Two patients with brain abscess and four phantoms containing aqueous solutions of alanine, valine, leucine, and iso-leucine were evaluated on a 3T scanner using quadrature body coil for transmission and an eight-channel phased-array SENSE head coil for receiving. The transmission RF pulse had a bandwidth of about 590 Hz. Localized 1H spectra were acquired with TR=2000 and TE=40, 144, and/or 288 ms, using 128 acquisitions. For the in vivo spectra, the voxel was placed in the abscess cavity to avoid contamination from surrounding brain parenchyma and the wall of the abscess.
In vivo spectra from the brain abscesses showed clearly visible alanine doublet at 1.47 ppm with TE=40ms and TE=288ms and little phase inversion with TE=144ms, consistent with presence of prominent anomalous J-modulation effects. The peaks of amino acid (AA) groups were at around 0.9 ppm corresponding to valine, leucine, and iso-leucine, showing clearly inverted spectral lines at TE=144ms without noticeable reduction in signal intensity, consistent with their smaller chemical shift difference between the methyl and methine protons. Phantom experiments confirmed the in vivo findings.
Get your grade
or your money back
using our Essay Writing Service!
The anomalous J-modulation effects at TE=144ms reported for lactate at 3T is also present for alanine to a similar degree. Care should be exercised on clinical interpretation of spectra upon alanine with current system settings at clinical 3T scanners.
Keywords: alanine, brain abscess, PRESS spectroscopy, high-field systems, anomalous J-modulation effects.
Magnetic resonance spectroscopy now gives access to information on cerebral metabolite changes which are of importance for differentiating pathogenetic mechanisms in human brain diseases. Since this technique is expected to be substantiately exploited in routine clinical use, the reliable determination and interpretation of the metabolites is essential. However, even if the results obtained so far are promising, there are still some considerable problems associated with its implementation in clinical practice, especially under the clinical setting of high field MR systems. One frequently encountered challenge at high field MR systems is the anomalous J-modulation effect on the weakly coupled AX3 system which arises from the enlarged chemical dispersion and the commonly used finite bandwidth of the slice-selective transmitted pulses under high field circumstances [1-3]. The anomalous J-modulation is known as a chemical shift displacement artifact caused by the higher field MR scanners with limited transmitted RF bandwidth [1-5]. While the investigations of the clinical detection on the anomalous J-modulation effects have been centering on the lactate [1-3,8], the possible occurrence arising from other AX3 spin systems such as alanine or amino acids (AA) found in brain abscess, to our knowledge, has never been investigated.
Alanine is potentially a pathological marker of abscess . It has also been observed following ischemia  or in meningioma . This nonessential amino acid containing a methyl and a methine group forms a weakly coupled AX3 system with a TE-dependent phase evolution of the methyl doublet (1.47 ppm), theoretically leading to a negative in-phase doublet at TE=144ms and a positive in-phase doublet at TE=288ms. It evolves in a manner similar to the 6.93Hz lactate doublet that also has a phase evolution depending on the TE used. The doublet in the methyl protons of alanine at 1.47 ppm is known to be split by coupling to the methine protons, with the coupling constant J = 7.23Hz . Due to similar coupling constant to that of lactate, the inverted phase relative to the rest of the spectrum at TE=144ms and the fully positive in phase at TE=288 ms are commonly used to distinguish alanine from overlapping lipids, bearing promise to provide differential diagnosis between brain abscess and tumors . Similar to alanine, valine contains two methyl groups with coupling constants J = 6.97Hz and 7.07Hz from methine protons . Therefore, inverted phase at TE=144ms can be expected as well, albeit with a complicated two-doublet pattern. In clinical MR spectra obtained from patients with brain abscess, the valine peaks are often indistinguishable from other amino acids such as the resonances around 0.90 ppm from leucine or iso-leucine which also exhibit inverted doublet at TE=144ms.
In this study, we focused on the phenomena of anomalous J-modulation effects on the metabolites of interest in brain abscess, namely alanine, valine, leucine and isoleucine by demonstrating examples of signal under-estimation at TE=144ms, using spectra acquired from in vivo human brain abscess and phantoms under the clinical setting of limited-bandwidth RF pulses in a PRESS pulse sequence on a 3T scanner.
Always on Time
Marked to Standard
We recast the theory of AX3 system to derive the essential features of weakly coupled spin mechanism. For a PRESS sequence, it started from the excitation of one 900 and two 1800 pulses with three slice selection gradients, Gz, Gx and Gy in three orthogonal directions, respectively. The intersection of these three planes defines the voxel of interest with the transverse magnetization created by the first pair of RF pulse and gradient in the planes perpendicular to the laboratory z axis, and the refocused magnetization by the second and third pairs of RF pulses and encoding gradients in laboratory direction of x and y axes.
The spatial offsets in three axes of two nuclei with different chemical shifts are given by the amount of
where the is the chemical shift difference of the coupled nuclei and BW is the RF pulse bandwidth. x, y, and z are the voxel dimensions along three axes. Then the voxels corresponding to these nuclei will have their centers shifted by the ratio of the difference in Larmor frequencies to the RF pulse bandwidth as
When the spectral width for the 1800 RF pulses is limited compared to the chemical shift difference between the methyl and methine resonances, a critical spatial mismatch for both lactate  and alanine may result. Such phenomena are particularly evident at 3.0 Tesla for two reasons . At 3.0 Tesla, the difference in chemical shift between the methyl and methine resonance is approximately 360 Hz (2.78 ppm) for lactate and 296 Hz (2.31 ppm) for alanine spin systems, respectively . In addition, due to RF amplitude restrictions at 3.0 Tesla, the 1800 refocusing function is often achieved by a lengthening in time durations of the RF pulses, leading to limited RF bandwidth [8,9]. On our system where the refocusing RF bandwidth is 590 Hz, more than 50% of the methine protons within the slice profile may thus exceed the spectral width without experiencing RF refocusing for each of the two 1800 pulses [3,4]. Depending on whether the methine protons within the excitation volume experience the second and/or the third RF pulses in the PRESS sequence, the excitation volume thus is divided into four partial volumes, each with different phase evolution that leads to destructive interference and consequent signal loss particularly prominent at TE=144 ms . In comparison, the signal is relatively less affected for TE=288 ms because of less phase difference among the four sub-regions [3, 4, 8].
In contrast to the situation of alanine, the chemical shift difference between the methyl and methine resonances is approximately 157 Hz (1.23 ppm) for valine and 89 Hz (0.70 ppm) for leucine, respectively , more than a factor of two smaller than that found for lactate. Consequently, the signal cancellation effect due to different phase evolution in the four sub-regions is anticipated to be substantially weaker than lactate and alanine. In other words, spectra acquired with TE=144ms may not show clearly noticeable anomalous J-modulation for the amino acids near the 0.90 ppm region in the spectra.
Materials and Methods
All experiments were performed on a 3.0T Philips Achieva scanner (Philips Medical System, Best, The Netherlands) acquired with the quadrature body coil for transmission and an eight-channel phased-array SENSE head coil for receiving. Two patients with abscess underwent MR spectroscopic examinations on this 3.0T MR system. The quadrature body coil had a maximum transmission B1 field of 13.5 Î¼T (~590 Hz). Localized 1H spectra were obtained by the double spin-echo method PRESS. Single-voxel MR spectral data were obtained with TR/TE=2000/40, 2000/144, and/or 2000/288 ms to identify the resonance with short T2 values (TE=40 ms) and to confirm the phase reversal shown by the J-coupled multiplets (TE=144 and 288 ms), respectively. 128 acquisitions were performed for each experiment with a total scan time of 4 min 56 seconds. The voxel was placed in the abscess cavity to avoid contamination from the surrounding brain parenchyma and the wall of the abscess cavity .
To validate the theoretical predictions of anomalous J-modulation effect on alanine at the clinical 3T environment, additional measurements were performed on four 1.5-L bottles of aqueous solution phantom containing 10 mM of alanine, leucine, isoleucine, and valine, respectively. These phantom measurements were acquired to verify the signal cancellation deduced from the limited RF bandwidth under the 3T environment. The phantom experiments were performed using the identical following parameters as for the patients, except that the voxel size was fixed at 25Ã-25Ã-25 mm3.
This Essay is
a Student's Work
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.Examples of our work
Figure 1 shows the data from a 19-year-old man with brain abscess. Single voxel spectra were acquired from the lesion (Fig.1a) with TE=40 (Fig.1b) and 144ms (Fig.1c), respectively. The doublet at 1.47 ppm due to the presence of alanine was clearly visible in the 40-msec TE MR spectrum in Fig.1b. However, in Fig.1c there was also a positive peak centered on the same resonance which, if from alanine, is expected to exhibit an opposite sign of resonance (i.e., inverted doublet) at TE=144 ms. This anomalous J-modulation of the methyl signal of alanine is consistent with similar phenomena found for lactate, due to interaction of the chemical shift displacement effects with the methine protons when using a finite bandwidth for the 1800 pulses at 3T.
Another example of anomalous J-modulation in alanine was also present in a 42-year female brain abscess patient (Fig.2). For this patient (Fig.2), the general pattern of lactate, lipids, amino acid groups, and alanine was similar to the other patient (Fig.1), except that acetate at 1.9 ppm was absent. The spectrum acquired with TE=288ms (Fig.2c) showed substantially larger and clearer alanine resonances than with TE=144ms (Fig.2b) where the expected inverted doublet was hardly detectable. Without the clearly identifiable phase inversion around 1.33 ppm for lactate and around 1.47 ppm for alanine in the TE=144ms spectra, misinterpretation of spectral data might take place. In contrast, note that the amino acid groups near 0.9 ppm for the two subjects both showed clearly inverted spectral lines in the TE=144ms spectra without noticeable reduction in signal intensity.
Phantom experiments provided further evidence supporting the in vivo observations. Figure 3 shows the incomplete inverted phase at TE=144 ms for the methyl protons in alanine with reduced amplitude, as opposed to the strong signals in TE=40ms and TE=288ms spectra, which could lead to considerable errors if quantification were to be made from the TE=144ms spectrum. For valine, on the other hand, the CH3 resonances centered at 0.9 ppm exhibited two clearly inverted doublets at TE=144 ms with magnitude comparable to that shown in the TE=40ms spectrum (Fig.4). Leucine and iso-leucine demonstrated phenomena similar to that found for valine, i.e., clear inverted peaks at TE=144ms for the methyl protons with comparable amplitude to TE=40ms and TE=288 spectra (data not shown).
High field MR systems such as 3T strength are expected to benefit the detection of brain metabolites by increased sensitivity. However, the improved spectral resolution of the clinical 3T system also increased the spatial mismatch of slice selective RF pulses due to larger chemical shift difference at higher fields and reduced RF pulse bandwidth [1,3]. The issue of limited RF bandwidth is especially prominent with the increasing popularity of parallel imaging with array coils whose receive-only function demands RF pulse transmission from the generally inefficient body coil. Consequently, the anomalous J-modulation effect that could hamper the identification of the ideally inverted lactate doublet at TE=144ms becomes clinically noticeable [1-3,8,9].
In this study, we demonstrated using both in vivo and in vitro experiments that the anomalous J-modulation effect analogous to lactate is also present for alanine on a clinical 3T scanner. With similar coupling constants exerted on the methyl protons around 7Hz for lactate, alanine, valine, and leucine, acquisition of spectra at TE=144ms for the inverted doublets is often essential to help visual interpretation. However, by the same token, the anomalous J-modulation effects that have been reported for lactate may also exist for alanine, valine, and leucine, all being important metabolites in differential diagnosis for brain abscess . Our inference suggests that the degree of reduction in sensitivity for alanine detection at TE = 144ms is similar to that of lactate detection, due to similar chemical shift difference between the methyl and the methine group protons . As opposed to lactate and alanine spin systems, the amino acid groups with methyl resonances around 0.9 ppm (valine, leucine, and isoleucine) showed substantially less effects of signal cancellation, due to prominently smaller chemical shift difference between methine and methyl resonances.
The potential impact of the destructive interference effects on clinical alanine behaviors should not be ignored without attention. Alanine is not only of special interest in providing complementary information for the differential diagnosis between tumors and abscess , but may also be potentially useful in monitoring metabolic alterations following brain injuries, as already demonstrated in a number of animal studies [12,13]. Furthermore, the concentration of alanine in the brain, if detectable by in vivo MRS, is usually small compared with lactate in the presence of prominent anaerobic glycolysis. This means that, with destructive signal loss due to anomalous J-modulation effects, the identification of alanine at TE=144ms using PRESS is even more prone to error than the situation expected for lactate. Furthermore, note that although the effect of phase evolution due to limited RF transmission bandwidth is theoretically predictable, a recovery of signal loss is however impossible when the distribution of metabolites is spatially heterogeneous in the lesion.
Aside from strategies to alleviate the anomalous J-modulation induced signal loss by modifying the spectroscopic sequence or transmission coil [2,3,9], suggestions for reducing possible misinterpretations include to perform the MR spectroscopic examinations at TE=288 ms or by much shorter TE acquisitions, in addition to TE=144 ms. In a clinical setting, the guidelines to avoid severe anomalous J-modulation effects in brain abscess hence follow those reported for lactate . However, since these remedies have specific trade-offs (such as reduced SNR at TE=288ms and increased background at short TE), careful spectral interpretation on the alanine doublet currently remains the most important and effective means to minimize the impact of these pitfalls.