Optimization of DSC MRI Echo Times for CBV Measurements Using Error Analysis in a Pilot Study of High-Grade Gliomas.

BACKGROUND AND PURPOSE: The optimal TE must be calculated to minimize the variance in CBV measurements made with DSC MR imaging. Simulations can be used to determine the influence of the TE on CBV, but they may not adequately recapitulate the in vivo heterogeneity of precontrast T2*, contrast agent kinetics, and the biophysical basis of contrast agent-induced T2* changes. The purpose of this study was to combine quantitative multiecho DSC MRI T2* time curves with error analysis in order to compute the optimal TE for a traditional single-echo acquisition. MATERIALS AND METHODS: Eleven subjects with high-grade gliomas were scanned at 3T with a dual-echo DSC MR imaging sequence to quantify contrast agent-induced T2* changes in this retrospective study. Optimized TEs were calculated with propagation of error analysis for high-grade glial tumors, normal-appearing white matter, and arterial input function estimation. RESULTS: The optimal TE is a weighted average of the T2* values that occur as a contrast agent bolus transverses a voxel. The mean optimal TEs were 30.0 ± 7.4 ms for high-grade glial tumors, 36.3 ± 4.6 ms for normal-appearing white matter, and 11.8 ± 1.4 ms for arterial input function estimation (repeated-measures ANOVA, P < .001). CONCLUSIONS: Greater heterogeneity was observed in the optimal TE values for high-grade gliomas, and mean values of all 3 ROIs were statistically significant. The optimal TE for the arterial input function estimation is much shorter; this finding implies that quantitative DSC MR imaging acquisitions would benefit from multiecho acquisitions. In the case of a single-echo acquisition, the optimal TE prescribed should be 30-35 ms (without a preload) and 20-30 ms (with a standard full-dose preload).

Simulations on the influence of myelin water in diffusion-weighted imaging.

While myelinated axons present an important barrier to water diffusion, many models used to interpret DWI signal neglect other potential influences of myelin. In this work, Monte Carlo simulations were used to test the sensitivity of DWI results to the diffusive properties of water within myelin. Within these simulations, the apparent diffusion coefficient (D app) varied slowly over several orders of magnitude of the coefficient of myelin water diffusion (D m), but exhibited important differences compared to D app values simulated that neglect D m (=0). Compared to D app, the apparent diffusion kurtosis (K app) was generally more sensitive to D m. Simulations also tested the sensitivity of D app and K app to the amount of myelin present. Unique variations in D app and K app caused by differences in the myelin volume fraction were diminished when myelin water diffusion was included. Also, expected trends in D app and K app with experimental echo time were reduced or inverted when accounting for myelin water diffusion, and these reduced/inverted trends were seen experimentally in ex vivo rat brain DWI experiments. In general, myelin water has the potential to subtly influence DWI results and bias models of DWI that neglect these components of white matter.

Development, simulation, and validation of NMR relaxation-based exchange measurements.

Two-dimensional (2D) nuclear magnetic resonance correlation experiments have recently been proposed as a means for studying exchange in porous media. Most notable of these is the T(2)-T(2) relaxation exchange spectroscopy (REXSY) experiment. Unfortunately, quantifying exchange with this method requires a relatively long, three-dimensional acquisition. To reduce acquisition times, novel 2D methods for quantifying exchange were developed. For each method, model equations were derived (for an arbitrary N-pool system), tested via simulation studies, and validated via experimental studies in an aqueous urea model system. Results indicate that the novel methods outperform REXSY-in terms of uncertainty per unit time for the fitted exchange rate-for certain model systems. The relative merits of each method are discussed in the text.

Disruption of Foxg1 expression by knock-in of cre recombinase: effects on the development of the mouse telencephalon.

The cre/loxP system is used routinely to manipulate gene expression in the mouse nervous system. In order to delete genes specifically from the telencephalon, the Foxg1-cre line was created previously by replacing the intron-less Foxg1 coding region with cre, resulting in a Foxg1 heterozygous mouse. As the telencephalon of heterozygous Foxg1 mice was reported to be normal, this genotype often has been used as the control in subsequent analyses. Here we describe substantial disruption of forebrain development of heterozygous mice in the Foxg1-cre line, maintained on the C57BL/6J background. High resolution magnetic resonance microscopy reveals a significant reduction in the volume of the neocortex, hippocampus and striatum. The alteration in the neocortex results, in part, from a decrease in its tangential dimension, although gross patterning of the cortical sheet appears normal. This decrease is observed in three different Foxg1 heterozygous mouse lines, independent of the method of achieving deletion of the Foxg1 gene. Although Foxg1 is not expressed in the diencephalon, three-dimensional magnetic resonance microscopy revealed that thalamic volume in the adult is reduced. In contrast, at postnatal day 4, thalamic volume is normal, suggesting that interactions between cortex and dorsal thalamus postnatally produce the final adult thalamic phenotype. In the Foxg1-cre line maintained on the C57BL/6J background, the radial domain of the cerebral cortex also is disrupted substantially, particularly in supragranular layers. However, neither Foxg1 heterozygous mice of the Foxg1-tet (tetracycline transactivator) line, nor those of the Foxg1-lacZ and Foxg1-cre lines maintained on a mixed background, displayed a reduced cortical thickness. Thus Cre recombinase contributes to the radial phenotype, although only in the context of the congenic C57BL/6J background. These observations highlight an important role for Foxg1 in cortical development, reveal noteworthy complexity in the invocation of specific mechanisms underlying phenotypes expressed following genetic manipulations and stress the importance of including appropriate controls of all genotypes.

The relationship of problems in biomedical MRI to the study of porous media.

The NMR methods that are used to characterize inanimate porous media measure relaxation times and related phenomena and material transport, fluid displacement and flow. Biological tissues are comprised of multiple small, fluid-filled compartments, such as cells, that restrict the movement of the bulk solvent water and whose constituents influence water proton relaxation times via numerous interactions with macromolecular surfaces. Several of the methods and concepts that have been developed in one field of application are also of great value in the other, and it may be expected that technical developments that have been spurred by biomedical applications of MR imaging will be used in the continuing study of porous media. Some recent specific studies from our laboratory include the development of multiple quantum coherence methods for studies of ordered water in anisotropic macromolecular assemblies, studies of the degree of restriction of water diffusion in cellular systems, multiple selective inversion imaging to depict the ratios of proton pool sizes and rates of magnetization transfer between proton populations, and diffusion tensor imaging to depict tissue anisotropies. These illustrate how approaches to obtain structural information from biological media are also relevant to porous media. For example, the recent development of oscillating gradient spin echo techniques (OGSE), an approach that extends our ability to resolve apparent diffusion changes over different time scales in tissues, has also been used to compute surface to volume measurements in assemblies of pores. Each of the new methods can be adapted to provide spatially resolved quantitative measurements of properties of interest, and these can be efficiently acquired with good accuracy using fast imaging methods such as echo planar imaging. The community of NMR scientists focused on applications to porous media should remain in close communication with those who use MRI to study problems in biomedicine, to their mutual benefits.

Polymer gels for magnetic resonance imaging of radiation dose distributions at normal room atmosphere.

Polymer gels whose NMR and optical properties change when irradiated offer unique advantages for measuring radiation dose distributions. To date, all acrylic polymer gel dosimeters must be manufactured, stored and irradiated in hypoxic conditions which severely limits their use and stability. A new formulation of acrylic dosimeter gel has been developed that responds well in normal atmosphere and which we have named MAGIC (Methacrylic and Ascorbic acid in Gelatin Initiated by Copper). To produce dosimeter gels, an aqueous solution of gelatin, open to the atmosphere, is mixed with methacrylic acid, copper(II) ions, ascorbic acid and hydroquinone. It is believed that the copper(II) and ascorbic acid form a complex with oxygen which (with radiolysis of water) serves as a free radical source for the initiation of the polymerization of methacrylic acid. At room air the water proton spin relaxation rate R2 in MAGIC gels is proportional to absorbed dose though the precise relationship depends on the composition of the gel and the initiating complex. For example, in the range 0-30 Gy the slope of the response of R2 versus dose at 20 MHz was 0.300, 0.519 and 0.681 s(-1) Gy(-1), respectively, when the concentration of MAA was 3, 6 and 9%. The slopes increased to 0.310, 0.567 and 0.868 s(-1) Gy(-1) at 85 MHz. An important determinant of the sensitivity to detect small dose changes is shown to be the slope-to-intercept ratio of the dose-response curve. These varied from 0.08 to 0.17, comparable to hypoxic gels described earlier. MAGIC gels can be manufactured and used much more easily than the previous formulations and can be imaged by magnetic resonance imaging or optical scanning, and thus they will likely be of considerable interest to radiation physicists.

Measurements of restricted diffusion using an oscillating gradient spin-echo sequence.

An oscillating gradient spin-echo (OGSE) pulse sequence was used to measure the apparent diffusion coefficient (D(app)) of water in the short diffusion time regime in the presence of restrictions. The diffusion coefficients of water in a simple water sample and a water and oil mixture were measured to be the same for different periods of the gradient oscillation, as expected when there are no restriction effects. The D(app) of water in the spaces between closely packed beads was also measured as a function of the gradient oscillation periods in the range 11 to 80 ms. The D(app) of water in restricted systems varies with the period of the gradient oscillation and the dispersion depends on the scale of the restriction. For a sample of packed beads of diameter 9.1 +/- 0.7 microm, the pore surface-to-volume ratio was estimated experimentally by this method to be 1.3 +/- 0.1 microm(-1), corresponding to a mean pore diameter of 6.4 +/- 0.7 microm. A Monte Carlo computer simulation of the NMR OGSE signal from the spins diffusing in a system of compartments was also implemented and the D(app) demonstrated similar behavior with gradient oscillation periods.

Rapid acquisition transverse relaxometric imaging.

Segmented echo-planar acquisitions have been incorporated into a multiecho imaging sequence to produce a MRI method for rapid transverse relaxometry. The method is demonstrated on gel phantoms and rat brain and found to produce unbiased estimates of T(2). Gradient performance can be a limiting factor for the implementation of this technique and there is a cost in signal-to-noise ratio resulting from the higher bandwidth required, as is typical for echo-planar acquisitions.

Compartmental study of diffusion and relaxation measured in vivo in normal and ischemic rat brain and trigeminal nerve.

The correlation between the apparent diffusion coefficient (ADC) and T(2) of water in rat brain and trigeminal nerve was investigated using a hybrid diffusion-weighted-CPMG imaging sequence. Little dependence of ADC on T(2) was found in brain regions of interest, which is postulated to be due to rapid exchange between intra- and extracellular water. Conversely, the ADC of water in trigeminal nerve was found to change significantly with echo time (TE). Parallel to the nerve and with a constant diffusion time (t(diff) = 10.8 ms), the ADC increased by approximately 30% between TEs of 25 ms and 185 ms; perpendicular to the nerve, the ADC decreased by a similar amount over the same range of TE. Measurements made following the onset of global ischemia yielded lower ADCs, with similar dependence on TE. Observations that transverse relaxation of water in nerves is multiexponential have previously been interpreted in terms of microanatomical compartments in slow exchange. In the context of this interpretation, our data suggest that diffusional anisotropy is greater outside than within the myelinated axons. Further, data following the onset of global ischemia suggest that the mechanism(s) by which ADC is reduced affect most or all microanatomical environments of nerve, at least insofar as they are represented over the TE domain investigated. Magn Reson Med 43:837-844, 2000.

Complications of nonlinear echo time spacing for measurement of T (2).

Some consequences of using nonlinear echo spacing in multi-echo sequences for measuring T(2) were investigated under the conditions of imperfect RF refocusing or diffusion losses. Although using nonlinear echo spacing has previously been shown to estimate T(2) more accurately, the effect of such spacing is shown to be detrimental when sequences use imperfect RF refocusing pulses. The progressive loss of transverse magnetization that results from imperfect refocusing will alter estimates of T(2) regardless of the echo spacing. However, when the echo spacing is nonlinear, this loss of magnetization also introduces non-mono-exponential T(2) components. Such an effect may distort relative amplitudes of a multi-component T(2) distribution or generate multiple T(2) components where they do not exist. Diffusion through inhomogeneous magnetic fields results in a similar loss of magnetization and T(2) distortion. For these reasons, the use of nonlinearly spaced echoes, while providing in theory a more appropriate sampling of transverse relaxation, is not appropriate for many imaging situations.

In vivo measurement of ADC change due to intravascular susceptibility variation.

The apparent diffusion coefficient (ADC) of extravascular tissue water in rat brains was measured in response to step-wise injections of the superparamagnetic intravascular contrast agent AMI-227. These data were normalized and compared with measured changes in R2* and blood magnetic susceptibility. Linear regression showed that ADC changes 33%/ppm shift of intravascular susceptibility and 0.43% s(-1) change in R2*. These changes correspond to a predicted ADC change of approximately 6% for a change between fully oxygenated and fully deoxygenated blood. The source of these ADC changes was confirmed to be background gradients within the sample by the use of diffusion weighting with bipolar gradients of odd symmetry designed to cancel such background gradient effects on ADC. The results suggest that diffusion-weighted imaging is sensitive to blood-oxygenation and may provide a means of measuring changes in blood oxygen. They also provide estimates of the potential contribution of susceptibility changes to changes in ADC that occur, for example, in stroke and seizure.

Multi-component T1 relaxation and magnetisation transfer in peripheral nerve.

We report here a study of longitudinal relaxation (T1) and magnetisation transfer (MT) in peripheral nerve. Amphibian sciatic nerve was maintained in vitro and studied at a magnetic field strength of 3 T. A CPMG pulse sequence was modified to include either a saturation pulse to measure T1 relaxation or an off-resonance RF irradiation pulse to measure MT. The resulting transverse relaxation (T2) spectra yielded four components corresponding to three nerve compartments, taken to result from myelinic, axonal, and inter-axonal water, and a fourth corresponding to the buffer solution water in which the nerve sample was bathed. Each nerve component was analysed for T1 relaxation and MT. All three nerve T2 components exhibited unique T1 relaxation and MT characteristics, providing further support for the assignment of the components to unique physical compartments of water. Numerical investigation of T1sat measurements of each of the three nerve T2 components indicates that while the two shorter-lived exhibit similar steady-state magnetisation transfer ratios (MTRs), their respective MT properties are quite different. Simulations demonstrate that mobile water exchange between these two components is not necessary to explain their similar steady-state MTR. In the context of the assignment of these two components to signal from myelinic and axonal water, this is to say that these two microanatomical regions of nerve may exhibit similar steady-state MTR characteristics despite possessing widely different MT exchange rates. Therefore, interpreting changes in MTR solely to reflect a change in degree of myelination could lead to erroneous conclusions.

Multiecho imaging with suboptimal spoiler gradients.

Although multiecho imaging may be used to measure transverse relaxation (T2), B1 and B0 inhomogeneity generally gives rise to unwanted coherence pathway signals which result in erroneous T2 measurements. One approach to suppressing this unwanted signal is to center each rf refocusing pulse between spoiler gradients which dephase the unwanted signal; however, hardware limitations often dictate the use of suboptimal spoiler gradients, that is, gradients that cannot provide sufficient dephasing strength. Using simulations, this work demonstrates that by means of a small additional spoiler gradient prior to the first rf refocusing pulse it is possible to reduce substantially the contribution from unwanted coherence pathways in multiecho imaging studies that use suboptimal spoiler gradients. This reduction of unwanted signal results in measured T2 values within approximately 1% of values obtained using spoiler gradients of optimal strength. These results were found for a wide range of biologically relevant T1 and T2 values, missettings of the rf refocusing pulse as large as 5%, and frequency offsets of up to 25 Hz. Multiecho image data agreed with the simulations. Using the additional spoiler gradient it is possible to reduce spoiler gradient strengths by up to 75%.

Changes in water diffusion due to Wallerian degeneration in peripheral nerve.

The authors report NMR measurements of the changes in water diffusion brought about by in vivo Wallerian degeneration due to either crush- or tie-injuries in the sciatic nerve of the frog. Using a pulsed-gradient spin-echo sequence with a diffusion measurement time of 28 ms, the degree of diffusion coefficient anisotropy ¿D(longitudinal)/D(transverse)¿ 4 weeks after injury in both crush- and tie-injured nerves (2.3 +/- 0.4 and 1.7 +/- 0.1, respectively) is significantly less than in normal frog sciatic nerve (3.9 +/- 0.4). The decrease of anisotropy in the degenerated nerves is due to both a decrease in longitudinal diffusion and an increase in transverse diffusion. The changes in diffusion coefficients are compared with the degree of axonal and myelin breakdown observed in light and electron micrographs of the nerves.

Multiexponential T2 relaxation in degenerating peripheral nerve.

The multiexponential T2 relaxation spectrum of peripheral nerve undergoing Wallerian degeneration has been measured both in vivo and in vitro. Degeneration of the sciatic nerve of the amphibian Xenopus laevis was induced by crush injury, and T2 relaxation spectra of the nerve were measured at several times up to 35 days following injury. Histologic evidence verified that the nerve underwent Wallerian degeneration. Relaxation spectra were observed to undergo measurable changes as degeneration progressed, the most evident being a reduction from three well-resolved T2 components to one and a decline in the fraction of the spectra associated with the shortest T2 component. The former appears to reflect the collapse and loss of myelinated fibers, while the latter a combination of interstitial edema and myelin loss.

T2 relaxation of peripheral nerve measured in vivo.

It is demonstrated that multi-exponential transverse (T2) relaxation components can be estimated from multi-echo images of peripheral nerve. Three T2-relaxation components with T2 values +/- standard deviations (populations +/- standard deviations) of 19 +/- 7 ms (26 +/- 9%), 63 +/- 31 ms (29 +/- 11%) and 241 +/- 24 ms (45 +/- 7%) have been identified in vivo in the sciatic nerve of the amphibian Xenopus laevis. The longer-lived component, not identified previously in vivo, provides a significant contrast-to-noise ratio (CNR) between nerve and muscle in the latter-echo images. It is shown that the CNR can be further improved by the averaging of selected images from the multi-echo set.