Time

Prog #

Phillip Zhe Sun¹, Yoshihiro Murata², Jie Lu², ³, Xiao Ying Wang², Eng H. Lo², Alma Gregory Sorensen¹

¹A. A. Martinos Center for Biomedical Imaging, Charlestown, Massachusetts, USA; ²Massachusetts General Hospital, Charlestown, Massachusetts, USA; ³XuanWu Hospital, Beijing, People's Republic of China

Amide proton transfer (APT) imaging is a variant of chemical exchange saturation transfer (CEST) imaging that has been shown capable of detecting acute ischemic tissue acidosis, and may serve as a surrogate metabolic imaging marker complementary to perfusion and diffusion MRI. Given that pathologies such as stroke and cancer are spatially heterogeneous, APT imaging with sufficient volume coverage is necessary in order to fully evaluate the diagnostic power. Using a 2-pool exchange model, we showed that the CEST contrast decrease after RF irradiation is governed by its intrinsic relaxation constant of bulk water. Here, we propose a fast volumetric APT imaging approach that acquires multi-slice CEST images immediately after a single long CW RF irradiation, and the relaxation-induced loss of APT contrast is compensated during post-processing. The proposed technique is verified by numerical simulation and validated with a tissue-like dual pH phantom. When translated to image acute animal stroke, the proposed technique detected heterogeneous distribution of PWI, pH and DWI lesions, permitting future study to fully elucidate the diagnostic value of pH-weighted APT MRI.

Jochen Keupp¹, Shelton D. Caruthers², ³, Dirk Burdinski⁴, Sander Langereis⁴, Jeroen A. Pikkemaat⁴, Rolf Lamerichs⁴, Holger Gruell⁴, Samuel A. Wickline², Gergory M. Lanza², Patrick M. Winter²

¹Philips Research Europe, Hamburg, Germany; ²Washington University, St. Louis, Missouri, USA; ³Philips Medical Systems, Andover, Massachusetts, USA; ⁴Philips Research Europe, Eindhoven, Netherlands

Responsive chemical-shift saturation transfer (CEST) agents have been developed for mapping of diagnostic parameters, like local pH for oncology applications. Because the ¹H-CEST signal strength depends as well on the physiological parameters as on agent concentration, a common challenge is the determination of the local concentration for calibration. If a CEST agent is provided, which is additionally labeled by a fluorine marker, the calibration can be performed via ¹⁹F-MRI. In this work, a simultaneous turbo-spin-echo imaging sequence is investigated, which allows to acquire ¹H-CEST and ¹⁹F images at the same time, offering improved time efficiency and precision of the calibration.

Phillip Zhe Sun¹, Christian T. Farrar¹, Gregory Sorensen¹

¹A. A. Martinos Center for Biomedical Imaging, Charlestown, Massachusetts, USA

Chemical exchange saturation transfer (CEST) imaging provides an indirect detection mechanism that allows quantification of certain labile groups unobservable using conventional MRI. However, CEST contrast is often only a few percent, and therefore, it is important to optimize experimental conditions for reliable and quantitative CEST imaging. In particular, CEST imaging is sensitive to B0 and B1 field, while on the other hand; field inhomogeneities persist despite recent advances in magnet technologies, especially for in vivo imaging at high fields. Consequently, correction algorithms that can compensate for field inhomogeneity-induced measurement errors in CEST imaging might be very useful. In this study, the dependence of CEST contrast on field distribution was solved and a correction algorithm was developed to compensate for field inhomogeneity-induced CEST imaging artifacts. In addition, the proposed algorithm was verified with both numerical simulation and experimental measurements, and showed nearly complete correction of CEST imaging measurement errors caused by moderate field inhomogeneity.

Timo Liimatainen¹, Michael Garwood¹, Dennis J. Sorce¹, Shalom Michaeli¹

¹University of Minnesota, Minneapolis, Minnesota, USA

A new contrast based on the frequency swept pulses with sine and cosine amplitude and frequency modulations was developed, referred to as relaxation along a fictitious field (RAFF). Increase of the relaxation rate with increased viscosity was smaller with RAFF than with adiabatic T_1ρ and T_2ρ. In a fast exchanging system, RAFF relaxation rate was found to be in the same range as adiabatic T_2ρ. Relaxation in the dipolar and exchanging systems suggests that the RAFF method appears to slow down dipolar mechanisms as compared to T_1ρ and T_2ρ, exhibits sensitivity to exchange processes, and produces artifact free relaxation maps.

Yoshinori Kato¹, Arvind P. Pathak¹, Dmitri Artemov¹

¹Johns Hopkins University School of Medicine, Baltimore, USA

We developed the activated MR contrast agent using the dual contrast technique.  The concept behind this strategy is that strong negative signal enhancement due to the T₂/T₂^* effects of iron oxides dominates the positive T1 contrast generated by a Gd-based contrast agent when these agents are in close proximity, such as within an intact nanocarrier encapsulating GdDTPA/SPIO, and positive T₁ contrast becomes evident upon release of Gd-based contrast agent from the carrier once the distance between Gd-based contrast agents and SPIO molecules is beyond the T₂/T₂^* enhancement range.  We corroborated the feasibility of this technique in vitro and in vivo.

Andreas Schäfer¹, Samuel Wharton¹, Richard Bowtell¹

¹University of Nottingham, Nottingham, UK

The phase of gradient echo images carries useful anatomical information resulting from field perturbations due to the variation of susceptibility across tissues.  Generating quantitative information from phase images is not straightforward, because of the non-local relationship between field perturbations and susceptibility. Here we show how this problem may be overcome via the calculation of a 3D map of the magnetic susceptibility from phase data. The iterative approach which is proposed takes account of the sphere of Lorentz and external field sources. Its operation is demonstrated via successful application to simulated brain images and phase maps measured from a structured phantom.

Karin Shmueli¹, Peter van Gelderen¹, Tie-Qiang Li¹, Jeff H. Duyn¹

¹National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, USA

The phase in susceptibility-weighted MRI shows excellent contrast but varies with tissue orientation and is affected beyond areas of altered susceptibility. Magnetic susceptibility is an intrinsic tissue property, closely reflecting tissue composition. Recently-proposed Fourier Transform methods for calculating susceptibilities from phase data were applied to high-resolution 7 Tesla phase images of the human brain (ex- and in-vivo). Some cortical layers were most conspicuous in the susceptibility maps, having consistent contrast independent of their orientation relative to B0, unlike in the phase images where contrast varied with orientation. 3D rather than 2D susceptibility calculations showed most promise for revealing fine-scale brain structure.

Tian Liu¹, ², Pascal Spincemaille², Ludovic de Rochefort², Bryan Michael Kressler¹, ², Yi Wang¹, ²

¹Cornell University, Ithaca, New York, USA; ²Weill-Cornell Medical College, New York, New York, USA

Quantification of susceptibility is practically challenging because the model of this linear system is ill-conditioned. A multiple-orientation measurement is implemented to stabilize this problem. Two reconstruction techniques based on multiple orientation measurement are presented. Both methods provide linearity. Streaking artifacts are eliminated in the reconstruction.

Nilesh R. Ghugre¹, ², Pippa Storey, Cynthia K. Rigsby², ³, Alexis A. Thompson², ³, Christine L. Carqueville³, Thomas D. Coates¹, John C. Wood¹

¹Childrens Hospital Los Angeles, Los Angeles, California , USA; ²Northwestern University, Chicago, Illinois, USA; ³Childrens Memorial Hospital, Chicago, Illinois, USA

At 1.5T, relaxivities R2 and R2* can accurately predict hepatic iron concentration in iron overload syndromes. With increasing popularity of 3T scanners, there is need to translate relaxivity-iron relationships to higher field strengths. In this regard, we followed a computational approach by generating realistic liver anatomies and simulating R2 and R2* imaging experiments. We also performed R2 and R2* imaging in the livers of 16 patients at 1.5T and 3T to validate the model. Results demonstrated that R2* scaled linearly with field strength while R2 had a curvilinear relationship. A model-based approach will eliminate the need to recalibrate in patients for changes in sequence type, sequence parameters and imaging conditions.

Nilesh R. Ghugre¹, ², Ignacio Gonzalez¹, Ellen Butensky², Roland Fischer², ³, Roger Williams², Paul Harmatz², Thomas D. Coates¹, John C. Wood¹

¹Childrens Hospital Los Angeles, Los Angeles, California , USA; ²Childrens Hospital and Research Center at Oakland, Oakland, California , USA; ³University Clinic Hamburg-Eppendorf, Hamburg, Germany

R2(1/T2) calibration curve has been established for predicting hepatic iron concentration(HIC) with clinical accuracy. However, there is indication that this relationship is affected by nature of iron overload(transfusion-dependent, hereditary) as well as by type of chelation therapy(subcutaneous, oral). Differences can be attributed to altered iron distribution within two liver compartments viz. hepatocyte and sinusoid. We characterized the relative iron loading in the two compartments(43 patients) and employed a computational approach to systematically evaluate their relative effects on R2. For low HIC, R2 dropped by 15% due to static refocusing while at higher HIC, effects were less apparent. A model-based approach will eliminate the need to recalibrate in patients for changes in therapy and syndrome-type.