| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
RESEARCH REPORT |
1 Advanced Biomaterials,
2 Oral and Maxillofacial Radiology, Graduate School, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo, 113-8549, Japan; and
3 Dept. of Oral and Maxillofacial Radiology, School of Dentistry, The University of Tokushima;
* corresponding author, shafii.orad{at}tmd.ac.jp
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS & METHODSRESULTSDISCUSSIONREFERENCES |
|---|
KEY WORDS: magnetic resonance imaging • metal artifact • dental casting alloy • titanium • implant
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS & METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Despite these imaging characteristics, MRI has the shortcoming of being prone to magnetic susceptibility difference artifacts caused by the presence of metallic materials such as dental or orthopedic implants, dental cast restorations, and aneurysm clips (Bui et al., 2000). Artifacts in MR images can be defined as the pixels that do not faithfully represent the tissue components being studied (Brown and Semelka, 1995). All substances when placed in a magnetic field are magnetized to a degree which varies according to their magnetic susceptibility. There are three types of substances with different magnetic susceptibilities that need to be considered in MRI, namely, paramagnetic, diamagnetic, and ferromagnetic. Although the artifacts caused by ferromagnetic substances such as iron, nickel, and cobalt have been widely reported (Rudisch et al., 1998; Suh et al., 1998), the effects of dental casting alloys with compositions of various diamagnetic and paramagnetic substances or combinations of both have not yet been analyzed in detail. The purpose of our study was to test the hypothesis that, with selection of dental casting alloys or implant materials with specific composition of diamagnetic and paramagnetic substances, it is possible to minimize the metal artifacts in MRI.
|
MATERIALS & METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS & METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
|
). Commercially pure titanium (Titan A) and titanium alloy were cast with the use of an argon arc melting centrifugal casting machine (Vulcant-T, Shofu, Kyoto, Japan) and the others by an electric resistance furnace motor drive centrifugal casting machine (Tiegelschleuder TS3, Degussa, Germany). After the casting process, all cast rings were kept at room temperature. All castings were divested, then separated from the sprue and ground to 5.0 mm from the edge by means of a cutting resin disc (#0, Ruby dental-products, Osaka, Japan). In all MR sequences, the cutting disc had been confirmed not to cause any artifacts. A cube (5-mm edge) was cut from the acrylic resin plate that was used as a control sample.
Phantom
We prepared a cylindrical acrylic resin phantom (outer diameter, 50 mm; inner diameter, 44 mm; depth, 50 mm) with a cover 5 mm thick and a positioning post 10 mm high at the center. Each sample was attached to the post, and the phantom was filled with distilled water and closed without any bubbles inside.
MR Imaging
The phantom with a sample on the positioning post was placed on the table of the MRI apparatus (1.5 T, Magnetom Vision, Siemens, Germany). The laser beam of the MR system was centered on the sample. With a body coil, all samples were imaged in 3 pulse sequences. Specific conditions for the sequences were: T1 fast-spin echo (FSE) [repetition time (TR) 500 msec, echo time (TE) 12 msec, echo train length (ETL) 3, number of acquisitions (AC) 6, flip angle (FA) 180°], T2 FSE (TR 3000 msec, TE 96 msec, ETL 7, AC 2, FA 180°), and gradient echo (GE) (TR 500 msec, TE18 msec, ETL 2, AC 2, FA 180°). Slice thicknesses of 5 mm, field of view 135 mm, and matrix size of 228 x 256 were used for all sequences. Longitudinal sections through the center of each sample, parallel to the Y-Z plane, were imaged. Imaging of each sample was repeated 5 times with parameters of imaging sequences re-set each time.
A region of interest (ROI) was drawn around the image of sample with the proprietary software installed for data acquisition, data processing, and image viewing in the MRI system (Fig. 1). The ROI consisted of approximately 4000 pixels, and signal intensity (SI) distribution among pixels of the ROI was measured. Mean value and standard deviation (SD) of the SI for the entire ROI were obtained from the 5 independent imaging sessions and measurements for each sample. We calculated the coefficient of variation (CV) to evaluate the heterogeneity in the SI distribution. The reproducibility was evaluated by the mean CV and 95% confidence interval from the 5 determinations of each sample.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS & METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Artifacts appear as signal intensifications, signal voids, and blurring on nonlinear distortions (Figs. 1c, 1f). Both the commercially pure titanium (Titan A) and titanium alloys show artifacts in all 3 sequences.
Distribution of SI per pixel in ROI is shown in Fig. 2. The SI is narrowly distributed in the ROI of acrylic resin control in both T2 FSE and GE sequences (Figs. 2a, 2d). However, the SI for Cobaltan is widely distributed (Figs. 2c, 2f), where the number of pixels increased at the SI around lower and higher values than the peak value corresponding to the areas of the signal void and intensification, respectively (Figs. 1c, 1f), with the decrease in pixel number at the peak value.
|
. The mean SI per pixel for Cobaltan is significantly different from that of acrylic resin control, but the difference was small in T1 and T2 FSE compared with a large difference in the CV. The mean value of KIK-Noble and Bior17 in T1 FSE sequences was significantly different in the SI from that of control but not different in CVs. The differences of SI in T1 FS and T2 FS between BiO17 and control and those between KIK-Noble and the control were not large; however, the differences were significant because the number of pixels was great. Nevertheless, the differences of CV in T1 FS and T2 FS between BiO17 and control and those between KIK-Noble and the control were not significant. However, the mean value of New Silver in T2 FSE was not different from control but was significantly different in CVs. In T1 FSE, there was no statistical difference in the CV between control and each of 6 dental casting alloys (New Silver, PGA3, Bior17, KIK-Noble, KIK-Wing, K14 Inlay) (p > 0.05). However, in T2 FSE and GE, only KIK-Wing showed no significant difference in the mean CV from control in both sequences and showed minimum metal artifact in the images.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS & METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Magnetic susceptibility is one of the physical properties of material and can be defined as the ratio of magnetic response of a material to the applied magnetic field (White, 1970). There are three types of substances, according to their magnetic susceptibility. Diamagnetic substances have no unpaired orbital electrons. When such a substance is placed in an external magnetic field, a weak magnetic field is induced in the direction opposite the magnetic field. Thus, diamagnetic substances have a small negative magnetic susceptibility and are basically non-magnetic.
Paramagnetic substances have unpaired orbital electrons. Their induced magnetic field, under the external magnetic field, has the same direction relative to the external magnetic field. Consequently, their presence causes an increase in the effective magnetic field. Ferromagnetic substances are strongly attracted by a magnetic field and thus have high potential for MRI artifacts. Iron, cobalt, and nickel are three types of ferromagnets (Hashemi and Bradley, 1997). Currently, most clips and many implants are made of non-ferromagnetic materials such as titanium. Even so, when a patient with a "non-ferromagnetic" metal in the body is subjected to MRI, an artifact is produced which causes a drop-out of signal near the metallic surface (Bennett et al., 1996). Dental implants, orthopedic screws, and aneurysm clips are some examples of titanium alloys that produce metal artifacts on MR images, thus obscuring the images of tissues near the metallic objects.
To overcome the "metal artifacts" problem and to improve image resolution around the metallic object, scientists have investigated several new sequences or signal manipulations which are not yet able to solve the problem completely (Rudisch et al., 1998; Viano et al., 2000). To our knowledge, no quantitative studies evaluating the magnitude of metal artifacts in MRI have been reported. There have been some paradoxical results reported in the literatures about metal artifacts from titanium alloys in spin-echo sequences. Some studies reported that titanium had no significant metal artifacts in spin-echo sequences (Teitelbaum et al., 1988; Vaccaro et al., 1994). In contrast, several authors have reported that titanium alloys produced high- to moderate-magnitude artifacts in the spin-echo sequence (Suh et al., 1998; Malik et al., 2001; Ganapathi et al., 2002). In the present study, we confirmed the appearance of moderate-magnitude artifacts from commercially pure titanium and titanium alloys in the T1 FSE sequence, which is least sensitive to metal artifacts (Suh et al., 1998). We also confirmed the appearance of high-magnitude artifacts in T2 FSE and GE sequences, which are highly sensitive to metal artifacts.
The contradictory results reported in the literature might be due to differences in the parameters used in MRI, such as magnetic field intensity and specific sequences, trace amounts of ferromagnetic substances from the samples, and geometric factors in the imaging. All of these factors are known to have a significant influence on the metal artifacts in MR. One reason for the absence of artifacts in spin-echo sequences from titanium (Vaccaro et al., 1994) may be that the sample was with an extremely small amount of metallic debris. It seems difficult to detect any signal void or signal intensification in spin-echo sequences for that amount. In fact, even a stainless steel sample did not show any artifacts in spin-echo sequences, but apparently did in GE. Titanium alloy imaged in a low-magnetic field intensity of 0.35 T also showed no artifact (Teitelbaum et al., 1988). However, diagnostic efficacy of MRI in the low-magnetic field is generally limited. Trace amounts of iron (0.15–0.25%) in the commercially pure titanium sample used in the present study, confirmed by the manufacturer, could be another reason for differences in the magnitude of artifacts.
Six dental casting alloys in the T1 FSE sequence had shown minimum artifacts. These 6 alloys were composed of one paramagnetic element with other diamagnetic substances (KIK-Noble, KIK-Wing, Bior17, PGA3) or only pure diamagnetic substances (New Silver and K14 Inlay). New Silver and K14 Inlay showed artifacts in both T2 FSE and GE sequences. Bior17, titanium with a high percentage of gold, and PGA3 platinum, with high percentages of gold and silver, have showed minimum artifacts in T1 FSE but showed artifacts of only moderate magnitude in T2 FSE and GE. The main elemental composition of 2 other dental casting alloys (KIK-Noble, KIK-Wing) is palladium. Pure palladium is one of the elements that has the highest positive magnetic susceptibility among paramagnets (Weast et al., 1984). However, in the present study, a combination of palladium with diamagnetic substances such as indium, antimony (KIK-Wing), and gold (KIK-Noble) showed little artifact in T1 FSE. Only KIK- Wing showed no artifact in two other sequences, T2 FSE and GE.
In the GE sequence—which superimposes a small magnetic field onto the main field and is therefore the most metal-sensitive sequence (Hashemi and Bradley, 1997)—all dental casting alloys and implant materials showed high-magnitude artifacts except KIK-Wing (Fig. 1e). Therefore, we can conclude that the specific composition of KIK-Wing, which is currently utilized in restorative and prosthodontics dentistry, does not significantly disturb the magnetic field. It thus seems an ideal composition regarding metal artifacts in all three sequences of MRI.
There would thus be no difficulty in the diagnostic interpretation of MRIs from head and neck regions in patients with dental casting alloys that do not disturb the magnetic field. Therefore, materials for prosthetic restoration should be selected based not only on their biological compatibility and functional and esthetic qualities, but also on whether they generate minimum artifacts in MRI.
|
ACKNOWLEDGMENTS |
|---|
Received October 15, 2002; Last revision March 31, 2003; Accepted May 28, 2003
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS & METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Brown MA, Semelka RC (1995). MRI: basic principles and applications. 2nd ed. New York: Wiley-Liss.
Bui FM, Bott K, Mintchev MP (2000). A quantitative study of the pixel-shifting, blurring and nonlinear distortions in MRI images caused by the presence of metal implants. J Med Eng Technol 24:20–27.[ISI][Medline]
Ganapathi M, Joseph G, Savage R, Jones AR, Timms B, Lyons K (2002). MRI susceptibility artefacts related to scaphoid screws: the effect of screw type, screw orientation and imaging parameters. J Hand Surg 27(B):165–170.[Medline]
Hashemi RH, Bradley WG Jr (1997). MRI: the basics. 1st ed. Baltimore: Williams & Wilkins.
Malik AS, Boyko O, Aktar N, Young WF (2001). A comparative study of MR imaging profile of titanium pedicle screws. Acta Radiol 42:291–293.[ISI][Medline]
Rudisch A, Kremser C, Peer S, Kathrein A, Judmaier W, Daniaux H (1998). Metallic artifacts in magnetic resonance imaging of patient with spinal fusion. A comparison of implant materials and imaging sequences. Spine 23:692–699.[ISI][Medline]
Shellock FG, Kanal E (1998). Aneurysm clips: evaluation of MR imaging artifacts at 1.5T. Radiology 209:563–566.
Suh JS, Jeong EK, Shin KH, Cho JH, Na JB, Kim DH, et al. (1998). Minimizing artifacts caused by metallic implants at MR imaging: experimental and clinical studies. AJR Am J Roentgenol 171:1207–1213.
Teitelbaum GP, Bradley WG Jr, Klein BD (1988). MR imaging artifacts, ferromagnetism, and magnetic torque of intravascular filters, stents, and coils. Radiology 166:657–664.
Vaccaro AR, Chesnut RM, Scuderi G, Healy JF, Massie JGB, Garfin SR (1994). Metallic spinal artifacts in magnetic resonance imaging. Spine 19:1237–1242.[ISI][Medline]
Viano AM, Gronemeyer SA, Haliloglu M, Hoffer FA (2000). Improved MR imaging for patients with metallic implants. Magn Reson Imaging 18:287–295.[ISI][Medline]
Vlaardingerbroek MT, den Boer JA (1996). Magnetic resonance imaging. Berlin, Heidelberg: Springer-Verlag.
Weast RC, Astle MJ, Beyer WH, editors (1984). Handbook of chemistry and physics. 65th ed. Boca Raton, FL: CRC Press.
White MR (1970). Quantum theory of magnetism. New York: McGraw-Hill Book Company.
This article has been cited by other articles:
|
|
 |
|
  T. M. J. Harris, M. R. Faridrad, and J. A. S. Dickson The benefits of aesthetic orthodontic brackets in patients requiring multiple MRI scanning. J. Orthod., June 1, 2006; 33(2): 90 - 94. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G Eggers, M Rieker, J Fiebach, B Kress, H Dickhaus, and S Hassfeld Geometric accuracy of magnetic resonance imaging of the mandibular nerve Dentomaxillofac. Radiol., September 1, 2005; 34(5): 285 - 291. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| IADR Journals | Advances in Dental Research ® |
| Journal of Dental Research ® | Critical Reviews (1990-2004) |
