Liyong Chen^1,2, Alexander Beckett^1,2, and David A Feinberg^1,2
1University of California, Berkeley, CA, United States, ²Advanced MRI Technologies, LLC, Sebastopol, CA, United States

Background suppression (BS) can increase SNR of ASL images by suppressing the background signal and raise temporal SNR. Recently ASL with simultaneous multi-slice (SMS) or multiband (MB) EPI has been successfully implemented which shortens the total time of the EPI readout trains, which can be used to reduce inversion time variations in BS. In this work, background suppression is implemented and evaluated in pseudo continuous ASL (pCASL) SMS-EPI to increase the temporal SNR and SNR.

Jan Petr¹, Georg Schramm¹, and Jörg van den Hoff^1
1Institute of Radiopharmaceutical Cancer Research, Helmholtz-zentrum Dresden Rossendorf, Dresden, Germany

Influence of vessel angulation and presence of vessel bends in the labeling plane on the resulting CBF was studied for a pseudo-continuous ASL through numerical simulations. Experimental comparison by repositioning and angulation of the labeling plane was done in five healthy volunteers. CBF change of less than 5% for plane angulations of up to 30° was expected from simulations and confirmed by experiments. Up to 30% difference in labeling efficiency was expected on tortuous vessels and 8.4% and 16.9% decrease in CBF was observed in the posterior cerebral artery and vertebral artery vascular territories.

Zhensen Chen^1,2, Xingxing Zhang², Andrew G. Webb², Xihai Zhao¹, and Matthias J.P. van Osch^2
1Center for Biomedical Imaging Research, Department of Biomedical Engineering, School of Medicine, Tsinghua University, Beijing, China, ²C.J. Gorter Center for High Field MRI, Department of Radiology, Leiden University Medical Center, Leiden, Zuid-holland, Netherlands

A novel method based on multi-phase pseudo-continuous ASL (pCASL) imaging distal to the labeling plane was proposed to estimate labeling efficiency. The acquired in vivo ASL signals were consistent with the expected curves from simulations of pCASL imaging using laminar flow model. The labeling efficiencies calculated from in vivo data were in line with theoretical simulation in spite of some measurement errors that might be due to cardiac pulsation and partial volume effect. The method is fast, promising and requires more investigation for improvement of robustness and quantification.

Mustafa Cavusoglu¹, Lars Kasper¹, Johanna S. Vannesjo², Benjamin E. Dietrich¹, Simon Gross¹, and Klaas P. Pruessmann^1
1Biomedical Engineering, ETH Zurich, Zurich, Zurich, Switzerland, ²FMRIB centre, Oxford University, Oxford, United Kingdom

Arterial spin labeling requires fast coverage of k-space which is often achieved by using gradient-echo-EPI as the readout sequence. This demands the gradients to be employed at the limits of the system capabilities and to be switched rapidly during the course of the acquisitions making acquisition highly vulnerable to gradient field imperfections. In this work, we directly measured the gradient field evolution monitored using a dynamic field camera and reconstructed the images based on gradient impulse response function predicted and concurrently monitored k-space trajectories. We explored the effects of gradient field imperfection driven artifacts on the absolute perfusion weighted images.

Yulin V Chang^1,2, Marta Vidorreta¹, Ze Wang^3,4, Maria A Fernandez-Seara⁵, and John A Detre^1
1Neurology, University of Pennsylvania, Philadelphia, PA, United States, ²Radiology, University of Pennsylvania, Philadelphia, PA, United States, ³Center for Cognition and Brain Disorders, Hangzhou Normal University, Hangzhou, Zhejiang, China, ⁴Psychiatry, University of Pennsylvania, Philadelphia, PA, United States,⁵Functional Neuroimaging Laboratory, CIMA, University of Navarra, Navarra, Spain

Single-shot, 3D spiral ASL is incensitive to motion and susceptibility artifacts and high temporal SNR and temporal resolution. However, the prolonged read-out time due to 3D acquisition reduces SNR because of T2 relaxation. In this work we present a 2D in-plane accelerated acquisition scheme using non-Cartesian parallel imaging that helps reduce the read-out time and remain image quality similar to multi-shot acquisitions.

Stephen James Wastling¹, Gareth John Barker¹, Jonathan Ashmore², and Fernando Zelaya^1
1Department of Neuroimaging, King's College London, London, United Kingdom, ²Department of Neuroradiology, King's College Hospital, London, United Kingdom

We show that it is possible to acquire 3D-PCASL data with the same effective in-plane resolution and very similar signal-to-noise ratio in a considerably shorter time by judicious choice of the readout parameters.

Paul Kyu Han¹, Jong Chul Ye¹, Eung Yeop Kim², Seung Hong Choi³, and Sung-Hong Park^1
1Department of Bio and Brain Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Korea, ²Department of Radiology, Gachon University Gil Medical Center, Incheon, Korea, ³Department of Radiology, Seoul National University College of Medicine, Seoul, Korea

In this study, we tested feasibility of segmented 3D approach and also optimized compresses sensing (CS) algorithms, to increase spatial coverage of pCASL-bSSFP. The segmented 3D approach efficiently increased spatial coverage of pCASL-bSSFP with no penalty in scan time and SNR. The pCASL-bSSFP-CS approach also showed good results with down-sampling factor of 4 that was converted to increased spatial coverage while no change in temporal resolution. Segmented 3D pCASL-bSSFP can be a good solution for high-resolution whole brain perfusion mapping and pCASL-bSSFP-CS may be useful for perfusion mapping with limited scan time and/or high temporal resolution such as fMRI.

Yash S Shah¹, Luis Hernandez-Garcia¹, Hesamoddin Jahanian¹, and Scott J Peltier^1
1University of Michigan, Ann Arbor, Michigan, United States

Machine learning has gained tremendous popularity in fMRI data analysis. This study presents an application of support vector machines for temporal brain state classification using multiple acquisition techniques (Blood Oxygenation Level Dependent, Perfusion-weighted Arterial Spin Labeling and Arterial Volume-weighted Arterial Spin Tagging) and highlights the advantages offered by AVAST. Arterial volume-weighted arterial spin tagging (AVAST) is a variant of pseudo continuous ASL technique. In this study, we demonstrate that AVAST exhibits superior detection sensitivity and temporal resolution comparable to BOLD while still retaining desirable properties of standard perfusion-weighted ASL techniques.

Toralf Mildner¹, Kathrin Lorenz^1,2, and Harald E. Möller^1
1Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Saxony, Germany, ²Faculty of Physics and Earth Sciences, University of Leipzig, Saxony, Germany

Mapping of Arterial Transit time by Intravascular Signal SElection (MATISSE) was introduced to obtain arterial transit times directly from temporal shifts of ASL time series acquired with short TR intervals. The RF power applied to a neck labeling coil was modulated for successive repetitions. The current work adopts this strategy to single-coil pseudo-continuous ASL by modulating the RF phase offset accordingly and, thereby, creating a smooth transition between zero and maximum labeling efficiency. Model-independent evaluation of the resulting ASL time series yielded arterial transit time maps that show reasonable patterns of human vascular territories including their border zones.

Jianxun Qu¹, Bing Wu¹, Min Chen², Yingkui Zhang¹, and Zhenyu Zhou^1
1GE Healthcare China, Beijing, Beijing, China, ²Beijing Hospital, Beijing, China

This abstract introduces estimating arterial transit time (ATT) with spatial specificity by using vessel selective ASL (veASL). Both conventional ASL and veASL were performed for comparison of the derived ATT map. Experiments showed that transit time derived from veASL was longer, suggesting the interested region is likely being fed by both ICAs and VAs with blood in ICAs arrived earlier. In this abstract, ATT estimation with blood supply specificity was introduced. It's helpful for choosing optimal PLD in veASL and studying hemodynamic properties of artery.

Yang Li¹, Deng Mao¹, and Hanzhang Lu^1
1Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, Texas, United States

The current recommended protocol for ASL is to use pseudocontinuous labeling with 3D acquisition after background suppression. With the recommended long labeling time (1.8s) and post-labeling delay time (1.8s), it is generally thought that the effect of cardiac pulsation should be minimal. However, we often observe the existence of large fluctuations in the difference image (control-label) intensity from one volume to another. Here, we show that the global signal fluctuations is largely attributed to cardiac phase differences from one volume to another, and that cardiac-gated pCASL can increase the SNR by 79% at a modest cost of scan time.

Andrew D. Robertson¹ and Bradley J. MacIntosh^1,2
1Heart & Stroke Foundation Canadian Partnership for Stroke Recovery, Sunnybrook Research Institute, University of Toronto, Toronto, ON, Canada, ²Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada

This project examined how large and small vessel characteristics contribute to cerebral blood flow (CBF) in chronic stroke. Magnetic resonance imaging methods were used to estimate CBF and cerebrovascular arterial transit time (ATT), as well as the area and blood flow velocity of the bilateral carotid and vertebral arteries. Anterior and posterior cerebrovascular beds were analyzed separately. In the anterior circulation, arterial size and ATT independently contributed to the CBF estimate. In the posterior circulation, only ATT contributed to CBF. This approach to characterizing cerebral hemodynamics may help identify mechanisms related to CBF change during prospective monitoring.

Ki Hwan Kim¹, Seung Hong Choi², and Sung-Hong Park^3
1Department of Bio and Brain Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Korea, ²Department of Radiology, Seoul National University College of Medicine, Korea, ³Department of Bio and Brain Engineering, Korea Advanced Institute of Science and Technology, Korea

In this study, we performed multiple time-phase multi-slice bSSFP imaging following the recently-proposed inter-slice perfusion technique termed alternate ascending/descending directional navigation (ALADDIN). The signal characteristics of the multi-phase ALADDIN perfusion images were consistent with simulations based on the general kinetic model. Quantitative perfusion values could be derived by fitting the multi-phase data to the general kinetic model, and the measured cerebral blood flow values were close to the known values. Contributions of arterial blood volume to the measured cerebral blood volume may be high, but further studies are necessary to understand the signal sources of the measured perfusion signals.

Luisa Ciobanu¹, Lydiane Hirschler^2,3, Tomokazu Tsurugizawa¹, Denis Le Bihan¹, Clément Debacker^2,3, and Emmanuel L. Barbier^3,4
1Neurospin, CEA, Gif-sur-Yvette, France, ²Bruker Biospin, Wissembourg, France, ³University Grenoble Alpes, Grenoble, France, ⁴U836, Inserm, Grenoble, France

The availability of ultra-high field (UHF) preclinical scanners renders the use of pCASL attractive for cerebral perfusion measurements in small animals, benefiting from increased signal to noise-ratio and longer blood longitudinal relaxation times. However, the stronger magnetic field can result in higher physiological noise and increased B0 inhomogeneity at the labeling plane, thereby limiting the inversion efficiency. The purpose of this work was to implement pCASL and test its feasibility at 17.2 T. We obtained inversion efficiencies higher than 70% and cerebral blood flow values in agreement with previous reports, demonstrating the potential of the technique for UHF applications.

Yasuhiro Fujiwara¹, Hirohiko Kimura², Tsuyoshi Matsuda³, Masayuki Kanamoto⁴, Tatsuro Tsuchida², Kazunobu Tsuji², Nobuyuki Kosaka², and Toshiki Adachi^4
1Department of Medical Imaging, Faculty of Life Sciences, Kumamoto University, Kumamoto, Kumamoto, Japan, ²Department of Radiology, University of Fukui, Fukui, Japan, ³Global MR Applications and Workflow, GE Healthcare Japan, Tokyo, Japan, ⁴Radiological Center, University of Fukui Hospital, Fukui, Japan

The purpose of this study was to optimize imaging parameters for accurate rCBF using long labeling duration (LD) and long PLD with higher SNR, which may be acceptable for clinical practice. First, the perfusion signal was simulated using a single-compartment model in each LD, and theoretical SNR efficiency was calculated. Next, in vivo studies were performed on a 3.0T MRI for fifteen volunteers. Based on our results, optimal imaging parameters suggested a combination of LD 3.0 s and PLD 2.0 s to improve rCBF quantification and allow sufficient SNR in elderly individuals with long ATT.

Virginia Mato Abad¹, Pablo García-Polo^2,3, Juan Álvarez-Linera⁴, Ana Frank⁵, Fernando Zelaya⁶, and Juan Antonio Hernández-Tamames^1
1LAIMBIO, Universidad Rey Juan Carlos, Móstoles, Madrid, Spain, ²Martinos Center, MGH, M+Visión Advanced Fellowship, Charlestown, Massachusetts, United States, ³Center for Biomedical Technology (CTB-UPM), Madrid, Spain, ⁴Hospital Ruber Internacional, Madrid, Spain, ⁵Hospital Universitario La Paz, Madrid, Spain,⁶Centre for Neuroimaging Sciences, Institute of Psychiatry, King's College London, London, United Kingdom

Partial volume effects (PVE) are a consequence of limited spatial resolution in ASL, where the low signal-to-noise ratio leads to the need to employ large voxels, being the value in each voxel the sum of the contributions of grey matter, white matter and CSF rather than a single tissue. Although the need for PVE correction for ASL applications in neurodegenerative diseases has been well-established; in this work, we quantitatively demonstrate the effect of PVE correction in a well-characterised MCI cohort. Our results show how PVE correction is essential to maximise the predictive value of ASL in this field of research.

Zungho Zun^1,2, R. Marc Lebel³, Ajit Shankaranarayanan⁴, and Greg Zaharchuk^1
1Stanford University, Stanford, CA, United States, ²Children's National Medical Center, Washington, DC, United States, ³GE Healthcare, Calgary, AB, Canada, ⁴GE Healthcare, Menlo Park, CA, United States

Modeling the dispersion of labeled blood flow may be critical in multi-delay arterial spin labeling (ASL) for accurate cerebral blood flow (CBF) measurement. In this study, a vascular transport function is used to represent arterial transit time and flow dispersion, and is modeled as a Gaussian function with a mean of transit time and standard deviation proportional to the transit time for each imaging voxel. Twelve Moyamoya disease patients were scanned using multi-delay pseudocontinuous ASL. ASL signal curve fitting to the measured ASL signal was improved using flow dispersion modeling, particularly for regions with longer transit delays. CBF maps with dispersion showed corrections of regional underestimation of CBF.

Henri Mutsaerts¹, Torbjørn Elvåshagen², Lars Westlye³, Atle Bjørnerud², and Inge Groote^3
1Academic Medical Center, Amsterdam, Netherlands, ²Oslo University Hospital, Norway, ³University of Oslo, Norway

The current study investigates the effect of sleep deprivation on cerebral blood flow (CBF) as measured by arterial spin labeling. 39 healthy male volunteers, were scanned on a first morning after sufficient sleep and on a second morning after either normal sleep (n=20, age 22.7 yrs) or 24 hours supervised sleep deprivation (n=19, age 21.8 yrs). Whereas the 24 hour test-retest showed nearly no CBF changes for the normal sleepers, there were wide-spread perfusion changes for the sleep deprived. This is the first study to show significant changes in CBF after sleep deprivation that are not observed after normal sleep.

Jason K Mendes¹, Christopher J Hanrahan¹, Jeff L Zhang¹, Gwenael Layec², Corey Hart³, Russell Richardson^3,4, Sarang Joshi⁵, and Vivian S Lee^4
1Radiology, University of Utah, Salt Lake City, Utah, United States, ²Medicine, University of Utah, Salt Lake City, Utah, United States, ³Exercise and Sports Science, University of Utah, Salt Lake City, Utah, United States, ⁴Medicine, University of Utah, Utah, United States, ⁵Bioengineering, University of Utah, Salt Lake City, Utah, United States

An exercise-recovery protocol in the MR scanner acts as a “stress” test to measure muscle ischemic physiology to improve diagnosis and management of peripheral arterial disease (PAD). Evidence suggests changes in the spatial homogeneity of capillary recruitment may be an earlier predictor of PAD progression than measures of vascular reserve. Using an exercise-recovery regimen and arterial spin labeling we demonstrate varying spatial homogeneity of capillary recruitment corresponding to disease progression and exercise induced hyperemia.

Jose A Palomares¹, Danny JJ Wang^2,3, Bumhee Park¹, Sudhakar Tummala¹, Mary A Woo⁴, Daniel W Kang⁵, Keith S St Lawrence⁶, Ronald M Harper⁷, and Rajesh Kumar^1,3
1Anesthesiology, University of California at Los Angeles, Los Angeles, CA, United States, ²Neurology, University of California at Los Angeles, Los Angeles, CA, United States, ³Radiological Sciences, University of California at Los Angeles, Los Angeles, CA, United States, ⁴School of Nursing, University of California at Los Angeles, Los Angeles, CA, United States, ⁵Medicine, University of California at Los Angeles, Los Angeles, CA, United States, ⁶Lawson Health Research Institute, London, Canada, ⁷Neurobiology, University of California at Los Angeles, Los Angeles, CA, United States

Obstructive sleep apnea (OSA) subjects show brain injury in sites that control autonomic, cognitive, and mood functions, deficient in OSA. However, the pathological processes contributing to brain damage in OSA are unclear, but may include altered blood brain barrier (BBB) function. We assessed global BBB function and large artery integrity in OSA and found that BBB function is compromised, but integrity of large arteries is intact. The BBB alteration in OSA can contribute to neural damage that contribute to abnormal functions in the syndrome. These findings suggest requirement to repair BBB function in OSA, with strategies commonly-used in other fields.

Andrea Federspiel¹, Simon Schwab¹, Mirjam R. Heldner², Urs Fischer², Jan Gralla³, and Roland Wiest^3
1Psychiatric Neurophysiology, University Hospital of Psychiatry, Bern, Bern, Switzerland, ²Inselspital, University of Bern, Department of Neurology and Stroke Center, Bern, Bern, Switzerland, ³Inselspital, University of Bern, Institute of Diagnostic and Interventional Neuroradiology, Bern, Bern, Switzerland

In the present study perfusion territories are assessed by vessel-encoded (VE) pseudo-continuous arterial spin labeling (p-CASL) in a cohort of 45 patients with intracranial stenosis. The features of a classification analysis performed with the framework of “searchlight” multi-voxel pattern analysis (MVPA) were investigated. The hypothesis was tested how accurate the classification (above chance level of 0.5) can be performed in these patients. Classification was high in left-, right and posterior perfusion territories with an overlap on the stenotic side. These findings suggest that the present classification method is sensitive to collateral flow.

Minhui Ouyang¹, Peiying Liu¹, Hanzhang Lu¹, Tina Jeon¹, Lina Chalak², Jonathan M Chia³, Andrea Wiethoff¹, Nancy K Rollins⁴, and Hao Huang^1
1Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, TX, United States, ²Department of Pediatrics, University of Texas Southwestern Medical Center, Dallas, TX, United States, ³Philips Healthcare, Cleveland, Ohio, United States, ⁴Radiology, Children's Medical Center, Dallas, TX, United States

Arterial spin labeling (ASL) is a noninvasive perfusion imaging method for quantifying regional CBF using labeled blood as an endogenous tracer. However, accurate measurement of regional CBF for the preterm human brains remains challenging due to extremely slow blood velocity and no standardized protocol for ASL of preterm brains. The purposes of our study are 1) to optimize 3D GRASE pseudo-continuous ASL (pCASL) protocol for highly reproducible and well validated ASL in preterm human brain; and 2) to explore the spatiotemporal CBF distribution and relationship of regional CBF to regional microstructural changes during the 3rd trimester.

Virginia Mato Abad¹, Pablo García-Polo^2,3, Owen O’Daly⁴, Juan Antonio Hernández-Tamames¹, and Fernando Zelaya^4
1LAIMBIO, Universidad Rey Juan Carlos, Móstoles, Madrid, Spain, ²Martinos Center, MGH, M+Visión Advanced Fellowship, Charlestown, Massachusetts, United States, ³Center for Biomedical Technology (CTB-UPM), Madrid, Spain, ⁴Centre for Neuroimaging Sciences, Institute of Psychiatry, King's College London, London, United Kingdom

In this work, we describe an Automatic Software for ASL Processing (ASAP) that can automatically process several ASL datasets, from their raw image format to a spatially normalised, smoothed (if desired) version and includes options like quantification, skull-stripping, co-registration and partial volume correction. ASAP requires minimal user intervention, minimising the possibility of random and systematic errors, and produces perfusion data that is ready for statistical group analysis.

Iris Asllani¹, Ajna Borogovac², Dylan Bruening², Sophie Schmid³, Wouter M. Teeuwisse³, and Matthias J.P. van Osch^3
1RIT, Rochester, NY, United States, ²RIT, NY, United States, ³Leiden University Medical Center, Leiden, Netherlands

Jimi Huh¹, Kyung Won Kim², Jisuk Park³, Jae Ho Byun², In Seong Kim⁴, Berthold Kiefer⁵, and Moon-Gyu Lee^6
1radiology, Seoul Asan Medical Center, Seoul, Seoul, Korea, ²radiology, Seoul Asan Medical Center, Seoul, Korea, ³radiology, Seoul Asan Medical Center, SEOUL, Korea, ⁴Siemens Healthcare, Seoul, Korea, ⁵Siemens Healthcare, Erlangen, Germany, ⁶Seoul Asan Medical Center, Seoul, Korea

Test-bolus DCE-MRI with CAIPIRINHA-VIBE using 2 cc of contrast agent is feasible for perfusion analysis of pancreas tumor with high spatial resolution (1x1 mm) and temporal resolution (3 sec). Indeed, in our series, there was significant difference in perfusion parameters between neuroendocrine tumors and pancreas ductal adenocarcinomas on test-bolus DCE-MRI. The timing accuracy of test-bolus DCE-MRI was excellent to estimate proper timing of arterial-phase images in 88.9% and portal-venous-phase images in 100% of patients. Therefore, incorporating the test-bolus DCE-MRI into the routine dynamic MRI can provide tumor vascularity information and estimation of proper scan timing of routine dynamic MRI.

Hamed Moradi¹, Azimeh Noorizadeh Dehkordi^2,3, Siamak P Nejad-Davarani⁴, Reza Faghihi¹, Brent Griffith⁵, Ali S Arbab⁶, Tom Mikkelsen⁷, Hamid Soltanian-Zadeh⁵, Lisa Scarpace⁷, and Hassan Bagher-Ebadian^5,8
1Mechanical Engineering, Shiraz University, Shiraz, Fars, Iran, ²Nuclear Engineering, Shahid Beheshti University, Tehran, Iran, ³Nuclear Engineering and Science, Azad University of Najafabad, Najafabad, Isfahan, Iran, ⁴Neurology, Henry Ford Hospital, Detroit, Michigan, United States, ⁵Radiology and Research Administration, Henry Ford Hospital, Detroit, Michigan, United States, ⁶GRU Cancer Center, Georgia Regents University, Atlanta, Georgia, United States,⁷Neurological Surgery, Henry Ford Hospital, Detroit, Michigan, United States, ⁸Physics, Oakland University, Rochester, Michigan, United States

The purpose of this pilot study was to investigate the role of Nested Model Selection (NMS) technique in Dynamic Contrast Enhanced MRI (DCE-MRI) data analysis for predicting patient survival. This study investigates the predictive power of different permeability parameters from different nested models for survival of patients with Glioblastoma Multiforme. 20 treatment naïve patients with GBM were studied. A Cox proportional hazards regression (CPHR) model was used to analyze the survival time of the patients. This study suggests an association between Ktrans, Kep and Ve of model 3 and the patient survival that may be of considerable clinical importance.

Ka-Loh Li¹, Alan Jackson¹, and Xiaoping Zhu^1
1WMIC, University of Manchester, Manchester, Great Manchester, United Kingdom

Clinical DTR-DCE data were collected from 12 patients with NF2 receiving anti-angiogenic therapy for rapidly growing vestibular schwannoma, to test potential of the ‘pre-bolus’ low (1/5 of full) dose high temporal resolution (LDHT)-DCE-MRI. Pharmacokinetic parameters from LDHT and the ‘major’ full dose high spatial resolution (FDHS) were jointly used as predictors to identify responders, who might benefit from the treatment. Results showed better separation of intra- and extra vascular space in brain on LDHT derived images than FDHS. The use of pre-bolus LDHT improved both assessment and prediction of vestibular schwannoma response to anti-angiogenic therapy in neurofibromatosis type 2 patients.

Yi Guo¹, R. Marc Lebel², Yinghua Zhu¹, Mark S. Shiroishi³, Meng Law³, and Krishna S. Nayak^1
1Department of Electrical Engineering, University of Southern California, Los Angeles, CA, United States, ²GE Healthcare, Calgary, Alberta, Canada, ³Department of Radiology, University of Southern California, Los Angeles, CA, United States

Here we present the first prospective clinical evaluation of whole-brain DCE MRI, based on sparse sampling and constrained reconstruction, in brain tumor patients. Results were compared with the conventional clinical approach in the same patients during the same examination. While performed in the same amount of time, experimental DCE MRI is able to realize near-isotropic resolution and whole-brain coverage without compromise of image quality in the ROI as the conventional approach.

Aprajita Mehta¹, Krishan K Jain¹, Prativa Sahoo², Bhaswati Roy¹, Ritu Tyagi¹, Ram K S Rathore³, Rana Patir⁴, Sandeep Vaishya⁴, Neeraj Prakash⁵, Nandini Vasudev⁵, and Rakesh K Gupta^1
1Radiology, Fortis Memorial Research Institute, Gurgaon, India, ²Philips Healthcare, Philips India Ltd, Gurgaon, India, ³Dept of Mathematics, Indian Institute of Technology, Kanpur, India, ⁴Neurosurgery, Fortis Memorial Research Institute, Gurgaon, India, ⁵Pathology, Fortis Memorial Research Institute, Gurgaon, India

It is important to differentiate high and low grade gliomas as treatment strategies are different for different grade of gliomas. Prospective grading of gliomas is helpful in explaining the prognosis to patient and where surgery or biopsy of tumor is difficult, clinical decision may be taken on the basis of perfusion imaging. We planned this study to evaluate the relative sensitivity and specificity of Dynamic Contrast Enhanced MR perfusion with single dose contrast in prospective evaluation of glioma grading. Without additional dose of contrast, perfusion study could be added to the routine brain tumor imaging protocol in differentiating the grade of gliomas.

Yinghua Zhu¹, Yi Guo¹, Sajan Goud Lingala¹, Samuel Barnes², R. Marc Lebel³, Meng Law¹, and Krishna Nayak^1
1University of Southern California, Los Angeles, CA, United States, ²California Institute of Technology, Pasadena, CA, United States, ³GE Healthcare, Calgary, Canada

Brain DCE MRI is a powerful technique for evaluating blood-brain-barrier leakage in tumors, multiple sclerosis lesions, and other neurologic disorders. DCE-MRI is an active area of research but lacks a gold standard making it difficult to evaluate novel image acquisition, reconstruction, and processing techniques. We introduce the use of patient-derived digital phantoms that provide ground-truth and the ability to generate data with arbitrary temporal resolution, and perform sensitivity analysis over a broad range of inputs. We demonstrate its application to the evaluation of sparse sampling and constrained reconstruction methods.

Raf H.M. van Hoof^1,2, Evelien Hermeling^1,2, Nicky J.A. Wijnen¹, Floris H.B.M. Schreuder^1,3, Martine T.B. Truijman^1,3, Stefan A. Voo^2,4, Jack P.M. Cleutjens^2,5, Judith C. Sluimer^2,5, Sylvia Heeneman^2,5, Robert J. van Oostenbrugge^2,3, Jan-Willem H. Daemen⁶, Mat J.A.P. Daemen⁷, Joachim E. Wildberger^1,2, and M. Eline Kooi^1,2
1Radiology, Maastricht University Medical Center, Maastricht, Netherlands, ²Cardiovascular Research Institute Maastricht (CARIM), Maastricht University, Maastricht, Netherlands, ³Neurology, Maastricht University Medical Center, Maastricht, Netherlands, ⁴Nuclear Medicine, Maastricht University Medical Center, Maastricht, Netherlands, ⁵Pathology, Maastricht University Medical Center, Maastricht, Netherlands, ⁶Surgery, Maastricht University Medical Center, Maastricht, Netherlands, ⁷Pathology, Academic Medical Center, Amsterdam, Netherlands

Pharmacokinetic quantification of the microvasculature from DCE-MRI in carotid plaques is of great interest. Currently, various studies focused on different regions of the vascular wall. In our study, DCE-MRI was performed and CEA specimens were collected. MRI data was quantified using the Patlak model and CD31 immunohistochemistry was performed on CEA specimens for quantification of endothelial microvasculature. Results showed strongest correlation between adventitial K^trans and histology, coinciding with smaller parameter estimation uncertainties. This indicates that adventitial K^trans may be a better measure for plaque microvasculature compared to other vascular regions.

Chunhao Wang^1,2, Fang-Fang Yin^1,2, John P Kirkpatrick^1,2, and Zheng Chang^1,2
1Radiation Oncology, Duke University Medical Center, Durham, North Carolina, United States, ²Medical Physics Graduate Program, Duke University, Durham, North Carolina, United States

In this pilot study, we investigated the feasibility of using an iterative MR reconstruction method with a total generalized variation (TGV) regularization term DCE-MRI study. The clinical brain DCE-MRI data was undersampled with a simulated acceleration factor of 8 in MR acquisition, and the quantitative permeability and perfusion information were jointly estimated from the reconstructed images. Results showed that by visual comparison, the parametric maps calculated from reconstructed images were very close to the corresponding ones calculated from original data. The quantitative evaluation within the selected region-of-interest indicated that permeability and perfusion parameters were accurately calculated using the undersampled data.

Luke Xie¹, Nian Wang², Chunlei Liu^1,2, and G. Allan Johnson^1
1Center for In Vivo Microscopy, Duke University Medical Center, Durham, North Carolina, United States, ²Brain Imaging Analysis Center, Duke University Medical Center, Durham, North Carolina, United States

Ultrashort echo time (UTE) based DCE-MRI alleviates T₂* blooming artifacts caused by high contrast agent concentrations. However, we showed here that even at 20-μs TE, we still see a significant signal drop off in the kidney. To overcome this limitation, we applied quantitative susceptibility mapping (QSM) to directly quantify the contrast agent concentration. The resulting positive susceptibility enhancement confirmed the very high concentration of the agent. The temporal dynamics revealed the extraordinary mechanism of the kidney to concentrate and clear the agent. UTE-QSM can complement magnitude UTE and offer a powerful tool to study renal physiology.

Deqiang Qiu¹, Amit Saindane¹, Xiaodong Zhong², and Seena Dehkharghani^1
1Radiology and Imaging Sciences, Emory University, Atlanta, GA, United States, ²Siemens Healthcare, GA, United States

In this paper, we developed a Multi-band Multi-echo EPI (M2-EPI) sequence and applied the method in dynamic susceptibility contrast perfusion imaging. Feasibility of the technique was studied with 12-channel and 32-channel head coil.

Matt N Gwilliam¹, David J Collins¹, Martin O Leach¹, and Matthew R Orton^1
1Institute of Cancer Research, London, Greater London, United Kingdom

In order to obtain a patient-specific AIF for DCE-MRI, it has been proposed to measure the T1 of blood in a large vessel. This work demonstrates through the use of physical and virtual phantoms that these measurements are highly compromised by flow at physiological rates and that this strategy may be unfeasible.

Ina Nora Kompan^1,2, Benjamin Richard Knowles³, Cristoffer Cordes¹, and Matthias Guenther^1,2
1Fraunhofer MEVIS, Bremen, Bremen, Germany, ²mediri GmbH, Heidelberg, Baden-Württemberg, Germany, ³Universitätsklinikum Freiburg, Freiburg, Baden-Württemberg, Germany

In dynamic contrast-enhanced MRI, pharmacokinetic modeling is used to quantify tissue physiology. The maximal allowed spatial resolution to still guarantee accurate model fitting depends on the shape of the signal time curve S(t), which is not a priori known. In this work, a 3D golden angle (GA) radial sequence is used to retrospectively adapt spatial resolutions to S(t) throughout the time interval to achieve the maximal feasible spatial resolutions. Using a perfusion phantom, it is shown that, compared to a high temporal and low spatial resolution sequence, the spatial resolution of parameter maps is improved at the same fitting accuracy.

Anne Kluge¹, Mathias Lukas², Vivien Toth³, Stefan Förster², Claus Zimmer¹, and Christine Preibisch^1,4
1Department of Neuroradiology, Klinikum rechts der Isar, TU München, Munich, Germany, ²Department of Nuclear Medicine, Klinikum rechts der Isar, TU München, Munich, Germany, ³Department of Radiology, Klinikum rechts der Isar, TU München, Munich, Germany, ⁴Department of Neurology, Klinikum rechts der Isar, TU München, Munich, Germany

Aim of this study was to investigate the influence of contrast agent leakage on relative CBV, using clinical scanning protocols. Three post-processing methods were compared, concerning CBV and K2 (permeability parameter) for data acquired with and without pre-bolus. Because of tumor heterogeneity CBV was separately analyzed for T1 (K2>0) and T2* (K2<0) effects. Generally all methods increased CBV for T1 and decreased CBV for T2* effects, but differently pronounced. Future investigations should combine the advantages of each method to improve CBV estimation.

Dina Sikpa¹, Réjean Lebel¹, Vincent Turgeon¹, Lisa Whittingstall¹, Jérémie Fouquet¹, Marc-André Bonin¹, Luc Tremblay¹, and Martin Lepage^1
1Centre d'imagerie moléculaire de Sherbrooke, Université de Sherbrooke, Sherbrooke, Québec, Canada

The Tofts model is commonly used to parameterize tissue perfusion, vessel permeability, extracellular extravascular space and vascular volume. However, these quantities are difficult to confirm independently. Here, we propose using intravital microscopy to monitor the extravasation of a fluorescent contrast agent, and its accumulation in the cortex of live animals. We compare our fluorescence imaging measurements with classic DCE-MRI pharmacokinetic modeling results.

J. Tim Marcus¹, Barry Ruijter¹, Taco Kind², Rudolf Verdaasdonk¹, and Anton Vonk Noordegraaf^2
1Physics & Medical Technology, VU University Medical Center, Amsterdam, Netherlands, ²Pulmonology, VU University Medical Center, Amsterdam, Netherlands

The aim of this study is to explore the effect of the delay between arterial input function and lung tissue enhancement curve on the observed pulmonary blood flow. 3D dynamic contrast-enhanced MRI was applied in healthy control subjects (n=5), and patients with pulmonary hypertension (n=14) or pulmonary artery atresia (n=5). Median bolus delay time ranged from 1.5 s for healthy subjects to 3.0 s for pulmonary hypertension patients. Not correcting for delay resulted in an average Pulmonary Blood Flow underestimation of 47 percent. Thus, correction for bolus delay is essential for accurate estimation of pulmonary perfusion parameters.

Siamak Nejad-Davarani^1,2, Hassan Bagher-Ebadian^3,4, Douglas Noll², Tom Mikkelsen⁵, Lisa Scarpace⁵, Azimeh Noorizadeh Vahed Dehkordi⁶, James R. Ewing^1,4, Michael Chopp^1,4, and Quan Jiang^1,4
1Department of Neurology, Henry Ford Hospital, Detroit, MI, United States, ²Department of Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States, ³Department of Radiology, Henry Ford Hospital, Detroit, MI, United States, ⁴Department of Physics, Oakland University, Rochester, MI, United States,⁵Department of Neurosurgery, Henry Ford Hospital, Detroit, MI, United States, ⁶Department of Nuclear Engineering, Shahid Beheshti University, Tehran, Iran

One of the problems in Dynamic Contrast Enhanced MRI (DCE-MRI) and Dynamic Contrast Enhanced Computed Tomography (DCE-CT) studies is estimation of the Arterial Input Function at the tissue level. In this abstract, we introduce a novel vascular transfer function based on laws of fluid dynamics, cerebral vascular morphology and extravasation of the contrast agent to the extravascular-extracellular space. By using this transfer function, the MRI tissue response signal sampled from brain regions with leaky vessels can be decomposed into the intra-vascular and extra-vascular components. Using these signals can lead to more accurate estimation of the permeability parameters.

Yonggang Lu¹, Yousef Mazaheri¹, Vaois Hatzoglou¹, Hilda Stambuk¹, Ashok Shaha¹, Joseph O. Deasy¹, R. Michael Tuttle¹, and Amita Shukla-Dave^1
1Memorial Sloan-Kettering Cancer Center, NEW YORK, New York, United States

Intravoxel microvascular kinetics heterogeneity is not addressed by the commonly used tracer kinetic model (the Tofts’s model) in dynamic contrast enhanced magnetic resonance imaging (DCE-MRI). In this study, we present a modified tracer capillary-tissue exchange model with multiple compartments (oCXM). The model was evaluated on DCE-MRI data collected retrospectively from 7 patients with biopsy proven papillary thyroid carcinoma (PTC). The results demonstrate that the oCXM outperforms the Tofts’s model in characterizing contrast kinetics, providing a more comprehensive interpretation of tumor kinetics. Once validated, this model could provide additional imaging biomarkers for assessing tumor characteristics including tumor aggressiveness.

Chunhao Wang^1,2, Fang-Fang Yin^1,2, and Zheng Chang^1,2
1Radiation Oncology, Duke University Medical Center, Durham, North Carolina, United States, ²Medical Physics Graduate Program, Duke University, Durham, North Carolina, United States

This work proposed an efficient method that can accelerate the pharmacokinetics parameter calculation in the T1w dynamic contrast enhanced MRI (DCE-MRI) permeability study with high temporal resolution. The results of the simulation studies in this work indicated that when comparing with the currently used calculation methods, the proposed calculation methods achieved comparable or improved calculation accuracies with considerably improved calculation efficiency.

Shiyang Wang¹, Gregory S. Karczmar¹, Xiaobing Fan¹, Federico Pineda¹, Milica Medved¹, Ambereen Yousuf¹, and Aytek Oto^1
1Radiology, University of Chicago, Chicago, IL, United States

To assess ‘shutter speed’ and T2* effects on AIF, we compared AIF_LD, measured following a low contrast dose (0.015 mmol/kg), to AID_HD, measured following a high dose (0.085 mmol/kg), from DCE-MRI scans of prostate cancer patients. A new empirical mathematical model for AIF_LD is introduced. The amplitude of AIF_LD, normalized by dose, was 6.45 times larger than that of AIF_HD. The shapes of AIF_LD and AIF_HD were substantially different: the first pass peak was absent from AIF_HD in all cases, and a second pass peak was typically detected in AIF_LD but not in AIF_HD.

Zaki Ahmed¹ and Ives R Levesque^1,2
1Medical Physics Unit, McGill University, Montreal, Quebec, Canada, ²Research Institute of the McGill University Health Center, Montreal, Quebec, Canada

We describe a new analysis method for DCE-MRI which is based on shape analysis and uses the Tofts model to define the classification shapes. Non-negative least-squares (NNLS) is used to identify the weight of these shapes in measured data. We apply this method to a dataset of breast cancer patients undergoing neoadjuvant chemotherapy, and show that our method can predict pathologic complete response (pCR) in images from pre-treatment or after one cycle of therapy. Our results also suggest that the method is robust to inaccuracies in the T1 and arterial input function (AIF).

Tian-Yu Su¹, Sheng-Min Huang¹, Cheng-He Li¹, Kung-Chu Ho², and Fu-Nien Wang^1
1Biomedical Engineering and Environmental Sciences, National Tsing Hua University, Hsinchu, Taiwan, ²Nuclear Medicine, Chang Gung Memorial Hospital, Taoyuan, Taiwan