Magnetic Resonance in Medicine

Published by Wiley
Online ISSN: 1522-2594
Print ISSN: 0740-3194
The purpose of this study was to develop a standard 0.014-inch intravascular magnetic resonance imaging guidewire (MRIG), a coaxial cable with an extension of the inner conductor, specifically designed for use in the small vessels. After a theoretical analysis, the 0.014-inch MRIG was built by plating/cladding highly electrically conductive materials, silver or gold, over the inside and outside of the coaxial conductors. The conductors were made of superelastic, nonmagnetic, biocompatible materials, Nitinol or MP35N. Then, in comparison with a previously designed 0.032-inch MRIG, the performance of the new 0.014-inch MRIG in vitro and in vivo was successfully evaluated. This study represents the initial work to confirm the critical role of highly conductive and superelastic materials in building such small-size MRIGs, which are expected to generate high-resolution MR imaging of vessel walls/plaques and guide endovascular interventional procedures in the small vessels, such as the coronary arteries.
Both normal and experimentally hydronephrotic rabbits were imaged at 0.02 T using partial saturation (PS 160/30) and inversion recovery (IR 1000/200/40) sequences. The signal intensity of normal renal medulla and cortex markedly increased after the injection of 0.1 mmol/kg of Gd-DOTA. In the unilateral total hydronephrosis the dilated renal pelvis did not contrast enhance after 15 and 35 min of Gd-DOTA injection. The enhancement pattern was similar in 1- and 3-week-old hydronephrosis. The effect of Gd-DOTA on renal T1 times at 0.02 T was studied using rats. Fifteen minutes after the Gd-DOTA injection (0.1 mmol/kg) the T1 times of excised rat kidneys decreased from 311 to 90 ms. The authors conclude that the enhancement of the MR signal of the kidney by Gd-DOTA at an ultralow magnetic field (0.02 T) is similar to its enhancement at higher fields (greater than 0.15 T).
The use of inversion recovery sequences to highlight intracranial tumors in children is illustrated. The effect of changing the inversion time (TI) to produce the best spatial resolution and to highlight the contrast resolution between different tumors and normal brain is analyzed. The normal appearances and clinical examples in the central nervous system are used to illustrate the options that are available using IR sequences. Variation of TI for providing a means of differentiating short T1 lesions from long T1 lesions is discussed, short TI sequences being best for demonstrating gliomas and astrocytomas whilst medium/long TI sequences are best for demonstrating vascular abnormalities and short T1 tumors. Inversion recovery imaging is considered to be an accurate alternative to spin-echo imaging as currently applied.
This work reports the use of single-shot spin echo sequences to achieve in vivo diffusion gas measurements and ultrafast imaging of human lungs, in vivo, with hyperpolarized (3)He at 0.1 T. The observed transverse relaxation time of (3)He lasted up to 10 s, which made it possible to use long Carr-Purcell-Meiboom-Gill echo trains. Preliminary NMR studies showed that the resolution of lung images acquired with hyperpolarized (3)He and single-shot sequences is limited to about 6 mm because of the diffusion of the gas in applied field gradients. Ultrafast images of human lungs in normal subjects, achieved in less than 0.4 s with the equivalent of only 130 micromol of fully polarized (3)He, are presented. Comparison with other studies shows that there is no SNR penalty by using low fields in the hyperpolarized case. Advantage was taken of the self diffusion-weighting of the rapid acquisition with relaxation enhancement (RARE) sequence to acquire apparent diffusion coefficient (ADC) images of the lungs. Time scales of seconds could be explored for the first time because there is no hindrance from T(*)(2) as with the usual approaches. At 0.1 T, 180 degrees RF pulses can be repeated every 10 ms without exceeding specific absorption rate limits, which would not be the case for higher fields. Moreover, at low field, susceptibility-induced phenomena are expected to be milder. This supports the idea that low-field imagers can be used for hyperpolarized noble gas MRI of lungs and may be preferred for ADC measurements.
A 3-4% change in signal intensity correlated with visual stimulation was observed in the occipital lobes of three normal volunteers examined with MRI at 0.15 T using fluid attenuated inversion recovery pulse sequences. Similar results were observed at 1.0 T. A double difference technique in which difference images are themselves opposed provided an increase in sensitivity.
Coil development is fundamental to the optimization of imaging systems at any field; however, it is especially critical for low-field systems where signal strength is limited. Here we demonstrate a type of surface coil design which makes possible the acquisition of high-quality images of the neck and extremities.
Perfusion measurements in lung tissue using arterial spin labeling (ASL) techniques are hampered by strong microscopic field gradients induced by susceptibility differences between the alveolar air and the lung parenchyma. A true fast imaging with steady precession (True-FISP) sequence was adapted for applications in flow-sensitive alternating inversion recovery (FAIR) lung perfusion imaging at 0.2 Tesla and 1.5 Tesla. Conditions of microscopic static field distribution were assessed in four healthy volunteers at both field strengths using multiecho gradient-echo sequences. The full width at half maximum (FWHM) values of the frequency distribution for 180-277 Hz at 1.5 Tesla were more than threefold higher compared to 39-109 Hz at 0.2 Tesla. The influence of microscopic field inhomogeneities on the True-FISP signal yield was simulated numerically. Conditions allowed for the development of a FAIR True-FISP sequence for lung perfusion measurement at 0.2 Tesla, whereas at 1.5 Tesla microscopic field inhomogeneities appeared too distinct. Perfusion measurements of lung tissue were performed on eight healthy volunteers and two patients at 0.2 Tesla using the optimized FAIR True-FISP sequence. The average perfusion rates in peripheral lung regions in transverse, sagittal, and coronal slices of the left/right lung were 418/400, 398/416, and 370/368 ml/100 g/min, respectively. This work suggests that FAIR True-FISP sequences can be considered appropriate for noninvasive lung perfusion examinations at low field strength.
Rapid T(2) weighted (T(2)W) images would facilitate physicians being able to distinguish normal tissues, vessels, tumors, and thermal lesions from therapeutic devices throughout interventional MRI procedures commonly performed in open low-field scanners (e.g., 0.2 T). Conventional diagnostic MRI techniques have not been successful at low-field strength for fast T(2)W imaging during the guidance phase of interventional MRI (I-MRI) procedures. FISP and true-FISP methods yield T(1)/T(2)-weighted images and do not always provide sufficient contrast for device guidance or lesion assessment. As such, a variant of PSIF (a gradient reversed form of FISP) which collects the T(2)-weighted spin echo of the SSFP signal was developed and implemented at 0.2 T for use in I-MRI procedures. The sequence has a balanced readout gradient to reduce motion sensitivity. Asymmetric sampling toward the end of the TR cycle reduces T(2)* decay of the spin echo component in the SSFP signal. The sequence gives one image in 5-7 s in vivo with adequate SNR and T(2) contrast for interventional applications. Patient studies showed that the PSIF sequence variant demarcates many tumors not detectable by either FISP or true-FISP. Results from animal experiments suggested that it has potential to monitor thermal lesions during interstitial thermal ablation procedures. Magn Reson Med 42:335-344, 1999.
The purpose of this study was to characterize T(1), particularly in the hyperthermia temperature range (ca. 37-44 degrees C), in order to control regional hyperthermia with MR monitoring using 0.2 Tesla, and to improve T(1) mapping. A single-slice and a new multislice "T One by Multiple Read-Out Pulses" (TOMROP) pulse sequence were used for fast T(1) mapping in a clinical MRI hyperthermia hybrid system. Temporal stability, temperature sensitivity, and reversibility of T(1) were investigated in a polyamidacryl gel phantom and in samples of muscle and adipose tissues from turkey and pig, and verified in patients. In the gel phantom a high linear correlation between T(1) and temperature (R(2) = 0.97) was observed. In muscle and adipose tissue, T(1) and temperature had a linear relationship below a breakpoint of 43 degrees C. Above this breakpoint muscle tissue showed irreversible tissue changes; these effects were not visible in adipose tissue. The ex vivo results were confirmed in vivo under clinical conditions. T(1) mapping allows the characterization of hyperthermia-related tissue response in healthy tissue. T(1), in combination with fast mapping, is suitable for controlling regional hyperthermia at 0.2 T within the hybrid system.
Relaxation times (T(1) and T(2)) of the bone marrow protons and trabecular bone volume fraction (TBVF) in the calcaneus were measured for 100 female volunteers using a compact MRI system at 0.2 T field strength. The speed of sound (SOS) through the calcaneus was measured also for the same subjects using a quantitative ultrasound system. Both relaxation times were found to have positive correlations with age (R = 0.40; P < 0.0001 and R = 0.31; P < 0.002, respectively) and negative correlations with SOS (R = -0.38; P < 0.0001 and R = -0.38; P < 0.0001, respectively). Although TBVF had a fairly high positive correlation with the SOS (R = 0.67), neither T(1) nor T(2) were correlated with TBVF (R = -0.062 and -0.024, respectively). These results suggest that the age dependence of both T(1) and T(2) is caused by the microdynamic properties of the lipid molecules in bone marrow observed using acoustic or elastic modalities.
Based on the principle of selective inversion of the fat or water signal, we have developed a method which allows the simultaneous acquisition of separated fat and water images. In combination with multiecho techniques, this allows the determination of shift selective relaxation times. A detailed analysis of the relaxation behavior of a human leg at 0.23 T shows, that T2 of fat and water can differ by a factor of two or more. We think, that the potential of MRI for differential diagnosis can be greatly enhanced using our method.
MRI can be used for monitoring temperature during a thermocoagulation treatment of tumors. The aim of this study was to demonstrate the suitability of a 3D steady-state free precession sequence (3D Fast Imaging with Steady-State Precession, 3D TrueFISP) for MR temperature measurement at 0.23 T, and to compare it to the spin-echo (SE) and spoiled 3D gradient-echo (3D GRE) sequences. The optimal flip angle for the TrueFISP sequence was calculated for the best temperature sensitivity in the image signal from liver tissue, and verified from the images acquired during the thermocoagulation of excised pig liver. Factors influencing the accuracy of the measured temperatures are discussed. The TrueFISP results are compared to the calculated values of optimized SE and 3D GRE sequences. The accuracy of TrueFISP in the liver at 0.23 T, in imaging conditions used during thermocoagulation procedures, is estimated to be +/-3.3 degrees C for a voxel of 2.5 x 2.5 x 6 mm(3) and acquisition time of 18 s. For the SE and GRE sequences, with similar resolution and somewhat longer imaging time, the uncertainty in the temperature is estimated to be larger by a factor of 2 and 1.2, respectively.
Flow-suppressed FLASH MR images of the human heart have been recorded within a measuring time of 0.3 s using a 2.0-T whole-body research system (Siemens Magnetom) equipped with a conventional 10 mT m-1 gradient system. Subsecond imaging times have been achieved by reducing the repetition time to TR = 4.8 ms and by lowering the spatial resolution to 64 X 128 measured data points. The flip angle of the slice-selective radiofrequency (rf) pulses was adjusted to 10 degrees. Cardiac chambers, ventricular walls, and valves are well delineated in images from a single cardiac cycle using a field of 250 mm and a slice thickness of 8 mm. No motion artifacts were observed as a consequence of the short echo time of TE = 2.8 ms. Distinction between flowing blood and solid structures has been achieved by spatial presaturation of adjacent slices using two slice-selective 90 degrees rf pulses preceding the entire imaging sequence.
An MRI installation (Magnetom, Siemens, software version B1 of NUMARIS) working at 0.35 T was used to estimate T1, T2, and relative proton density in the spleen, liver, adipose tissue, and vertebral body in 14 healthy volunteers. Two double-echo sequences were applied for all subjects: TR = 500 ms, TE1 = 35 and TE2 = 70 ms; and TR = 1600 ms, TE1 = 35 and TE2 = 70 ms. The images were sampled in regions of interest and appropriate relaxation expressions fitted to the ROI data yielding relaxation parameters and relative proton densities. Relaxation expressions, included in standard software (Siemens), were compared to more elaborate functions, developed in parallel to this study. The latter were found more appropriate, especially for high T1 values, and gave the following mean values for the four tissues (estimated uncertainty of mean in parentheses) T1 (ms) 915(36), 428(5), 261(7), and 501(11); T2 (ms) 79.7(8.8), 51.0(0.2), 59.8(1.0), and 64.7(0.8); and corresponding relative proton density (rho, arbitrary units) 2088(136), 2182(10), 2915(49), and 2136(21). The uncertainty in the values was estimated in the fitting procedure and does not include systematic errors. The relative noise in the ROIs was about 9% and the reproducibility of the ROI mean values about 8%.
Proton electron double resonance imaging (PEDRI) uses the Overhauser effect to image the distribution of free-radicals in biological samples and animals. Standard MRI hardware and software is used, with the addition of hardware to irradiate the free-radical-of-interest's EPR resonance. For in vivo applications it must be implemented at a sufficiently low magnetic field to result in an EPR irradiation frequency that will penetrate the sample but will not cause excessive nonresonant power deposition therein. Many clinical MRI systems use resistive magnets that are capable of operating at 10-20 mT, and which could thus be used as PEDRI imagers with the addition of a small amount of extra hardware. This article describes the conversion of a 0.38 T whole-body MRI system for operation as a 20.1 mT small-animal PEDRI imager. The magnet power supply control electronics required a small modification to operate at the lower field strength, but no permanent hardware changes to the MRI console were necessary, and no software modification was required. Frequency down- and up-conversion was used on the NMR RF system, together with a new NMR/EPR dual-resonance RF coil assembly. The system was tested on phantoms containing free-radical solution, and was also used to image the distribution of a free-radical contrast agent injected intravenously into anesthetized mice.
Pulse sequences based on FID signals and projection reconstruction (PR) were investigated for lung MRI at 0.5 T and evaluated for artifacts caused by: (1) k-space mismapping due to either delay or distortion of the readout gradient waveform, (2) cardiac motion and pulsatile flow, and (3) respiratory motion. Nonstructured artifacts were described, simulated, and experimentally confirmed for the first time. Nonstructured artifacts did not impair the demonstration of structures of high signal-to-noise ratio (SNR) but generated quantitative errors in the image intensity analysis over the lung parenchyma. The use of FID-based PR techniques for lung MRI is not justified at 0.5 T.
The application of segmented 3D gradient echo EPI at 0.5 T for coronary artery imaging is described. Experiments were performed using fat suppression, ECG triggering, and a patient-controlled breath-holding scheme. This approach provides a sufficient signal-to-noise ratio for thin contiguous slices in conjunction with a phased array cardiac receive coil. Wide 3D volumes, covering the proximal branches of the coronary tree, were measured with a high spatial resolution. Such data sets can be used for subsequent vessel segmentation. Furthermore, data out of narrow 3D volumes were obtained containing fewer slices angulated in the direction of a selected coronary artery. This provides a good visualization of the selected vessel over several centimeters without the need for segmentation.
Echo-planar imaging using a magnetic field strength of 0.5 T has resulted in an improvement in image quality compared with recent images published at 0.1 T. The sensitivity of the technique to main magnetic field inhomogeneity and transient eddy currents has necessitated innovations in gradient and radiofrequency coil design. These improvements are described, and new variations in the echo-planar pulse sequence which provide better contrast and allow separate imaging of water and fat distributions are presented.
Improvements in the echo-volumar imaging (EVI) technique and its application in whole-body studies are described. Using an in-house built 0.5 T echo-planar imaging system, complete modulus three-dimensional image data are acquired in 102 ms with real-time display. Hardware limitations have restricted the maximum array size achievable to 64 x 32 x 8 voxels. Representative voxel dimensions are x = 6.0 mm; y = 4.0 mm, and z = 10.0 mm. Results on human volunteers are presented, showing cardiac, liver, and bladder images. Also shown are the first EVI gastric filling and emptying results.
Coronary angiography techniques have been implemented on a 0.5-Tesla scanner with a view to performing coronary artery imaging. Slice-followed, segmented k-space FLASH sequences and interleaved-spiral sequences have been employed with acquisitions under real-time navigator echo control with patient feed back, enabling poor signal-to-noise levels to be overcome by averaging data acquired over multiple, variable-length, reproducible breath holds. Good-quality, millimetre-resolution coronary images were obtained in ten normal subjects with both techniques. The mean percent of data segments or interleaves acquired with the navigator echo within the 5-mm diaphragm acceptance window was 57% [standard deviation (S.D.), 11%; range, 38-85%], and the average image-acquisition times were 123+/-22 sec and 71+/-14 sec for segmented FLASH and interleaved-spiral imaging, respectively. In addition to shorter acquisition times, the interleaved-spiral sequence has superior temporal resolution, allowing the acquisition of limited, multislice data sets. However, the sequence is particularly sensitive to the off-resonance effects of residual epicardial fat surrounding the artery and to field nonuniformities, both of which lead to image blurring and, unlike segmented FLASH acquisitions (which are very robust), the spiral data sets generally require postprocessing.
This paper presents the first in vivo measurements of perfusion in the human placenta from 20 weeks gestational age until term, using the non-selective/selective inversion recovery echo-planar imaging sequence, in which data is alternately acquired following a selective and non-selective inversion pulse. Twenty pairs of images were collected, two each at the following inversion times: 20, 310, 610, 910, 1110, 1410, 1910, 2810, 3310, and 4510 ms with the sequence being repeated with a repetition time (TR) of 10 s. The results of these measurements were used to suggest the optimum sequence for future work in terms of the signal to noise ratio in the measured perfusion rate in a given measurement time. The sequence was also analyzed to determine the expected variability in the measurements. In normal pregnancies the average value of perfusion rate was found to be 176 (standard error = +/-24) ml/100 mg/min. (n = 16, standard deviation = 96 ml/100 mg/min). The expected variability in the measured parameters due to signal to noise ratio considerations alone was calculated to be 71%. For a maximum scanning time of 400 s, the optimum sequence for measuring placental perfusion was found to require 8 repetitions at each of 10 inversion times which were geometrically spaced (given by a(o), a(o)r, a(o)r2, a(o)r3, . . .), with a(o) = 850 ms, r = 1.073 and TR = 5 s, giving a pixel variability of 38%. Other timing schemes are recommended for measuring perfusion in other anatomical regions with different values of perfusion rate and longitudinal relaxation time.
This paper presents the first in vivo measurements of intravoxel incoherent motion in the human placenta, obtained using the pulsed gradient spin echo (PGSE) sequence. The aims of this study were two-fold. The first was to provide an initial estimate of the values of the IVIM parameters in this organ, which are currently unknown. The second aim was then to use these results to optimize the sequence timings for future studies. The moving blood fraction (f), diffusion coefficient (D), and pseudo-diffusion coefficient (D*) were measured. The average value of f was 26 +/- 6 % (mean +/- SD), D was 1.7 +/- 0.5 x 10(-3) mm2/sec, and D* was 57 +/- 41 x 10(-3) mm2/sec. For the optimized values of b, the expected percentage uncertainty in the fitted values of f, D, and D* for the placenta were sigmaf/f = 14.9%, sigmaD/D = 14.3%, sigmaD*/D* = 44.9%, for an image signal-to-noise of 20:1, and a total imaging time of 800 sec.
The construction and application of eight different MRI surface coils is described. The coils consist of an anatomically shaped copper wire loop as an antenna and a printed circuit board containing electronic components for tuning and matching. The electronic device for tuning and matching is interchangeable between the various coils. Surface coils for signal detection yield images with high signal-to-noise ratio in comparison to the usual saddle-shaped head or body coils. The sensitivity of a surface coil decreases with increasing distance between the coil and the object of interest and therefore the coils are constructed to fit the anatomical structure under examination as well as possible. The application of dedicated surface coils for superficial structures in the body extends the possibilities of the MRI system. Photographs of the coils positioned on the body and MR images of volunteers and patients are shown.
The echo planar imaging (EPI) method and related variants of this technique can produce complete two-dimensional images from the data collected in a single experiment lasting a fraction of a second. EPI methods are used at 0.5 T to produce snapshot images of the human head with a spatial resolution of less than 2 mm.
Hyperpolarized (3)He MRI of the human lung was performed at 0.54 and 1.5 T using identical software and hardware (except for RF coils) at both field strengths. The T(*) (2) of (3)He gas in the lung was measured, and the effects of magnetic-susceptibility-induced field inhomogeneities on the appearance of interleaved-spiral and interleaved-echo-planar lung images at 1.5 T were compared to those at 0.54 T. Mean T(*) (2) values for (3)He gas in the healthy human lung were 26.8 +/- 1.5 ms and 67.9 +/- 1.3 ms at 1.5 and 0.54 T, respectively. At 0.54 T, interleaved-spiral images showed markedly less blurring due to susceptibility effects compared to images acquired at 1.5 T. At both 0.54 and 1.5 T, interleaved-echo-planar images appeared essentially identical to corresponding GRE images, even though the data-sampling period per echo and echo time were substantially longer for the interleaved-echo-planar images acquired at 0.54 T.
Midfield proton magnetic resonance spectroscopy (MRS) provides a noninvasive method to monitor glutamate and glutamine (Glx) levels in vivo. Experiments to detect the gamma and beta resonances of Glx have been performed by using commercial 0.5 T and 1.5 T MR scanners on seven patients with elevated blood ammonia and eight normal volunteers. Compared with the spectral sensitivity obtained on an otherwise identical system operating at 1.5 T, the singlet resonance of N-acetyl aspartate (NAA) was decreased by a factor of 1.48, which is significantly less than expected using the ratio of Boltzman populations at the two field strengths. However, the resonances of Glx at 0.5 T increased in signal-to-noise ratio (SNR) by a factor of 2. The increased SNR of Glx is principally due to improved B0 main-field homogeneity and collapse of the strongly J-coupled Glx resonances. Our preliminary results suggest that midfield proton MRS will provide significant clinical utility in the detection of Glx levels in human brain.
Echo-planar imaging is used in combination with a spin preparation phase to produce a T1-weighted image. The small additional time penalty in this procedure does not detract significantly from the ultrahigh-speed imaging capability of EPI, allowing a rapid real-time optimization of tissue parameters in the image display. Results obtained on a head demonstrate this technique.
The aim of this study was twofold: First, to establish the normal range of fetal lung diffusion values measured during healthy pregnancy; and second, to determine whether fetal lung diffusion could be used as an indication of fetal lung maturity. The apparent diffusion coefficient (ADC), averaged over all 26 subjects with an average gestational age of 29 +/- 6 weeks (mean +/- sd), was found to be 2.0 +/- 0.6 x 10(-9) m(2)/sec (mean +/- sd), but a trend was found indicating that ADC increased with gestational age at the rate of 0.07 x 10(-9) m(2)/sec per week (P = 4 x 10(-5)). To determine the usefulness of this data in predicting lung maturity, a simple three-compartment model was proposed which was comprised of intra-lung amniotic fluid, intra-tissue water, and vascular blood. The relative proportions of each compartment were taken from the literature, and exchange between the compartments was assumed to be minimal. This model predicted the in vivo data reasonably well, and indicated that MR measurements of fetal lung diffusion are a marker for the degree of vascularization of the terminal tubules. Magn Reson Med 45:247-253, 2001.
Cerebral in vivo proton magnetic resonance spectroscopy of 13 newborn infants displaying seizures and receiving phenobarbitone, in one case supplemented by phenytoin, showed signals from propan-1,2-diol (the injection vehicle for both these anticonvulsants). Subsequent in vitro spectroscopy of cerebro spinal fluid (CSF) from one of these infants also showed signals from this substance. The estimated in vivo propan-1,2-diol concentration (approximately 3 mM) was less than that measured in the CSF sample (14.4 mM). These observations suggest that propan-1,2-diol may accumulate in cerebral tissue and misidentification of its signals in both in vivo and in vitro proton spectra may confuse diagnoses of metabolic or other disorders.
1H NMR spin-lattice relaxation times (T1) of the N-CH3 proton resonances of phosphocreatine (PCr) and creatine (Cr) in water solutions were obtained using the 1,3,3,1 pulse sequence. These T1 values were equivalent to those obtained in D2O and water using either the conventional inversion-recovery experiment or the 1,3,3,1 pulse sequence. Thus, the 1,3,3,1 sequence of proton NMR can provide an independent means along with phosphorous NMR for assess PCr and for the study of the creatine kinase reaction (PCr + ADP in equilibrium ATP + Cr) in aqueous solutions and perhaps in biological preparations.
Noninvasive techniques to monitor temperature have numerous useful biomedical applications. However, MR thermometry techniques based on the chemical shift, relaxation rates, and molecular diffusion rate of the water 1H signal suffer from poor thermal resolution. The feasibility of MR thermometry based on the strong temperature dependence of the hyperfine-shifted 1H signal from the paramagnetic lanthanide complex thulium-1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetate (TmDOTA-) was recently demonstrated. The use of paramagnetic lanthanide complexes for MR thermometry can be further enhanced by improving the signal-to-noise ratio (SNR) of the observed signal. In this study, the use of lanthanide complexes of a methyl-substituted analog of DOTA4-, 1,4,7,10-tetramethyl 1,4,7,10-tetra azacyclodoecane-1,4,7,10-tetraacetic acetate (DOTMA4-) was evaluated. DOTMA4- complexes have 12 magnetically equivalent methyl protons, which provide an intense and sharper resonance compared to the corresponding DOTA- complexes. Experiments with paramagnetic Pr3+, Yb3+, Tb3+, Dy3+, and Tm3+ complexes of DOTMA4- showed that the Tm3+ complex is most favorable for MR thermometery because of the high temperature dependence of its chemical shift and its relatively narrow linewidth. The chemical shift of the methyl 1H signal from TmDOTMA- was approximately 60 times more sensitive to temperature than the water 1H shift and was insensitive to changes in concentration, pH, [Ca2+], or the presence of other ions and macromolecules. The application of TmDOTMA- for measuring temperature in a subcutaneously implanted tumor model was demonstrated. Lastly, the feasibility of obtaining 3D images from the methyl 1H resonance of TmDOTMA- was demonstrated in phantom and live animal experiments. Overall, TmDOTMA- appears to be a promising probe for MR thermometry in vivo.
MR thermometry based on the water (1)H signal provides high temporal and spatial resolution, but it has low temperature sensitivity (approximately 0.01 ppm/degrees C) and requires monitoring of another weaker signal for absolute temperature measurements. The use of the paramagnetic lanthanide complex, thulium 1,4,7,10- tetraazacyclododecane-1,4,7,10-tetramethyl-1,4,7,10-tetraacetate (TmDOTMA(-)), which is approximately 60 times more sensitive to temperature than the water (1)H signal, is advanced to image absolute temperatures in vivo using water signal as a reference. The temperature imaging technique was developed using gradient echo and asymmetric spin echo imaging sequences on 9.4 Tesla (T) horizontal and vertical MR scanners. A comparison of regional temperatures measured with TmDOTMA(-) and fiber-optic probes showed that the accuracy of imaging temperature is <0.3 degrees C. The temperature imaging technique was found to be insensitive to inhomogeneities in the main magnetic field. The feasibility of imaging temperature of intact rats at approximately 1.4 mmol/kg dose with approximately 1-mm spatial resolution in only 3 min is demonstrated. TmDOTMA(-) should prove useful for imaging absolute temperatures in deep-seated organs in numerous biomedical applications.
Localized, water-suppressed (1)H-[(13)C]-NMR spectroscopy was used to detect (13)C-label accumulation in cerebral metabolites following the intravenous infusion of [1,6-(13)C(2)]-glucose (Glc). The (1)H-[(13)C]-NMR method, based on adiabatic RF pulses, 3D image-selected in vivo spectroscopy (ISIS) localization, and optimal shimming, yielded high-quality (1)H-[(13)C]-NMR spectra with optimal NMR sensitivity. As a result, the (13)C labeling of [4-(13)C]-glutamate (Glu) and [4-(13)C]-glutamine (Gln) could be detected from relatively small volumes (100 microL) with a high temporal resolution. The formation of [n-(13)C]-Glu, [n-(13)C]-Gln (n = 2 or 3), [2-(13)C]-aspartate (Asp), [3-(13)C]-Asp, [3-(13)C]-alanine (Ala), and [3-(13)C]-lactate (Lac) was also observed to be reproducible. The (13)C-label incorporation curves of [4-(13)C]-Glu and [4-(13)C]-Gln provided direct information on metabolic pathways. Using a two-compartment metabolic model, the tricarboxylic acid (TCA) cycle flux was determined as 0.52 +/- 0.04 micromol/min/g, while the glutamatergic neurotransmitter flux equaled 0.25 +/- 0.05 micromol/min/g, in good correspondence with previously determined values.
The in vivo pain treatment was successfully performed with the patient in a prone position. The PD-weighted TSE with echo time = 10 ms rendered contrast-to-noise-ratio values of 27 ± 10 for needle/fat, 1.6 ± 5 for needle/muscle, and 4 ± 4.7 for needle/nerve tissue. The mean diameter of the needle artifact was 1.2 ± 0.2 mm. In the T(1)-weighted gradient echo, the needle's artifact diameter was 6 ± 2 mm; the needle's contrast-to-noise ratio relative to muscle tissue was 4 ± 2, 7.6 ± 1.5 for needle/fat, and 5 ± 1 for needle/nerve tissue. With the PD-weighted TSE (echo time = 10 ms) and the T(1)-weighted gradient echo, the needle was imaged reliably throughout the intervention. The butterfly surface coil is feasible for the guidance of spinal interventions in a prone patient.
Spectra obtained with phase-encoding techniques show phase-shifts varying from voxel to voxel. The procedure allowing voxel-dependent phase-shifts to be compensated is presented. The method has been applied to the characterization of the 1.3-ppm resonance observed in intracerebral tumors in the rat.
A single-voxel proton NMR J-difference editing method for discriminating between the 1.31 ppm resonances of lactate (Lac) and threonine (Thr) in human brain in vivo at 3 T is reported. One double-band and two triple-band Gaussian 180 degrees RF pulses, all with a bandwidth of 15 Hz, were employed within an adiabatic-refocused double-echo localization sequence to induce the target signals of Lac and Thr and simultaneously acquire a creatine singlet in each subscan. The optimum echo time and the editing efficiency were obtained by numerical analysis of the filtering performance. The Lac and Thr signals were extracted, without lipid contamination, from three subspectra. Using the calculated yields, the concentrations of Lac and Thr in the human occipital cortex were estimated to be 0.47+/-0.07 and 0.56+/-0.06 mM (mean+/-SD, N=7), respectively, with reference to Cr at 8 mM.
We measured proton magnetic longitudinal (R(1)) and transverse (R(2)) relaxation rates at 1.4T, iron concentrations, water contents, and amyloid plaque densities in postmortem brain tissue samples from three Alzheimer's disease (AD), two possible AD, and five control subjects. Iron concentrations and R(1) were significantly higher in the temporal cortex region of our AD group compared to the controls. Frequency analyses showed that the observed trends of higher iron, R(1), and R(2) in AD gray matter regions were statistically significant. Simple regression models indicated that for AD and control gray matter the iron concentrations and water contents have significant linear correlations with R(1) and R(2). Multiple regression models based on iron concentrations and water contents were highly significant for all groups and tissue types and suggested that the effects of iron become more important in determining R(1) and R(2) in the AD samples. At 1.4T R(1) and R(2) are strongly affected by water content and to a lesser extent by variations in iron concentrations. The AD plaque density did not correlate with iron concentrations, water contents, R(1), or R(2), suggesting that increases in AD brain iron are not strongly related to the accumulation of amyloid plaques.
The signal-to-noise ratio in hyperpolarized noble gas MR imaging is expected to be independent of field strength at frequencies typical of clinical systems (e.g., 1.5 T), where body noise dominates over coil noise. Furthermore, at higher fields (e.g., 3 T), the SNR of lung images may decline due to decreases in T(2) originating from increases in susceptibility-induced field gradients at the air-tissue interface. In this work, the SNR of hyperpolarized (3) He lung imaging at two commonly used clinical field strengths (1.5 T and 3 T) were compared in the same volunteers. Thermally polarized and hyperpolarized (3) He phantoms were used to account for differences in MR imaging system and (3) He polarizer performance, respectively, at the two field strengths. After correcting for T(2) values measured at 1.5 T (16 ± 2 ms) and 3 T (7 ± 1 ms), no significant difference in image SNR between the two field strengths was observed, consistent with theory.
In 31P-(1H) MR experiments of humans in a 1.5-T whole-body system, signal intensity enhancements of 31P resonances of up to 68 +/- 4% (for phosphocreatine of the calf muscle) have been observed upon irradiation at proton frequency. This observation is explained as a nuclear Overhauser effect due to the dipolar coupling between 1H and 31P spins.
Applying an active intravascular MR catheter device that allows signal transmission from the catheter tip requires special means to avoid radiofrequency-induced heating. This article presents a novel, miniaturized all-optical active MR probe to use with real-time MRI in minimally invasive interventions for catheter guidance and intravascular imaging. An optical link transmits the received MR signals from the catheter tip to the MR receiver with inherently radiofrequency-safe optical fibers. Furthermore, power is supplied optically to the transmitter as well. The complete integration into a small tube of 6-Fr (2-mm diameter) size with a 7-Fr (2.33-mm diameter) rigid tubing was realized using chip components for the optical modulator and a novel miniaturized optical bench fabricated from silicon substrates with 3D self-aligning structures for fiber integration. In MRI phantom measurements, projection-based tip tracking and high-resolution imaging were successfully performed with the optical link inside a 1.5-T MRI scanner. Images were obtained in a homogeneous phantom liquid, and first pictures were acquired from inside a kiwi that demonstrates the potential of the MR-safe optical link. The signal-to-noise ratio has significantly improved compared with former systems, and it is demonstrated that the novel optical link exhibits a signal-to-noise ratio comparable to a direct electrical link.
Cardiac echo-planar imaging suffers invariably from regions of severe distortion and T*2 decay in the myocardium. The purpose of this work was to perform local measurements of T*2 and field inhomogeneities in the myocardium and to identify the sources of focal signal loss and distortion. Field inhomogeneity maps and T*2 were measured in five normal volunteers in short-axis slices spanning from base to apex. It was found that T*2 ranged from 26 ms (SD = 7 ms, n = 5) to 41 ms (SD = 11 ms, n = 5) over most of the heart, and peak-to-peak field inhomogeneity differences were 71 Hz (SD = 14 Hz, n = 5). In all hearts, regions of severe signal loss were consistently adjacent to the posterior vein of the left ventricle; T*2 in these regions was 12 ms (SD = 2 ms, n = 5), and the difference in resonance frequency with the surrounding myocardium was 70-100 Hz. These effects may be caused by increased magnetic susceptibility from deoxygenated blood in these veins.
Reproducibility of functional MRI (fMRI) data has been controversial. This issue was examined in this study by evaluating a strictly controlled voluntary force-matching handgrip task. Handgrip force, electromyogram (EMG), and fMRI data of brain activity were simultaneously recorded during the task performance. The task was repeated three times in each of the two experimental sessions. While force remained unchanged and EMG showed little variation across trials and sessions, the results revealed that fMRI-measured brain signals varied significantly among individual trials. However, the averaged fMRI signals over the three trials did not show significant difference between the two sessions. Our data suggest that fMRI is better at defining brain activation qualitatively than quantitatively, i.e., the locations of the activation areas could be reproduced quite reliably while their sizes fluctuated substantially.
When patients with metallic prosthetic implants undergo an MR procedure, the interaction between the RF field and the prosthetic device may lead to an increase in specific absorption rate (SAR) in tissues surrounding the prosthesis. In this work, the distribution of SAR(10g) around bilateral CoCrMo alloy hip prostheses in situ in anatomically realistic voxel models of an adult male and female due to RF fields from a generic birdcage coil driven at 64 or 128 MHz are predicted using a time-domain finite integration technique. Results indicate that the spatial distribution and maximum values of SAR(10g) are dependent on body model, frequency, and the position of the coil relative to the body. Enhancement of SAR(10g) close to the extremities of a prosthesis is predicted. Values of SAR(10g) close to the prostheses are compliant with recommended limits if the prostheses are located outside the coil. However, caution is required when the prostheses are within the coil since the predicted SAR(10g) close to an extremity of a prosthesis exceeds recommended limits when the whole body averaged SAR is 2 W kg(-1) . Compliance with recommended limits is likely to require a reduction in the time averaged input power.
Aerosol toxicology and drug delivery through the lungs, which depend on various parameters, require methods to quantify particle deposition. Intrapulmonary-administered MRI contrast agent combined with lung-specific imaging sequences has been proposed as a high performance technique for aerosol research. Here, aerosol deposition is assessed using ultra-short echo (UTE) sequences. Before and after administration of Gd-DOTA-based aerosol delivered nose-only in free-breathing healthy rats, a T1 -weighted 3D UTE sequence was applied in a clinical 1.5 Tesla scanner. Administration lasted 14 min, and the experiment was performed on six rats. A contrast-enhanced quantitative analysis was done. Fifty percent signal enhancement was obtained in the lung parenchyma. Lung clearance of the contrast agent was evaluated to be 14% per h (corresponding to a characteristic clearance time of 3.6 h) and aerosol deposition was shown to be homogeneous throughout the lung in healthy rats. The total deposited dose was estimated to be 1.05 µmol/kg body weight, and the concentration precision was 0.02 mM. The UTE protocol with nebulized Gd-DOTA is replicable to significantly enhance the lung parenchyma and to map aerosol deposition. This functional strategy, applied in a clinical system with a clinical nebulization setup and a low inhaled dose, suggests a feasible translation to human. Magn Reson Med, 2015. © 2015 Wiley Periodicals, Inc. © 2015 Wiley Periodicals, Inc.
Wideband steady-state free precession (WB-SSFP) is a modification of balanced steady-state free precession utilizing alternating repetition times to reduce susceptibility-induced balanced steady-state free precession limitations, allowing its use for high-resolution myelographic-contrast spinal imaging. Intertissue contrast and spatial resolution of complete-spine-coverage 3D WB-SSFP were compared with those of 2D T(2) -weighted fast spin echo, currently the standard for spine T(2) -imaging. Six normal subjects were imaged at 1.5 and 3 T. The signal-to-noise ratio efficiency (SNR per unit-time and unit-volume) of several tissues was measured, along with four intertissue contrast-to-noise ratios; nerve-ganglia:fat, intradural-nerves:cerebrospinal fluid, nerve-ganglia:muscle, and muscle:fat. Patients with degenerative and traumatic spine disorders were imaged at both MRI fields to demonstrate WB-SSFP clinical advantages and disadvantages. At 3 T, WB-SSFP provided spinal contrast-to-noise ratios 3.7-5.2 times that of fast spin echo. At 1.5 T, WB-SSFP contrast-to-noise ratio was 3-3.5 times that of fast spin echo, excluding a 1.7 ratio for intradural-nerves:cerebrospinal fluid. WB-SSFP signal-to-noise ratio efficiency was also higher. Three-dimensional WB-SSFP disadvantages relative to 2D fast spin echo are reduced edema hyperintensity, reduced muscle signal, and higher motion sensitivity. WB-SSFP's high resolution and contrast-to-noise ratio improved visualization of intradural nerve bundles, foraminal nerve roots, and extradural nerve bundles, improving detection of nerve compression in radiculopathy and spinal-stenosis. WB-SSFP's high resolution permitted reformatting into orthogonal planes, providing distinct advantages in gauging fine spine pathology. Magn Reson Med, 2012. © 2012 Wiley Periodicals, Inc.
Quantitative interpretation of BOLD fMRI signal changes has predominantly employed empirical models for the whole parenchyma and a calibration step is usually needed to determine the physiological parameters during activation. Although analytical expressions are available for the extravascular and intravascular components of the BOLD effects, it is difficult to experimentally separate tissue from blood signal contributions at the low magnetic fields in which most fMRI studies are performed. Even if this can be achieved, an additional problem that remains is the separation of two types of extravascular BOLD effects, namely those around microvasculature (in the parenchyma close to the site of activation) and those around draining macrovasculature (e.g., in tissue and CSF more remote from the site of activation). In the recently developed vascular space occupancy technique, blood signals are nulled and the activations are localized predominantly in gray matter, allowing experimental measurement of parenchymal extravascular R(2)* and its changes accompanying activation. When comparing such data with total parenchymal R(2)* changes in BOLD fMRI, the extravascular fractions were found to be 47 +/- 7% (mean +/- SEM, n = 4) and 67 +/- 6% at 1.5 and 3.0 T, respectively, in line with expectations that intravascular BOLD contributions are reduced at higher field. The present approach provides a noninvasive means to determine parenchymal oxygen extraction fraction (OEF) in situ. During visual stimulation, OEF values measured at 1.5 and 3.0 T were in good agreement, giving 0.23 +/- 0.01 and 0.21 +/- 0.01, respectively.
This manuscript describes a number of sources of nonuniformity for spin echo images at 1.5 T. Both coil tuning and crosstalk can have significant effects on image nonuniformity. For short repetition times, nonuniformity increases with decreasing TR, possibly due to gradient eddy currents. In sections of RF coils with poor RF uniformity, image nonuniformity varies with both echo time and the number of echoes in a multiecho sequence. For the particular imager used, there are small differences between transverse and sagittal/coronal nonuniformity. The temporal stability of image nonuniformity is very good. The use of uniform oil phantoms is shown to be superior to low pass filtered images for correction of image nonuniformity.
Purpose: Voltage-based device-tracking (VDT) systems are commonly used for tracking invasive devices in electrophysiological cardiac-arrhythmia therapy. During electrophysiological procedures, electro-anatomic mapping workstations provide guidance by integrating VDT location and intracardiac electrocardiogram information with X-ray, computerized tomography, ultrasound, and MR images. MR assists navigation, mapping, and radiofrequency ablation. Multimodality interventions require multiple patient transfers between an MRI and the X-ray/ultrasound electrophysiological suite, increasing the likelihood of patient-motion and image misregistration. An MRI-compatible VDT system may increase efficiency, as there is currently no single method to track devices both inside and outside the MRI scanner. Methods: An MRI-compatible VDT system was constructed by modifying a commercial system. Hardware was added to reduce MRI gradient-ramp and radiofrequency unblanking pulse interference. VDT patches and cables were modified to reduce heating. Five swine cardiac VDT electro-anatomic mapping interventions were performed, navigating inside and thereafter outside the MRI. Results: Three-catheter VDT interventions were performed at >12 frames per second both inside and outside the MRI scanner with <3 mm error. Catheters were followed on VDT- and MRI-derived maps. Simultaneous VDT and imaging was possible in repetition time >32 ms sequences with <0.5 mm errors, and <5% MRI signal-to-noise ratio (SNR) loss. At shorter repetition times, only intracardiac electrocardiogram was reliable. Radiofrequency heating was <1.5°C. Conclusion: An MRI-compatible VDT system is feasible.
The literature has conflicting reports concerning the effect of static magnetic fields on body and skin temperatures in mammals. Since temperature changes induced by static magnetic fields would have important safety implications for clinical magnetic resonance imaging body (sublingual pocket) and skin (abdomen, forehead, chest, upper arm, forearm, thigh, and calf) temperatures were determined in six normal subjects using a fluoroptic thermometry system during a 20-min exposure to a 1.5-T static magnetic field. Ambient conditions were controlled and held constant. An analysis of variance for repeated measures revealed that there were no statistically significant changes in body or any of the skin temperatures recorded. We conclude that exposure for 20 min to a 1.5-T static magnetic field does not alter body and skin temperatures in man.
Top-cited authors
James Hyde
  • Medical College of Wisconsin
Bharat Biswal
  • New Jersey Institute of Technology
Zerrin f Yetkin
  • University of Texas Southwestern Medical Center
Victor M Haughton
  • University of Wisconsin–Madison
Bruce Rosen
  • Massachusetts General Hospital