Data obtained from in-vivo and in-vitro measurements on human skin and skin samples are presented. In-vivo measurements at 3 T on the forearm of three female volunteers show an increase of signal intensity, after application of cream, between 30 and 58%, depending on the cream type. In plane resolution was 90 m, with slice thickness of 1 mm. Chemical-shift-selective in-vitro microimaging studies were performed on a 9.4-T spectrometer using human skin samples from the abdomen. In these experiments, in-plane resolution was 6 m and slice thickness was 100 m. Results demonstrate clearly the moisturizing effect of beauty care products on the stratum corneum, which can be identified in in-vitro experiments. No ingress into the skin of the lipid components of the emulsions could be observed in-vitro. However, skin occlusion and consequently a long-lasting moisturizing effect cannot be excluded. These results prove the capability of MRI to investigate human skin in-vivo and in-vitro and to follow the skin penetration of beauty-care products and pharmaceutical ointments.
There is a considerable need for therapies that could prevent restenosis following angioplasty procedures in humans. Several different animal models of neointimal formation are used to evaluate efficacy of novel pharmacologic and physical (stents) interventions. The standard method of evaluation of blood vessels from these models is histomorphometric analysis. The determinant index for measuring the extent of injury and response to treatment is the intima/media ratio, obtained from media and intima area measurements of 2-6 sections made at fixed distances along the length of the vessel. We have used high resolution 3D MR microscopy to evaluate the remodeling (favorable and/or unfavorable) in the coronary(pig), iliac(rabbit) and carotid (rat) arteries following endothelial injury (stent or balloon) and treatment. MR images provide clear visualization of the vessel wall as well as lumen and a quantitative evaluation of various parameters such as total area and volume measurements of wall and lumen, maximal wall thickness, lumen to wall ratios, etc. In addition, critical information on the location and extent of damage can be accurately determined. All MR experiments were carried out on a Bruker AMX-400 equipped with a sample changer. Volume RF coils of various sizes, depending on the diameter of the vessel, were used for data acquisition. Images were acquired using 3D spin-echo sequences with varying TE/TR values, optimized for best contrast. Quantitative measurements were made using "Analyze" on a SGI workstation. All vessels were then processed for histological analysis. The results demonstrate that MR microscopy offers unique advantages in evaluating these disease models.
Multidimensional NMR methods are most commonly applied to structural studies of macromolecules in solution, yet little use has been made of such methods in metabolic studies involving stable isotopes. In this study, we have applied a well known 2D version of the HMQC-TOCSY experiment to monitor intermediary metabolism by 13C isotopomer analysis of glutamate. The HMQC scheme selects1H coherences correlated with labeled aliphatic carbons and the isotropic mixing period in the TOCSY scheme allows nearly homogeneous coherence transfer between all protons in the glutamate molecule. In order to extend the amount of isotopomer information derived from the HMQC-TOCSYexperiment, glutamate was reduced to the 2-amino-1,5-pentanodiol derivative. By converting all glutamate carbons to aliphatic carbons, this has allowed us to conveniently extend the spin network to glutamate C1 and C5. Isotopomer quantitation was provided by integration and normalization of the HMQC-TOCSY cross peaks. Normalizing factors were obtained from the HMQC-TOCSY spectra of a known mixture of [U-13C5]-, [1-13C]- and [4-13C]-2-amino-1,5-pentanodiol (1:1:1). Finally, a complete isotopomer analysis was performed using HMQC-TOCSY data on glutamate isolated from perchloric acid extracts of rat hearts perfused with a mixture of 12C-, 1-13C-, 2-13C and [1,2-13C2]-acetate (1:1:1:1).
Several methods have been developed for reducing the effects of motion artifacts in MRI including echo planar imaging, respiratory and cardiac gating, and the use of navigator pulses. This work presents an alternative method for reducing artifacts that uses feedback control. The motion of the object is predicted based upon measurements of the NMR signal. The spectrum of the first spin echo is used as a reference signal. Object motion relative to the reference signal for subsequent spin echos is evaluated using an autocorrelation function. The location of the object is predicted using a one-step-ahead predictor based on a Nth - order autoregressive model whose parameters are updated following each spin echo acquisition using a recursive least-squares algorithm. The predicted motion is then used to automatically adjust the magnetic field gradients so that the object remains centered in the field of view. Initial results based on computer simulations are presented and show that the algorithm is effective at reducing motion artifacts.
Studies of tumor angiogenesis by MR imaging can benefit from the use of blood pool-specific contrast agents that remain in circulation for several days. No such intravascular contrast agent, however, is currently available for MRI. Radioactive 51Cr, which enters the erythrocyte as chromate and remains there as Cr(III) after in situ reduction, has long been used to tag human red blood cells in nuclear medicine. We evaluated the potential of non-radioactive, paramagnetic Cr(III) as an intravascular contrast agent for studying tumor angiogenesis in rat brain by proton MRI. Temperature, pH and incubation time for optimum uptake of chromate into rat erythrocytes were established. Changes in water proton relaxation times of erythrocytes after chromium-labeling were measured. Water T1 of unlabeled erythrocytes is 1.5s at 200 MHz. T1 decreased linearly with increasing concentration of chromium used for labeling: from 320 msec for [Cr] = 1mM, to 60 msec for [Cr] = 7mM, and leveled off for [Cr] > 7mM. Water T2 of erythrocytes also decreased upon Cr-labeling. The lower T1 of Cr-labeled erythrocytes in the blood pool was exploited to provide contrast in T1-weighted images. Erythrocytes treated with 3-5mM Cr provided adequate reduction in T1 of systemic blood upon reinjection into the rat, and at the same time avoided harmful levels of oxidation. Good contrast between unlabeled blood (T1 = 1.35s), chromium-labeled blood (T1 = 700 msec) and rat brain (T1 = 1.35s) were seen in vitro in inversion recovery images. Rate of depletion of chromium from circulation and accumulation of chromium in different organs were measured. No measureable accumulation of Cr was detected in brain.
In spite of extensive research on changes in the human eye lens that occur with age, the causes of presbyopia and senile cataract remain unclear. In this project magnetic resonance micro-imaging has been used to study the kinetics of water transport in intact human eye lenses. Transport of water, together with nutrients and anti-oxidant species, from the cortex to the nucleus, is vital for maintaining the structures of the proteins (crystallins) in the lens nucleus, because the nucleus (being metabolically inactive), is unable to re-generate them. A deficiency in this transport may therefore be responsible for degradation or disruption of the crystallins, leading to the onset of presbyopia (in most individuals) and cataract (in some cases) with age. While conventional pulsed field gradient (PFG) methods allows study of "short range" diffusion (distances comparable to cell dimensions) it does not allow the study of time averaged "long range" diffusion (distances comparable to lens dimensions), important for the transport of lens anti-oxidants. Presented is a novel approach to studying long range water transport and diffusion in human eye lenses. Lenses were stored at 34.5 oC in artificial aqueous humour (AAH) containing nutrients and metabolites similar to those present in vivo. MR images were acquired over approximately a twenty hour period following replacement of H2O based AAH with deuterium oxide (D2O) based AAH. NMR signal intensity from the lenses decreased with time corresponding to a decrease in concentration of H2O within the lenses. A statistically significant correlation (p <0.001) was found between the rate of NMR signal loss from the lens nuclei and increased age of the lenses. The results show that as lenses age there is a reduction in the rate at which water (and presumably also water soluble low molecular weight metabolites) can enter the cells of the lens nucleus via the epithelium and cortex. A decrease in the rate of transport of water, nutrients and anti-oxidant species would be expected to lead to increased damage to lenses with age, and ultimately a potential cause of presbyopia and senile cataract.
The physiology of tumors and the response to therapy is determined by a continuum of factors, but changes in blood flow and vessel architecture can be the ultimate source of other microenvironmental changes such as pH and pO2. We have investigated the application of motion sensitive pulsed-gradient spin-echo (PGSE) MR experiments (IVIM method) as a means to continuosly monitor tumor blood flow during the course of therapeutic manipulations. The endogenous water signal is not suitable for use in tumors because water transport by flow is not very different from transport by diffusion. The 19F PGSE signal (stimulated-echo in most cases) in experimental tumors following i.v. administration of perfluorocarbon (PFC) vescicle suspensions is observed to decay with increased flow-weighting in all of nine tumors examined, and transport by diffusion is insignificant. The largest flow-weighting applied is large enough to supress 80% of the signal from capillaries with a mean velocity of 0.085 mm/s. S:N ratios of the double-projection images are typically 40-80:1, and we estimate <0.5ul of pure PFC in each 50ul voxel. The data provide the opportunity to examine distributions of blood velocities. Much of the apparent distribution in blood velocity can be explained by simply considering a distribution of blood vessel diameters, as Henkelman, et al. have done previously in brain [1], although the data from some tumors suggest additional sources of distributions. Over the 24-48 hour period following administration of PFC, the apparently non-flowing fraction (that portion of the signal remaining in the presence of large flow-weighting) is observed to increase, corresponding to extravasation of the PFC. Extravasation is sufficiently slow to allow continuous observation of intravascular-PFC flow for several hours. Changing the breathing gas to 95%O2/5%CO2, which is known to increase tumor blood flow, caused the signal decay rate to increase by about 30% for the more rapidly flowing components, but not for the slower components. 1) R.M. Henkelman,J.J. Neil,Q.-S. Xiang Magn. Res. Med. 32:464 (1994). Supported by NIH CA36856, CA52586, CA11198, CA68409
Non-invasive MRI techniques using the Rotating Ultra-Fast Imaging Sequence (RUFIS) have been applied to characterize curved flow inside a U-tube. Longitudinal images (with no slice selection) were acquired after tagging spins in a 0.5 cm slice and allowing their migration during a set delay to demonstrate the variation of spin velocities across the internal diameter of the curved flow tube. Velocity (cosine) encoded images were acquired for a 0.5 cm slice transverse to the flow direction. The RUFIS images were acquired in 22 ms and processed using Singular Value Decomposition to obtain the 1D radial projections from the FID data, followed by filtered back-projection. Velocity contours were obtained by taking the arccosine of the ratio of images in the presence and absence of flow. Velocity contours of curved flow obtained by MRI were compared with velocity contours obtained by others using Laser Doppler Anemometry (LDA) for flow at various Dean Numbers, where the Dean Number = 2(Reynolds Number)(ID/turn diameter)0.5. The velocity contours were obtained at the 00, 900, and 1800 observation points for a 14 mm ID glass U-tube phantom with a 34 mm turn diameter. The MRI images and velocity contours confirm that as fluid moves through a curved channel, faster moving fluid moves away from the center of the channel toward the outer channel wall. The migration of fluid toward the outer wall of the curve is a result of secondary flows caused by centrifugal forces.
MR images of the thorax often contain blurring and ghosting caused by respiration and cardiac motion. Segmented acquisition, gating, and motion-insensitive k-space trajectories have been used in the past to minimize these motion artifacts. Alternatively, ultra-fast imaging sequences with short acquisition times in comparison with the rate of motion can effectively freeze the desired object. Recently, we have begun demonstrating the utility of the Rotating Ultra-Fast Imaging Sequence (RUFIS) for in vivo imaging of the heart and lungs of a Sprague Dawley rat. The rat was anaesthetized and paralyzed, and placed on a ventilator to permit respiratory gating. Signal interference associated with using ECG wires inside the magnet prevented ECG gating. A 0.5 cm transverse slice was imaged through the rat thorax. Images were acquired in 22 ms using a 5 cm ID passively-shielded Parallel Cosine (PCOS) current distribution RF coil. The imager included a 2.35 T, 31-cm bore Magnex magnet with a homebuilt 8.5 cm internal diameter actively-shielded gradient set. Images were reconstructed using Singular Value Decomposition to obtain 1D projections followed by filtered back-projection to obtain the 2D image. The resultant images contained detail observed in spin warp images obtained from the same rat but without the distortions and artifacts caused by motion.
MRI ultra-fast imaging techniques are employed to characterize flow emerging from both streamlined and abrupt stenoses located inside 8 mm internal diameter cylindrical channels. Reattachment lengths of the shear boundary to the channel wall were measured using the Rotating Ultra-Fast Imaging Sequence (RUFIS). RUFIS uses low flip angle RF pulses, samples k-space using radial projections, and acquires FIDs. RUFIS images were acquired in less than 15 ms and reconstructed using Singular Value Decomposition to obtain 1D projections and filtered back-projection to obtain a 2D image. Reattachment lengths were measured from longitudinal images of in-flowing spins without slice selection. Velocity profiles of flow were created by obtaining velocity (sine and cosine) encoded RUFIS images for a 0.5 cm slice transverse to the primary flow direction. The sine-encoded images permitted us to identify reverse flows (i.e., eddies) that arise within the region of flow reattachment. The ratios of peak velocities (downstream/upstream of the stenosis) derived from the cosine-encoded images were used to identify the transition from the laminar to the turbulent regime. Based on these experiments, the transition from the laminar to turbulent regime occurs at a stenotic Reynolds Number of 350 while fully developed turbulence occurs at a stenotic Reynolds Number of 2600. These results were compared with the results from invasive flow studies.
The two most powerful medical imaging techniques are positron emission tomography (PET) and MRI. The major strengths and weaknesses of these are rather complementary. The PET technique is extremely sensitive, quantitative, and chemically targeted: MRI is not chemically sensitive, and usually not quantitative. The spatial and temporal resolution of PET is rather poor (a few mm, and minutes): MR images made from the strong, ubiquitous 1H2O signal can have sub-mm spatial and sub-second temporal resolution. These two modalities can be combined so as to take advantage of the strengths of each. We have been exploring the intrinsic combination of real space or Fourier space PET and MR data. 18F-labeled deoxyglucose (FDG) PET and 1H2O MR brain images of the same volunteer subject were acquired, and co-registered using the GALAXY method.1 The PET image shows predominant intensity in grey matter pixels because of a greater degree of metabolic activity. The grey matter anatomy is, however, not very evident. A relaxographic MR image that quantitatively segments grey matter water with high resolution,2 was combined with the PET image. The pet pixel color was used to color the underlying MRI gray-scale pixels to give a PETAMR image. The resultant grey matter-edited FDG image shows the metabolic information of the PET image and the high spatial resolution of grey matter from the MR image. 1. A. V. Levy, D. L. Alexoff, F. Hode, M. Denis, D. Bertollo, A. P. Dhawan, J. Logan, A. B. Andrews and N. D. Volkow Proc. IEEE/EMBS 18 452 (1996). 2. I. Palyka, J-H. Lee, K. Ugurbil, M. G. Garwood, and C. S. Springer Proc. ISMRM 4 1642 (1996).
Information about biological tissue compartmentalization cannot be readily obtained from the frequency of the isochronous water proton NMR signal. Other NMR properties, like relaxation times (Tis), must be used for discriminating sub-voxel resonances. It has long been known that the (formal) inverse Laplace transformation (ILT) of an NMR time domain signal produces the distribution of Ti values - what we call the relaxogram1 - describing that signal. Images edited for Ti information from discrete portions of the relaxogram exhibit natural segmentation.2 Due to the often effectively continuous Ti distribution in the relaxogram, it is difficult to discern discrete Ti peaks. Thus, we have been developing curve-resolving approaches to relaxographic analyses that introduce minimal error. We have found this to be quite useful in the comparison of relaxograms to detect changes. For example, we have compared relaxograms and relaxographic images of the same brain slice of the same individual obtained at 1.5 T and at 4.0 T. Since the mechanistically informative field-dependence of relaxation times has long been termed relaxometry, we refer to this as relaxometric relaxographic imaging. (1) C. Labadie, J-H. Lee, G. Vetek, and C. S. Springer, J. Magn. Reson. B , 105 99-112 (1994) (2) I. Palyka, J-H. Lee, K. Ugurbil, M. G. Garwood, and C. S. Springer, Proc. Int. Soc. Magn. Reson. Med. 4 1642 (1996)
A pulsed Electron Paramagnetic Resonance spectrometer operating at 300 MHz frequency corresponding to a resonant magnetic field of 10 mT is described. The transmit arm is capable of generating pulses of 10-100 ns duration which are further amplified using a 200 Watt amplifier. The resonators used has Q-values typically in the range of 25-30 to encompass the spectral bandwidths of >10 MHz. Resonators capable of accommodating small experimental animals were designed and tested. In the detection scheme, two configurations are used. The first configuration involves mixing the induction signal with the carrier (300 MHz) reference requiring the use of quadrature mixer and dual channel data acquisition to generate pure absorption line shapes and retain sign information. CYCLOPS phase cycling is required to eliminate artifacts associated with quadrature detection and systematic noise from the spectrometer. In the second configuration, two frequencies namely 300 MHz and 350 MHz are derived from a single phase locked source and the induction signal is detected as intermediate frequency centered around 50 MHz. A band width of up to 100 MHz can be provided by such configuration, without the requirement of CYCLOPS phase cycling since all resonance frequencies below 100 MHz would be positive. The efficiencies and the relative advantages of the two configurations in the detection scheme is presented. We have successfully tested this spectrometer to image phantoms and small experimental animals infused with the paramagnetic spin probes which have their spin-spin relaxation times in the order of several microseconds. Imaging of these spin probes was carried out using filtered back projection using static gradients. The use of this EPR imaging in vivo to obtain physiologic information such as tissue oxygenation status will also be presented.
The diffusion coefficient and relaxivity of paramagnetic compounds in gels and cartilage is measured using a magnetic resonance 1-dimensional imaging technique. The paramagnetic compound to be studied can be either an element such as copper or gadolinium, or the paramagnetic element can be bound to a larger molecule such as a protein(1). The technique relies on the effect of the paramagnetic on the T1 relaxation properties of the surrounding water. Thus as the concentration of the paramagnetic rises, due to diffusion of the paramagnetic into a material, the NMR signal experiences a faster relaxation rate. Acquisition of the NMR signal is performed using an inversion recovery pulse sequence. The technique is an extension of a previously published NMR technique (2,3), with the additional capability of measuring the relaxivity of the paramagnetic in the gel or cartilage matrix. The accuracy of the technique is verified using mathematical simulations and experimental diffusion of CuSO4 into agarose. Data on small and large paramagnetic molecule diffusion in articular cartilage has also been obtained. (1) Lauffer, RB, Brady TJ. Magn. Reson. Imag. 3: 11-16, 1985. (2) Balcom, BJ, Fischer, AE, Carpenter, TA, Hall, LD. J. Am. Chem. Soc. 115: 3300-3305, 1993. (3) Potter, K, Spencer, RGS, McFarland, EW. Biochim. Biophys. Acta 1334: 129-139, 1997.
Signal attenuation curves from NMR diffusion measurements in yeast suspensions show non-monoexponential signal decay which presumably represents the separate contributions from intra- and extracellular water.1 However, signal-attenuation curves having a similar appearance can also be obtained fromsingle compartment models, such as polystyrene bead packs saturated with water,2 where restriction effects alone account for the deviation from monoexponential behavior. In an attempt to deconvolve these restriction effects, we are investigating combined relaxography and diffusion measurements to unequivocally differentiate between compartmental diffusion constants. Intra- and extracellular compartments in yeast suspensions are separated on the basis of T1 relaxation by adding a contrast reagent to the extracellular space. Data from a series of diffusion-weighted inversion-recovery experiments are used to calculate the diffusion coefficients from the individual compartments. In preliminary studies, diffusion coefficients measured using this approach differ significantly from those obtained from direct multiexponential fitting of the diffusion attenuation curve alone. These differences are believed to be due to a convolution of restriction effects with the compartmental contributions to diffusion. Combined relaxography and diffusion measurements thus allow determination of separate compartmental diffusion in the absence of confounding restriction effects. 1. Gabor Vetek, Ildiko Palyka, Christopher H. Sotak, Charles S. Springer Jr., CR-Free Discrimination of Intra- and Extracellular 1H20 Signals from Yeast Cell Suspensions by Diffusion-Space Relaxography (Diffusigraphy), Abstract, Second Meeting of the Society of Magnetic Resonance, San Francisco, CA, August, 1994; also published in Proc. Soc. Magn. Reson. 2, 1051 (1994). 2. Karl G. Helmer, Bernard J. Dardzinski, Christopher H. Sotak, The Application of Porous-Media Theory to the Investigation of Time-Dependent Diffusion in In Vivo Systems, NMR Biomed. 8, 297-306 (1995).
H217O in is a freely diffusable tracer attractive for in vivo blood flow measurements. Disadvantages include: low sensitivity of oxygen 17 NMR, low natural abundance of H217O (0.035%), high cost of oxygen 17 enriched compounds. Improved sensitivity has been reported with proton detection using H217O as a T2 shortening contrast agent (via an exchange modulated scalar coupling through the hydrogen bond). This interaction is suppressed by 17O decoupling, and the proton NMR signal with/without decoupling can be directly related to H217O concentration. We have exploited this effect using an alternating "on-off" decoupling scheme, similar to the periodic stimulation paradigm used in fMRI. We used an fMRI type cross correlation analysis to image very low concentrations of H217O that could be masked by signal drift, motion or scanner instability. Experiments were performed on a rat stroke model (suture MCAO) using a 2T GE/Bruker Omega system with a proton transmit/receive surface coil over the brain, and an orthogonal Helmhotz oxygen coil. CW decoupling was applied either side of the 180o pulse in a spin-echo snapshot echo-planar imaging (EPI) sequence (TR 2s, TE 90ms). 1ml of 10% or 20% atm. H217O was injected i.v. as a bolus. A series of 128 EPI images were acquired with the decoupler power switched on or off every 8 images. Signal vs. time curves at each pixel position were cross correlated with a square wave (period 16 images) and a map of the correlation coefficient 'r' generated (the higher the oxygen 17 concentration the better the correlation). The 'r' maps clearly detected H217O in normal brain and could differentiate ischemic regions with very low blood volume and flow. A small correlation was observed even before injection, indicating the detection of natural abundance H217O. This method can detect very low H217O concentrations and may be useful in studies of cerebral perfusion and metabolism using oxygen 17 labeled water and gas.
Microimaging experiments using a 5 mm high-temperature superconductive (HTS) probe and, with identical imaging parameters, a conventional probe, were performed at the National Cancer Institute - Frederick Cancer Research and Development Center (NCI-FCRDC), Frederick, MD, on a 300 MHz NMR system (89 mm bore magnet, Unity INOVA console and, on loan from Varian, microimaging hardware: 50 G/cm, unshielded gradients.) The promise of HTS probes is in significantly increased sensitivity over conventional imaging probes allowing for faster data acquisition and/or increased resolution. The HTS probe is based on the Conductus/Bruker Instruments 5 mm spectroscopy probe, utilizing a superconductive Helmholtz-style pair of coils patterned from thin-film Y1Ba2Cu3O7-d, and modified to reduce eddy currents. The Q of the coils, matched to a 50 ohm load, is > 25,000. The 25 K operational temperature was achieved through a simple open-cycle liquid He flow system, consuming about 1 L/hr. The coils and probe were manufactured at Conductus in cooperation with Bruker Instruments. The conventional coil (Q ~ 400) is a 12 mm diameter, 25 mm long, single-turn solenoid constructed at the NCI-FCRDC from microwave substrate, and is inductively coupled to both a separate receive and tuning loop. Signal to noise ratio measurements were made on images from 3-D acquisitions of a section of rat cerebellum (TR 250 ms, TE 30 ms, 63 micron resolution, 1.1 hour acquisition time) and a mouse spinal column (TR 300 ms, TE 15,30 ms, 352 x 70 micron resolution, 1.4 hour acquisition time). In the rat cerebellum, SNR values ranged from 1.9 to 6.6 with the HTS probe, and from 1.0 (noise level) to 1.5 with the solenoid. In the spinal column, SNR values were ~ 1.0 (noise level) with the solenoid and were ~ 4 (TE = 15 ms), and 4 - 5 (TE = 30 ms) in white matter, and between 3 and 4 (TE = 15 ms), and 7-9 (TE = 30 ms) in gray matter. Thus, the HTS probe showed as much as a six-fold increase in SNR for the rat cerebellum and a nine-fold increase in the mouse spinal column images. This work was partially supported by NIH grant 2 R44 RR09244-02.
Although the metabolic information from short echo in-vivo 1H spectroscopy is invaluable in the study of many disease processes, precise quantification of such specrta is limited by severe spectral overlap and low signal to noise ratio (SNR). Since spectral peak separation, SNR, susceptibility artifacts, and T1 relaxation all increase with magnetic field strength, it is not clear if significant benefit can be gained by acquiring data at high fields (4.0 Tesla) compared to lower fields (1.5 Tesla). In this study the precision of metabolite estimates from STEAM spectra acquired from the parietal lobe of the same individual at 1.5 Tesla (2x2x2 cc, TR=1.5 s, TE=20 ms, TM=30 ms, 200 acquisitions, N=8, Siemens SP4000 Helicon) and 4.0 Tesla (1.5x1.5x1.5 cc, TR=2.0 s, TE=20 ms, TM=30 ms, 128 acquisitions, N=12, Varian/Siemens Unity Inova) were compared. Following QUALITY deconvolution with an unsuppressed water spectrum, suppressed spectra were quantified in the time domain using fully automated software developed in our lab which incorporates prior-knowledge obtained from the in-vitro spectra of 12 metabolites (listed below) and three macromolecules. Results are shown as mean(% standard deviation). The FWHM of unsuppressed water peaks were 4.34(0.5%) Hz at 1.5 Tesla and 9.34(5.4%) Hz at 4.0 Tesla, while SNR (NAA level divided by time domain RMS noise) was 6.0(7%) and 7.4(10%) respectively. Metabolite levels at 1.5 and 4.0 Tesla respectively were: N-acetylaspartate 14.2(3.5%) 13.5(4.5%), glutamate 9.9(11.4%) 7.5(14.9%), glutamine 5.1(19.2%) 6.2(8.7%), gaba-aminobutyric acid 1.5(66.4%) 2.4(21.2%), aspartate 3.4(20.7%) 4.8(28.5%), N-acetylaspartylglutamate 2.3(23.0%) 1.0(59.0%), taurine 2.4(19.9%) 2.9(16.5%), glucose 1.7(61.9%) 0.4(106.1%), choline 3.0(6.5%) 2.5(6.1%), phosphocreatine/creatine 9.6(6.4%) 7.2(1.7%), scyllo-inositol 0.3(80.9%) 0.7(31.5%), and myo-inositol 5.6(52.1%) 9.25(36.2%). Although preliminary, these 4.0 Tesla results suggest the precision of in-vivo metabolite level measurements may be superior at this field strength. Also, by trading some SNR for a smaller volume, partial volume effects and susceptibility problems can be reduced.
The water signal which dominates localized proton spectroscopy is the source of base-line artifacts, and the suppression or avoidance of this signal is the cause of many other spectral distortions. On the positive side, the water signal is required for full quantitative analysis, and for phase and low frequency eddy current correction. The ability to collect in vivo spectra without water suppression would be a major advantage. The method introduced here, over-sampled J-resolved PRESS, overcomes this problem by sampling a localized 2D J-resolved spectrum to F1=F2 thus isolating the coherent base-line artifacts outside the JHH bandwidth in F1. The unsuppressed water is used to correct shot to shot phase error and a two-dimensional water reference is used to achieve a phase sensitive display of the metabolites.
In previous in situ PRESS 3D 1H MR spectroscopic imaging studies, we have demonstrated that the ratio of cytosolic prostatic metabolites can non-invasively discriminate prostate cancer from surrounding healthy tissue. However, in these studies we have observed inadequate water and lipid suppression from the CHESS and STIR techniques. In order to improve this suppression, as well as the spatial coverage, we have implemented the newly developed T1 insensitive dual BASING Bandstop Filter and dual phase-compensating spectral/spatial spin-echo (EPSE) pulses. In a study of 40 prostate cancer patients, the use of either BASING or EPSE suppression increased spatial coverage 2.5 fold (p<0.04) over CHESS or STIR. As well, lipid and water suppression were improved 2010 and 74 fold, respectively. In a comparison of the BASING and EPSE techniques, improved suppression of water (99.80.2% and 99.80.1%) and lipid (98.51.6% and 97.52.5%) were achieved without a significant difference in the prostatic metabolite ratios. EPSE suppression has the added advantage of reducing the chemical shift dependence of the PRESS volume, but requires high performance gradients and is more susceptible to eddy currents. In summary, not only did the use of BASING and EPSE pulses increase coverage of the prostate and eliminate spectral baseline artifacts, it also saved time in data acquisition and processing. This increased coverage, without the loss of metabolic specificity for detecting cancer, greatly improves our ability to assess the presence and spatial extent of prostate cancer.
The NMR frequency displacement, D, of any nucleus varies with the shape and orientation of its compartment and with the values of BMS inside (Xi) and outside (Xo), typically over several ppm. Columnar structures aligned along Bo (an NMR tube, or capillaries or straight young maize roots inside an NMR tube) show narrow lines at D = Xi/3, independent of Xo. Infiltration of cell-membrane-impermeable paramagnetic lanthanide species into the extracellular spaces of roots creates separate intra- (i) and extracellular (e) populations for each nucleus (1H, 13C, 23Na, etc.) showing spectral lines at the predicted values of Di and De (1) ; here, e.g. (De - Di) = (Xi(e) - Xi(i))/3 = 1.02ppm, for 10mM GdDTPA. MESS images (2) at De or Di for 1H2O show the characteristic distribution of extracellular and intracellular spaces, in particular the agent-accessible xylem channels in the core of the root (in-plane resolution 23 x 23 micron by 1 mm thick). Localized spectroscopy with PRESS allows the quantitative assignment of local intracellular and extracellular volumes, and allows study of rates of lateral permeation or of axial flow along the channels. 1. Y. Shachar-Hill, D. E. Befroy, P. E. Pfeffer and R. G. Ratcliffe, J. Magn. Reson. 127, 17-25 (1997). 2. Y. Xu, J. A. Balschi and C. S. Springer, Magn. Reson. Med. 16, 89-90 (1990).
Yaotang Wu,a,b Melvin J. Glimcher,b David A. Chesler,a Breege A. Concannon,a Leoncio Garrido,a Denise P. Hinton,a Aiping Jiang,a Martin J. Lizak,a Hong Jiang,c Jun Li,d Robert M. Neer,d Chandrasekhar Ramanathan,a Bettina Pfleiderer,a Jinxi Wangb aNMR Center, Massachusetts General Hospital, Charlestown, MA; bChildren's Hospital,Boston, MA; cCenter for Imaging and Pharmaceutical Research, MGH, Boston, MA; dEndocrine Unit, MGH, Boston, MA Because high power RF pulses and magic angle spinning are inappropriate for living subjects, in vivo solid state NMR must work under conditions of broad spectral linewidths, and must rely on other methods for accomplishing cross polarization and for discriminating among resonances of various chemical species. However, special solid state measurement techniques we have developed show great promise for eventual use in human subjects. Three dimensional solid state FID projection imaging of phosphorus yields a direct quantitative measurement of bone mineral density, and as opposed to conventional radiologic imaging methods, does not rely on models of the x-ray attenuation of a variable mixture of bone, soft tissue and fat. Cross polarization may be accomplished by means of adiabatic demagnetization in the rotating frame, a methods which involves very low total RF dose. Spectroscopic analysis or chemically selective imaging can provide information on the chemical composition of bone mineral or synthetic calcium phosphate prosthetic material. This composition may be of interest in evaluating the maturity of the mineral (as reflected in the [HPO4-2]/[PO4-3] ratio, for example) in a healing fracture or in the rapid growing bones of a child. A second application of compositional measurement is in the determination of the amount of residual synthetic calcium phosphate present in a prosthetic material which is subject to remodeling. Several examples of these applications will be presented.