Background: Current clinical standard for diagnosing carotid atherosclerosis is angiography, which determines the location and severity of luminal stenosis. Whereas vessel wall magnetic resonance imaging (MRI) depicts atherosclerotic plaque composition, which has been correlated with increased risk of cerebrovascular events [1, 2]. As such, the ideal is to assess both luminal stenosis and plaque vulnerability comprehensively during a signal session. Previous studies found that a 3D based black blood sequence is feasible for assessing carotid artery [3, 4]. However, whether its capability for identification carotid stenosis and plaque characteristics remains unknown. Purpose: We hypothesized that the 3D black blood, a single sequence, can quantify carotid stenosis and plaque composition simultaneously. To test this hypothesis, we sought to (1) compare images from the 3D black blood with contrast-enhanced MR angiography (MRA) in measuring the carotid stenosis; (2) compare the 3D black blood with corresponding histology in identifying major plaque components: lipid core (LC), intraplaque hemorrhage (IPH), and calcification (CA). Methods: Nine subjects with known or suspected carotid atherosclerosis underwent carotid MRI examinations on a 3T scanner. High-resolution 3D SPGR (TR/TE = 6.9/3.2ms, FA = 8 ̊, FOV = 140mm x 112mm, matrix = 256 x 205. NEX=0.75, slice thickness = 20, 20 slices) sequence was prepared with a flow suppression technique (Spatial LabEling with multiple invErsion pulses, SLEEK, TI = 500 for inflow blood suppression), and followed by a waiting period (TW = 900ms) for relaxation. The total scan time was 4min52sec. An additional 3D contrast enhanced MRA was acquired in the coronal plane with 3D FSPGR sequence (TR/TE: 2.5/ 1.2 ms; Flip angle: 40°; slice thickness: 2 mm; partitions per 3D slab: 36; matrix: 192 X 320; field of view: 320 mm; number of excitations: 1). Two experienced radiologists independently interpreted all bilateral carotid MRA images and 3D SPGR images. Luminal stenosis was quantified in both carotid arteries using the NASCET criterion: (1 luminal diameter at the point of maximal narrowing / the diameter of the normal distal internal carotid artery) x 100% [5]. All diameter measurements were acquired using GE software (GE medical system advantage workstation 4.2.) and were oriented on MIP/MPR images. One trained reviewer, blinded to histological results, determined plaque components using criteria that have been validated with 2D multi-contrast weighted images [6]. The criteria for identifying plaque components on 3D SPGR are (1) CA is no signal; (2) IPH is hyper-intense signal; (3) LC is hypo-intense signal. Spearman’s rank correlation coefficient was used to determine agreement between 3D SPGR and MRA in measuring luminal stenosis. Results: Bilateral carotid arteries (n=16) were included in the final analysis after 2 arteries were excluded due to occlusion. There was an excellent correlation of measuring stenosis between 3D SPGR and MRA (r = 0.95, p < 0.001). For those arteries with corresponding histology (n = 5), 3D SPGR identified all CA accurately (Table 1). It was less sensitive for IPH detection (Table 1). There was good agreement between 3D SPGR and histology for LC detection (Table 1). Representative cases are shown in figures 1-2.

H. GUO, E. X. WU, W. WU, X. MA, G. WANG, AND C. YUAN BIOMEDICAL ENGINEERING DEPARTMENT, TSINGHUA UNIVERSITY, BEIJING, BEIJING, CHINA, PEOPLE'S REPUBLIC OF, CENTER FOR BIOMEDICAL IMAGING RESEARCH, TSINGHUA UNIVERSITY, BEIJING, BEIJING, CHINA, PEOPLE'S REPUBLIC OF, LABORATORY OF BIOMEDICAL IMAGING AND SIGNAL PROCESSING, THE UNIVERSITY OF HONG KONG, HONG KONG, HONG KONG, 4DEPARTMENT OF ELECTRICAL AND ELECTRONIC ENGINEERING, THE UNIVERSITY OF HONG KONG, HONG KONG, HONG KONG, PHILIPS RESEARCH ASIA, SHANGHAI, CHINA, PEOPLE'S REPUBLIC OF, DEPARTMENT OF RADIOLOGY, UNIVERSITY OF WASHINGTON, SEATTLE, WA, UNITED STATES

Introduction Diffusion-weighted magnetic resonance imaging (DWI) appears promising for the functional assessment of kidneys (1) and the detection and characterization of renal lesions (2). Recent studies have shown that, in highly vascular tissue, perfusion effect (or “intravoxel incoherent motion (IVIM)” effect) can be identified separately from diffusion through a bi-exponential signal analysis (3). However, bi-exponential analysis is numerically challenging, as resulting parameters have large variability. We explore the precision of the biexponential parameters as a function of diffusion weightings (b values). We calculated optimal b values for DWI of renal lesions, using analytic expressions for noise propagation and Monte Carlo simulations. We hypothesized that, by sampling at the optimal b values, bi-exponential parameters can be estimated with higher precision than at commonly used uniformly distributed b values. We also tested if benign and malignant renal lesions can be accurately differentiated with DWI. Methods The bi-exponential model for DWI data is given as )] exp( ) exp( ) 1 [( 0 P i P T i P i D b F D b F S S ⋅ − ⋅ + ⋅ − ⋅ − ⋅ = , [1]

Introduction: Concerns about nephrogenic systemic fibrosis (NSF) have spurred a search for alternatives to first-pass gadolinium-enhanced MRA in patients with renal insufficiency. Among the proposed alternatives for peripheral MRA are non-contrast ECG-gated techniques [1-3] that exploit differences in flow velocity between diastole and systole. Due to flow-related dephasing, arteries appear dark in systole when the flow is fast, but brighter in diastole, when the flow is slow. Bright-blood images of the arteries can therefore be obtained by subtraction. These techniques are capable of providing highly detailed angiograms in healthy subjects. The goal of this work was to examine the performance of ECG-gated 3D FSE [1,2] in patients with vascular disease, who may have atypical flow patterns. It has earlier been shown that the flow sensitivity of FSE increases as the flip angle of the refocusing pulses decreases [4]. This provides some ability to adapt to individual flow patterns. However, it is not yet known whether it provides sufficient control to obtain accurate depiction of vascular pathology over a wide range of disease states. Methods: Imaging was performed on a Siemens 1.5T Avanto system using a peripheral phased array coil. 26 patients and 9 healthy subjects were included in the study, of whom 15 patients were enrolled prospectively (8 men, 7 women, ages 68.5±17.5 years, range 22–88). Retrospective analysis was conducted on data from an additional 11 patients (6 men, 5 women, ages 69.5±12.1 years, range 51–92) and 9 healthy subjects (6 men, 3 women, ages 34.1±13.7 years, range 23–63). All subjects provided written informed consent. Patients were recruited from among those scheduled for a routine contrast-enhanced MRA. Reasons for the exam included claudication (21 patients), non-healing ulcers (2), DVT (2) and Leriche syndrome (1). The contrast-enhanced MRA included time-resolved imaging of the calves followed by a multi-station bolus-chase. Non-contrast sequences were run prior to contrast administration. In the prospective study, an ECG-gated 3D FSE-based MRA was performed in two stations (calf and thigh) using the following parameters: FOV=450mm, in-plane voxel size 1.7x1.4mm, nominal slice thickness 1.5mm (calf) or 2mm (thigh), slice resolution 75%, echo spacing 2.76ms, 2 echo trains per partition, GRAPPA 2, constant FA of 60°–120°, TE 19–104 ms. Phase-contrast imaging was conducted in multiple axial planes to calculate appropriate trigger delays for the MRA acquisition and to document flow characteristics. Data were also analyzed retrospectively from 11 patients enrolled in a study to compare ECG-gated FSE with diffusion-prepared SSFP-based MRA [3] in the calf. Diffusion weighting was applied only during the systolic acquisition; during diastole a 90°-180°-90° preparation with identical timing but without the diffusion gradients was used. Phasecontrast and FSE-based MRA data from the calves of 9 healthy subjects were also analyzed retrospectively. Images from the prospective patient study were read by a board-certified radiologist, who graded vessel visualization in the non-contrast FSE-based MRA as compared to contrast-enhanced MRA on a 5-point scale: better/same/worse/much worse/not visualized. Separate grades were assigned to each of the following arterial segments in the calves: popliteal, tibioperoneal trunk, and proximal and distal segments of the anterior tibial, peroneal, and posterior tibial. In the thighs, the following arteries were graded: common femoral, superficial femoral (proximal and distal segments), profunda femoris and proximal popliteal. Average grades were also assigned to collaterals and small branch arteries. Results: In the healthy subjects, peak systolic flow in the popliteal arteries was 43.8±7.8cm/s. Mean diastolic flow was 1.8±2.9cm/s (including negative values). The highest diastolic flow measured in any of the healthy subjects was 6.5±3.1cm/s; for this subject, optimal arterial conspicuity over both large and small vessels in the calves was achieved with FA=120°. Results from the prospective patient study are shown in Fig. 1. Non-contrast FSE-based MRA performed better or the same as contrast-enhanced MRA in 73% and 52% of named arterial segments in the calves and thighs respectively. Reasons for poor vessel visualization included patient motion and vessel orientation perpendicular to the frequency-encoding direction. Signal loss in the right groin, probably due to B1 inhomogeneity [5], contributed to the inferior overall performance in the thigh compared to the calf station. However, faster flow in the thigh may also warrant flip angles higher than 120°.

Introduction Gadolinium-enhanced MRA is routinely used for evaluation of the abdominal vasculature. However, non-contrast-enhanced MRA is less invasive and especially desirable in patients with renal impairment who are at risk for Nephrogenic Systemic Fibrosis. Visualization of the abdominal aorta with non-contrast techniques remains challenging, because of demands for large field of view coverage and respiratory motion suppression, and the need to suppress veins and heterogeneous static tissue. 3D balanced steady-state free precession MRA with a slab-selective inversion (IR SSFP-MRA) exploits inflow effects to provide contrast between arteries and background. The inversion time (TI) is chosen to suppress the background and venous signals and allow non-inverted arterial blood to enter the imaging slab to generate high arterial signal. The TI and arterial blood flow velocity determine the arterial segmental length that can be visualized. Prior groups have demonstrated the promise of this approach for renal artery evaluation at 1.5T, using fixed TI of 325ms [1-2] or 800-1200ms [3], and so far validated in healthy subjects and in patients with renal artery stenosis [2] and renal transplant [3]. The feasibility of coronal plane imaging has been demonstrated in healthy volunteers at 1.5T with TI of 1100-1700ms [4-5]. In prior studies, the TI times have been determined a priori based on background suppression considerations, and similar values have been applied across all subjects. Patient-specific TI values have been reported [5], but no study has explored the relationship between TI and subject vascular hemodynamics. Our hypothesis is that an optimal TI can be chosen for each patient by measuring the arterial velocities within the arteries of interest prior to the MRA acquisition. We evaluate this hypothesis with respect to abdominopelvic MRA, aiming for comprehensive superior-inferior coverage from the suprarenal aorta to the external iliac arteries with coronal IR SSFP-MRA.

Introduction In spite of high perfusion, tissue oxygenation in the kidney, in particular in the renal medulla, is low. This is presumed to be the consequence of low oxygen delivery efficiency to medullary tissue (arterio-venous shunting, low medullary perfusion) and high metabolic activity in the medulla secondary to active transport of solutes (1). The resulting low level of O2 in the medulla makes it particularly vulnerable to hypoxic injury. Techniques such as blood oxygenation level dependent (BOLD) imaging to measure renal oxygenation noninvasively show promise for functional assessment of the kidneys (2). The kidney has unique features that affect its BOLD contrast. First, because of its filtration function, each kidney typically receives 300-500 ml/min of blood, but overall, only a small fraction of oxygen carried by the blood (~8%) is consumed. Second, the renal cortex and medulla have very different perfusion rates and different oxygen extraction fraction. The renal medulla is less perfused than cortex, but its oxygen extraction is high (~80%) secondary to active solute transport, particularly in the ascending limb of Loop of Henle. Third, in renal medulla the blood vessels are oriented radially, following the configuration of loops of Henle, unlike a more random vascular orientation in the cortex. Fourth, both passive and active types of water transit occur in the kidney to enable fluid homeostasis. With BOLD imaging, the partial pressure of oxygen (pO2) measured by microprobes has been shown to correlate with transverse relaxation rate R2* (2). Clinically detected changes in renal R2* have been reported in renal artery stenosis, diabetic nephropathy, transplant rejection, and other diseases. However, the mechanism of renal R2* contrast is complex and not well understood. In addition to tissue oxygenation, R2* is influenced by vascular density, vessel diameter, blood flow, water diffusion rate, field strength, acquisition parameters, and possibly tubular flow in the kidneys. While factors that may contribute to brain R2* have been studied extensively (3, 4), the applicability of existing R2* models to renal cortex and medullary tissue has never been established. We investigated the BOLD mechanism in the kidney and compared predictions made by the simulation technique of Martindale et al (4) with experimental results obtained in human kidney and with experimental results in rat kidney reported by Santos et al (5). Methods We used a Monte Carlo technique to simulate blood vessels, red blood cells and diffusing protons in a voxel to examine the relationship between the oxygen saturation and the resulted BOLD signal or R2*. A cubic voxel is modeled as the combination of intra-vascular (IV) and extra-vascular (EV) spaces. EV: The blood vessels (containing deoxyhemoglobins) are simulated as randomly oriented cylinders with magnetic susceptibility different from surrounding EV space. The magnetic field inhomogeneity ΔBZ, induced by such a cylinder, has been derived as (4),

H. GUO, I. ATANASOVA, R. P. LIM, P. STOREY, J. XU, Q. CHEN, H. RUSINEK, Z. FAN, D. LI, AND V. S. LEE DEPARTMENT OF RADIOLOGY, NEW YORK UNIVERSITY SCHOOL OF MEDICINE, NEW YORK, NY, UNITED STATES, COLUMBIA UNIVERSITY, NEW YORK, NY, UNITED STATES, SIEMENS MEDICAL SOLUTIONS USA, INC., MR R&D COLLABORATION, NEW YORK, UNITED STATES, DEPARTMENTS OF RADIOLOGY AND BIOMEDICAL ENGINEERING, NORTHWESTERN UNIVERSITY, CHICAGO, IL, UNITED STATES