[Show abstract][Hide abstract] ABSTRACT: Many MRI applications such as dynamic contrast-enhanced MRI of the breast require high spatial and temporal resolution and can benefit from improved gradient performance, e.g., increased gradient strength and reduced gradient rise time. The improved gradient performance required to achieve high spatial and temporal resolution for this application may be achieved by using local insert gradients specifically designed for a target anatomy. Current flat gradient systems cannot create an imaging volume large enough to accommodate both breasts; further, their gradient fields are not homogeneous, dropping off rapidly with distance from the gradient coil surface. To attain an imaging volume adequate for bilateral breast MRI, a planar local gradient system design has been modified into a superellipse shape, creating homogeneous gradient volumes that are 182% (Gx), 57% (Gy), and 75% (Gz) wider (left/right direction) than those of the corresponding standard planar gradient. Adding an additional field-modifying gradient winding results in an additional improvement of the homogeneous gradient field near the gradient coil surface over the already enlarged homogeneous gradient volumes of the superelliptical gradients (67%, 89%, and 214% for Gx, Gy, and Gz respectively). A prototype y-gradient insert has been built to demonstrate imaging and implementation characteristics of the superellipse gradient in a 3 T MRI system.
Magnetic Resonance in Medicine 03/2011; 65(3):863-72. DOI:10.1002/mrm.22639 · 3.57 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Many MRI applications such as dynamic contrast-enhanced MRI of the breast require high spatial and temporal resolution and can benefit from improved gradient performance, e.g., increased gradient strength and reduced gradient rise time. The improved gradient performance required to achieve high spatial and temporal resolution for this application may be achieved by using local insert gradients specifically designed for a target anatomy. Current flat gradient systems cannot create an imaging volume large enough to accommodate both breasts; further, their gradient fields are not homogeneous , dropping off rapidly with distance from the gradient coil surface. To attain an imaging volume adequate for bilateral breast MRI, a planar local gradient system design has been modified into a superellipse shape, creating homogeneous gradient volumes that are 182% (G x), 57% (G y), and 75% (G z) wider (left/right direction) than those of the corresponding standard planar gradient. Adding an additional field-modifying gradient winding results in an additional improvement of the homogeneous gradient field near the gradient coil surface over the already enlarged homogeneous gradient volumes of the superelliptical gradients (67%, 89%, and 214% for G x , G y , and G z respectively). A prototype y-gradient insert has been built to demonstrate imaging and implementation characteristics of the superellipse gradient in a 3 T MRI system. Magn Reson Med 65:863–872, 2011. V Many imaging tasks, such as dynamic contrast-enhanced MRI for breast lesion characterization, can benefit from high spatial and temporal resolution. The improved gradient performance required for high spatial and temporal resolution may be achieved by local gradient coils such as planar insert gradients. With higher gradient strength and slew rate, the local planar gradient can attain higher spatial and temporal resolution than the body gradients. In general, local gradient coils are designed to produce a smaller homogeneous gradient volume (HGV) compared with that of the MRI system whole-body gradients. For local planar gradients, the x-gradient, which requires four fingerprint patterns per plane rather than two fingerprints as in the y-gradient, results in the most limited imaging volume. As a result, for certain applications such as breast imaging, the HGV may be too small to cover the entire desired imaging region (i.e., both breasts). Because of the small imaging volume, there have been very few attempts to image both breasts simultaneously with a local gradient system (1). MRI of the female breast is performed typically with the patient in a prone position to reduce breathing-induced motion artifacts and increase reproducibility. This position places the breast near the planar insert gradient. However, the magnetic field as well as gradient strength of uniplanar gradient systems falls off nearly exponentially with distance from the coil surface. As a result, the gradient varies continuously and even an approximately linear region is narrow in the direction of each gradient. Image distortion caused by the nonlinear gradient can be corrected to some extent by using a non-linear compensating transformation (2,3). Nonplanar surface geometries might be a good alternative for homogeneous breast gradient designs that can cover the volume of both breasts. The boundary element method, which was introduced by Pissanetzky (4), allows gradient coil design of irregular surfaces by using a finite element mesh with a great deal of geometrical freedom, and further research has followed (5,6). However , practical applications of unusually shaped gradient coil designs (7) remain unanswered because of manufacturing issues. In this study, to create wider homogeneous imaging volumes, the planar insert gradient geometry was extended laterally, and the edges were bent vertically using a superelliptical curvature on the left and right sides, creating a so-called box-in or superellipse shape to fit in the magnet bore. Bending the wire pattern on the vertical edges increases its effects on the imaging volume by increasing the gradient uniformity in the left/right and anterior/posterior directions. To increase the gradient strength and to further improve the gradient homoge-neity in all directions, an additional field-modifying (FM) layer of current windings was added. Wire patterns are designed using stream functions (8,9). We present design and optimization methods for a multilayered superelliptical local insert gradient designed for breast imaging with a homogeneous imaging
Magnetic Resonance in Medicine 01/2011; 65:863-872. · 3.57 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The tradeoff between gradient performance factors, size of the imaging region, and physiological factors such as nerve stimulation typically leads to compromises in gradient design and ultimately suboptimal imaging performance. Local gradient systems can add some performance flexibility, but are cumbersome to set up and remove. In nearly all conventional MRI systems, the use of local gradients precludes the use of the more homogeneous whole body gradients. This paper presents the concept of dynamically selectable composite gradient systems where local gradients and whole body gradients can be selected independently and simultaneously. The relative performance of whole body, insert, and composite gradients is predicted for echoplanar (EPI), turbo spin echo (TSE), and steady state free precession (SSFP). A realization of the concept is presented.
Concepts in Magnetic Resonance Part B Magnetic Resonance Engineering 04/2009; 35(2):89-97. DOI:10.1002/cmr.b.20134 · 0.69 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: We have constructed a small-bore insertable gradient coil with two linear gradient imaging regions and interfaced it with an MRI scanner. We have also constructed an RF system capable of transmitting or receiving in both regions simultaneously.Designs for conductor placement for two-region X-, Y- and Z-gradient coils were optimized by simulated annealing. Wire patterns for each axis were chosen that gave low inductance, reasonable homogeneity over a large imaging volume and high efficiency (gradient field per-unit-current).Imaging was performed on a Siemens 3T TIM Trio scanner equipped with three additional gradient amplifier channels and a second RF/gradient array controller. Phantoms were placed in the two imaging regions as well as the central non-imaging region to test gradient homogeneity and crosstalk between regions. Images acquired simultaneously in the two regions showed very little signal crosstalk between imaging regions and even less signal from the central, non-imaging region.When combined with an overlapping single-region gradient insert, extended field-of-view (FOV) imaging will be possible without moving the table or the subject and without increasing nerve stimulation. Construction and testing of a two-region gradient coil insert is a necessary intermediate step as a proof of concept for an extended field of view, contiguous, three-region human-sized gradient system.
Concepts in Magnetic Resonance Part B Magnetic Resonance Engineering 04/2009; 35B(2):98-105. DOI:10.1002/cmr.b.20138 · 0.69 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Chronic hemodialysis requires a vascular access that provides high blood-flow rates for the extracorporeal recirculation of blood. Synthetic arteriovenous (AV) grafts often fail because of clotting caused by underlying hyperplasia formation. The authors report the use of magnetic resonance (MR) imaging (MRI) without contrast agent to monitor tissue hyperplasia formation as well as luminal area in a porcine model of AV graft stenosis.
Expanded reinforced polytetrafluoroethylene grafts were surgically placed between the common carotid artery and the external jugular vein, bilaterally, in pigs. Animals underwent MRI in a 3-T scanner at 3, 4, or 6 weeks after graft placement, followed by euthanasia and the collection of grafts and adjacent tissues for histologic analysis. Two animals underwent sequential scanning at 1, 2, 3, 5, and 7 weeks after graft placement, followed by histologic analysis.
Measurements of hyperplasia obtained from the MR images were compared with, and correlated well with, measurements obtained from the histologic cross-sections (r = 0.932, P = .02). The MR images provided a more complete view of the venous hyperplasia throughout the graft compared with histology. The MR images could be examined from multiple angles and were unaffected by histologic preparation artifacts.
Unlike histology, MRI provided longitudinal 3-dimensional views of hyperplasia within the AV grafts. This ability of MRI to more completely identify the geometry of hyperplasia and to quantify the tissue volume in vivo could provide benefits over histologic analysis in assessing the pathology of AV graft failure and the efficacy of antihyperplasia interventions.
[Show abstract][Hide abstract] ABSTRACT: A methodology using magnetic resonance angiography (MRA) is presented for identifying thermally significant blood vessels in isolated kidneys, specifically for use in biothermal model development with application to high intensity focused ultrasound (HIFU). A combination of a proven preservation technique, newly developed MR-compatible experimental procedures and the refinement of MR pulse sequence parameters was used to determine vascular characteristics using high-resolution three-dimensional time-of-flight MRA image of flow through isolated kidneys. Results presented are twofold. First, improved vessel visibility was attained through decreasing the magnetic resonance imaging bandwidth from 150 to 30 Hz/pixel while simultaneously increasing the echo time, repetition time, and flip angle; vascular center line extraction showed an 18% improvement in the number of vessel segments detected and a 23% increase in length of the terminal segments over a base line technique without improvements. Second, the overall system was shown to be practical to determine vascular flow effects during HIFU heating; testing results from heating the kidney with HIFU are presented, showing a decrease of average kidney temperature with an increase of flow rate through the kidney with localized cooling demonstrated surrounding known vessel locations.
Medical Physics 11/2008; 35(10):4426-34. DOI:10.1118/1.2975226 · 2.64 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Computer simulations and measurements on human volunteers were used to test the extent to which the quality of carotid imaging might be improved by coil arrays that are not limited by a constraint on the number of RF coil receiver ports.
Analytic near-field equations for the magnetic and electric fields of a rectangular loop resonator were used to estimate the relative signal-to-noise ratio (rSNR) along the length of a simulated carotid artery as a function of loop size, loop position and vessel depth. The sizes, positions and number of elements in a linear coil array that resulted in the maximum composite SNR along the length of a simulated carotid artery were then estimated. The linear array results were used to predict the total number of elements needed for optimal imaging of the carotid arteries. Also, three normal volunteers were imaged with a variety of RF coils, and the rSNR measurements along the lengths of the carotid artery were evaluated for each coil combination.
The analytic simulation and the human volunteer measurements both show that improved SNR (e.g., >300% at the bifurcation) can be obtained with coils tailored to each specific region of the carotid artery in comparison to that obtained with four-element arrays designed and used to image the entire carotid artery.
The resulting number of coil ports, 16 to 24, required for full coverage of the carotid arteries is consistent with the number of channels just becoming available on recently developed clinical scanners.
Magnetic Resonance Imaging 07/2005; 23(5):629-39. DOI:10.1016/j.mri.2005.04.009 · 2.09 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The structures of polydisulfide-based biodegradable macromolecular Gd(III) complexes were modified to improve their in vivo retention time and MRI contrast enhancement. Steric hindrance was introduced around the disulfide bonds to control their access to free thiols in order to alter the degradation rate of the copolymers. Two new macromolecular agents, (Gd-DTPA)-cystine copolymers (GDCP) and (Gd-DTPA)-cystine diethyl ester copolymers (GDCEP), were prepared. Both agents were readily degraded in vitro and in vivo by the disulfide-thiol exchange reaction, but at a slow rate. The introduction of COOH and COOEt groups slowed down the degradation of the copolymers in the incubation with 15 microM cysteine. Metabolic degradation products were identified by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry in the urine samples from rats injected with the agents. The T(1) relaxivity (r(1)) was 5.43 mM(-1)s(-1) for GDCP, and 5.86 mM(-1)s(-1) for GDCEP, respectively, at 3T. MRI contrast enhancement of both agents was studied in nude mice bearing MDA-BM-231 human breast carcinoma xenografts, on a Siemens Trio 3T scanner. The modified agents resulted in more significant contrast enhancement in the blood pool and tumor periphery than (Gd-DTPA)-cystamine copolymers (GDCC) and a low-molecular-weight control agent, Gd-(DTPA-BMA), at a dose of 0.1 mmol-Gd/kg. The results demonstrate that the structural modification of the biodegradable macromolecular Gd(III) complexes resulted in a relatively slow degradation of the macromolecules and significantly improved in vivo contrast enhancement. The modified agents show promise for use in investigations of blood pool and cancer by contrast-enhanced (CE) MRI.
Magnetic Resonance in Medicine 04/2005; 53(4):835-42. DOI:10.1002/mrm.20402 · 3.57 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Sensitivity encoding (SENSE) was combined with keyhole and selective line acquisition mode (SLAM) techniques to acquire a time series of images during contrast passage. The acquisition speed of the dynamic time frames was improved by a factor of 8 in total. The high spatial frequencies were sampled during the steady state and combined with the dynamic time frames to construct a series of high-resolution time-resolved contrast-enhanced 3D images. Filtered temporal correlation analysis was used to separate the arteries and veins.
Magnetic Resonance Imaging 12/2004; 22(9):1161-8. DOI:10.1016/j.mri.2004.08.025 · 2.09 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Biodegradable PEGylated Gd-DTPA l-cystine copolymers, PEG-g-poly(GdDTPA-co-l-cystine), were prepared and tested as a blood pool contrast agent in mice. The biodegradable macromolecular agent was designed to be broken down into smaller Gd complexes by endogenous thiols via the disulfide-thiol exchange reaction to facilitate the clearance of Gd complexes after the contrast-enhanced MRI examination. Gd-DTPA l-cystine copolymers were synthesized by condensation polymerization of l-cystine and DTPA-dianhydride in water followed by chelating with Gd(OAc)(3). MPEG-NH(2) (MW = 2000) was then conjugated to the polymeric backbone in different ratios. The macromolecular contrast agent was readily degraded with the incubation of l-cysteine. It also demonstrated superior contrast enhancement in the heart and blood vessels as compared to a low molecular weight control agent, Gd-(DTPA-BMA). At 1 h postcontrast, the PEGylated macromolecular agent still showed prominent enhancement, while little contrast enhancement was detectable in the blood pool by the control agent. PEG-g-poly(GdDTPA-co-l-cystine) shows promise as an MR blood pool imaging agent.
[Show abstract][Hide abstract] ABSTRACT: To study the accumulation of macromolecules into the arthritic joints and the possible applications of such phenomenon.
The accumulation of plasma albumin in the joints of adjuvant-induced arthritis (AIA) rat model was first visualized with Evans blue injection. A N-(2-hydroxypropyl)methacrylamide (HPMA) copolymer contrast agent was then synthesized and injected into the AIA rats to allow qualitative examination of biodistribution and pharmacokinetics of the injected macromolecule with magnetic resonance imaging (MRI). Vital organs and the diseased joints were isolated and examined histologically to correlate with the MRI findings.
Deep blue color developed around the arthritic joints of the AIA rat a few hours after the injection of Evans blue. MR imaging of the AIA rats injected with polymer contrast agent demonstrated a gradual but very strong accumulation of the injected polymer in the arthritic joints, which lasted for 1-2 days. Observed differences in the concentration of the injected polymer in the joints correlated with disease severity as assessed histologically.
Profound arthrotropism of macromolecules in the AIA rat model was demonstrated with various imaging tools. These observations should help in the conceptual and practical design of novel macromolecular delivery systems for the imaging and treatment of rheumatoid arthritis.
Pharmaceutical Research 11/2004; 21(10):1741-9. DOI:10.1023/B:PHAM.0000045232.18134.e9 · 3.42 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: To evaluate the ability of high-resolution MRA to monitor changes in intracranial aneurysm volume, and devise a highly reliable technique for obtaining these measurements.
To obtain a baseline estimate of the repeatability of MRA scans and validate the statistics-based technique for aneurysm volume measurement, multiple scans were obtained on individual subjects over a period of up to 1 year. These 3D MRA data sets were coregistered and then analyzed using the volumetric analysis of segmented data and the proposed statistical method.
It was shown that high-resolution MRA provides highly repeatable data sets. Both methods used for the aneurysm volume measurements showed consistent results. However, the proposed statistical method had lower error and was much less sensitive to the choice of segmentation parameter than the volumetric analysis of segmented data. A change of 1 mm in the average radius of the aneurysm was detectable with the statistics-based technique.
This study demonstrates that the statistical method of aneurysm volume measurement in high-resolution MRA allows reliable and accurate assessments of aneurysm volume changes.
Journal of Magnetic Resonance Imaging 08/2004; 20(2):340-6. DOI:10.1002/jmri.20108 · 3.21 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The clinical application of macromolecular gadolinium (Gd) complexes as MRI contrast agents is limited by the slow excretion of Gd(III) complexes and consequent long-term tissue accumulation of toxic Gd ions. To alleviate the problem of slow excretion, biodegradable polydisulfide-based macromolecular Gd(III) complexes were designed and prepared based on the disulfide-thiol exchange to allow degradation of the macromolecules by endogenous thiols and to facilitate excretion of Gd(III) complexes after the MRI examination. The in vitro degradation study showed that the polydisulfide agent was readily degraded by cysteine at plasma thiol concentrations. No cross-reaction was observed between the cysteine-34 on human serum albumin (HSA) with the agent. Concentration-dependent blood pool contrast enhancement was observed for the polydisulfide agents. The agents of both high molecular weight (35,000 Da) and low molecular weight (17,700 Da) produced significant contrast enhancement in the heart and aorta in rats at relatively high doses. Except for the bladder, the signal intensities gradually decreased over time. Significant blood pool contrast enhancement was also observed for the high molecular weight agent at a low dose (0.03 mmol-Gd/kg), but not for the agent with a lower molecular weight. The contrast enhancement in the urinary bladder increased over time for the polydisulfide agents and Gd(III)-(DTPA-BMA). Degradation products were identified by mass spectrometry in the urine samples from the rats administered with both polydisulfide agents, which confirmed that the polydisulfide agents were degraded in vivo and excreted through renal filtration. The preliminary results demonstrated the in vitro and in vivo degradability, superior blood pool contrast enhancement, and rapid clearance through renal filtration of the novel biodegradable macromolecular agent. This agent has a great potential for further preclinical and clinical development with application in contrast-enhanced blood pool and cancer MR imaging.
Magnetic Resonance in Medicine 01/2004; 51(1):27-34. DOI:10.1002/mrm.10656 · 3.57 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: To demonstrate that the time delay between phase and frequency encoding and the presence of pulsatile blood flow in high-resolution time-of-flight (TOF) imaging of the intracranial arteries (especially near the circle of Willis) can distort the appearance of blood vessels and result in a cross-hatch-appearing artifact in surrounding tissue.
Two techniques to reduce the artifact, tri-directional flow compensation (3DFC) and elliptical-centric (EC) phase-encoding order, are investigated in five volunteer studies.
3DFC eliminates the pulsation-related artifacts and the vessel distortion. A residual amplitude variation artifact is observed. EC phase encoding nearly eliminates the pulsatile motion-related artifact, but it does not eliminate vessel distortion.
The combination of 3DFC and EC phase encoding appears to provide the greatest artifact reduction in the five volunteer studies performed.
Journal of Magnetic Resonance Imaging 07/2003; 18(1):121-7. DOI:10.1002/jmri.10322 · 3.21 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The clinical application of macromolecular Gd(III) complexes as MRI contrast agents is impeded by their slow excretion and potential toxicity due to the release of Gd(III) ions caused by the metabolism of the agents. A polymer Gd(III) chelate conjugate with a cleavable spacer has been designed to solve this problem. Poly(l-glutamic acid)-cystamine-[Gd(III)-DOTA] was prepared by the conjugation of DOTA to PGA (MW = 50,000) via cystamine, a cleavable disulfide spacer, followed by the complexation with GdCl(3). A Gd(III) DOTA chelate derivative was readily released from the polymer conjugate in the incubation with cysteine, an endogenous plasma thiol. The conjugate produced significant MRI blood pool contrast enhancement in nude mice bearing OVCAR-3 human ovarian carcinoma xenographs. Less significant contrast enhancement was observed for a small molecular contrast agent, Gd(DTPA-BMA). The pharmacokinetic MRI study showed that the Gd(III) chelate from the conjugate accumulated in the urinary bladder in a similar kinetic pattern to Gd(DTPA-BMA), suggesting that the chelate was released by the endogenous thiols and excreted through renal filtration. The preliminary results suggest that this novel design has a great potential to solve the safety problem of macromolecular MRI contrast agents.
[Show abstract][Hide abstract] ABSTRACT: Synopsis In order to determine reproducibility of MR Angiography acquisition and segmentation, MRA images must be co-registered. MR Angiograms were interpolated, segmented and a landmark visible in all data sets was used as the center of rotation. Random angles within a user-defined range are generated and used to rotate the image. A binary (1=vessel, 0=not vessel) multiplication of the 3D volumes is summed and used as the figure of merit. The process is repeated with a diminishing range of random angles until the stop criterion is met. The algorithm obtains excellent registration of 3D segmented angiograms in a reasonable time. Methods Five patient volunteers have been serially scanned over a 3 year period at least four times each, in order to measure reproducibility of MR Angiography acquisition and segmentation. All MR Angiograms were acquired on a 1.5T GE NV/CVi scanner with the LX 8.4 operating system using an endcap head coil and a single volume Time of Flight spoiled gradient echo (SPGR) pulse sequence with abbreviated magnetization transfer (1). During the 3 year period many pulse sequence improvements were implemented, most notably variable TE (2), so a majority of the images were acquired with a TE on the order of 4-5msec, but with variable TE the echo time was reduced to about 2.2 ms. Images were acquired with a 22 cm FOV with an imaging matrix of 512x256x64. Raw data was saved and zero-filled interpolated offline to an image matrix of 1024x768x128. After vessel segmentation using the ZBS algorithm (3) an anatomical landmark is chosen which is visible in all MRA sets. Typically this landmark is the tip of the basilar artery, but could be any feature in the image. This landmark is used as the center of rotation for each data set. Since user choice is not reproducible (often +-1 or 2 pixels) the region around the anatomical feature is windowed, filtered to reduce boundary discontinuities, and cross-correlated to more precisely choose the same point of the feature in all sets. To register two MRA image sets, the algorithm requires starting values for the three angles of rotation, the initial range of angles for the search, and parameters to control number of iterations at each range, the factor for range reduction, and the stopping criteria. The main algorithm is essentially 2 nested loops. The outer loop simply reduces the range of the 3 random angles until the stopping range is reached. The inner loop generates 3 random angles which are used to rotate the image to be registered. A binary (1=vessel, 0=not vessel) multiplication of the 3D volumes is summed and used as the figure of merit (FOM). The FOM is tested after each rotation and if it is greater than any previous FOM it is saved along with the corresponding random angles which are then used as the range center for subsequent random angles. When reasonable good angles are found, the neighborhood near the center of rotation is tested to see if the FOM can be improved by small translations. If an improvement is found, the random angle loops are revisited. The Code is written in IDL 5.3 (Interactive Data Language, Research Systems, Inc. Boulder Co) and C and run on a SUN ultra 80 with 2G RAM and dual 450Mhz CPU. Results The example registration going from Figure 1 to 2 took less than 10 minutes on the SUN ultra 80. This included 180 sets of random angles, searching the neighborhood of the center of rotation for improvement in the FOM and then repeating the random angle search with the improved center of rotation. Conclusion This Monte-Carlo algorithm is a robust technique capable of registering 3D segmented angiograms in a reasonable time with minimal user input. References 1. Parker DL, Buswell HR, Goodrich KC, Alexander A, Keck N, Tsuruda J: The application of Magnetic transfer to MR angiography with reduced total power. Magn. Reson. Med. 34:283-286, 1995. 2. Jeong EK, Parker DL, Tsuruda JS, Won JY,Reduction of flow-related signal loss in flow-compensated 3D TOF MR angiography, using variable echo time (3D TOF-VTE). Magn Reson Med. 2002 Oct;48(4):667-76 3. Parker DL, Chapman BE, Roberts JA, Alexander AL and Tsuruda JS: Enhanced image detail using continuity in the MIP Z-buffer: applications to magnetic resonance angiography. JMRI. 11(4):378-88, 2000.