- Email: [email protected]
Assessment of signal-to-noise ratio and contrast-to-noise ratio in 3 T magnetic resonance imaging in the presence of zirconium, titanium, and titanium-zirconium alloy implants
ARTICLE IN PRESS Vol. 00 No. 00 && 2019
Assessment of signal-to-noise ratio and contrast-to-noise ratio in 3 T magnetic resonance imaging in the presence of zirconium, titanium, and titanium-zirconium alloy implants Husniye Demirturk Kocasarac, DDS, MS, PhD,a Emine Sebnem Kursun-Cakmak, DDS, PhD,b Gulbahar Ustaoglu, DDS, MS,c Seval Bayrak, DDS, PhD,d Kaan Orhan, DDS, PhD,e,f and Marcel Noujeim, DDS, MSg Objective. We quantitatively compared the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) in T1 weighted imaging (WI) and T2 WI sequences in 3 Tesla (T) magnetic resonance imaging (MRI) using zirconium, titanium (grades 4 and 5), and titanium-zirconium alloy implants to evaluate the effect of implant type and imaging sequence. Study Design. MRI was acquired using a 3 T magnet with a 16-channel head coil. Implants of each type were mounted in gel and scanned in axial, coronal, and sagittal planes using fast spin echo sequences in T1 WI (TR = 600, TE = 12 milliseconds) and T2 WI (TR = 3000, TE = 80 milliseconds) sequences. Data were transferred to Synapse 3-D software, and images were measured twice by an oral and maxillofacial radiologist blinded to the type of implants. Results. Zirconium implants resulted in the lowest SNR and CNR values (P < .05). No significant differences were identified between titanium (grades 4 and 5) and titanium-zirconium implants. The T2 WI sequence had a significantly higher SNR and CNR than T1 WI. There was no difference in intraobserver agreement between T1 WI and T2 WI. Conclusions. CNR and SNR at 3 T MRI are dependent on implant type and imaging sequence. Titanium (grades 4 and 5) and titanium-zirconium implants and the T2 WI sequence produced higher SNR and CNR values. (Oral Surg Oral Med Oral Pathol Oral Radiol 2019;000:17)
Magnetic resonance imaging (MRI) has extensive applications in the head and neck region based on its excellent physical characteristics and superlative visualization of bone marrow and soft tissues without the use of ionizing radiation.1 The quality of a magnetic resonance (MR) image is affected by the efficacy of the pulse sequence or imaging protocol used, voxel size, inherent contrast difference between tissues in the MRI sequence (i.e., T1, T2), strength of the MR signal and noise, and scan time.2 The signal-to-noise ratio (SNR) defines the signal in a single tissue in relation to pixel-to-pixel variations in this tissue or pixel-to-pixel variations (noise) in a background tissue and defines the performance of MRI a
Department of General Dental Sciences, Marquette University, Milwaukee, WI, USA. b Ministry of Health, Turkey Public Hospitals Agency, Ankara, Turkey. c Department of Periodontology, Bolu Abant Izzet Baysal University, Bolu, Turkey. d Department of Oral and Maxillofacial Radiology, Bolu Abant Izzet Baysal University, Bolu, Turkey. e Professor, Department of Oral and Maxillofacial Radiology, Ankara University, Ankara, Turkey. f OMFS-IMPATH Research Group, Department of Imaging & Pathology, Faculty of Medicine, University of Leuven and Oral & Maxillofacial Surgery, University Hospitals Leuven, Leuven, Belgium. g Professor, Oral and Maxillofacial Radiology, Private practice, San Antonio, Texas, USA. Received for publication Jan 20, 2019; returned for revision Aug 9, 2019; accepted for publication Aug 31, 2019. Ó 2019 Elsevier Inc. All rights reserved. 2212-4403/$-see front matter https://doi.org/10.1016/j.oooo.2019.08.020
systems. SNR is mostly used for assessment of the image, quality assurance, comparison of radiofrequency coil and pulse sequence, and measurement of contrast enhancement.3,4 Changing magnetic field strength, scan parameters, field of view, and slice thickness can change SNR in MR images because of their effect on the background noise.5 SNR quantitatively shows how much the signal increases above pixel-topixel variations and therefore is a reliable measure for image quality. Generally, the higher the SNR, the more appealing and smoother the image is to the reader.3 The most commonly used technique in calculating SNR in MRI is the 2-region method, which is based on the signal statistics in 2 different regions of interest (ROIs) from a single image: one in the tissue of interest to measure the signal intensity and one in the image background to detect the noise intensity.4 The contrast in the image is the difference in the MR signal strength from one tissue to another, resulting from differences in T1, T2, the density of hydrogen, or a combination of these parameters, depending on the pulse sequence.2 The contrast-to-noise ratio (CNR) is
Statement of Clinical Relevance A T2 weighted imaging sequence is more suitable in the presence of zirconium, titanium, and titaniumzirconium alloy implants, and titanium and titanium-zirconium alloy implants give better quality images compared with zirconium. 1
ARTICLE IN PRESS ORAL AND MAXILLOFACIAL RADIOLOGY 2 Demirturk Kocasarac et al.
OOOO && 2019
described as the ratio of signal difference (contrast) to the noise level in an image and is used as a measure of image quality that depends on contrast rather than the raw signal. Whereas SNR is used to compare the quality of the imaging sequence or scanner hardware, CNR may be indicative of the quality (i.e., detectability) of the contrast of interest.3,5 CNR is a valuable indicator of the ability to detect low-contrast lesions. It shows the difference in signal between background and lesion increasing above variation between adjacent pixels, either in the background or in both tissues. Image quality increases with larger SNR and CNR values.3 The great advantage of MRI after dental implant insertion is the capability to track bone formation because the initial healing phases are composed of unmineralized structures and water; these changes are therefore undetectable by X-ray examination.6 Unfortunately, it is well known that metallic objects cause artefacts in MRI, reducing image quality. Multiple types of dental implants can be found in the market, titanium (Ti) being the most widely used by itself or as an alloy. Two kinds of titanium implants are used by clinicians: commercially pure (CP) and titanium alloys. CP titanium can be found in 4 different grades depending on the degree of purity or the content of oxygen, iron, and carbon. Grade 1 has the least impurities and grade 4 has the greatest amount. Independently of the grade, CP titanium is still the most commonly used material in implant dentistry.7,8 Grade 5, also available in the market, is called as Ti-6 AL-4 V because it has 6% aluminum and 4% vanadium in its composition.9 This alloy presents with improved mechanical characteristics. Nontitanium implants, including zirconium, have been recently introduced into practice. A new titaniumzirconium alloy containing 13%-17% zirconium, revealed better mechanical features compared with titanium in terms of elongation and fatigue strength.7 It has been reported that zirconium implants induce less severe artifacts in MRI than titanium.10 Thus, MRI can be a radiation-free technique to evaluate the position of zirconium implants after implant placement.11 Although titanium alloys are not considered as contraindications during MRI acquisition, they may cause susceptibility artifacts because of their paramagnetic properties.12
The objective of this research was to compare the SNR and CNR in magnetic resonance images containing zirconium, titanium grade 4, titanium grade 5, and titanium-zirconium implants with T1 and T2 weighted images using a 3 Tesla (T) magnet. The null hypothesis stated that there were no significant differences in SNR or CNR based on implant type and pulse sequence.
MATERIALS AND METHODS This study included 4 dental implants: zirconium, titanium grade 4, titanium grade 5, and titanium-zirconium alloy. Specifications of the implants are presented in Table I. Implants were individually embedded in a 3% gelatin (ultrasound gel) in a cylindrical plastic container (15 centimeters [cm] diameter £ 10 cm length). The gelatin phantoms were used to mimic soft tissues and standardize the immediate environment of the implants for all 3 techniques. To prevent the airgelatin boundaries in proximity to the implants, sodium chloride (0.9%) was added within residual spaces in containers. Standard clinical protocols that are applied for head and neck imaging in MRI in our institutes were followed for maximum representation of clinical settings. MRI was carried out using a 3.0 Tesla Discovery (GE Medical systems LLC, Waukesha, WI) with a 16-channel head coil. The dental implants were mounted in a gel box to ensure stabilization. Moreover, an MR phantom containing distilled water that is used for uniformity measurements in MR studies was placed under the gel box to create the standardization needed to measure signal intensity ratio (SIR). Mathematically, SIR is the quotient of the signal intensity from a region of interest when divided by the standard deviation of the signal intensity from a region situated outside the examined area (i.e., a tissue region not producing a signal).13 All implants were scanned in all 3 directions (sagittal, coronal, and axial) using T1 weighted imaging (WI) and T2 WI fast spin echo sequences. Imaging settings were as follows. T1 WI images: TR = 600, TE = 12 milliseconds (ms), number of excitations = 2, 256 £ 256 matrix, echo train length = 10, 1 millimeter (mm) slice thickness, and field of view (FOV) = 120 mm; T2 WI images: TR = 3000, TE = 80 ms, number
Table I. Implant trade names, composition, manufacturer, length, and diameter Material
Manufacturer
PURE Ceramic Implant Octagon
Zirconium Titanium grade 4
Biotech Dental Kontact Roxolid
Titanium grade 5 Titanium-zirconium (85% titanium, 15% zirconium)
Straumann (Basel, Switzerland) ACE Surgical Supply (Brockton, USA) Biotech Dental (Salon de Provence, France) Straumann (Basel, Switzerland)
Length including platform (mm)
Diameter of body (mm)
16 8
4.1 4.1
6 8
4.1 4.1
ARTICLE IN PRESS OOOO Volume 00, Number 00
of excitations = 2, 256 £ 256 matrix, echo train length = 10, 1 mm slice thickness, and FOV = 120 mm. The MRI data were acquired in DICOM 3.0 format and transferred to Synapse 3-D software (Fujifilm Holdings Corporation, Tokyo, Japan). The images were measured 2 times by an oral and maxillofacial radiologist who was blinded to the type of dental implant. Using Synapse 3-D software, the dental implant ROI was selected automatically by applying a region grow threshold, the boundaries were automatically marked on MRI images, and the area within the boundaries was shaded in color. SIR measurements were done by means of the software in axial slices (Figure 1). In addition, SIR from the
ORIGINAL ARTICLE Demirturk Kocasarac et al. 3
gel and SIR from the center of the cubic phantom (Figure 2) were measured to normalize differences in signal intensity in axial slices. SIR from the gel was measured from a fixed ROI (30 mm2) that was placed in a uniform gel layer close to the center of the image, whereas SIR from the phantom was measured in an approximately 100 mm2 ROI from the center of the phantom. SIR measurements were made both from T1 WI and T2 WI images. The final SNR and CNR of the dental implants were calculated as follows: SNR ¼ SIRðimplantÞ SIRðphantomÞ = SIRðgelÞ SIRðphantomÞ CNR ¼ 20 log10 SIR½implant = SIR½gel
Fig. 1. (a) T1 weighted implant image; (b) color shading for automatic identification of the implant boundaries; (c) automatic selection of region of interest by applying region grow threshold; (d) signal intensity ratio (SIR) measurements; (e) SIR measurement from the gel that surrounds the implant from the T1 weighted image.
ARTICLE IN PRESS ORAL AND MAXILLOFACIAL RADIOLOGY 4 Demirturk Kocasarac et al.
OOOO && 2019
Fig. 2. Signal intensity ratio measurement from the center of the cubic phantom from a T1 weighted image.
Statistical analysis A power analysis was done indicating the differences between study groups obtained with 4 different implants, which were scanned 5 times, in different MR sequences, to provide statistical power of 0.8 and an a value of .05. One head and neck radiologist with 16 years of experience and special training in MRI evaluated the images. The observer was calibrated by doing measurements on scans that were not part of the research but used only for calibration purposes. During the evaluation of the images in the research, the observer was blinded to the type of dental implant. All measurements were done 2 times by the same observer and their mean values were statistically analyzed. The same observer repeated the measurements after 2 weeks to allow calculation of intraobserver agreement. All statistical analyses were performed using SAS Version 9.4 (SAS Institute, Cary, NC), with the PROC ANOVA (analysis of variance) function. Analysis of the measurements was conducted using ANOVA and post hoc Tukey tests to assess (1) whether one or more materials resulted in a better SNR and/or CNR than the others and (2) whether one sequencing method resulted in a better SNR and/or CNR than the other. Intraobserver agreement was assessed by calculating the percentage difference in SNR (absolute value of the difference, divided by the average value) and
conducting a 2-way ANOVA (for material and sequence) to control for variation between materials. An identical procedure was done for CNR. This analysis was used to calculate intraobserver agreement. Significance for all tests was established at P = .05.
RESULTS The average values of SNR and CNR for each implant type according to the 2 pulse sequences are presented in Table II. The calculated SNR on T1 weighted images was between 0.579 and 1.024. The lowest SNR was found for zirconium, and the highest was for the titanium grade 5 implant. The CNR of T1 weighted images had similar results, with the lowest value for zirconium (8.275) and highest for titanium grade 5 (20.100). On T2 weighted images, the calculated SNR values were between 0.726 and 1.830. The lowest SNR was recorded for zirconium and the highest was in titanium grade 4. The CNR values of T2 weighted images were also lowest for zirconium (7.750), but titanium-zirconium had the highest value (27.580), as shown in Table II. The ANOVA results indicated that zirconium implants resulted in significantly lower SNR and CNR values compared with the other 3 implant types (P < .05). No statistically significant differences were found between titanium (grades 4 and 5) and titanium-zirconium alloy implants (P > .05). There was enough statistical evidence
ARTICLE IN PRESS OOOO Volume 00, Number 00
ORIGINAL ARTICLE Demirturk Kocasarac et al. 5
Table II. Mean values of SNR and CNR measurements for the 4 implant types according to pulse sequences Implant types
T1-WI
T2-WI
P
T1-WI
T2-WI
P
Zirconium Titanium grade 4 Titanium grade 5 Titanium-zirconium
SNR 0.579 a,1 0.912 b,1 1.024 b,1 0.832 b,1
SNR 0.726 a,1 1.830 b,2 1.497 b,2 1.536 b,2
(implant types) P < .05 P > .05 P > .05 P > .05
CNR 8.275 a,1 16.130 b,1 20.100 b,1 12.000 b,1
CNR 7.750 a,1 26.960 b,2 22.615 b,2 27.580 b,2
(pulse sequences) P > .05 P < .05 P < .05 P < .05
SNR, signal-to-noise ratio; CNR, contrast-to-noise ratio; WI, weighted image. *P < .05 indicates statistical significance between implant types. Different letters indicate significant differences for SNR and CNR among implant types. Different numbers indicate significant differences between pulse sequences.
to conclude that the T2 WI sequence had significantly higher SNR and CNR values than the T1 WI sequence for all implant types (P < .05) except zirconium. The average percentage difference in SNR across the 4 materials was 2.4% for T1 WI, and 4.1% for T2 WI (after excluding the large discrepancy of zirconium values). The average percentage difference in CNR across the 4 materials was 5.4% for T1 WI and 7.6% for T2 WI. Regarding intraobserver agreement, in spite of a large disparity in zirconium SNR measurements, ANOVA revealed no significant differences (P = .44 for SNR and .61 for CNR), indicating that T1 WI was no different from T2 WI regarding agreement between the observer’s 2 readings.
DISCUSSION MRI has been used for many purposes in medicine and dentistry and is gaining interest for the assessment of patients before and after implant treatment. However, the presence of some implant materials provokes artifacts because of their physical properties. Signal void is often present at the site of implants, making assessment of the tissues and detection of possible pathosis near the implants difficult. This is particularly important in cancer patients, who need to undergo frequent restaging processes in oncologic imaging. For patients with cancer or a risk of cancer, images with high quality, despite the presence of implants, are indispensable for a correct diagnosis of primary, or detection of recurrent, lesions. In this study, zirconium implants produced significantly lower SNR and CNR values compared with the other three implant types, which did not differ significantly between themselves. SNR and CNR values were significantly higher in the T2 WI sequence than T1 WI for titanium (grades 4 and 5) and titanium-zirconium alloy implants, whereas the values were almost the same for the zirconium implant. There were no significant intraobserver differences for the measurements both in T1 WI and T2 WI sequences (P > .05). All measurements were greatly reproducible, and no significant difference was found between 2 measurements done by the observer (P > .05).
To our knowledge, this is the first study assessing SNR and CNR of images of dental implants with different compositions in a standardized model at 3 T. Hypothetically, increasing magnetic field strength improves SNR in a linear trend. However, different scan parameters as well as many other factors, in addition to the external magnetic field strength, might have a possible effect on SNR.14,15 Moreover, different pulse sequence parameters, such as spatial resolution, FOV, bandwidth, echo time, and repetition time, can affect SNR and CNR as well. Sequence timing details can also be important.16 T1 relaxation time differences are known to be associated with the amount of iron and hydrogen in a structure, whereas T2 relaxation time is known to decrease substantially at higher frequencies, principally as a result of chemical exchange between bulk and bound hydrogen molecules in tissue.17,18 The prolonged translation relaxation time in this study might be the reason for the increase in calculated SNR and CNR as a result of MRI signal saturation. Additionally, SNR and CNR benefit from using a surface coil for signal detection; unfortunately, this was not possible in our study because of the phantom size used. Other studies19,20 have been done regarding artifact generation of dental materials on MRI. However, only a few investigations have evaluated the artifact production resulting from dental implant materials.10,11,21 Moreover, there has been no study concerning the SNR and CNR of dental implants on MRI images and our research is, to our knowledge, the first study to deal with this topic. Repetition and echo times, spatial resolution, FOV, and bandwidth are among many sequence parameters that can influence changes in SNR and CNR. Using identical FOV and spatial resolution may exclude their influence. As a result of variations in tissue relaxation parameters such as longitudinal relaxation time resulting from different external magnetic field strengths (as anticipated by relaxation theory), the sequence timing might have more influence.22 Previous studies, conducted in clinical settings, have reported advantages for 3 T magnets, specifically in brain and musculoskeletal conditions.22,23 Theoretically,
ARTICLE IN PRESS ORAL AND MAXILLOFACIAL RADIOLOGY 6 Demirturk Kocasarac et al.
SNR increases linearly as the magnetic field increases and therefore should double with 3 T relative to 1.5 T. This gives the operator the options of either reducing the scan time or increasing the spatial resolution.22,24 Lu et al.25 examined multiple anatomic structures of the brain and the CNR and SNR of white and gray matter using T1 and T2 relaxation times at 1.5 and 3 T. They found a statistically significant improvement in SNR of white and gray matter of 36% and 41%, respectively, on T1 WI and T2 WI at 1.5 and 3 T. However, only 21% improvement of CNR was noted. DiFrancesco et al.24 studied CNR and SNR in mouse brain imaging at 3 and 7 T in vivo, and their results verified that increases in magnetic field strength yielded predicted increases in the SNR and therefore resolution in MR images of small animals. If the SNR is large enough at lower field strengths, the CNR will improve to the same degree. However, more challenging circumstances resulting in poor SNR or resolution at lower field strength will find the CNR improving even more markedly as field strength is increased. Magnotta et al.26 measured image quality differences using CNR and SNR across sites, vendors, and field strengths in structural fast spin echo T2 WI scans of the brain. Results of the study verified that both CNR and SNR were higher in 3 and 4 T magnetic fields compared with 1.5 T. Although CNR and SNR between different vendors and coils were almost uniform in the 1.5 T field strength scanners, higher field scanners (3 T and 4 T) significantly improved the SNR nonuniformly. Gradl et al.27 compared the effect of a standard head and neck coil with a surface dental coil for SNR of multiple anatomic structures such as osseous and muscular tissues and dental pulp and hard tissues. They detected a 3.5 times increase, on average, in SNR with the dental coil compared with head and neck coil (between 2.7 times increase for masseter muscle and 4.6 for pulp tissue). Using a dental coil increased the SNR for all evaluated structures, which led to a better image quality detected in a subjective assessment. The limitation of this study was the evaluation of SNR in a small ROI (pulp chamber) influenced by volume averaging in standard sequences. Despite the substantial results obtained using a standard approach, our study presents with some limitations. As an in vitro study, it was deprived of the anatomic landmarks that might play a role in image formation and quality. Second, although the imaging settings for the fast spin echo sequences used are similar to the clinical imaging settings, they may vary in different investigations. Third, there is a lack of similar results in the literature with which to compare our results. To our knowledge, this is the first research into the effect of implants and pulse sequences on MRI image quality in the dental field. Lastly, it is vital to emphasize that in vitro studies cannot
OOOO && 2019
mimic actual clinical conditions. However, in vitro research permits the investigators to compare different materials and protocols in ways that would not always be possible for research projects involving patients, especially in expensive and time-consuming modalities such as MRI. There may be a health care need for MR imaging of regions with implants within the field of view. The assessment may be for the detection of pathoses, staging of malignancy, or malignant posttreatment restaging. However, the presence of implants may significantly compromise the image quality in terms of SNR and CNR. Therefore, the main goal of our study was to explore the influence of implant type and imaging sequence on image quality to find the best implant materialimaging sequence combination. The results of our study may facilitate the radiologist’s or clinician’s decision to give the best care to his or her patient. In conclusion, this comparative study found that changes in SNR and CNR at 3 T MRI depend on the implant type and imaging sequence. The results revealed that although titanium (grades 4 and 5) and titaniumzirconium implants imaged with the T2 WI sequence theoretically produce higher SNR and CNR, in clinical conditions the increase might be limited but still significant. However, extension of translation relaxation time might be the reason behind the CNR and SNR improvement in this study. Moreover, it can be recommended that calculation of small volumes or materials that have similar SNR and CNR values can be done by estimating the value of activation strength. This study may provide such values for generalization in further studies. Our findings indicate that the T2 WI sequence may be preferred to T1 WI while evaluating implant integration and inflammatory/infectious processes around or in proximity of implants or for the purpose of cancer restaging in patients who have implants.
REFERENCES 1. Assaf AT, Zrnc TA, Remus CC, Schonfeld M, Habermann CR, Riecke B, Friedrich RE, Fiehler J, Heiland M, Sedlacik J. Evaluation of four different optimized magnetic-resonance-imaging sequences for visualization of dental and maxillo-mandibular structures at 3 T. J Craniomaxillofac Surg. 2014;42:1356-1363. 2. Hart H.R. Jr, Bottomley PA, Edelstein WA, et al. Nuclear magnetic resonance imaging: contrast-to-noise ratio as a function of strength of magnetic field. AJR Am J Roentgenol. 1983;141:1195-1201. 3. Signal, noise, signal-to-noise, and contrast-to-noise ratios. In: Hendrick RE, ed. Breast MRI: Fundamentals and Technical Aspects, New York, NY: Springer; 2008:93-111. 4. Dietrich O, Raya JG, Reeder SB, Reiser MF, Schoenberg SO. Measurement of signal-to-noise ratios in MR images: influence of multichannel coils, parallel imaging, and reconstruction filters. J Magn Reson Imag. 2007;26:375-385. 5. Welvaert M, Rosseel Y. On the definition of signal-to-noise ratio and contrast-to-noise ratio for FMRI data. PloS One. 2013;8: e77089.
ARTICLE IN PRESS OOOO Volume 00, Number 00 6. Korn P, Elschner C, Schulz MC, Range U, Mai R, Scheler U. MRI and dental implantology: two which do not exclude each other. Biomaterials. 2015;53:634-645. 7. Bruschi M, Steinm€uller-Nethl D, Goriwoda W, Rasse M. Composition and modifications of dental implant surfaces. J Oral Implants. 2015;2015. 8. Mouhyi J, Dohan Ehrenfest DM, Albrektsson T. The periimplantitis: implant surfaces, microstructure, and physicochemical aspects. Clin Implant Dent Relat Res. 2012;14:170-183. 9. Elias CN, Oshida Y, Lima JH, Muller CA. Relationship between surface properties (roughness, wettability and morphology) of titanium and dental implant removal torque. J Mech Behav Biomed Mater. 2008;1:234-242. 10. Kocasarac HD, Ustaoglu G, Bayrak S, et al. Evaluation of artifacts generated by titanium, zirconium and titaniumzirconium alloy dental implants in MRI, CT and CBCT images: a phantom study. Oral Surg Oral Med Oral Pathol Oral Radiol. 2019;127:535-544. 11. Smeets R, Schollchen M, Gauer T. Artefacts in multimodal imaging of titanium, zirconium and binary titanium-zirconium alloy dental implants: an in vitro study. 2017;46:20160267. 12. Schenck JF. The role of magnetic susceptibility in magnetic resonance imaging: MRI magnetic compatibility of the first and second kinds. Med Phys. 1996;23:815-850. 13. Weishaupt D, K€ochli VD, Marincek B. How Does MRI Work? An Introduction to the Physics and Function of Magnetic Resonance Imaging. New York, NY: Springer Science & Business Media; 2008. 14. Biswas J, Nelson CB, Runge VM, et al. Brain tumor enhancement in magnetic resonance imaging: comparison of signal-tonoise ratio (SNR) and contrast-to-noise ratio (CNR) at 1.5 versus 3 Tesla. Invest Radiol. 2005;40:792-797. 15. Orhan K, Nishiyama H, Tadashi S, Murakami S, Furukawa S. Comparison of altered signal intensity, position, and morphology of the TMJ disc in MR images corrected for variations in surface coil sensitivity. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2006;101:515-522. 16. Bottomley PA, Foster TH, Argersinger RE, Pfeifer LM. A review of normal tissue hydrogen NMR relaxation times and relaxation mechanisms from 1-100 MHz: dependence on tissue type, NMR frequency, temperature, species, excision, and age. Med Phys. 1984;11:425-448. 17. Gelman N, Gorell JM, Barker PB, et al. MR imaging of human brain at 3.0 T: preliminary report on transverse relaxation rates and relation to estimated iron content. Radiology. 1999;210:759-767.
ORIGINAL ARTICLE Demirturk Kocasarac et al. 7 18. Zhong JH, Gore JC, Armitage IM. Relative contributions of chemical exchange and other relaxation mechanisms in protein solutions and tissues. Magn Reson Med. 1989;11:295-308. 19. Eggers G, Rieker M, Kress B, Fiebach J, Dickhaus H, Hassfeld S. Artefacts in magnetic resonance imaging caused by dental material. MAGMA. 2005;18:103-111. 20. Idiyatullin D, Garwood M, Gaalaas L, Nixdorf DR. Role of MRI for detecting micro cracks in teeth. Dentomaxillofac Radiol. 2016;45:20160150. 21. Duttenhoefer F, Mertens ME, Vizkelety J, Gremse F, Stadelmann VA, Sauerbier S. Magnetic resonance imaging in zirconia-based dental implantology. Clin Oral Implants Res. 2015;26:1195-1202. 22. Biswas J, Nelson CB, Runge VM, et al. Brain tumor enhancement in magnetic resonance imaging: comparison of signal-tonoise ratio (SNR) and contrast-to-noise ratio (CNR) at 1.5 versus 3 tesla. Invest Radiol. 2005;40:792-797. 23. Frayne R, Goodyear BG, Dickhoff P, Lauzon ML, Sevick RJ. Magnetic resonance imaging at 3.0 Tesla: challenges and advantages in clinical neurological imaging. Invest Radiol. 2003;38:385-402. 24. DiFrancesco MW, Rasmussen JM, Yuan W, et al. Comparison of SNR and CNR for in vivo mouse brain imaging at 3 and 7 T using well matched scanner configurations. Med Phys. 2008;35:3972-3978. 25. Lu H, Nagae-Poetscher LM, Golay X, Lin D, Pomper M, van Zijl PC. Routine clinical brain MRI sequences for use at 3.0 Tesla. J Magn Reson Imag. 2005;22:13-22. 26. Magnotta VA, Friedman L. Measurement of signal-to-noise and contrast-to-noise in the fBIRN Multicenter Imaging Study. J Digital Imag. 2006;19:140-147. 27. Gradl J, Horeth M, Pfefferle T, et al. Application of a dedicated surface coil in dental MRI provides superior image quality in comparison with a standard coil. Clin Neuroradiol. 2017;27:371-378. Reprint requests: Husniye Demirturk Kocasarac, DDS, MS, PhD Department of General Dental Sciences Division of Oral and Maxillofacial Radiology Marquette University School of Dentistry PO Box 1881 Milwaukee WI 53201-1881 [email protected]
Assessment of signal-to-noise ratio and contrast-to-noise ratio in 3 T magnetic resonance imaging in the presence of zirconium, titanium, and titanium-zirconium alloy implants Husniye Demirturk Kocasarac, DDS, MS, PhD,a Emine Sebnem Kursun-Cakmak, DDS, PhD,b Gulbahar Ustaoglu, DDS, MS,c Seval Bayrak, DDS, PhD,d Kaan Orhan, DDS, PhD,e,f and Marcel Noujeim, DDS, MSg Objective. We quantitatively compared the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) in T1 weighted imaging (WI) and T2 WI sequences in 3 Tesla (T) magnetic resonance imaging (MRI) using zirconium, titanium (grades 4 and 5), and titanium-zirconium alloy implants to evaluate the effect of implant type and imaging sequence. Study Design. MRI was acquired using a 3 T magnet with a 16-channel head coil. Implants of each type were mounted in gel and scanned in axial, coronal, and sagittal planes using fast spin echo sequences in T1 WI (TR = 600, TE = 12 milliseconds) and T2 WI (TR = 3000, TE = 80 milliseconds) sequences. Data were transferred to Synapse 3-D software, and images were measured twice by an oral and maxillofacial radiologist blinded to the type of implants. Results. Zirconium implants resulted in the lowest SNR and CNR values (P < .05). No significant differences were identified between titanium (grades 4 and 5) and titanium-zirconium implants. The T2 WI sequence had a significantly higher SNR and CNR than T1 WI. There was no difference in intraobserver agreement between T1 WI and T2 WI. Conclusions. CNR and SNR at 3 T MRI are dependent on implant type and imaging sequence. Titanium (grades 4 and 5) and titanium-zirconium implants and the T2 WI sequence produced higher SNR and CNR values. (Oral Surg Oral Med Oral Pathol Oral Radiol 2019;000:17)
Magnetic resonance imaging (MRI) has extensive applications in the head and neck region based on its excellent physical characteristics and superlative visualization of bone marrow and soft tissues without the use of ionizing radiation.1 The quality of a magnetic resonance (MR) image is affected by the efficacy of the pulse sequence or imaging protocol used, voxel size, inherent contrast difference between tissues in the MRI sequence (i.e., T1, T2), strength of the MR signal and noise, and scan time.2 The signal-to-noise ratio (SNR) defines the signal in a single tissue in relation to pixel-to-pixel variations in this tissue or pixel-to-pixel variations (noise) in a background tissue and defines the performance of MRI a
Department of General Dental Sciences, Marquette University, Milwaukee, WI, USA. b Ministry of Health, Turkey Public Hospitals Agency, Ankara, Turkey. c Department of Periodontology, Bolu Abant Izzet Baysal University, Bolu, Turkey. d Department of Oral and Maxillofacial Radiology, Bolu Abant Izzet Baysal University, Bolu, Turkey. e Professor, Department of Oral and Maxillofacial Radiology, Ankara University, Ankara, Turkey. f OMFS-IMPATH Research Group, Department of Imaging & Pathology, Faculty of Medicine, University of Leuven and Oral & Maxillofacial Surgery, University Hospitals Leuven, Leuven, Belgium. g Professor, Oral and Maxillofacial Radiology, Private practice, San Antonio, Texas, USA. Received for publication Jan 20, 2019; returned for revision Aug 9, 2019; accepted for publication Aug 31, 2019. Ó 2019 Elsevier Inc. All rights reserved. 2212-4403/$-see front matter https://doi.org/10.1016/j.oooo.2019.08.020
systems. SNR is mostly used for assessment of the image, quality assurance, comparison of radiofrequency coil and pulse sequence, and measurement of contrast enhancement.3,4 Changing magnetic field strength, scan parameters, field of view, and slice thickness can change SNR in MR images because of their effect on the background noise.5 SNR quantitatively shows how much the signal increases above pixel-topixel variations and therefore is a reliable measure for image quality. Generally, the higher the SNR, the more appealing and smoother the image is to the reader.3 The most commonly used technique in calculating SNR in MRI is the 2-region method, which is based on the signal statistics in 2 different regions of interest (ROIs) from a single image: one in the tissue of interest to measure the signal intensity and one in the image background to detect the noise intensity.4 The contrast in the image is the difference in the MR signal strength from one tissue to another, resulting from differences in T1, T2, the density of hydrogen, or a combination of these parameters, depending on the pulse sequence.2 The contrast-to-noise ratio (CNR) is
Statement of Clinical Relevance A T2 weighted imaging sequence is more suitable in the presence of zirconium, titanium, and titaniumzirconium alloy implants, and titanium and titanium-zirconium alloy implants give better quality images compared with zirconium. 1
ARTICLE IN PRESS ORAL AND MAXILLOFACIAL RADIOLOGY 2 Demirturk Kocasarac et al.
OOOO && 2019
described as the ratio of signal difference (contrast) to the noise level in an image and is used as a measure of image quality that depends on contrast rather than the raw signal. Whereas SNR is used to compare the quality of the imaging sequence or scanner hardware, CNR may be indicative of the quality (i.e., detectability) of the contrast of interest.3,5 CNR is a valuable indicator of the ability to detect low-contrast lesions. It shows the difference in signal between background and lesion increasing above variation between adjacent pixels, either in the background or in both tissues. Image quality increases with larger SNR and CNR values.3 The great advantage of MRI after dental implant insertion is the capability to track bone formation because the initial healing phases are composed of unmineralized structures and water; these changes are therefore undetectable by X-ray examination.6 Unfortunately, it is well known that metallic objects cause artefacts in MRI, reducing image quality. Multiple types of dental implants can be found in the market, titanium (Ti) being the most widely used by itself or as an alloy. Two kinds of titanium implants are used by clinicians: commercially pure (CP) and titanium alloys. CP titanium can be found in 4 different grades depending on the degree of purity or the content of oxygen, iron, and carbon. Grade 1 has the least impurities and grade 4 has the greatest amount. Independently of the grade, CP titanium is still the most commonly used material in implant dentistry.7,8 Grade 5, also available in the market, is called as Ti-6 AL-4 V because it has 6% aluminum and 4% vanadium in its composition.9 This alloy presents with improved mechanical characteristics. Nontitanium implants, including zirconium, have been recently introduced into practice. A new titaniumzirconium alloy containing 13%-17% zirconium, revealed better mechanical features compared with titanium in terms of elongation and fatigue strength.7 It has been reported that zirconium implants induce less severe artifacts in MRI than titanium.10 Thus, MRI can be a radiation-free technique to evaluate the position of zirconium implants after implant placement.11 Although titanium alloys are not considered as contraindications during MRI acquisition, they may cause susceptibility artifacts because of their paramagnetic properties.12
The objective of this research was to compare the SNR and CNR in magnetic resonance images containing zirconium, titanium grade 4, titanium grade 5, and titanium-zirconium implants with T1 and T2 weighted images using a 3 Tesla (T) magnet. The null hypothesis stated that there were no significant differences in SNR or CNR based on implant type and pulse sequence.
MATERIALS AND METHODS This study included 4 dental implants: zirconium, titanium grade 4, titanium grade 5, and titanium-zirconium alloy. Specifications of the implants are presented in Table I. Implants were individually embedded in a 3% gelatin (ultrasound gel) in a cylindrical plastic container (15 centimeters [cm] diameter £ 10 cm length). The gelatin phantoms were used to mimic soft tissues and standardize the immediate environment of the implants for all 3 techniques. To prevent the airgelatin boundaries in proximity to the implants, sodium chloride (0.9%) was added within residual spaces in containers. Standard clinical protocols that are applied for head and neck imaging in MRI in our institutes were followed for maximum representation of clinical settings. MRI was carried out using a 3.0 Tesla Discovery (GE Medical systems LLC, Waukesha, WI) with a 16-channel head coil. The dental implants were mounted in a gel box to ensure stabilization. Moreover, an MR phantom containing distilled water that is used for uniformity measurements in MR studies was placed under the gel box to create the standardization needed to measure signal intensity ratio (SIR). Mathematically, SIR is the quotient of the signal intensity from a region of interest when divided by the standard deviation of the signal intensity from a region situated outside the examined area (i.e., a tissue region not producing a signal).13 All implants were scanned in all 3 directions (sagittal, coronal, and axial) using T1 weighted imaging (WI) and T2 WI fast spin echo sequences. Imaging settings were as follows. T1 WI images: TR = 600, TE = 12 milliseconds (ms), number of excitations = 2, 256 £ 256 matrix, echo train length = 10, 1 millimeter (mm) slice thickness, and field of view (FOV) = 120 mm; T2 WI images: TR = 3000, TE = 80 ms, number
Table I. Implant trade names, composition, manufacturer, length, and diameter Material
Manufacturer
PURE Ceramic Implant Octagon
Zirconium Titanium grade 4
Biotech Dental Kontact Roxolid
Titanium grade 5 Titanium-zirconium (85% titanium, 15% zirconium)
Straumann (Basel, Switzerland) ACE Surgical Supply (Brockton, USA) Biotech Dental (Salon de Provence, France) Straumann (Basel, Switzerland)
Length including platform (mm)
Diameter of body (mm)
16 8
4.1 4.1
6 8
4.1 4.1
ARTICLE IN PRESS OOOO Volume 00, Number 00
of excitations = 2, 256 £ 256 matrix, echo train length = 10, 1 mm slice thickness, and FOV = 120 mm. The MRI data were acquired in DICOM 3.0 format and transferred to Synapse 3-D software (Fujifilm Holdings Corporation, Tokyo, Japan). The images were measured 2 times by an oral and maxillofacial radiologist who was blinded to the type of dental implant. Using Synapse 3-D software, the dental implant ROI was selected automatically by applying a region grow threshold, the boundaries were automatically marked on MRI images, and the area within the boundaries was shaded in color. SIR measurements were done by means of the software in axial slices (Figure 1). In addition, SIR from the
ORIGINAL ARTICLE Demirturk Kocasarac et al. 3
gel and SIR from the center of the cubic phantom (Figure 2) were measured to normalize differences in signal intensity in axial slices. SIR from the gel was measured from a fixed ROI (30 mm2) that was placed in a uniform gel layer close to the center of the image, whereas SIR from the phantom was measured in an approximately 100 mm2 ROI from the center of the phantom. SIR measurements were made both from T1 WI and T2 WI images. The final SNR and CNR of the dental implants were calculated as follows: SNR ¼ SIRðimplantÞ SIRðphantomÞ = SIRðgelÞ SIRðphantomÞ CNR ¼ 20 log10 SIR½implant = SIR½gel
Fig. 1. (a) T1 weighted implant image; (b) color shading for automatic identification of the implant boundaries; (c) automatic selection of region of interest by applying region grow threshold; (d) signal intensity ratio (SIR) measurements; (e) SIR measurement from the gel that surrounds the implant from the T1 weighted image.
ARTICLE IN PRESS ORAL AND MAXILLOFACIAL RADIOLOGY 4 Demirturk Kocasarac et al.
OOOO && 2019
Fig. 2. Signal intensity ratio measurement from the center of the cubic phantom from a T1 weighted image.
Statistical analysis A power analysis was done indicating the differences between study groups obtained with 4 different implants, which were scanned 5 times, in different MR sequences, to provide statistical power of 0.8 and an a value of .05. One head and neck radiologist with 16 years of experience and special training in MRI evaluated the images. The observer was calibrated by doing measurements on scans that were not part of the research but used only for calibration purposes. During the evaluation of the images in the research, the observer was blinded to the type of dental implant. All measurements were done 2 times by the same observer and their mean values were statistically analyzed. The same observer repeated the measurements after 2 weeks to allow calculation of intraobserver agreement. All statistical analyses were performed using SAS Version 9.4 (SAS Institute, Cary, NC), with the PROC ANOVA (analysis of variance) function. Analysis of the measurements was conducted using ANOVA and post hoc Tukey tests to assess (1) whether one or more materials resulted in a better SNR and/or CNR than the others and (2) whether one sequencing method resulted in a better SNR and/or CNR than the other. Intraobserver agreement was assessed by calculating the percentage difference in SNR (absolute value of the difference, divided by the average value) and
conducting a 2-way ANOVA (for material and sequence) to control for variation between materials. An identical procedure was done for CNR. This analysis was used to calculate intraobserver agreement. Significance for all tests was established at P = .05.
RESULTS The average values of SNR and CNR for each implant type according to the 2 pulse sequences are presented in Table II. The calculated SNR on T1 weighted images was between 0.579 and 1.024. The lowest SNR was found for zirconium, and the highest was for the titanium grade 5 implant. The CNR of T1 weighted images had similar results, with the lowest value for zirconium (8.275) and highest for titanium grade 5 (20.100). On T2 weighted images, the calculated SNR values were between 0.726 and 1.830. The lowest SNR was recorded for zirconium and the highest was in titanium grade 4. The CNR values of T2 weighted images were also lowest for zirconium (7.750), but titanium-zirconium had the highest value (27.580), as shown in Table II. The ANOVA results indicated that zirconium implants resulted in significantly lower SNR and CNR values compared with the other 3 implant types (P < .05). No statistically significant differences were found between titanium (grades 4 and 5) and titanium-zirconium alloy implants (P > .05). There was enough statistical evidence
ARTICLE IN PRESS OOOO Volume 00, Number 00
ORIGINAL ARTICLE Demirturk Kocasarac et al. 5
Table II. Mean values of SNR and CNR measurements for the 4 implant types according to pulse sequences Implant types
T1-WI
T2-WI
P
T1-WI
T2-WI
P
Zirconium Titanium grade 4 Titanium grade 5 Titanium-zirconium
SNR 0.579 a,1 0.912 b,1 1.024 b,1 0.832 b,1
SNR 0.726 a,1 1.830 b,2 1.497 b,2 1.536 b,2
(implant types) P < .05 P > .05 P > .05 P > .05
CNR 8.275 a,1 16.130 b,1 20.100 b,1 12.000 b,1
CNR 7.750 a,1 26.960 b,2 22.615 b,2 27.580 b,2
(pulse sequences) P > .05 P < .05 P < .05 P < .05
SNR, signal-to-noise ratio; CNR, contrast-to-noise ratio; WI, weighted image. *P < .05 indicates statistical significance between implant types. Different letters indicate significant differences for SNR and CNR among implant types. Different numbers indicate significant differences between pulse sequences.
to conclude that the T2 WI sequence had significantly higher SNR and CNR values than the T1 WI sequence for all implant types (P < .05) except zirconium. The average percentage difference in SNR across the 4 materials was 2.4% for T1 WI, and 4.1% for T2 WI (after excluding the large discrepancy of zirconium values). The average percentage difference in CNR across the 4 materials was 5.4% for T1 WI and 7.6% for T2 WI. Regarding intraobserver agreement, in spite of a large disparity in zirconium SNR measurements, ANOVA revealed no significant differences (P = .44 for SNR and .61 for CNR), indicating that T1 WI was no different from T2 WI regarding agreement between the observer’s 2 readings.
DISCUSSION MRI has been used for many purposes in medicine and dentistry and is gaining interest for the assessment of patients before and after implant treatment. However, the presence of some implant materials provokes artifacts because of their physical properties. Signal void is often present at the site of implants, making assessment of the tissues and detection of possible pathosis near the implants difficult. This is particularly important in cancer patients, who need to undergo frequent restaging processes in oncologic imaging. For patients with cancer or a risk of cancer, images with high quality, despite the presence of implants, are indispensable for a correct diagnosis of primary, or detection of recurrent, lesions. In this study, zirconium implants produced significantly lower SNR and CNR values compared with the other three implant types, which did not differ significantly between themselves. SNR and CNR values were significantly higher in the T2 WI sequence than T1 WI for titanium (grades 4 and 5) and titanium-zirconium alloy implants, whereas the values were almost the same for the zirconium implant. There were no significant intraobserver differences for the measurements both in T1 WI and T2 WI sequences (P > .05). All measurements were greatly reproducible, and no significant difference was found between 2 measurements done by the observer (P > .05).
To our knowledge, this is the first study assessing SNR and CNR of images of dental implants with different compositions in a standardized model at 3 T. Hypothetically, increasing magnetic field strength improves SNR in a linear trend. However, different scan parameters as well as many other factors, in addition to the external magnetic field strength, might have a possible effect on SNR.14,15 Moreover, different pulse sequence parameters, such as spatial resolution, FOV, bandwidth, echo time, and repetition time, can affect SNR and CNR as well. Sequence timing details can also be important.16 T1 relaxation time differences are known to be associated with the amount of iron and hydrogen in a structure, whereas T2 relaxation time is known to decrease substantially at higher frequencies, principally as a result of chemical exchange between bulk and bound hydrogen molecules in tissue.17,18 The prolonged translation relaxation time in this study might be the reason for the increase in calculated SNR and CNR as a result of MRI signal saturation. Additionally, SNR and CNR benefit from using a surface coil for signal detection; unfortunately, this was not possible in our study because of the phantom size used. Other studies19,20 have been done regarding artifact generation of dental materials on MRI. However, only a few investigations have evaluated the artifact production resulting from dental implant materials.10,11,21 Moreover, there has been no study concerning the SNR and CNR of dental implants on MRI images and our research is, to our knowledge, the first study to deal with this topic. Repetition and echo times, spatial resolution, FOV, and bandwidth are among many sequence parameters that can influence changes in SNR and CNR. Using identical FOV and spatial resolution may exclude their influence. As a result of variations in tissue relaxation parameters such as longitudinal relaxation time resulting from different external magnetic field strengths (as anticipated by relaxation theory), the sequence timing might have more influence.22 Previous studies, conducted in clinical settings, have reported advantages for 3 T magnets, specifically in brain and musculoskeletal conditions.22,23 Theoretically,
ARTICLE IN PRESS ORAL AND MAXILLOFACIAL RADIOLOGY 6 Demirturk Kocasarac et al.
SNR increases linearly as the magnetic field increases and therefore should double with 3 T relative to 1.5 T. This gives the operator the options of either reducing the scan time or increasing the spatial resolution.22,24 Lu et al.25 examined multiple anatomic structures of the brain and the CNR and SNR of white and gray matter using T1 and T2 relaxation times at 1.5 and 3 T. They found a statistically significant improvement in SNR of white and gray matter of 36% and 41%, respectively, on T1 WI and T2 WI at 1.5 and 3 T. However, only 21% improvement of CNR was noted. DiFrancesco et al.24 studied CNR and SNR in mouse brain imaging at 3 and 7 T in vivo, and their results verified that increases in magnetic field strength yielded predicted increases in the SNR and therefore resolution in MR images of small animals. If the SNR is large enough at lower field strengths, the CNR will improve to the same degree. However, more challenging circumstances resulting in poor SNR or resolution at lower field strength will find the CNR improving even more markedly as field strength is increased. Magnotta et al.26 measured image quality differences using CNR and SNR across sites, vendors, and field strengths in structural fast spin echo T2 WI scans of the brain. Results of the study verified that both CNR and SNR were higher in 3 and 4 T magnetic fields compared with 1.5 T. Although CNR and SNR between different vendors and coils were almost uniform in the 1.5 T field strength scanners, higher field scanners (3 T and 4 T) significantly improved the SNR nonuniformly. Gradl et al.27 compared the effect of a standard head and neck coil with a surface dental coil for SNR of multiple anatomic structures such as osseous and muscular tissues and dental pulp and hard tissues. They detected a 3.5 times increase, on average, in SNR with the dental coil compared with head and neck coil (between 2.7 times increase for masseter muscle and 4.6 for pulp tissue). Using a dental coil increased the SNR for all evaluated structures, which led to a better image quality detected in a subjective assessment. The limitation of this study was the evaluation of SNR in a small ROI (pulp chamber) influenced by volume averaging in standard sequences. Despite the substantial results obtained using a standard approach, our study presents with some limitations. As an in vitro study, it was deprived of the anatomic landmarks that might play a role in image formation and quality. Second, although the imaging settings for the fast spin echo sequences used are similar to the clinical imaging settings, they may vary in different investigations. Third, there is a lack of similar results in the literature with which to compare our results. To our knowledge, this is the first research into the effect of implants and pulse sequences on MRI image quality in the dental field. Lastly, it is vital to emphasize that in vitro studies cannot
OOOO && 2019
mimic actual clinical conditions. However, in vitro research permits the investigators to compare different materials and protocols in ways that would not always be possible for research projects involving patients, especially in expensive and time-consuming modalities such as MRI. There may be a health care need for MR imaging of regions with implants within the field of view. The assessment may be for the detection of pathoses, staging of malignancy, or malignant posttreatment restaging. However, the presence of implants may significantly compromise the image quality in terms of SNR and CNR. Therefore, the main goal of our study was to explore the influence of implant type and imaging sequence on image quality to find the best implant materialimaging sequence combination. The results of our study may facilitate the radiologist’s or clinician’s decision to give the best care to his or her patient. In conclusion, this comparative study found that changes in SNR and CNR at 3 T MRI depend on the implant type and imaging sequence. The results revealed that although titanium (grades 4 and 5) and titaniumzirconium implants imaged with the T2 WI sequence theoretically produce higher SNR and CNR, in clinical conditions the increase might be limited but still significant. However, extension of translation relaxation time might be the reason behind the CNR and SNR improvement in this study. Moreover, it can be recommended that calculation of small volumes or materials that have similar SNR and CNR values can be done by estimating the value of activation strength. This study may provide such values for generalization in further studies. Our findings indicate that the T2 WI sequence may be preferred to T1 WI while evaluating implant integration and inflammatory/infectious processes around or in proximity of implants or for the purpose of cancer restaging in patients who have implants.
REFERENCES 1. Assaf AT, Zrnc TA, Remus CC, Schonfeld M, Habermann CR, Riecke B, Friedrich RE, Fiehler J, Heiland M, Sedlacik J. Evaluation of four different optimized magnetic-resonance-imaging sequences for visualization of dental and maxillo-mandibular structures at 3 T. J Craniomaxillofac Surg. 2014;42:1356-1363. 2. Hart H.R. Jr, Bottomley PA, Edelstein WA, et al. Nuclear magnetic resonance imaging: contrast-to-noise ratio as a function of strength of magnetic field. AJR Am J Roentgenol. 1983;141:1195-1201. 3. Signal, noise, signal-to-noise, and contrast-to-noise ratios. In: Hendrick RE, ed. Breast MRI: Fundamentals and Technical Aspects, New York, NY: Springer; 2008:93-111. 4. Dietrich O, Raya JG, Reeder SB, Reiser MF, Schoenberg SO. Measurement of signal-to-noise ratios in MR images: influence of multichannel coils, parallel imaging, and reconstruction filters. J Magn Reson Imag. 2007;26:375-385. 5. Welvaert M, Rosseel Y. On the definition of signal-to-noise ratio and contrast-to-noise ratio for FMRI data. PloS One. 2013;8: e77089.
ARTICLE IN PRESS OOOO Volume 00, Number 00 6. Korn P, Elschner C, Schulz MC, Range U, Mai R, Scheler U. MRI and dental implantology: two which do not exclude each other. Biomaterials. 2015;53:634-645. 7. Bruschi M, Steinm€uller-Nethl D, Goriwoda W, Rasse M. Composition and modifications of dental implant surfaces. J Oral Implants. 2015;2015. 8. Mouhyi J, Dohan Ehrenfest DM, Albrektsson T. The periimplantitis: implant surfaces, microstructure, and physicochemical aspects. Clin Implant Dent Relat Res. 2012;14:170-183. 9. Elias CN, Oshida Y, Lima JH, Muller CA. Relationship between surface properties (roughness, wettability and morphology) of titanium and dental implant removal torque. J Mech Behav Biomed Mater. 2008;1:234-242. 10. Kocasarac HD, Ustaoglu G, Bayrak S, et al. Evaluation of artifacts generated by titanium, zirconium and titaniumzirconium alloy dental implants in MRI, CT and CBCT images: a phantom study. Oral Surg Oral Med Oral Pathol Oral Radiol. 2019;127:535-544. 11. Smeets R, Schollchen M, Gauer T. Artefacts in multimodal imaging of titanium, zirconium and binary titanium-zirconium alloy dental implants: an in vitro study. 2017;46:20160267. 12. Schenck JF. The role of magnetic susceptibility in magnetic resonance imaging: MRI magnetic compatibility of the first and second kinds. Med Phys. 1996;23:815-850. 13. Weishaupt D, K€ochli VD, Marincek B. How Does MRI Work? An Introduction to the Physics and Function of Magnetic Resonance Imaging. New York, NY: Springer Science & Business Media; 2008. 14. Biswas J, Nelson CB, Runge VM, et al. Brain tumor enhancement in magnetic resonance imaging: comparison of signal-tonoise ratio (SNR) and contrast-to-noise ratio (CNR) at 1.5 versus 3 Tesla. Invest Radiol. 2005;40:792-797. 15. Orhan K, Nishiyama H, Tadashi S, Murakami S, Furukawa S. Comparison of altered signal intensity, position, and morphology of the TMJ disc in MR images corrected for variations in surface coil sensitivity. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2006;101:515-522. 16. Bottomley PA, Foster TH, Argersinger RE, Pfeifer LM. A review of normal tissue hydrogen NMR relaxation times and relaxation mechanisms from 1-100 MHz: dependence on tissue type, NMR frequency, temperature, species, excision, and age. Med Phys. 1984;11:425-448. 17. Gelman N, Gorell JM, Barker PB, et al. MR imaging of human brain at 3.0 T: preliminary report on transverse relaxation rates and relation to estimated iron content. Radiology. 1999;210:759-767.
ORIGINAL ARTICLE Demirturk Kocasarac et al. 7 18. Zhong JH, Gore JC, Armitage IM. Relative contributions of chemical exchange and other relaxation mechanisms in protein solutions and tissues. Magn Reson Med. 1989;11:295-308. 19. Eggers G, Rieker M, Kress B, Fiebach J, Dickhaus H, Hassfeld S. Artefacts in magnetic resonance imaging caused by dental material. MAGMA. 2005;18:103-111. 20. Idiyatullin D, Garwood M, Gaalaas L, Nixdorf DR. Role of MRI for detecting micro cracks in teeth. Dentomaxillofac Radiol. 2016;45:20160150. 21. Duttenhoefer F, Mertens ME, Vizkelety J, Gremse F, Stadelmann VA, Sauerbier S. Magnetic resonance imaging in zirconia-based dental implantology. Clin Oral Implants Res. 2015;26:1195-1202. 22. Biswas J, Nelson CB, Runge VM, et al. Brain tumor enhancement in magnetic resonance imaging: comparison of signal-tonoise ratio (SNR) and contrast-to-noise ratio (CNR) at 1.5 versus 3 tesla. Invest Radiol. 2005;40:792-797. 23. Frayne R, Goodyear BG, Dickhoff P, Lauzon ML, Sevick RJ. Magnetic resonance imaging at 3.0 Tesla: challenges and advantages in clinical neurological imaging. Invest Radiol. 2003;38:385-402. 24. DiFrancesco MW, Rasmussen JM, Yuan W, et al. Comparison of SNR and CNR for in vivo mouse brain imaging at 3 and 7 T using well matched scanner configurations. Med Phys. 2008;35:3972-3978. 25. Lu H, Nagae-Poetscher LM, Golay X, Lin D, Pomper M, van Zijl PC. Routine clinical brain MRI sequences for use at 3.0 Tesla. J Magn Reson Imag. 2005;22:13-22. 26. Magnotta VA, Friedman L. Measurement of signal-to-noise and contrast-to-noise in the fBIRN Multicenter Imaging Study. J Digital Imag. 2006;19:140-147. 27. Gradl J, Horeth M, Pfefferle T, et al. Application of a dedicated surface coil in dental MRI provides superior image quality in comparison with a standard coil. Clin Neuroradiol. 2017;27:371-378. Reprint requests: Husniye Demirturk Kocasarac, DDS, MS, PhD Department of General Dental Sciences Division of Oral and Maxillofacial Radiology Marquette University School of Dentistry PO Box 1881 Milwaukee WI 53201-1881 [email protected]