| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Experimental Studies |
1 From the Neuroradiology Section (D.A.F., C.P.D.) and Neuroimaging Laboratory (C.T.H., C.P.D.), Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 S Kingshighway Blvd, St Louis, MO 63110; and Research and Development Division, Siemens Medical Systems, Iselin, NJ (K.W.). Received August 10, 2000; revision requested October 3; revision received November 16; accepted December 11. Address correspondence to C.P.D. (e-mail: derdeync@mir.wustl.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Two aneurysm models were constructed: one from porcine carotid artery and the other from a pharmaceutical capsule. Both were filled with Guglielmi detachable coils. The models were tested with a 3-T MR imager for heating, deflection, and imaging artifact. Testing for heating and deflection was performed (a) at static points both inside and outside the bore, (b) during movement into the imager, and (c) during clinical imaging sequences.
RESULTS: No change in temperature was measured during movement into the imager bore or at different points within the bore. No differences in heating from radio-frequency energy were found between aneurysm models and controls. Similarly, no evidence of deflection of the coil mass (capsule model) was found. Minor susceptibility artifacts were found in the readout direction during gradient-echo sequences. Magnetic field mapping showed no induced field inhomogeneity.
CONCLUSION: MR imaging at field strengths of 3 T in patients with aneurysms treated with Guglielmi detachable coils is safe. Imaging artifacts are likely to be minimal.
Index terms: Aneurysm, cerebral, 17.73 • Head, MR, 10.121411, 10.121412, 10.12142 • Magnetic resonance (MR), safety, 10.121411, 10.121412, 10.12142 • Radiology and radiologists, iatrogenic injury
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Guglielmi detachable coils (GDCs) (Boston Scientific/Target Therapeutics, Boston, Mass), used for the endovascular occlusion of intracranial aneurysms, have been less extensively evaluated than other intravascular devices, particularly at high field strengths (6,7). To our knowledge, researchers in prior studies have not investigated several important factors that might pose dangers to patients with intracranial aneurysms treated with GDCs, such as (a) heating at all points from the imaging center to points outside the portal, (b) heating during movement into and out of the imager, (c) heating during imaging with high radio-frequency (RF) energy commonly used to image the human head (relative to the small amount needed to image a small coil alone), and (d) horizontal and longitudinal deflection forces at these static positions during movement into and out of the imager and during imaging sequences. This study was conducted to evaluate safety-related issues and imaging artifacts of GDCs in vitro with 3-T MR imaging.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
Temperature Probe Placement
Heating measurements were made by using both the capsule and the artery models. Temperatures were recorded with four simultaneous probes in real time by using an MR-compatible in situ fiberoptic temperature-sensing device (model 755 multichannel fluoroptic thermometer; Luxtron, Santa Clara, Calif) sensitive to 0.1°C. Temperature data were recorded by a single investigator (C.T.H.).
MR-compatible Box
To precisely locate the models during both the heating and deflection experiments, a foam core box was constructed to fit the RF head coil of a 3-T superconducting MR imager (Siemens Medical Systems, Iselin, NJ) (Fig 2). This MR-compatible box was 1 m in length, with a square cross-sectional shape and a diagonal measurement of 265 mm to match the inside diameter of the RF head coil. A sliding floor was constructed for the box to allow placement at exact coordinates in the x, y, and z planes (the z plane was along the length of the MR imager bore, including the space outside the portal). The box was then placed in the RF coil on the patient gantry and was driven partially into the bore up to and including the imaging volume (Fig 2). Twenty-five centimeters of the box extended outside of the bore, and 75 cm of the box was located in the bore and included the imaging volume.
|
For the imaging experiment, the four temperature probes were located (a) in the coil mass of the porcine model suspended in 4 mL of saline, (b) in a control artery (without coils) suspended in saline, (c) in a control syringe of the same volume, and (d) in the dry coil mass of the capsule model surrounded by air. Measurements were obtained during two sequences: MR angiographic sequence (29/5 [repetition time msec/echo time msec], 20° flip angle, 0.89-mm section thickness, duration of 6 minutes 33 seconds) and a fast spin-echo sequence (2,510/104, 180° flip angle, 1.0-mm section thickness, duration of 27 minutes). The model and the controls were placed in the center of the imaging volume and were surrounded by five 200-mL intravenous bags of saline to achieve the specific absorption rate limit of 3.3 W per kilogram of body weight for the human head. An RF strength of 119–160 V was achieved, maximizing RF energy deposition. Twenty temperature measurements were recorded for each probe during each imaging sequence.
Assessment of Attractive Forces: Static, during Motion, and during Imaging
Attractive forces were quantitatively measured by using a well-established method (1,8). The dry capsule model was suspended by a thread to allow deflection in the direction of the most rapidly increasing field or highest gradient, while the angle formed from the vertical axis was measured (Fig 1) by two investigators (C.T.H., K.W.). The force F acting on this mass orthogonal to gravity was calculated with the following equation: F = mg tan 
, where m is mass (in grams), g is the acceleration of gravity (980 cm/sec2), and is the angle made with the vertical axis. The attractive force (in dynes) can then be converted to a more intuitive unitless number by using the equation F(g) = F(dyne)/mg (9), where g is the force of gravity.
Attractive forces were measured (C.T.H., K.W.) at numerous static points in the x, y, and z planes, where x is the left-to-right horizontal direction in the bore, y is the vertical distance from the top to the bottom of the bore, and z is the distance from the portal outside the bore or distance from the portal into the bore and the imaging volume. For measurements along the z plane, x was 0 and y was 0; and for measurements along the x plane, y was 0 and x was 0.
Measurements of deflection were obtained at 1- and 10-cm increments along the x and z planes. The z plane was measured not only outside the bore and at the portal but also inside the bore. Since the deflection would be expected along the z direction when x and y are zero, it is impossible to visualize deflection deep in the bore. However, in this experiment, a 2 x 2 cm front-surface mirror mounted at 45° to a 1-m-long balsa rod allowed visualization of even the slightest movement deep within the magnet, preventing parallax error induced by the observer’s eye being placed offline with the string and the protractor (Fig 2).
While longitudinal z and horizontal x forces were measured with this method, vertical y forces were not directly measured, and the reason for this is twofold. First, vertical forces cannot be measured in the same way as the other forces because they are parallel to gravity. Second, horizontal forces, not unlike the vertical forces, are measured at increments approaching the wall of the magnet. Therefore, vertical forces could be estimated from horizontal forces.
The capsule model was observed for deflection for 30 seconds to 1 minute at each position during static measurements and imaging sequences. For the measurements obtained during gantry motion, the inertia of the capsule was accounted for by building a simple mechanism with a fixed-length string attached to a plate that held the capsule as the table accelerated but let the capsule loose once a constant velocity was achieved. These observations were made during four cycles of motion completely into and out of the imager at a rate of 0.1 m/sec.
Assessment of Artifacts during Imaging
Images were acquired with the porcine model suspended in both saline and gadodiamide (Nycomed Amersham, Princeton, NJ) at the appropriate concentration. The saline bags were used again and were placed out of the phase-encoding direction. The selected imaging sequences included the following: (a) field-mapping sequence, 300/100, 90° flip angle, one signal acquired, duration of 1 minute 30 seconds, 5.0-mm section thickness; (b) gradient-echo fast low-angle shot sequence, 250/18, 50° flip angle, two signals acquired, duration of 3 minutes 13 seconds, 2.0-mm section thickness, 40-kHz bandwidth; (c) fast spin-echo T2-weighted sequence, 3,000/96, 180° flip angle, one signal acquired, duration of 3 minutes 40 seconds, 2.0-mm section thickness, 33-kHz bandwidth; (d) spin-echo T1-weighted sequence, 290/14, 90° flip angle, six signals acquired, duration of 11 minutes, 1.0-mm section thickness, 143-kHz bandwidth; and (e) spin-echo T1-weighted sequence, 300/20, 90° flip angle, one signal acquired, duration of 4 minutes 32 seconds; 2.0-mm section thickness, 40-kHz bandwidth.
Higher-spatial-resolution T1-weighted images (sequence d compared with sequence c) were obtained by using thinner sections, higher bandwidths, and more repetitions. Images were visually inspected by all authors for subjective evidence of distortion of length or width, as well as for changes in signal intensity in or near the coil mass.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
|
|
Imaging Artifact
Magnetic field mapping sequences failed to demonstrate any visible alterations in field homogeneity caused by GDCs (Fig 5). Minor alterations in field strength were present at the interface between the syringe (containing the gadodiamide-saline solution) and the normal saline–filled cup in which it was suspended. Gradient-echo sequences caused a distortion in size. The coil mass and the surrounding artery wall were enlarged or drawn out (Fig 6) in the readout direction. This artifact was greatest at lower bandwidths. A narrow band of increased signal intensity was noted surrounding the coil mass on spin-echo images (Figs 7 –9). This high signal intensity was particularly apparent on T2-weighted fast spin-echo images (Fig 7), as compared with T1-weighted images (Figs 8, 9). It was also less visible on a standard T1-weighted image (Fig 9). Imaging of the dry capsule model was impossible, since no signal intensity from the capsule, coils, or air surrounding either could be visualized.
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
The potential for heating of GDCs is founded on the fact that current may be induced in an electric conductor, as the magnetic and RF field gradients are applied or as the conductor moves through the fringe region of the static field outside the imager. Furthermore, a circuit with variable resistance created by overlapping GDCs may lead to heating. Arcing and heat production between various cut ends in the coil mass are also possible. Arcing has been tested in copper and steel rings, without a positive result (8).
To date, no heating has been demonstrated in GDCs; however, to our knowledge, in only one study (6) has the heating in GDCs been tested. In this experiment, measurements were made in a static position (without movement into or out of the imager), and the position in the bore was not stated.
In the present study, we evaluated two additional variables that might lead to induced current and heating (motion in the magnetic field and image acquisition). According to the Faraday law, a voltage is induced in any conducting material placed in a time-varying magnetic field. If a circuit is produced, the current may create localized heating or arcing. MR imaging requires three electromagnetic fields: (a) a static and constant field used to align the protons, (b) a smaller rapidly time-varying second magnetic field for spatial localization, and (c) a time-varying RF field. In modern superconducting MR imagers, the static magnetic field corresponds to the stated type of imager (eg, 1.5 or 3 T). While this field is not time-varying and therefore should not induce a current in a conductor, it becomes time-varying when the patient is moved from outside the bore, where the field is weakest, to the inside of the bore, where it is strongest.
In fact, one of the few reported (11) cases of ferromagnetic complications of MR imaging occurred during the movement of the patient out of the magnet, when an intraocular foreign body caused damage and blindness. To our knowledge, no metallic implants have been tested for movement or heating while being moved into or out of the MR imager (1–12). The second and third time-varying electromagnetic fields are active during imaging sequences. In these prior studies (1–12), the effect of imaging sequences on heating was not systematically evaluated.
We observed no evidence of heating at any static location or with movement into or out of the field. Heating was not observed during a standard angiographic sequence. Heating was measured during a 27-minute sequence designed to deposit the maximum allowable RF energy. However, the observed increase in temperature was less than 1°C and was similar between the control saline-filled syringe, the control artery, and the coils (Fig 9). These data were not independent samples, and therefore, a parametric test for differences would be invalid. This experiment was designed to provide the maximum sensitivity to heating: no simulation of blood flow, a small fluid volume for the arterial model, and no fluid volume in the capsule model.
Platinum coils are thought to be nonferromagnetic and are, at most, paramagnetic. Paramagnetic objects have the potential to align along the direction of the lines of magnetic flux and display a force proportional to not only the gradient but also to the field strength (12). Therefore, a 3-T imager could have more effect on platinum coils than a 1.5-T imager, as the magnitude of the field strength is much greater.
An essential part of any attractive force is the size of the gradient. While the static field of an MR imager is uniform in the imaging portion of the magnet, a steep gradient exists outside of the isocenter (9). The point of highest gradient is different for different MR imagers. Kagetsu and Litt (9) have found that in resistive-type imagers, the highest gradient is at the portal, and in superconducting types (as was used in this experiment), the highest point is well within the bore. Despite this, most previous tests (1) have been performed only at the portal of the magnet. In this experiment, forces were measured at all points inside and outside of the bore. Furthermore, the gradient was expected to be steeper, regardless of its location, in a 3-T magnet than in a 1.5-T magnet.
Eddy currents are another potential hazard when moving a conducting material through a magnetic field. When any magnetic field, static or time-varying, is applied to a moving conductor, eddy currents are produced in the conductor. This phenomenon generates a retarding force, which opposes any further movement (9). This may be demonstrated by holding a conducting but nonferromagnetic aluminum plate of any dimension near the bore of an MR imager. As the experimenter attempts to move the plate, a considerable retarding force will be felt to any movement, which becomes more apparent as the plate is moved within the magnet, where the field is strongest. Eddy currents induced in the coils during movement across the fringe regions of a magnetic field were either insufficient or nonexistent, causing neither apparent heating nor deflection forces.
In this study, deflection was not encountered at any point in the length of the magnet outside the homogeneous imaging volume. Time-varying gradients of RF and magnetic fields during imaging were also ineffective in producing any measurable deflection. Neither magnetic nor RF-induced torque (the rotation of an object about an axis) was objectively tested in this experiment. Qualitatively, no torque was observed or felt during the experiment.
With field-mapping sequences (Fig 5), the field homogeneity can be visualized even with slight distortions of that field (each concentric line in Fig 5 represents a small 50-Hz increment). GDCs produced no discernable field inhomogeneity during this sequence. Alterations that did occur arose from the interface of the contrast material–filled syringe and the normal saline–filled cup, rather than the suspended coils themselves. Resolution of detail in the tight coil mass of the porcine model was difficult to achieve with only normal saline. The relatively low signal intensity that arose from the small quantities of normal saline was difficult to distinguish from the signal void of the coils themselves. The addition of gadodiamide increased the signal intensity from the fluid and allowed resolution of fine detail in the tight coil mass (Fig 9).
Two imaging artifacts were identified. Gradient-echo sequences (250/18, 50° flip angle) produced an elongation of the dimensions of the model only along the readout direction (Fig 6). The second artifact was a hyperintense rim of signal intensity that could be seen immediately adjacent to the coil mass (Figs 7–9). This bright artifact was most apparent on fast spin-echo images and at higher echo time values (Fig 7). This effect was also more pronounced during imaging sequences that used lower bandwidths (33 kHz in Fig 9 vs 143 kHz in Fig 8). This artifact has been previously described (6,13–15) with MR imaging of GDC-treated aneurysms at 1.5 T.
Practical application: We found no evidence for heating or deflection of GDCs in a comprehensive series of in vitro experiments at 3 T. We rigorously evaluated several experimental conditions not previously reported, such as eddy currents, phantoms used to increase RF energy, motion, position in the MR imager, and simultaneous temperature monitoring. We conclude that MR imaging at 3 T of patients with GDC-treated intracranial aneurysms is safe. Furthermore, imaging artifacts are minimal, and high-spatial-resolution structural and functional imaging is likely to be feasible.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, C.P.D.; study concepts, C.T.H., C.P.D., D.A.F.; study design, C.T.H., C.P.D.; literature research, C.T.H., C.P.D.; experimental studies, C.T.H., C.P.D.; data acquisition, C.T.H., K.W.; data analysis/interpretation, C.T.H., K.W., C.P.D.; statistical analysis, C.P.D.; manuscript preparation, C.T.H.; manuscript definition of intellectual content, C.P.D.; manuscript editing and revision/review, K.W., C.P.D., D.A.F.; manuscript final version approval, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. N. Levine, A. S. Gomes, A. E. Arai, D. A. Bluemke, S. D. Flamm, E. Kanal, W. J. Manning, E. T. Martin, J. M. Smith, N. Wilke, et al. Safety of Magnetic Resonance Imaging in Patients With Cardiovascular Devices: An American Heart Association Scientific Statement From the Committee on Diagnostic and Interventional Cardiac Catheterization, Council on Clinical Cardiology, and the Council on Cardiovascular Radiology and Intervention: Endorsed by the American College of Cardiology Foundation, the North American Society for Cardiac Imaging, and the Society for Cardiovascular Magnetic Resonance Circulation, December 11, 2007; 116(24): 2878 - 2891. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Costalat, E. Lebars, L. Sarry, A. Defasque, E. Barbotte, H. Brunel, G. Bourbotte, and A. Bonafe In Vitro Evaluation of 2D-Digital Subtraction Angiography versus 3D-Time-of-Flight in Assessment of Intracranial Cerebral Aneurysm Filling after Endovascular Therapy AJNR Am. J. Neuroradiol., January 1, 2006; 27(1): 177 - 184. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. B. L. M. Majoie, M. E. Sprengers, W. J. J. van Rooij, C. Lavini, M. Sluzewski, J. C. van Rijn, and G. J. den Heeten MR Angiography at 3T versus Digital Subtraction Angiography in the Follow-up of Intracranial Aneurysms Treated with Detachable Coils AJNR Am. J. Neuroradiol., June 1, 2005; 26(6): 1349 - 1356. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. G. Shellock, M. Gounis, and A. Wakhloo Detachable Coil for Cerebral Aneurysms: In Vitro Evaluation of Magnetic Field Interactions, Heating, and Artifacts at 3T AJNR Am. J. Neuroradiol., February 1, 2005; 26(2): 363 - 366. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Nehra, C. J. Moran, D. T. Cross III, and C. P. Derdeyn MR Safety and Imaging of Neuroform Stents at 3T AJNR Am. J. Neuroradiol., October 1, 2004; 25(9): 1476 - 1478. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. G. Shellock and J. V. Crues MR Procedures: Biologic Effects, Safety, and Patient Care Radiology, September 1, 2004; 232(3): 635 - 652. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. G. Shellock, J. A. Tkach, P. M. Ruggieri, T. J. Masaryk, and P. A. Rasmussen Aneurysm Clips: Evaluation of Magnetic Field Interactions and Translational Attraction by Use of "Long-Bore" and "Short-Bore" 3.0-T MR Imaging Systems AJNR Am. J. Neuroradiol., March 1, 2003; 24(3): 463 - 471. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
= probe between the coil and artery wall, 
= probe in the coils, 
= probe outside aneurysm wall, and 
= probe in the saline bath. In b, 
