Institute: MMRRCC, Dept. of Radiology, Univ. of Pennsylvania
Keywords:
Abstract:
Osteoarthritis (OA) is a progressive disease of cartilage associated with severe joint pain and eventual joint immobilization. Detection of proteoglycan (PG) loss in early OA is necessary for efficient therapeutic intervention. 23Na MRI holds promise as a means of detecting PG loss since extrusion of PG during cartilage degeneration results in a loss of sodium from the tissue as well. Recently, we showed the feasibility of indirectly measuring PG loss in cartilage by 23Na MRI of intact articular cartilage in vitro. However, in vivo sodium quantitation in joints is confounded by confusing the sodium signal from copious synovial fluid with the sodium signal from cartilage. Here we present a method of suppressing the fluid signal during 23Na MRI of patellar joint in an attempt to improve the accuracy of sodium quantitation in cartilage in vivo.

The MR behavior of 23Na varies depending upon its relaxation environment. Human patellar joint is made up of articular cartilage with a relatively high sodium concentration (~230mM in healthy human subjects, our own measurement), synovial fluid (sodium concentration ~150mM) and other tissues with smaller sodium content. There is also a large difference in sodium relaxation times (synovial fluid T1~60ms, T2~30ms and cartilage T1~30ms, T2~20ms). We can exploit this difference in relaxation times and concentration to obtain 23Na MR images with little or no fluid signal while preserving cartilage signal. To accomplish this, we employed an inversion recovery magnetization prepared gradient echo imaging sequence. By choosing an appropriate inversion time (TI) between the initial inversion pulse and the rest of the imaging sequence, sodium signal from synovial fluid will contribute little to longitudinal magnetization and thus is nulled in the subsequent image.

In an in vitro experiment, we simultaneously imaged a tube of saline and a piece of bovine cartilage. Significant saline signal suppression (decrease by 75%) was achieved by choosing a TI=40ms and TR=160ms compared to a control experiment where an inversion pulse was not used and TR was set to 120ms. This degree of signal change was in agreement with theoretical simulations of the Bloch equations. The total imaging time was kept the same by collecting fewer signal averages for the longer TR sequence. The resultant loss of SNR was more than compensated by the excellent fluid signal suppression.

We have demonstrated that it is possible to obtain adequate suppression of fluid signal in 23Na MR images of the human patella in vivo. This is useful for accurate quantitation of sodium in joint cartilage. We are in the process of applying this technique to quantify in vivo sodium concentration changes in animal models of osteoarthritis

This research was performed at a NIH supported research center (RR 02305) and grant (AR45242-01).

152
Session Assigned: W&Th P
Title: RAT ARTICULATION IN VIVO , 2D AND 3D NMR IMAGING AT 7 TESLAS

Institute: RMN-Bio, ICSN, CNRS
Keywords: 2D an 3D imaging, in vivo rat articulation, 7T
Abstract:
Our purpose was to set up a model and a strategy to study the diseased articulation of rat in vivo at 7T. MRI is a well adapted technique allowing the longitudinal follow-up in time. To date, images of articulation and of cartilage have mainly been carried out on humans at 1 or 2T. Our study was carried out on the rat at high field (7 T), on a Varian Inova spectrometer equipped with a horizontal magnet with a 20 cm bore diameter and a 24 mm surface coil. The animal was anaesthetized with halothane/02/N2O mixture and its temperature was monitored. NMR parameters were optimised to yield the best images in terms of contrast, sensitivity and resolution. These parameters have to be compatible with the timing of potential pharmacological protocols in vivo.

The T1 and T2* relaxation rates of the various structures of the articulation (cartilages, trabecular bone, bone marrow, muscle, skin) were measured. The T1 relaxation rate of the articular cartilage and the muscle are similar (1.8+0.1s), while their T2* rates are different: muscle T2 =20.2+0.1ms and cartilage T2 =32.5+0.7ms. These results directed us towards a T2 and rho weighting. These conditions are coherent with the visualisation of the cartilage since it is made of proteoglycanes and collagen, but especially of water (70%). To compare the images, we used the crossed ligaments as an anatomical reference mark. To favour sensitivity, short TE with gradient echo sequence were preferred to spin echo sequences. 2D multislice gradient-echo in the coronal and sagittal orientation images were acquired with the following parameters: 2s/5ms ,4 transients, matrix = 512*256, FOV = 30*30 mm2, slice thickness = 0.28 mm, experimental time = 35 minutes (23 sagittal slices and 27 coronal slices), voxel size 1.9.10^-3 mm³. The 3D gradient-echo sequence allows the coronal and sagittal slices to be obtained simultaneously and is more convenient for the global observation of the articulation. It was optimised to run in 35 minutes : 200ms/5ms, 1 transient, matrix = 512*172*64, FOV = 30*30*14 mm3, voxel size 2.2.10^-3 mm³.

The resulting images are of a high quality in terms of resolution and contrast. Fine details from the articular cartilage and the bone region were obtained allowing the study of articular diseases in rats. Our results show that the use of high field enables a longitudinal follow-up of potential treatment of the articular diseases on rats before the study on humans. This method avoids the drawback of the variability between animals compared with classical biological techniques.

153
Session Assigned: W&Th P
Title: MAGNETIC RESONANCE MICROIMAGING (MRM)OF HUMAN CAROTID PLAQUES

Institute: Baylor College of Medicine, Houston , TX , USA
Keywords: MAGNETIC RESONANCE MICROIMAGING (MRM)OF HUMAN CAROTID PLAQUES
Abstract:
Felicia Shay*, Gerald Lawrie*, Hazim Safi*, Joel Morrisett*, Ed Ezell**, Joe Schwartz**, and David Gorenstein** *Departments of Medicine and Surgery Baylor College of Medicine, Houston, TX **Sealy Center for Structural Biology University of Texas Medical Branch, Galveston, TX

MAGNETIC RESONANCE MICROIMAGING (MRM) OF HUMAN CAROTID PLAQUES

Carotid arteries are vital vessels for carrying blood to the brain. Occluded carotid arteries are a major cause of blurred vision, transient ischemic attacks, and stroke. Magnetic resonance (MR) imaging is a powerful non-invasive technique capable of visualizing the carotids and determining their degree of stenosis. The goal of this project was to develop correlations between ex vivo MR images acquired at 9.4 Tesla of occluding tissue resected by carotid endarterectomy (CERD) and corresponding histopathologic images. CERD tissues ~2-6 cm long and 0.5-1.5 cm wide were transferred to 15 ml plastic culture tubes containing either 10% formalin in TBS or 50% glycerol in TBS and microimaged under the following conditions: Ho=9.4T, Te=30ms, Tr=1.5s, matrix size=512 x 256, ns=25, slice thickness=l mm, fov=2.5 cm, in plane resolution=0.l mm. After imaging, the tissues (fixed or fresh frozen) were cut into 5 cm segments and processed (imbedded in paraffin or frozen in OCT), sectioned (5 mm), and stained (Movat or Trichome). The morphology of the 5 µ thin sections was compared to that of the 1 mm MR image slices. Good registration between sections and slices could be achieved despite the 200-fold difference in thickness. Total CERD cross sectional areas computed by MR imaging correlated well with areas determined from histologic sections (r2=0.699, p=0.005). For the lumen areas, the correlation was even better (r2=0.843, p=0.005). This level of correlation indicates the validity of MR microimaging in assessing morphologic features of carotid artery plaques, and suggests the potential of MR whole body in vivo imaging for non-invasive measurement of carotid artery patency, plaque volume, and plaque composition.

155
Session Assigned: W&Th P
Title: Evaluation of reduced encoding MRI for temperature mapping in phased array microwave hyperthermia

Institute: University of Illinois at Urbana-Champaign
Keywords: Electromagnetic Hyperthermia; generalized series reconstruction
Abstract:
This work presents an experimental evaluation of the RIGR [1,2] technique applied to temperature mapping in phantoms heated by phased array microwave hyperthermia applicators. Hyperthermia, which aims to elevate tumor temperature by 6^oC, is commonly used in conjunction with other modes of therapy for a more effective treatment. However, its clinical usage is limited due to a lack of an adequate and noninvasive temperature measurement system accompanied by closed loop control of the heating. Control systems previously reported for hyperthermia involve temperature measurements that are either few in number and invasive, or are reported from the surface and do not record the temperature inside the tissue [3,4]. Using MRI temperature mapping, such temperature measurements can be performed non-invasively and at many more locations inside the subject. However, there exists a trade off between spatial and temporal resolution of the temperature maps. It is thus important to eliminate any encoding of redundant information in order to improve the temporal resolution, with minimal degradation of spatial resolution. Here, using the proton resonance shift method [5] at 200 MHz, temperature maps obtained via the RIGR method are compared to those obtained by complete Fourier encoding and the accuracy measured via fiber-optic probes. Experimentally it is shown that using the RIGR method, single slice temperature images with pixel size of 0.8 mm x 0.8 mm can be acquired in 0.9 sec intervals with temperature errors of +/-0.5^oC. Compared to the regular Fourier encoding, it is shown that this method yields a time resolution reduced by a factor of 3 with an identical thermal resolution within experimental error. [1] Z. P. Liang, and P. C. Lauterbur, IEEE Trans. Med. Imag. 13, Number 4, 677 (1994).[2] A. G. Webb, Z. P. Liang, R. L. Magin, and P. C. Lauterbur, J. Magn. Reson. Imag. 3, 925 (1993).[3] A. J. Fenn, C. J. Diederich, and P. R. Stauffer, Lincoln Lab. J. 6, Number 2, 269(1993).[4] A. J. Fenn, and G. A. King, the Lincoln Lab J. 5, Number. 2, 223 (1992).[5] J. De Poorter, C. De wagner, Yves De deene, C. Thomsen, F. Stahlberg, and E. Achten, Magn. Reson. Med. , 33, 74 (1995).

156
Session Assigned: W&Th P
Title: PRAWN EPI: Proton detected ¹³C Imaging.

Institute: University of Nottingham, England
Keywords:
Abstract:
Direct ¹³C imaging suffers from low sensitivity and poor resolution. The use of spatially resolved, Cyclic Cross Polarisation (CYCLCROP¹) theoretically offers a 64 fold increase in sensitivity, with a high spectral selectivity. For in vivo, liquid state experiments this enhancement factor is hard to achieve, owing to the sensitivity of the technique to both radio frequency and main field inhomogeneities.

A new mixing sequence for the polarisation transfer, called PRAWN², offers greater insensitivity to radio frequency inhomogeneity and substantially reduced RF energy deposition. This sequence has already been used, in conjunction with 2DFT imaging to monitor carbohydrate transport in plants³, achieving a temporal resolution of 90 minutes.

For systems exhibiting a favourable T2 relaxation time and an enriched ¹³C concentration, PRAWN EPI imaging offers a significantly improved temporal resolution. Using our 11.7T NMR microscope, and an optimised dual resonant probe, PRAWN EPI phantom images have been generated in 800ms. (Sample: 12mm tube of 230mM (beta-1-¹³C-glucose_(aq). Image: half fourier EPI with a matrix size of 40x64 (two shot phase cycle ), giving a voxel dimension of 0.4x0.4x20mm)

PRAWN provides a sensitive tool for tracing isotopically labelled carbon molecules, in systems where other labelling approaches fail.

References:

1 Kunze C., Kimmich R., Proton-Detected ¹³C imaging using cyclic J cross Polarization, Mag. Reson. Imag., 12 no.5, pp805-810 (1994)
2 Chandrakumar N., Kimmich R., PRAWN: mixing sequences for selective heteronuclear J Cross Polarization, in press (JMR 1998).
3 Heidenreich M., et al., Investigation of Carbohydrate Metabolism and Transport in Castor Bean Seedlings by Cyclic J Cross Polarization Imaging and Spectroscopy, JMR 132, no.1, pp109-124 (1998)

157
Session Assigned: W&Th P
Title: Truncating High Spatial Frequency Data Does Not Lead To Truncation Artifacts In MRI

Institute: University of Florida, Gainesville, FL. USA
Keywords:
Abstract:
It is reported that truncation artifacts in MRI arise primarily in two circumtances: 1) because of data interplolation (zero filling), when a smaller acquisition matrix (256x128) is interpolated into a large display matrix (256x256), and 2) near edges where there are abrupt transition in signal intensity (1). But, the widely prevalent and well-documented explanation of truncation artifacts in MRI is that they arise when raw data do not contain all of the high spatial frequency information present in the imaged subject(1,2). In this report, it will be demonstrated that the truncation artifacts in MRI occur not because of truncation of the high spatial frequency data, but because of an abrupt transition of signal intensity present in the imaged subject. The structure of the data acquired in a finite time determines the pattern and structure of the truncation artifacts. This conclusion is consistent with the observation in imaging practice that the truncation artifacts occur whenever there is a strong signal discontinuity. By using the conclusion of this report, the above two circumstances will finally reduce to only one circumstance.

It is well known that increasing the image resolution lrads to the reduction of the truncation artifacts(2). This fact is an excellent experimental verifaction of the above conclusion. The main reason is that the signal intensity is reduced in each voxel if image resolution is increased. The difference of the signal intensities between two contiguous voxels is reduced also. In other words, the signal discontinuity is reduced. Another reason is that if more voxels are located at the area of an abrupt transition, the signal disconsinuity betwenn each two contiguous voxels will be reduced further. Thus, using the possibly highest image resolution is the most efficient way to reduce the truncation artifacts in MRI. Another possible way to reduce the truncation artifacts in a real image is to extrapolate the acquired data of an abrupt transition of signal intensity into the data of smooth transition of signal intensity, such as expressed by a sigmoid function. Recovering the lost signal data of high spatial frequency as proposed by some references may not be a good choice for reducing the truncation artifacts. Extrapolationg data into high spatial frequency without blurring the image can only change the pattern of the truncation artifacts, but not the amplitude.

One should not confusion the current situation with other Gibbs phenomena in frequency domain, for example the FID truncation in NMR spectroscopy and the low pass filter design in signal processing. Although the appearances of these different phenomenoa are quite similar and they are all referred as to the Gibbs phenomenon, it will demonstrated that their essences are different.

References:

1. G.Wesbey, etc., in "Clinical Magnetic Resonance Imaging" (R.R. Edelman, etc., Eds.). p.109, 2nd Ed. 1996.
2. R.T. Constable, etc., Magn.Reson.Med. 17,108-118 (1991).

158
Session Assigned: W&Th P
Title: Novel Data Acquisition System for Signal Enhancement in Time Domain EPR Imaging in vivo

Institute: Radiation Biology Branch, NCI, NIH, Bethesda, MD
Keywords: FT-EPR, imaging, in vivo
Abstract:
Design strategies, system configuration and operation of a dual channel data acquisition system for a radiofrequency (RF) time domain electron paramagnetic resonance (EPR) spectrometer/imager operating at 300 MHz are described. This system was configured to incorporate high speed analog to digital conversion (ADC) and summation capabilities with both internal and external triggering via GPIB interface. The sampling rate of the ADC is programmable up to a maximum of 1 Gs/s when operating in a quadrature mode or 2 Gs/s in single channel mode. By using high speed flash ADCs, a pipe-lined eight bit adder, and a 24-bit incrementer, a summing rate of 230 kHz is achieved for FIDs of 4K record length. The record length is programmable up to a maximum of 8 K points and a large number of data sets can be can be averaged without overflow (120, 4K data sets) before transfer to a host computer for further processing. The averager operates in a two channel (quadrature) receiver mode for the conventional mixing to base band. For detection using the single channel mode the resonance signals around the carrier frequency of 300 MHz were mixed with a local oscillator of appropriate frequency (350 MHz, in our case) leading to an IF which is sampled at the rate of 1Gs/s. Comparison of quadrature mode and an IF mode operation is presented by studying the FID signal intensity across a band width of 10 MHz and as a function of transmitter RF power.

Imaging of a large sized phantom accommodated in appropriately sized resonators, in the IF mode operation suggests that this mode can be used to obtain distortion free images in resonators of size 50 mm (dia.) X 50 mm (length). Some results obtained from in vivo experiments will be presented.

159
Session Assigned: W&Th P
Title: Spectral-spatial Imaging using Radiofrequency Time domain EPR

Institute: Radiation Biology Branch, NCI, NIH, Bethesda, MD
Keywords: FT-EPR, spectral-spatial, oxymetry
Abstract:
A radiofrequency EPR spectrometer operating in time domain is used to perform spectral-spatial imaging on phantom objects. The spectrometer operates in the frequency range of 280 - 300 MHz affording a penetration depth of about 5 cm in biologic objects. Excitation pulses of 70-100 ns were used in cylindrical resonators of 25 mm diameter and 50 mm length with Q of about 20. Following a dead time of 400 - 600 ns, the Free induction decays were collected using a high speed digitizer/averager. In this configuration, repetition rates of 100 kHz can be achieved to sample and sum FIDs of 4K record length.

Using phantom objects containing spin probes at different levels of oxygen, spectral-spatial EPR imaging experiments were carried out to determine the oxygen status. The algorithm for the image data collection and processing will be discussed. Results from preliminary experiments with solid spin probes as well as spin probes dissolved in physiologic medium will be presented. The method will be useful in performing oxymetry using infusible and implantable EPR probes whose line width is sensitive to the in vivo oxygen concentration.

160
Session Assigned: W&Th P
Title: SPRITE MRI of Materials

Institute: Physics Department, University of New Brunswick
Keywords:
Abstract:
A large variety of solid like 1H containing materials are attractive candidates for MRI but are challenging or impossible to image with traditional methods due to short T2* relaxation times. We have recently developed a new MRI pulse sequence which is a time efficient and very flexible version of Single Point or Constant Time Imaging (1). This sequence is called SPRITE (Single Point Ramped Imaging with T1 Enhancement). The sequence (2) is a pure phase encoding method which relies on broad band RF pulses applied in the presence of a ramped primary phase encode gradient and constant secondary phase encode gradients. The ramped gradient permits repetition of large or small flip angle RF pulses at arbitrarily short repetition times, with no gradient vibration. The technique permits the introduction of T1 contrast for samples which may have T1 relaxation times which range from hundreds of usec to many seconds. The method is ideally suited to the visualization of nuclei with short T2*, from tens of usec to hundreds of usec.

This lecture will describe the application of the method to visualization of cortical bone, rigid polymeric and concrete materials. All three classes of materials have T2* relaxation times under 200 usec. Techniques employed to manipulate image contrast will be discussed as will the extension of these methods to permit the determination of local T1,T2, T2* and T1r relaxation times. Relaxation time mapping in conjunction with variable temperature MRI for localized calorimetry measurements will be introduced.

1. S. Gravina and D.G. Cory, J. Mag. Reson. B 104, 53 (1994).
2. B.J. Balcom, R.P. MacGregor, S.D. Beyea, D.P. Green, R.L. Armstrong and T.W. Bremner, J. Mag. Reson. A 123, 131 (1996).

161
Session Assigned: W&Th P
Title: Quantitative Solid State NMR Imaging of Synthetic Calcium Phosphate Implants

Institute: Massachusetts General Hospital, Charlestown, MA
Keywords:
Abstract:
Synthetic calcium phosphates such as hydroxyapatite and tricalcium phosphate are used as orthopaedic implant materials as well as constituents of bone cements due to their chemical and crystallographic similarity to bone mineral. It is difficult to monitor the remodeling of implants in vivo using x-radiography as the synthetic calcium phosphates have similar x-ray absorptivities to bone mineral. NMR techniques hold promise due to their chemical sensitivity. We have used solid state phosphorus-31 NMR imaging to quantitatively measure the mass of hydroxyapatite in the presence of bone at 4.7 T. Three dimensional projection reconstruction imaging was used to produce solid state images from 998 free induction decays sampled in the presence of a fixed amplitude magnetic field gradient whose direction was varied uniformly over the unit sphere. Chemical selection was achieved using T1 contrast, as the T1 of the synthetic calcium phosphate (1.8 s at 4.7 T) is significantly shorter than that of the bone (approximately 15 s at 4.7T in vivo, 42 s ex vivo). It is therefore possible to saturate the bone mineral signal by using a shorter recycle time. Experimental results demonstrating the linear relationship between image intensity and hydroxyapatite density in phantoms containing hydroxyapatite and silicon (IV) oxide, and hydroxyapatite and bone have been obtained. The quantitation improved with increasing gradient field strength as expected. However, for a fixed acquisition time there is a loss of SNR with increasing gradient strength, making it necessary to find an optimal value for the gradient. In vivo images were acquired of the tibia of a New Zealand White rabbit with a hydroxyapatite phantom fixed adjacent to it, using a surface coil. Chemically pure images of bone mineral and synthetic hydroxyapatite were computed from images acquired in vivo at two different recycle times. The technique can be used to non-invasively follow the resorption and remodeling of calcium phosphate implants in vivo.

162
Session Assigned: W&Th P
Title: Quantitative Evaluation of Bone Mineral Density using 31P Solid State 3D Magnetic Resonance Imaging

Institute: Massachusetts General Hospital
Keywords: Densitometry; MRI; Bone Mineral Content
Abstract:
Current clinical schemes for evaluation of bone mineral density such as emission tomogragphy (ET), plane film and x-ray computed tomography, and dual energy X-ray Absorptiometry (DXA) involve exposure to ionizing radiation and in some cases only provide 2D projective views. Magnetic resonance imaging has a distinct advantage as a modality that can provide non-invasive, non-ionizing serial assessment which make the technique ideal for characterizing bone mineral and matrix in disease progression, e.g. osteoporosis, fracture healing and bone growth. Solid state ³¹P 3D MR imaging has been established as a method for both in vivo and ex vivo bone imaging. The goal of this work is to establish the quantitative ability of the 3D ³¹P technique for measuring bone mineral density. Ex vivo specimens of bovine trabecular and cortical bone were evaluated at 4.7 T using a 3D projection imaging sequence in which FIDs acquired in the presence of a readout gradient readily accommodate the short T₂ of the ³¹P mineral signal and produce a true volumetric image with spatial resolution on the order of 2 mm. Quantitation is achieved by comparison of bone image intensity to a calibration curve from images of a series of synthetic hydroxyapatite phantoms of known composition. The MRI results are compared to DXA measurements and gravimetric analysis for measuring the mineral density. Implications for a quantitative clinical technique based on the ³¹P MR solid state method will be discussed.

163
Session Assigned: W&Th P
Title: Mass Transfer of Alkanes in Zeolite Packed-Bed Reactors Studied by MRI

Institute: University of New Brunswick
Keywords:
Abstract:
The pulse sequence, Single Point Imaging with T₁ Enhancement (SPRITE) is a pure phase encode MRI technique with broadband RF-excitation pulses and a ramped primary phase encoding gradient (1). This method has been used to obtain in situ images of gas pulses passing through industrial grade packed beds of 13X and 5A zeolites at room temperature.

The experiments were carried out using a single pass flow reactor system placed in the MRI magnet. A pulse ~10^-3 mol of an alkane was injected into the carrier gas (He) flow. Concentration profiles were measured with the SPRITE technique. Three dimensional concentration-position-time profiles for a fixed flow rate of the gas through the reactor system were obtained and analysed by the method of Moments to determin parameters of gas dispersion in the zeolite bed (2). Experiments were conducted using different zeolites, alkanes and carrier gas flow rates. The results are presented in Van Deemter plots (3).

(1) B. J. Balcom, R. P. MacGregor, S. D. Beyea, D. P. Green, R. L. Armstrong, T. W. Bremner, J. Magn. Reson. A 123 (1996) 131.
(2) J. Kaerger, D. M. Ruthven, Diffusion in Zeolites and Other Microporous Solids, Wiley, N.Y., 1992.
(3) J. J. Van Deemter, F. J. Zuiderweg, A. Klinkenberg, A. Chem. Eng. Sci. 5 (1956) 271.

164
Session Assigned: W&Th P
Title: Synchrotron X-ray Microtomography and Solid-State NMR of Toluene Waste in Cement

Institute: Louisiana State University, Baton Rouge, LA, USA
Keywords: environment environmental imaging
Abstract:
Synchrotron X-ray microtomography shows vesicular structures for toluene/cement mixtures, prepared with 1.22 to 3.58 wt% toluene. Three-dimensional imaging of the cured samples shows spherical vesicles, with diameters ranging from 20 to 250 microns; a search with electron microscopy for vesicles in the range of 1-20 microns proved negative. However, the total vesicle volume, as computed from the microtomography images, accounts for less than 10% of initial toluene. Since the cements were cured in sealed bottles, the larger portion of toluene must be dispersed within the cement matrix. Evidence for toluene in the cement matrix comes from ²⁹Si MAS NMR spectroscopy, which shows a reduction in chain silicates with added toluene. Also, ²H NMR of d8-toluene/cement samples shows high mobility for all toluene and thus no toluene/cement binding.

A model that accounts for all observations follows: For loadings below about 3 wt%, most toluene is dispersed in the cement matrix, with a small fraction of the initial toluene phase separating from the cement paste and forming vesicular structures that are preserved in the cured cement. Furthermore, at loadings above 3 wt%, the abundance of vesicles formed during toluene/cement paste mixing leads to macroscopic phase separation (most toluene floats to the surface of the cement paste). Finally, tomography and solid-state NMR make a great analytica combination.

Institute: University of Florida
Keywords: NMR; ³¹P; spectroscopy; spinal cord; RF coil
Abstract:
³¹P NMR spectroscopy is a powerful tool for the study of in vivo nervous system metabolism. However, due to poor spatial selection and low signal-to-noise ratio (SNR), the only previously recorded ³¹P in vivo spectra from spinal cords were aquired from NMR experiments accompanied by surgical exposure of the spinal cord and were thus terminal experiments (1,2). Utilizing computer simulations and phantom tests, an inductively-coupled implantable coil was designed and implemented. This coil was employed to obtain the first in vivo ³¹P localized NMR spectra from a rat spinal cord using a chronically implanted inductively driven coil. A 4 x 5 mm inductively coupled coil with a resonance block circuit was coated with a bio-compatible acrylic and implanted above a T-13 vertebral level laminectomy. The implanted coils resonance block circuit allowed decoupling during RF transmission and optimal coupling to a 3 x 3 cm surface coil during reception. RF field excitation was thus achieved solely by the external surface coil, which was mounted on a Plexiglas cradle designed to support a supine rat. This design scheme provided uniform magnetic field excitation and a highly sensitive, chronically endured, inductively-coupled implanted reception coil. A 4 x 10 cm quadrature ¹H surface coil was also mounted on the ³¹P coil system cradle for precise field shimming and high SNR image acquisition of the region of interest (ROI) and surrounding anatomy. The use of a quadrature coil minimized coupling between the ¹H imaging coil and the ³¹P spectroscopy coils and provided well-resolved anatomical images. As the localization accuracy of the coil system alone is inadequate and permits contamination from surrounding tissues, an ISIS pulse sequence was employed to localize a ROI of spinal cord at the T-13 vertebral level. Using a 2 ms hyperbolic secant localization pulse a 1.25 x 1.00 x 5.85 mm (7 µL) voxel was selected within the spinal cord based on transverse and longitudinal pilot images. Based on minimum required parameter estimates, determined from phantom experiments, 2048 averages with a 2-second TR were collected. The resulting spectra had a SNR of approximately 3.5 to 1 with a PCr to b-ATP ratio of 2.4 to 1.

Using the combination of coil design and pulse sequence localization, the obtained ³¹P spectra demonstrated metabolite ratios indicative of healthy spinal cord tissue and showed no contamination from surrounding tissues. Furthermore, the application of inductively coupled implantable coils induced no adverse reactions in the rat, thus validating the potential for long term longitudinal studies. Using this methodology, changes in spinal cord metabolism can be monitored periodically through the chronic phase of spinal cord injury with optimal sensitivity and no contamination from adjacent tissues. The utilization of inductively coupled implanted coils, combined with gradient pulse localization, allows long-term in-vivo studies of phosphorus metabolites in both the normal and injured spinal cord. Future implementation of this protocol may benefit studies of pharmacological intervention in spinal cord injury.

1. M. Akino, et al., Invest.Radiol. 32, 382 (1997).
2. R. Vink, et al., Brain Res. 490, 144 (1989b).

166
Session Assigned: W&Th P
Title: An NMR investigation of the metabolism of an environmental iron-reducing microbial consortium

Institute: University of Manitoba, Winnipeg, Manitoba, Canada
Keywords: iron-dissimilatory respiration; fermentation; bacterial consortium
Abstract:
The products of iron-respiration and fermentation pathways of an environmental iron-reducing bacterial consortium were followed in real time by ¹Hand ¹³C nuclear magnetic resonance spectroscopy by simply incubating the bacteria with various substrates in NMR tubes and obtaining the spectra at regular intervals over a period of 6weeks. The metabolic pathways of dissimilatory iron-reduction and fermentation could be established by the identification of metabolites produced and from the loss of the substrate. The consortium utilized ferrous ammonium citrate for energy through dissimilatory iron respiration by reducing citrate-chelated Fe(III) to Fe(II). The acetate produced was then utilized as an electron donor but only after a delay of several days. As the rate of loss of the glucose NMR peaks was about the same in the presence or absence of iron, it seems that glucose is not involved in iron respiration. Fermentation or oxygen respiration of glucose produced lactate which was then used in iron respiration.

167
Session Assigned: W&Th P
Title: P-31 Studies of Cell Cultures Using a Bioreactor Incorporating a Horizonally Oriented Detection Region

Institute: University Of Hawaii, Honolulu, HI
Keywords:
Abstract:
Phosphate plays vital roles in energy transfer, metabolic regulation and as a structural constituent of biomolecules. As a result, phosphate assimilation, storage and metabolism become critical in the growth and development of plants. The use of P-31 NMR spectroscopy for the study of phosphate metabolism is an established technique. However, in In-Vivo studies of small organisms and cell cultures the methodology is often hampered by low sensitivity and the problem of supplying the subjects with sufficient nutrients and respiratory gases. The sensitivity problem can be minimized by using a horizontally mounted detection section. The most common method to supply the necessary nutrients is either filtered flow or perfusion techniques. In this work we have utilized the horizontal detection region as a basis for our apparatus and have incorporated a flow system that is based on a recirculating loop driven by hydrostatic motion generated by an air feed outside the detection area.

168
Session Assigned: W&Th P
Title: Methodological Standardization for a Multi-Institutional In Vivo Trial of Localized ³¹P MR Spectroscopy in Human Cancer Research

Institute: Fox Chase Cancer Center. Philadelphia, PA., USA
Keywords: In vivo human cancer; 31P MR spectroscopy; multi-institutional trial
Abstract:
Analysis of ³¹P MR spectra of human cancers in vitro has shown relatively high levels of phosphomonoesters (PME) and a reduction of these metabolites associated with early indication of treatment response. The finding of relatively large levels of PME has also been corroborated in vivo in two types of human cancer. Our nine institutions are collecting ³¹P MR spectra from several human cancers in vivo to demonstrate if the high levels of PME can be found in other human tumors and if their reduction as indication of treatment response can be reproduced in vivo. In order to demonstrate these assumptions, a multi-institutional trial with a high level of quality control has been started.

Two steps have been taken to achieve quality control. A custom-built dual-tuned probe (¹H/³¹P) was supplied to all institutions with a carefully calibrated ³¹P external scaling reference placed inside the probe housing, and a fast and easy quality control protocol has been implemented to monitor acquisition at each institution. Quality control of signal quantification has been tested using 9 phosphoric acid samples. When these samples were tested in one institution, the RMS error of the MR determination was 2.8% compared to the actual amount in each sample. One each of these samples was tested at each institution, recording a multi-institutional RMS value of 11.3% (n=6). Other parts of the quality control protocol check the performance of adiabatic (BIRP) pulses, chemical shift imaging localization, and ¹H decoupling of ³¹P signals. The careful and systematic performance of this combined quality control protocol will ensure comparable results between the different institutions, thereby decreasing the possible problems generated by a multi-institutional study and increasing the sensitivity of the data analysis.

169
Session Assigned: W&Th P
Title: Deconvolution of Restriction Effects Compartmental Diffusion Using Relaxography and Diffusion Measurements

Institute: Worcester Polytechnic Institute, Worcester, MA USA
Keywords:
Abstract:
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 from single conpartment models, such as 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 conbined 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 increasingly 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 combination of restriction effects with the compartmental dontributions to diffusion. The diffusion-time dependence of the diffusion-attenuation curves has also been measured and used to interpret the multiexponential fitting of diffusion constants.

1. G. Vetek, I. Palyka, C.H. Sotak, C.S. Springer Jr., CR-Free Discrimination of Intra- and Extracellular ¹H₂O Signals from Yeast Suspensions by Diffusion-Space Relaxography (Diffusigraphy), Abstract, 2nd Meeting of the Society of Magnetic Resonance, San Francisco, CA, 1994; also published in Proc. Soc. Magn. Reson. 2, 1051 (1994).
2. K.G. Helmer, B.J. Dardzinski, C.H. Sotak, The Application of Porous-Media Theory to the Investigation of Time-Dependent Diffusion in In Vivo Systems, NMR in Biomed. 8, 297-306 (1995).

Institute: Quantum Magnetics, San Diego, CA USA
Keywords: airline security; explosives detection
Abstract:
As everyone knows when entering an airport and boarding a plane, ones person and personal effects are screened for the presence of weapons and explosives to prevent aircraft hijacking and terrorist actions that may cause catastrophic airframe failure. Typically X-ray image systems are used to screen personal effects and metal detection portals to screen passengers. You may be surprised to find screening systems based on magnetic resonance technology are being introduced into the area of checkpoint security.

Nuclear Quadrupole Resonance (NQR) and Nuclear Magnetic Resonance (NMR) have been demonstrated to be effective techniques for detecting the presence of specific chemical compounds. Quantum Magnetics has developed a series of NQR-based baggage screening systems for explosives that are being used around the world to screen both carry-on and checked baggage. QR is similar to the widely used nuclear magnetic resonance (NMR) and magnetic resonance imaging (MRI) techniques, but has the considerable advantage that the item to be inspected does not need to be immersed in a steady, homogeneous magnetic field. The target compounds are identified unambiguously by their unique quadrupole resonance frequencies. Screening systems based on NMR have also been developed specifically for the purpose of screening liquids. The system developed is non invasive and can detect the presence of both liquid explosives and dissolved narcotics.

As well as screening personal effects, magnetic resonance, in particular NQR can be applied to screening passengers for the presence of concealed explosives. Tests have shown that threat quantities of explosives can be detected on the person in a short scan together with some gross localization information. The introduction of explosive screening of passengers is very important as no other method currently exists apart from manual search.

Magnetic resonance screening technologies also combine well with the conventional screening technologies such as X-ray, metal detection and vapor detection systems. Quantum Magnetics is actively developing combined systems that use sensor fusion to increase the performance of the system rather than the individual parts.

The work presented here describes the concept of an airport security checkpoint consisting of screening systems based on magnetic resonance. The concept will be discussed together with results from the various systems obtained operationally.

172
Session Assigned: W&Th P
Title: Landmine Detection by Nuclear Quadrupole Resonance

Institute: Quantum Magnetics Inc., San Diego, CA USA
Keywords: landmines; NQR
Abstract:
Non-metallic or plastic landmines pose a grave worldwide threat. Due to the very small amount of metal used in the construction of such mines, traditional landmine detection systems based upon metal detection do not work reliably. A variety of advanced technologies have been introduced to address this problem such as ground penetrating radar, infrared imaging and chemical vapor sensing. For a variety of reasons, all these technologies have proven of limited effectiveness for plastic landmine detection.

Under DARPA sponsorship, Quantum Magnetics is developing a nuclear quadrupole resonance (NQR) based landmine detection system. The current system uses ¹⁴N NQR to detect both RDX and TNT based non-metallic landmines. NQR combines the compound specific detection capability offered by chemical detection techniques with the spatial localization capability and convenience of an induction coil metal detector. In the 20 years since NQR was first applied to mine detection in the U.S., there has been tremendous improvement in the NQR detection sensitivity due to significant progress in hardware design and pulse sequence development. In recent field trials, Quantum Magnetics successfully demonstrated the detection capability of a mobile, unshielded system operating in a landmine test range at Fort Leonard Wood, MO. We will review the progress made by Quantum Magnetics in this area over the last two years, including an overview of the basic technique, a description of the system hardware, methods for coping with radio frequency interference, and a summary of the results from recent field trials.

173
Session Assigned: W&Th P
Title: Optimization of Offset Frequency in the SORC Pulse Sequence using Feedback

Institute: Penn State University, University Park, PA
Keywords: NQR, SORC, parametric model, signal norms, feedback
Abstract:
Nuclear quadruple resonance (NQR) provides a means for detecting energetic compounds. The low signal-to-noise ratio (SNR) of NQR measurements has motivated research on signal enhancement methods. Recently, it has been demonstrated that feedback can be used to automatically optimize pulse sequence parameters to maximize the SNR. We now present a feedback algorithm that automatically adjusts the offset frequency in the strong off-resonant comb (SORC) pulse sequence to maximize the NQR signal energy.

As an aid in choosing a signal metric and designing the feedback algorithm, a parametric model based on experimental observations is developed for the steady-state SORC waveform. Of the three signal metrics considered, peak-to-peak voltage, mid-signal amplitude, and signal energy, the signal energy is found to be the most sensitive to changes in offset frequency. Based on this result, a feedback algorithm that automatically adjusts the offset frequency to maximize the SORC signal energy is developed. Analysis and results are presented for a sample of sodium nitrite at both liquid nitrogen and room temperatures.

174
Session Assigned: W&Th P
Title: Some Practical Aspects of LC-NMR

Institute: G. D. Searle, Skokie, Illinois, USA
Keywords:
Abstract:
LC-NMR is rapidly becoming an important tool in solving many chemical problems. While some aspects of each technique are essentially independent, some aspects become more important when the two techniques are combined. For example, the presence of propionitrile in acetonitrile may not be a major problem in LC work, but the propionitrile is a hindrance in LC-NMR studies. While the LC flow rate is important in determining retention times, it becomes even more critical when the timing between the LC detector and NMR probe must be accurately known. In the same way, while the chromatography depends to some extent on temperature control, the NMR results are critically affected by any temperature variation, because the field is usually locked on deuterium in deuterium oxide, which is very temperature dependent. Gradient shimming on proton works well, but subsequent auto shimming on the deuterium lock can degrade the homogeneity. Direct comparison of the NMR spectrum of a component isolated by LC with the spectrum of a known reference compound requires that the two spectra be obtained in exactly the same solvent. Thus the exact composition of the solvent at the elution time must be known and used in determining the spectrum of the reference compound. In this poster we will discuss these and other practical problems involved in combining LC and NMR studies.

175
Session Assigned: W&Th P
Title: Recent Advances in the Application of LC-NMR and LC-NMR-MS to Drug Discovery

Institute: DuPont Pharmaceuticals, Co.
Keywords:
Abstract:
Complete characterization of metabolites usually requires time-consuming isolation from the biological matrix followed by off-line spectroscopic ananlysis. The use of LC-NMR and directly coupled LC-NMR-MS has greatly simplified this process by eliminating the isolation step. The structure elucidation of several metabolites of the anti-HIV drug Sustiva (TM) (Efavirenz) will be presented, demonstating the application and utility of LC-NMR and LC-NMR-MS as tools that enhance the efficiency of metabolite characterization. These tools also provide a means for structure elucidation of reactive metabolites that cannot be isolated. The characterization of the Chrysin metabolites will be presented as an example.

176
Session Assigned: W&Th P
Title: Characterizationof the Oxidative Degradation Products of SCH 56592 by LC-NMR and LC-MS

Institute: Schering-Plough Research Institute
Keywords:
Abstract:
LC-NMR and LC-MS were used to identify the structures of four major degradation products of SCH 56592, an antifungal drug in clinical trial. These compounds were formed under stress condition in which the bulk drug substance was heated in air at 150 degrees C for 12 days, and were separated from SCH 56592 using a semi-preparative HPLC method. LC-MS results indicated that the oxidation occurred at the piperazine ring in the center of the molecule. The structures of these degradation products were then determined from the 1H NMR spectra obtained via LC-NMR. A degradation pathway of SCH 56592 was proposed based on these structures.

177
Session Assigned: W&Th P
Title: Recent Developments and Applications of HPLC-NMR-MS for Structural Elucidation

Institute: Lilly Research Laboratories, Eli Lilly and Company
Keywords:
Abstract:
HPLC coupled to NMR has seen tremendous growth in recent years, especially in the pharmaceutical industry. A wealth of information can be obtained without the need for tedious off-line sample isolation. In a stopped-flow mode, an array of 2-D experiments can be performed to aid in structure elucidation. This methodology has been successfully used for the identification of drug metabolites, process contaminants, and a host of other analytes. Often LC-NMR only tells part of the picture, typically mass spectrometry is also used for unambiguous structure identification. This leads to the in-line coupling of LC-NMR-MS to obtain all of the necessary information on one sample. Two LC-NMR-MS systems will be demonstrated by coupling an HPLC-600 MHz NMR to either a time-of flight MS or an ion cyclotron resonance MS. Both configurations allow for the on-line determination of the analyte exact mass, leading to the molecular formula, and the latter affords additional valuable fragmentation information. The difficulties introduced into the MS analysis by the use of deuterated solvents will also be addressed.

178
Session Assigned: W&Th P
Title: Determination of Sites of Glucuronidation Using LC-NMR

Institute: Astra Charnwood, Loughborough, Leicestershire, UK
Keywords:
Abstract:
The application of LC-NMR at 500 MHz to the study of the site of glucuronidation of Compound Y will be described. Compound Y possesses three hydroxyl moieties, each of which is a possible site for hepatic modification. An initial LC-MS study indicated that a glucuronide was the major metabolite present in dog bile. Direct injection of a bile sample from a dosed animal onto the LC system was followed by NMR analysis using the stopped flow technique. The site of metabolism was deduced from the proton chemical shifts of protons adjacent to the hydroxyl groups as compared with the parent compound.

In a further study activated microsomes were used to produce glucuronides. Samples of the microsome incubation from three species (dog, rat and human) were separated on the LC-NMR system, either with or without prior purification. Again the stopped-flow technique was used, and proton and ROESY spectra collected on each sample. The site of glucuronidation was deduced by changes in proton NMR shifts and confirmed by the ROESY spectra. The predominant glucuronide produced by each species was found to be different, and all three possible glucuronides were ultimately identified.

These observations have critical implications for the choice of appropriate animal species in studies of this type, and the results of these LC-NMR studies were essential in elucidating this unexpected behaviour.

179
Session Assigned: W&Th P
Title: Study of slow rotational motion of particles by NMR

Institute: Dept. of Physics, MTU, Houghton,MI, USA
Keywords:
Abstract:
Nuclear Magnetic Resonance is a well-known technique to study the motion of fluids. Two novel methods are proposed here to measure slow rotational speeds of solid particles suspended in a fluid using a single or multi-pulse NMR techniques. A working model of a probe designed to check the feasibility of these proposed methods is described. Data obtained from such a probe for a Cu₂O powder sample using one of the techniques is presented where rotational speeds of approx. 1Hz were easily detected. The time dependence of the resonant frequency is exploited here to measure the rotational speeds and theoretically an expression is computed. The proposed technique should be effective method to observe slow Brownian rotational motion of particles suspended in a fluid. An attempt is made on Cu₂O powder particles suspended in a fluid is presented.

180
Session Assigned: W&Th P
Title: Groundwater relaxation and spin-echo, surface NMR-tomography of aquifers.

Institute: Inst. Chem. Kinet. & Combust. Russian Acad. Sci.
Keywords:
Abstract:
A geophysical surface-NMR method can be applied to investigate near-surface aquifers. Macroscopic "samples" of more than 100x100 m porous (or fractured) aquifers can be investigated using proton relaxation and spin-echo measurement in geomagnetic field. An antenna (usually a circular loop) with a diameter of about 100 m is laid out on the ground to excite and receive the NMR signal. An oscillating current with rectangular pulse-shape and the carrier-frequency being equal to the proton-resonance frequency in the Earths magnetic field (about 2 kHz) is passed through the antenna and followed by a nuclear induction emf due to free Larmor precession in the Earths field. Distribution of groundwater saturation with depth can be determined by free-induction decay (FID) measurement in dependence on the excitation-current pulse moment using inversion procedure to solve resulting integral equation. Therefore this method can be regarded as a surface-NMR tomography of subsurface aquifers [1]. The technique referred as the Hydroscope is employed for hydrogeological surveys down to a depth of about 100 m. Nevertheless, the vertical resolution of the surface-NMR method decreases with increasing depth. Finite ground conductivity can result in induced currents that can screen the NMR signal. High conductivity is inherent in both saline water-saturated layers and aquatards, e.g. clay-rich rocks. The resistivity of clay-rich soil and rock can be as low as 1-10 ohm-m. The resistivity of saline-water saturated rocks can amount the same value. The conductivity can affect both phase and amplitude of the NMR signal at resistivities of a few to a few tens of ohm-m depending on the depth of the water-saturated layers [2].

Unlike conventional geophysical prospecting tools used for the investigation of groundwater, selective characteristics of NMR permit new equipment to respond only to the water target, thus directly measuring free-water content. The surface NMR registers only free (movable) water, as well as bounded on rock-grains surface water can not be registered because of relaxation time being shorter than characteristic dead time of the equipment. Dynamical characteristics of water confined to porous aquifer correlate with nuclear-relaxation rate. This correlation can be described in terms of water-permeability. Nevertheless, surface-NMR field examples prove that correlation is not so simple as reported earlier phenomenological correlation between surface-NMR inhomogeneous spin-spin relaxation times and water-bearing rock grain-size. Groundwater FID and spin-echo experiments have been performed and relaxation times have been measured for porous-rock and fractured-rock aquifers as well as for bulk water under ice of frozen lake [3].

1. A. V. Legchenko, O. A. Shushakov Geophysics, 1998, v. 63, p. 75.
2. O. A. Shushakov Geophysics, 1996, v. 61, p. 998.
3. O. A. Shushakov Magnetic Resonance Imaging, 1996, v. 14, p. 959.