIND), VA Medical Center, San Francisco, California, United States, ⁴FB1, University Bremen, Bremen, Germany

A new method is presented to derive quantitative brain maps of water transfer from blood to tissue, based on non-invasive Arterial Spin Labeling (ASL) techniques using additional T2 measurements. A two-compartment perfusion model including T2 and permeability effects is derived based on the General Kinetic Model. Resulting whole-brain maps of transfer time, which is inversely proportional to capillary water permeability, are shown for healthy volunteers acquired on a 3 Tesla system.

Weiying Dai¹, Ajit Shankaranarayanan², and David C. Alsop^1
1Radiology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, United States, ²Global Applied Science Laboratory, GE Healthcare, Menlo Park, CA, United States

A good estimate of arterial transit time (ATT) is required to eliminate ATT effects in perfusion measurement with ASL. Multiple delay methods for measuring ATTs can be time consuming. Both a rapid, low resolution scan at multiple labeling delays and more efficient encoding of multiple delay signals using Hadamard encoding can reduce the time penalty. We sought to determine the optimal delays and labeling durations for estimating ATT with continuous labeling using a theoretical model for signal and a Monte Carlo optimization of parameters. Simulations suggest that only two delays and labeling durations are required to estimate ATT and that Hadamard encoding can improve the measurement speed and sensitivity. A low resolution scan at two optimal delays using Hadamard encoding is demonstrated to achieve in-vivo ATT measurement efficiently.

Kay Jann¹, Martinus Hauf², Frauke Kellner-Weldon², Marwan Mohamed El-Koussy², Claus Kiefer², Andrea Federspiel¹, and Gerhard Schroth^2
1Department of Psychiatric Neurophysiology, University Hospital of Psychiatry / University of Bern, Bern, Switzerland, ²University Institute of Diagnostic and Interventional Neuroradiology, Inselspital and University of Bern, Bern, Switzerland

The time needed by labelled blood to reach the cortical capillaries (parenchyma phase) has to be accounted for in ASL, therefore a delay between labelling and readout is introduced. The duration of this delay is limited by the fast T1 relaxation-time to only a few seconds. In patients with steno-occlusive arterial disease (SOAD) above the labelling plane (intracranial stenosis) this time window might be too narrow due to prolonged arrival times to the parenchyma. We estimated to what extent arrival is delayed in patients with SOAD and from which vascular compartment the ASL signal origins at a given delay time.

Fernando O Zelaya¹, Astrid Pauls², Owen O'Daly², Matthew Howard², David Alsop³, and Mitul Mehta^2
1Neuroimaging, Institute of Psychiatry, London, London, United Kingdom, ²Neuroimaging, Institute of Psychiatry, London, United Kingdom, ³Beth Israel Hospital, United States

Several physiological modulators have been proposed as a means of reducing the inter-subject variance of BOLD-fMRI signals in group comparisons. In this abstract, we demonstrate that rCBF maps can be used as a voxel-wise co-variate in standard statistical models that compare brain activity maps, obtained from 2 groups of subjects who execute the same cognitive task after administration of a placebo and a psycho-active substance. The rCBF maps allow us to successfully identify statistically significant, drug-dependent differences that would otherwise appear as a type 2 error (false negative) using conventional analysis models.

Patrick William Hales¹, and Chris A Clark^1
1Imaging and Biophysics Unit, UCL Institute of Child Health, London, United Kingdom

We present a methodology to measure flow, capillary permeability-surface product (PS) and capillary volume fraction (vbw) in the human brain, using combined diffusion weighted imaging (DWI) and arterial spin labelling (ASL). Local values of vbw were calculated from DWI data using the intra voxel incoherent motion theory. These data were used in a two-compartment model of time-series ASL data in the same subject. Mean values in three healthy subjects were flow = 42.6 ±2.7 ml/100g/min, PS = 58.6±8.1 ml/100g/min. A slight under-estimation of PS is caused by partial volume contamination from CSF, which we aim to eliminate in future studies.

Wayne Lee¹, Rafal Janik², Bojana Stefanovic^2,3, and John G Sled^1,3
1Hospital for Sick Children, Toronto, Ontario, Canada, ²Sunnybrook Health Sciences Center, Canada, ³Medical Biophysics, University of Toronto, Canada

Whole brain quantitative imaging of cerebral perfusion and transit time provides a means to assess and characterize neurodegeneration. Robust estimation of perfusion and transit time by arterial spin labelling requires adequate sampling of the labelled spins in the tissue preceding the difference signal peak. However, even with rapidly acquired slices, the top third of the brain may not be sampled before the signal peak, too late for estimation of tA and having lower SNR. To address the issue of poor time-delay sampling of distal slices in traditional multi-slice acquisitions, we propose a novel slice ordering approach to multi-slice pseudo-continuous ASL. This Round Robin approach ensures that all slices are equally acquired across the same range of delays, enabling reliable model fits for a wide range of transit times throughout the brain.

Neville D Gai¹, and John Butman^1
1Radiology & Imaging Sciences, National Institutes of Health, Bethesda, MD, United States

Brain perfusion asymmetry is thought to be associated with several cerebrovascular diseases. To a radiologist left-right asymmetry serves a critically important function as a visual cue to the presence of pathology. The ability to assess inter-hemispheric symmetry in the human brain with precision is therefore of importance. Arterial spin labeling is now of routine clinical relevance. Here we study the effect of B1 field inhomogeneity on a whole brain pulsed arterial spin labeling method and show that correcting for transmit and residual B1 inhomogeneity is essential before any inferences can be made based on CBF values particularly related to perfusion asymmetry.

Wayne Lee¹, Rafal Janik², Bojana Stefanovic^2,3, and John G Sled^1,3
1Hospital for Sick Children, Toronto, Ontario, Canada, ²Sunnybrook Health Sciences Center, Canada, ³Medical Biophysics, University of Toronto, Canada

Conventional practice for whole brain cerebral perfusion imaging using arterial spin labelling (ASL) is to acquire consecutive transverse slices from the inferior to the superior aspect of the brain. However, when consecutive slices are acquired at equivalent post-label-delay a confound associated with ascending slice order becomes evident. Labelled blood that is intravascular and transiting to a distal slice can be saturated during the imaging of proximal slices. This study demonstrates that this effect decreases label efficiency by an undetermined amount, which may lead to underestimated perfusion. This effect may be avoided by acquiring multislice data from superior-to-inferior acquisition using the Round Robin sampling approach.

Anne Klomp¹, Matthan W Caan¹, Aart J Nederveen¹, and Liesbeth Reneman^1
1Radiology, Academic Medical Center, Amsterdam, Netherlands

The purpose of this study was to verify feasibility and repeatability of ASL-based phMRI after an oral challenge in assessing cerebral serotonin function. Twelve subjects were scanned six times; twice before and after taking 16mg citalopram and once before and after taking placebo. Subjects were blinded to treatment and treatment order was (pseudo)randomized. Although within-session reproducibility of the ASL signal was good, effects of the citalopram challenge could not be repeated between sessions. We attribute these contradictory findings between the two citalopram sessions to the need for relatively large sample sizes due to variation in both ASL signal and serotonergic activity.

Michael Marxen^1,2, Gabriela Gan^1,2, Christine Monika Zimmermann^1,2, Maximilian Pilhatsch^1,2, Ulrich S Zimmermann¹, Matthias Guenther^3,4, and Michael N Smolka^1,2
1Department of Psychiatry and Psychotherapy, Technische Universität Dresden, Dresden, Germany, ²Neuroimaging Center, Technische Universität Dresden, Dresden, Germany, ³Fraunhofer MEVIS-Institute for Medical Image Computing, Bremen, Germany, ⁴Faculty of Physics and Electronics, Universität Bremen, Bremen, Germany

Perfusion changes (rCBF) during a 15 min. alcohol infusion up to a target alcohol concentration of 0.06 % were studied in 5 subjects using a pulsed arterial spin labelling technique with 3D GRASE readout. 5-25% changes in global gray matter perfusion were observed, which are above the level of change found with a placebo. Increases were seen in most of the brain, especially superiorly. Local decreases were also found in inferior regions such as the medial temporal lobe. The observed changes need to be accounted for in functional BOLD fMRI studies.

Hsiao-Ying Wey^1,2, Kihak Lee¹, Peter T Fox^1,2, and Timothy Q Duong^1,2
1Research Imaging Institute, University of Texas Health Science Center at San Antonio, San Antonio, TX, United States, ²Radiology, University of Texas Health Science Center at San Antonio, San Antonio, TX, United States

We cross validated quantitative (1) CBF at rest and during global and regional activations using pseudo-continuous arterial spin labeling MRI and 1H215O-PET, and (2) stimulus-evoked CMRO2 using calibrated fMRI and 15O2-PET on large nonhuman primates (baboons) when experimental differences were minimized in this study. Baseline CBF was higher as measured with MRI than PET, but comparable stimuli-evoked CBF and CMRO2 were found. Future studies will investigate CBF and CMRO2 measurements in stroke where CBF is compromised and to develop models to better determine CBF and CMRO2 under perturbed conditions.

Kathrin Lorenz¹, Toralf Mildner¹, Andre Pampel¹, and Harald E. Möller^1
1Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany

Inflow vascular-space-occupancy with dynamic subtraction (iVASO-DS) is a non-invasive MRI method for absolute arterial cerebral blood volume (aCBV) quantification. In the present study, mild flow-weighting (FW) gradients were added to the imaging readout. Experiments without FW gradients show a significant drop of the measured aCBV from more than 3% at TI=839ms to about 1.5% at TI=1143ms in regions which are dominated by signals from larger-scale arteries. This drop is explained by non-inverted blood reducing the iVASO-DS contrast at longer TI. It is shown that experiments with FW gradients effectively remove this source of error yielding aCBV values of about 1%.

Adam Martin Winchell^1,2, Kevin Krull³, Noah Sabin¹, Jan Sedlacik¹, Ruitian Song¹, Ralf B Loeffler¹, Melissa Hudson³, and Claudia M Hillenbrand^1
1Radiological Sciences, St. Jude Children's Research Hospital, Memphis, TN, United States, ²Biomedical Engineering, University of Memphis, Memphis, TN, United States, ³Epidemiology & Cancer Control, St. Jude Children's Research Hospital, Memphis, TN, United States

The purpose of our study was to explore the association between gray matter perfusion assessed by arterial spin labeling (ASL)-MRI and neurocognitive function in long-term adult survivors of ALL.