Perfusion assessment with bolus differentiation: a technique applicable to hyperpolarized tracers.

A new technique for assessing tissue blood flow using hyperpolarized tracers, based on the fact that the magnetization of a hyperpolarized substance can be destroyed permanently, is described. Assessments of blood flow with this technique are inherently insensitive to arterial delay and dispersion, and allow for quantification of the transit time and dispersion in the arteries that supply the investigated tissue. Renal cortical blood flow was studied in six rabbits using a 13C-labeled compound (2-hydroxyethylacrylate) that was polarized by the parahydrogen-induced polarization (PHIP) technique. The renal cortical blood flow was estimated to be 5.7/5.4 +/- 1.6/1.3 ml/min per milliliter of tissue (mean +/- SD, right/left kidney), and the mean transit time and dispersion in the renal arteries were determined to be 1.47/1.42 +/- 0.07/0.07 s and 1.78/1.93 +/- 0.40/0.42 s2, respectively.

Cerebral perfusion assessment by bolus tracking using hyperpolarized 13C.

Cerebral perfusion was assessed with 13C MRI in a rat model after intravenous injections of the 13C-labeled compound bis-1,1-(hydroxymethyl)-1-13C-cyclopropane-D8 in aqueous solutions hyperpolarized by dynamic nuclear polarization (DNP). Since the tracer acted as a direct signal source, several of the problems associated with techniques based on traditional dynamic susceptibility contrast (DSC) MRI contrast agents were avoided. Maps of cerebral blood flow (CBF), cerebral blood volume (CBV), and mean transit time (MTT) were calculated. The MTT was determined to be 2.8 +/- 0.8 sec. However, arterial partial-volume effects in the animal model prevented accurate absolute quantification of CBF and CBV. It was demonstrated that depolarization of the hyperpolarized 13C tracer via relaxation and the imaging sequence had little influence on CBF assessment when the time resolution of the imaging sequence was short compared to the MTT. However, CBV and MTT were increasingly underestimated as MTT or the depolarization rate increased if depolarization was not taken into account. With a modified bolus-tracking theory depolarization could be compensated for, assuming that the depolarization rate was known. Three separate compensation methods were investigated experimentally and by numerical simulations.

Molecular imaging using hyperpolarized 13C.

MRI provides unsurpassed soft tissue contrast, but the inherent low sensitivity of this modality has limited the clinical use to imaging of water protons. With hyperpolarization techniques, the signal from a given number of nuclear spins can be raised more than 100 000 times. The strong signal enhancement enables imaging of nuclei other than protons, e.g. (13)C and (15)N, and their molecular distribution in vivo can be visualized in a clinically relevant time window. This article reviews different hyperpolarization techniques and some of the many application areas. As an example, experiments are presented where hyperpolarized (13)C nuclei have been injected into rabbits, followed by rapid (13)C MRI with high spatial resolution (scan time <1 s and 1.0 mm in-plane resolution). The high degree of polarization thus enabled mapping of the molecular distribution within various organs, a few seconds after injection. The hyperpolarized (13)C MRI technique allows a selective identification of the molecules that give rise to the MR signal, offering direct molecular imaging.

Gradient echo imaging of flowing hyperpolarized nuclei: theory and phantom studies on 129Xe dissolved in ethanol.

The influence of flip angle and flow velocity on the signal intensity achieved when imaging a hyperpolarized substance with a spoiled gradient echo sequence was investigated. The study was performed both theoretically and experimentally using hyperpolarized xenon dissolved in ethanol. Analytical expressions regarding the optimal flip angle with respect to signal and the corresponding signal level are presented and comparisons with thermally polarized substances are made. Both experimentally and theoretically, the optimal flip angle was found to increase with increasing flow velocity. Numerical calculations showed that the velocity dependence of the signal differs between the cases of hyperpolarized and thermally polarized substances.

Echo-planar MR imaging of dissolved hyperpolarized 129Xe.

PURPOSE: The feasibility of hyperpolarized 129Xe for fast MR angiography (MRA) was evaluated using the echo-planar imaging (EPI) technique. MATERIAL AND METHODS: Hyperpolarized Xe gas was dissolved in ethanol, a carrier agent with high solubility for Xe (Ostwald solubility coefficient 2.5) and long relaxation times. The dissolved Xe was injected as a bolus into a flow phantom where the mean flow velocity was 15 cm/s. Ultrafast EPI images with 44 ms scan time were acquired of the flowing bolus and the signal-to-noise ratios (SNR) were measured. RESULTS: The relaxation times of hyperpolarized Xe in ethanol were measured to T1=160+/-11 s and T2 approximately 20 s. The resulting images of the flowing liquid were of reasonable quality and had an SNR of about 70. CONCLUSION: Based on the SNR of the obtained Xe EPI images, it was estimated that rapid in vivo MRA with 129Xe may be feasible, provided that an efficient, biologically acceptable carrier for Xe can be found and polarization levels of more than 25% can be achieved in isotopically enriched 129Xe.

Parahydrogen-induced polarization in imaging: subsecond (13)C angiography.

High nuclear spin polarization of (13)C was reached in organic molecules. Enhancements of up to 10(4), compared to thermal polarization at 1.5 T, were achieved using the parahydrogen-induced polarization technique in combination with a field cycling method. While parahydrogen has no net polarization, it has a high spin order, which is retained when hydrogen is incorporated into another molecule by a chemical reaction. By subjecting this molecule to a sudden change of the external magnetic field, the spin order is transferred into net polarization. A (13)C angiogram of an animal was generated in less than a second. Magn Reson Med 46:1-5, 2001.

Dynamic in vivo oxymetry using overhauser enhanced MR imaging.

A noninvasive method for in vivo measurement of the oxygen concentration has been developed. By introducing a novel contrast medium (CM) based on a single electron substance, it is possible to enhance the proton signal through the Overhauser effect. A low-field magnetic resonance scanner is used to image the proton nuclei of the object. The electron spin transition of the CM is saturated using rf irradiation. As a consequence, the nuclear polarization becomes enhanced through dipole-dipole interaction. The signal enhancement is a function of rf power and of the EPR line width of the substance, which is influenced by the oxygen concentration. The maximum in vivo enhancement has been measured to 60. Image data, generated with different scanning parameters, is used in a postprocessing method to generate images showing pO(2) and the contrast medium concentration, respectively. The mathematical foundation of the postprocessing algorithm is outlined. The results from phantom experiments and animal experiments, in which the oxygen content of the inspired gas was varied, are presented. The potential for human imaging is discussed. J. Magn. Reson. Imaging 2000;12:929-938.

Image artifacts due to a time-varying contrast medium concentration in 3D contrast-enhanced MRA.

The purpose of this work was to study image effects due to time-varying contrast medium concentration in contrast-enhanced three dimensional (3D) magnetic resonance angiography (MRA) images. Two different simulation models (1D and 3D) and two different contrast medium variation schemes were used. Phantom measurements were also performed. Experiments were performed for several different bolus timings. Similar sequence and image object parameters were used in both simulations and measurements (TE/TR 2. 1/7.8 mses, flip angle 30 degrees, T1/T2 1200-80/150-40 msec, flow velocity 100 cm/sec). A small variation in bolus timing yielded large variations in the appearance of the image effects, especially if the center of k-space was sampled in the vicinity of rapid contrast medium concentration variation. For a typical bolus injection in a patient, a severe signal loss but only minor ringing and edge artifacts appeared if the bolus injection was poorly timed. Effects of pulsatile flow were minor. The 3D model proved to be a useful tool in these studies. J. Magn. Reson. Imaging 1999;10:919-928.

EPR and DNP properties of certain novel single electron contrast agents intended for oximetric imaging.

Parameters of relevance to oximetry with Overhauser magnetic resonance imaging (OMRI) have been measured for three single electron contrast agents of the triphenylmethyl type. The single electron contrast agents are stable and water soluble. Magnetic resonance properties of the agents have been examined with electron paramagnetic resonance (EPR), nuclear magnetic resonance (NMR), and dynamic nuclear polarization (DNP) at 9.5 mT in water, isotonic saline, plasma, and blood at 23 and 37 degreesC. The relaxivities of the agents are about 0.2-0.4 mM-1s-1 and the DNP enhancements extrapolate close to the dipolar limit. The agents have a single, narrow EPR line, which is analyzed as a Voigt function. The linewidth is measured as a function of the agent concentration and the oxygen concentration. The concentration broadenings are about 1-3 microT/mM and the Lorentzian linewidths at infinite dilution are less than 1 microT in water at room temperature. The longitudinal electron spin relaxation rate is calculated from the DNP enhancement curves. The oxygen broadening in water is about 50 microT/mM O2 at 37 degreesC. These agents have good properties for oximetry with OMRI.

Overhauser-enhanced MR imaging (OMRI).

PURPOSE: To evaluate a new single-electron contrast agent for Overhauser-enhanced MR imaging. The contrast agents that are currently available give enhancement factors that are too low to make the technique a valid option for routine clinical use. MATERIAL AND METHODS: MR images were generated directly following the injection of the substance into rats. The MR scanner was operated at a main magnetic field of 0.01 T and equipped with a separate rf-transmitter tuned to the electron paramagnetic resonance frequency of the contrast agent. RESULTS: As expected, the images generated show a high level of enhancement in areas where the contrast agent was present, and a maximum enhancement of 60 times the normal proton signal was obtained in the vascular area. The signal-to-noise ratios in the images were superior to those previously attained. CONCLUSION: The new contrast agent makes it possible to generate MR images with both morphological and functional information at 0.01 T. The signal-to-noise ratios found in the generated images were of the same order as, or better than, those obtained with the standard clinical routine.

A multidimensional partition analysis of SSFP image pulse sequences.

The k-space description, of MRI pulse sequences, has been combined with a partition model in order to model the image reconstruction and the contrast behaviour found in SSFP pulse sequences. A partition represents the magnetisation created, due to excitation by a given rf pulse. In the present model, it is visualised as a set of parameters rather than a vector sum taken over a collection of spins. A multidimensional parameter space, where each dimension is associated with one of the partition parameters, is introduced in order to describe the interaction between partitions and pulse sequence events (e.g., rf pulses and gradients). The three k-space dimensions form the first three dimensions and higher orders are used to handle phase dispersions due to diffusion and main field inhomogeneities. The model makes it possible to perform fast simulation of images resulting from general SSFP pulse sequences. A computer implementation generates images (256 matrix), containing more than 10 different T1/T2 combinations, in less than 45 s on a 120 MHz Pentium computer. The contrast behaviour and signal intensities found in simulated images show excellent agreement with data generated using a clinical MRI scanner system.

MRI simulation using the k-space formalism.

An MRI simulation method, together with a corresponding computer program, using the k-space formalism has been developed. It uses a FFT algorithm to generate the ideal NMR signal from a user defined object. The k-space trajectory given by a pulse sequence is calculated. And it is used to select elements from the ideal NMR signal. This selection of elements mimic the sampling of the signal in an actual MRI experiment. During the sampling procedure changes in signal amplitude due to relaxation and excitation are introduced as well as signal phase changes due to movement or flow. Artifacts due to stimulated echoes and transversal magnetization that propagate through several repetition periods are also handled. The usefulness of the method is demonstrated by calculations using standard spin-echo sequence as well as modifications introduced in order to generate angiographical images and flow phase images. Further more a fast pulse sequence, echo planar imaging (EPI), is also simulated. The method is faster than previously presented ones. It is capable of generating images (128 x 128 matrix), including more than eight different T1 and T2 combinations, in less than 3 min on a standard 386/387 type IBM compatible PC.