evaluation of intraorbital prosthetic pigmentation using 0.3 and 1.5 tesla magnetic resonance...
TRANSCRIPT
Evaluation of intraorbital prosthetic pigmentation using 0.3 and 1.5Tesla magnetic resonance imaging and computed tomography
D. Dustin Dees,* Nicole E. MacLaren,* Kevin J. Fritz,† Michael R. Broome‡ and Douglas W. Esson†*Eye Care for Animals, 1021 E. 3300 S., Salt Lake City, UT, 84106, USA; †Eye Care for Animals, 3025 Edinger Avenue, Tustin, CA, 92780, USA;
and ‡Advanced Veterinary Medical Imaging, 3047 Edinger Avenue, Tustin, CA, 92780, USA
Address communications to:
D. D. Dees
Tel.: 801-942-3937
Fax: 801-942-4302
e-mail: ddees@eyecarefor
animals.com
AbstractPurpose To investigate the magnetic susceptibility artifact associated with pigmentedintraorbital prosthetics when performing magnetic resonance imaging (MRI) and com-
puted tomography (CT). Potential artifact reduction techniques were also investigated.Study Design Prospective study.
Methods Five different-colored 20-millimeter small animal silicone intraorbital pros-thetics and two equine prosthetics were evaluated using 0.3 and 1.5 Tesla (T) MRIand CT. MRI sequences included T1- (T1WI) and T2-weighted spin echo (T2WI),
T2 gradient echo (T2*), short tau inversion recovery (STIR), and fluid-attenuatedinversion recovery (FLAIR). When present, artifact size was measured using comput-
erized software by three separate observers. Artifact reduction techniques includedalterations in receiver bandwidth, field of view, slice thickness, and matrix size.
Results The ferrous brown-pigmented prosthetic resulted in a magnetic susceptibilityartifact with MRI. No artifact was observed on CT images. Interobserver variability
was not statistically significant. For both the 0.3T and 1.5T MRI, the T2* sequenceexhibited the largest artifact surface area followed by T2WI, T1WI, STIR, and
FLAIR. Decreasing slice thickness showed a decrease in artifact size; however, thisdifference was not statistically significant.Conclusions The ferrous substances in the brown intraorbital prosthetic resulted in a
significant magnetic susceptibility artifact when performing MRI. Artifact reductiontechniques did not significantly decrease artifact surface area. The use of ferrous
brown-pigmented prosthetics and their potential to affect future MR imaging studiesshould be adequately discussed with pet owners.
Key Words: computed tomography, intraocular silicone prosthetic, intraorbitalsilicone prosthetic, magnetic resonance imaging, magnetic susceptibility artifact
INTRODUCTION
Advanced imaging techniques, such as magnetic resonanceimaging (MRI) and computed tomography (CT), are fre-quently utilized in veterinary medicine, and new informa-tion is constantly gained with regards to their potentialuses and downfalls. Computed tomography is a techniquein which a thin beam of X-rays pass through the body asan X-ray tube rotates around the patient producing cross-sectional images with the use of computerized software.1
When the computer reconstructs images with CT, the rel-ative density of various tissues is expressed in Hounsfieldunits (HU). Different organs and materials within thebody have characteristic HU values, such as water (zero),
cortical bone (+3000), and air (�1000). Magnetic reso-nance imaging also generates cross-sectional images simi-lar to CT; however, the principles of image generationgreatly differ. The patient to be imaged is placed within ahigh-strength external magnetic field, with the magnitudeof the field strength measured in Tesla (T). Hydrogenatoms within the various tissue of the patient line up andafter stimulation by radiofrequency energy pulses emitvarying radio signals which are then used by computerizedsoftware to create an image.1 CT offers superior bonedetail and is better able to detect tissue calcification,whereas MRI has better soft-tissue discrimination.1
Various publications have reported advanced imagingsusceptibility artifacts as well as techniques to reduce these
© 2013 American College of Veterinary Ophthalmologists
Veterinary Ophthalmology (2014) 17, 3, 184–189 DOI:10.1111/vop.12064
artifacts.2–17 In a recent case report, a brown-pigmentedintraorbital prosthetic caused a significant susceptibilityartifact on MR images that precluded adequate visualiza-tion of intracranial structures in a patient with neurologicdisease.18 Since this publication, to the authors’ knowl-edge, no studies have evaluated the effects of other com-monly used intraorbital prosthetics on MRI or CT imagesor potential techniques to reduce this artifact.
The purpose of this study was to determine which ofthe commonly used commercially available intraorbitalprosthetics cause imaging artifact when performing MRIand CT. Potential artifact reduction techniques were alsoinvestigated.
MATERIALS AND METHODS
Five commercially available 20-mm small animal prosthet-ics (Fig. 1) (black and brown – Veterinary OphthalmicSpecialties, Inc., Moscow, ID, USA; clear, yellow, andblue – Jardon Eye Prosthetics, Inc., Southfield, MI, USA)and two equine top hat prosthetics (blue and green – Vet-erinary Ophthalmic Specialties, Inc.) were evaluated with0.3T MRI (Esaote Opera, Genova, Italy), 1.5T MRI (GESigna Infinity 1.5T LX High Speed CXK4; GE Health-care, Princeton, NJ, USA), and CT (Dual-detector-rowCT unit, GE Highspeed Nx/i, GE Healthcare). TwoMRI machines with differing magnetic field strengthswere used to evaluate whether this difference affected thedimensions of the imaging artifact. MRI sequencesincluded T1 (T1WI) and T2-weighted spin echo (T2WI),T2 gradient echo (T2*), short tau inversion recovery(STIR), and fluid-attenuated inversion recovery (FLAIR).Specific sequence parameters are listed in Table 1.These sequences were chosen as they are routinelyemployed for canine intracranial imaging.19,20 CT scans(Helical mode; Pitch of 1.5; kVp = 120; mA = 100; Fieldof view (FOV) = 18 cm, Matrix = 512 9 512, Slice thick-ness = 1 mm) were performed with each prosthetic pro-spectively reconstructed using three different algorithms(GE proprietary names STD, CHST, and BONE) andevaluated using a soft-tissue window (WL: 40 WW: 350).
Artifact reduction techniques were evaluated on the 1.5TMRI and included alterations in receiver bandwidth, FOV,slice thickness, and matrix size (Table 2).
To properly image the prosthetics, a phantom (Fig. 2)was constructed. The base was constructed of 4.5-cm thickwhite (nonpigmented) foam, 23.5 cm long by 15.9 cmwide. Two circles, 6.5 cm in diameter, were cut into thecenter of the foam block approximately 2.5 cm apart. Twowhite (nonpigmented) foam cups were placed into the cir-cular cutouts. The control prosthetic (clear) was placedinto one cup suspended with 4-0 polypropylene (Prolene;Ethicon, Somerville, NJ, USA) suture material. The othervarious pigmented prosthetics, suspended with suturematerial, were individually placed into the adjacent cup.The cups were filled with water to the point at which theprosthetics were completely submerged. The entire phan-tom was then placed into a human knee coil for scanacquisitions. Computer software (Sound-Eklin, Carlsbad,CA, USA) was used by three separate observers (Observer1 – DD; Observer 2 – NM; Observer 3 – KF) to objec-tively measure the horizontal and vertical diameters andsurface area of image artifact when present (Fig. 3).
To evaluate the effect of magnet size, MRI sequence,and reduction methods on artifact surface area, a mixedmodel analysis of variance was used in SAS v9 (SAS Insti-tute Inc., Cary, NC, USA). The observer was included inthe model as a random effect. An interaction term wasincluded in the model and comparisons done using Least
Table 1. MRI sequence parameters
Sequence
Magnetsize intesla (T)
Repetitiontime (TR)
Echotime(TE)
Inversiontime (TI) Echo
T1WI 0.3 T 800 26 n/a Spin echoT1WI 1.5 T 450 10 n/a Spin echoT2WI 0.3 T 3150 90 n/a Fast-spin echoT2WI 1.5 T 5000 123 n/a Fast-spin echoT2* 0.3 T 800 22 n/a Gradient echoT2* 1.5 T 234 15 n/a Gradient echoFLAIR 0.3 T 3840 80 960 Fast-spin echoFLAIR 1.5 T 6827 155 2100 Fast-spin echoSTIR 0.3 T 2600 30 90 Fast-spin echoSTIR 1.5 T 1200 47.7 150 Fast-spin echo
Table 2. Artifact reduction techniques included alterations in band-
width, field of view, slice thickness, and matrix size on 1.5T MRI.
Note that only one parameter was altered per sequence group to
identify if that technique could suppress the present artifact. Resul-
tant echo times correlated with decrease in artifact surface area.
Sequence parameter constants for all sequence groups, repetition time
(TR) = 3500 ms and echo time (TE) = 120 ms
BandwidthField ofview (cm)
Slicethickness(mm) Matrix
Echospacing(min TE)
Bandwidth1a 41 14 4 256 8.11b 62 14 4 256 8.21c 83 14 4 256 8.3
Field of view2a 41 10 4 256 8.72b 41 12 4 256 8.32c = 1a 41 14 4 256 8.12d 41 16 4 256 7.9
Slice thickness3a 41 14 2 256 7.63b 41 14 3 256 7.63c = 1a 41 14 4 256 8.1
Matrix4a = 1a 41 14 4 256 8.14b 41 14 4 512 12.0
MRI, magnetic resonance imaging.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 17, 184–189
e v a l u at i on o f p ro s t h e t i c p i gm ent a t i on u s i n g mr i a nd c t 185
Squares Means. A Friedman’s test compared differences inmeasurements made by three different observers. Signifi-cance level was set at P � 0.01.
RESULTS
The CT image densities of each prosthetic are providedin Table 3. No artifact was observed with any of thetested prosthetics on CT images (Fig. 4). Only thebrown-pigmented prosthetic resulted in a susceptibilityartifact with MRI. There was strong correlation of artifactsurface area and diameter measurements (0.97). For thisreason, statistical analysis was only performed using sur-face area measurements.
The overall artifact mean surface area was12.27 � 4.52 cm2 (Range = 8.06–27.73 cm2). The meansurface area for observer 1 was 11.59 � 4.08 cm2,12.87 � 5.16 cm2 for observer 2, and 12.34 � 4.36 cm2
for observer 3. The differences of mean surface area mea-surements between observers were not significant at the0.01 level (P = 0.0330).
The artifact surface area had a significant magnet sizeby sequence interaction (P = 0.0024). When comparingeffect of MRI magnet size on artifact surface area keepingindividual sequences fixed, artifact surface area with theT2* (P = 0.0004) and T2WI (P = 0.0076) sequences withthe 0.3T MRI was significantly larger than thosesequences with the 1.5T MRI. Surface area differenceswere of borderline significance for FLAIR (P = 0.0401)and T1WI (P = 0.0374). No significant difference wasnoted for the STIR sequence (P = 0.42). Artifact surfacearea size differences were estimated between fixedsequences of the two MRI machines. The surface area dif-ferences between the 0.3T MRI and 1.5T MRI were8.43 cm2 with the T2* sequence, 5.80 cm2 with theT2WI sequence, 4.33 cm2 with the T1WI sequence, and1.60 cm2 with the STIR sequence. The artifact surfacearea difference between the 1.5T MRI and 0.3T MRI was4.26 cm2 with the FLAIR sequence.
Figure 1. Intraorbital prosthetics evaluated with magnetic resonance
imaging and computed tomography including clear, black, brown,
yellow, and blue 20 mm small animal and blue and green equine Top
Hat prosthetics.
Figure 2. Constructed imaging phantom. The prosthetics (clear –
right; brown – left) are shown suspended in water by suture material.
Figure 3. MR image showing vertical and horizontal diameter and
surface area measurements.
Table 3. Comparison of prosthetic density on CT images
Prosthetic color Hounsfield units (HU)
SA – Clear (control) + 265SA – Brown + 245SA – Blue + 250SA – Black + 275SA – Yellow + 420EQ – Green + 295EQ - Blue + 285
SA, small animal prosthetic; EQ, equine prosthetic; CT, computedtomography.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 17, 184–189
186 d e e s E T A L .
The effect of sequence on artifact surface area was com-pared keeping MRI magnet size fixed. For the 0.3T MRI,the T2* sequence had a significantly larger artifact,whereas the FLAIR sequence had a significantly smallerartifact. For the 1.5T MRI, the T2* sequence had asignificantly larger artifact surface area. For both the 0.3Tand 1.5T MRI, the T2* sequence had the largest surfacearea, followed by T2WI, T1WI, STIR, and FLAIR(Fig. 5).
Using the T2WI sequence artifact surface area imagewith the 1.5T MRI as control, reduction techniques wereevaluated. The overall analysis showed no statistically sig-nificant decrease in artifact surface area (P = 0.0563). Ascompared to the control image, two reduction techniques(3A and 3B) did show a decrease in artifact size, but nei-ther of the differences were significant (P = 0.56 and 0.87,respectively).
DISCUSSION
Enucleation and evisceration are relatively commonly per-formed surgical procedures in veterinary ophthalmology.Enucleation is recommended for blind and painful eyessuch as those affected with chronic inflammation, trauma,intraocular neoplasia, endophthalmitis, or glaucoma.21–23
Evisceration is a potentially more cosmetic surgical proce-dure typically performed for end-stage glaucomatouseyes.21–24 In both procedures, a silicone prosthesis isplaced either within the corneoscleral shell or the intraor-bital space.24,25 Clinicians performing these proceduresare cautioned when considering use of the brown ferrous-pigmented prosthetic. Although the probability of encoun-tering this artifact in a clinical patient due to presence ofan ocular prosthetic is likely rare, the results may be dele-terious for the patient due to the increased anesthetic timeneeded to remove the prosthetic for accurate imaging.
In this study, only the brown prosthetic caused suscepti-bility artifact on MR images; however, no artifact wasobserved with any of the tested prosthetics with CT. Wepreviously reported the pigment composition of the brownprosthetic to contain numerous ferrous compounds.18 Ourresults show that although an artifact was observed withMRI, the ferrous materials were not sufficient enough inquantity or composition to cause artifact with CT. This isan important finding indicating CT, instead of MRI, canbe utilized to adequately image intracranial structures inpatients having a brown-pigmented prosthetic. Interest-ingly, the yellow, small animal prosthetic (HU = 420) wasconsiderably more dense than all other prosthetics (HUrange = 245–295) on CT images. The reason for this iscurrently unknown; however, it can be speculated thatbecause all of the prosthetics were similar in shape andmade of the same medical grade silicone, it is likely thatthe pigments used to color the prosthetic resulted in thisincreased density. It is also currently unknown if theincreased density of the yellow prosthetic could hinder
Figure 5. Visual comparison of different
magnetic resonance imaging (MRI) sequence
scans of the control clear prosthetic and brown
prosthetic. The top row represents images from
the clear prosthetic on 0.3T MRI with no artifact
observed. The middle and bottom rows represent
the brown prosthetic on 0.3T MRI and 1.5T
MRI, respectively. The columns from left to right
indicate representative sequence for each of
images below. Note the extent of artifact as
compared to control images and the differences in
artifact shape and size between the 0.3T and 1.5T
machines.
Figure 4. Computed tomography image of the control clear
prosthetic (left) and brown prosthetic (right). No artifact was
observed associated with any of the tested prosthetics. (Control
[HU = 265] and Brown [HU = 245]; WL: 40 WW: 350)
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 17, 184–189
e v a l u at i on o f p ro s t h e t i c p i gm ent a t i on u s i n g mr i a nd c t 187
CT imaging of a clinical patient. Further investigation ofthis statement is needed.
Also, our results indicate the artifact surface area wasaffected by specific MRI sequence and magnet size.Artifact surface area tended to be larger with the 0.3Tmachine versus the 1.5T machine. This is in directcontrast to a previously published report of metal-inducedsusceptibility artifact worsening with increased magnetstrength.10 The reason for this is unknown. Differencesin size, shape, and composition of the biomaterials testedin our study may have accounted for the contrastingresults. For both the 0.3T and 1.5T MRI machines, theT2* sequence had the largest surface area followed byT2WI, T1WI, STIR, and FLAIR. In a patient withbrown-pigmented prosthetic requiring intracranial imag-ing, these results, along with consultation with a boardcertified veterinary radiologist may help guide an imagingprotocol.
Movement of metallic substances or implants duringMR imaging is a significant concern that can result in fur-ther tissue damage.26 The aforementioned case report ofmagnetic susceptibility artifact caused by a brown intraor-bital prosthesis did not identify presence of motion artifacton MR images obtained in that study.18 Similarly in thisreport, the images obtained revealed no artifactual changesdue to movement of the prosthetics. Even slight magneticpull placed on the prosthetics should have been observedon obtained images as the prosthetics were suspended bythin suture material in water and could freely move if sucha magnetic attraction existed. A cadaver imaging studyevaluating prosthetic movement within the orbit would beinteresting and informative.
Various imaging protocols have been evaluated toreduce susceptibility artifacts using MRI.3,6,8,11,12,15–17,27
Methods in these studies used to decrease magnetic sus-ceptibility artifact included increasing receiver bandwidth,decreasing slice thickness, increasing matrix size, anddecreasing the FOV. We mirrored these same techniquesin an attempt to decrease artifact surface area; however,no statistically or clinically relevant reduction wasobserved. Although decreasing slice thickness resulted in aslight decrease in artifact surface area, the difference wasnot significant. These results suggest that prostheticremoval is likely the only means to adequately imageintracranial and surrounding structures with the MRImachines utilized in this study. Use of CT, if available,may allow for adequate intracranial imaging and detractthe need for surgical prosthetic removal.
The major goals of this study were: (i) Identification ofMRI/CT image artifacts caused by commonly used com-mercially available small and large animal veterinary in-traorbital prostheses and (ii) identifying sequenceparameter changes to correct or decrease the magnitudeof the artifact when present. Only the brown ferrous-pigmented prosthetic caused imaging artifact with MRIand not CT. None of the other prosthetics caused imag-
ing artifact with either modality; however, the yellow-pigmented prosthetic was considerably denser on CTimages. With the current techniques, machines, and soft-ware we used, no artifact reduction was achieved. Addi-tional and worthwhile information obtained from thisstudy included how differing MRI sequences can have alarger or smaller associated image artifact and this infor-mation may be utilized, along with consultation with aveterinary radiologist, to tailor an imaging protocol forthese patients. Limitations of this study include: (i) Notall known magnetic susceptibility artifact reduction tech-niques were utilized such as changes in phasing andencoding direction or changes in echo timing, and (ii)currently employed artifact reduction techniques werenot performed with both machines (0.3T and 1.5T MRImachines) and with all presently used sequences. Theauthors’ decision to perform artifact reduction techniquesusing the T2WI sequence on the 1.5T MRI machinewas based on a few factors. Firstly, the artifact surfacearea was the second largest using the T2WI sequenceand had clearly defined borders. We chose this sequenceas with the large artifact surface area even small decreasesin size in that surface area could be measured and com-pared. If other sequences were used, such as STIR, forexample, the original (no reduction technique applied)artifact surface area was likely too small to detect andmeasure even small decreases in that surface area due touse of specific artifact reduction techniques, thus notappropriately allowing for notice of change. Secondly,the capabilities of different MRI machines are based onthe size of the magnet and quality of software used withthe specific machine. In our study, the 1.5T machine wasused for artifact reduction techniques as the softwarewith the machine was of higher quality and allowedchanges of the specific parameters that we evaluated.Also, the CT images were not viewed on a brain windowwhich may have provided greater detail than a soft-tissuewindow, thus allowing for identification of minor imageartifacts. Futures studies, potentially utilizing cadaver tis-sues, are warranted to investigate these limitations. Inconclusion, the ferrous substances found within thebrown-pigmented intraorbital prosthetic resulted in a sig-nificant magnetic susceptibility artifact when performingcranial MRI. Artifact reduction techniques did not signif-icantly decrease artifact surface area. The use of ferrousbrown-pigmented prosthetics and their potential to affectfuture MR imaging studies should be adequately dis-cussed with pet owners.
ACKNOWLEDGEMENTS
The authors thank Dr. Amy M. Knollinger, DVM,DACVO and Jeffery P. Simmons, DVM, MS, DACVECCfor their assistance with phantom design and construction,and Brandon Estrada and Megan Pieper, RT (R) (MR) fortheir technical assistance with scan acquisitions.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 17, 184–189
188 d e e s E T A L .
REFERENCES
1. Penninck D, Daniel GB, Brawer R et al. Cross-sectional imaging
techniques in veterinary ophthalmology. Clinical Techniques in
Small Animal Practice 2001; 16: 22–39.2. Cooper JJ, Young BD, Hoffman A et al. Intracranial magnetic
resonance imaging artifacts and pseudolesions in dogs and cats.
Veterinary Radiology and Ultrasound 2010; 51: 587–595.
3. Herrick RC, Hayman LA, Taber KH et al. Artifacts and pitfalls
in MR imaging of the orbit: a clinical review. RadioGraphics 1997;
17: 707–724.4. Freer SR, Scrivani PV. Postoperative susceptibility artifact during
magnetic resonance imaging of the vertebral column in two dogs
and a cat. Veterinary Radiology and Ultrasound 2008; 49: 30–34.
5. Saito M, Ono S, Kayanuma H et al. Evaluation of the
susceptibility artifacts and tissue injury caused by implanted
microchips in dogs on 1.5T magnetic resonance imaging. Journal
of Veterinary Medical Science 2010; 72: 575–581.6. Peh WC, Chan JH. Artifacts in musculoskeletal magnetic
resonance imaging: identification and correction. SkeletalRadiology 2001; 30: 179–191.
7. Fry CL, Haik BG. Magnetic resonance imaging artifacts induced
by ocular prosthesis. American Journal of Ophthalmology 1994;
118: 262–263.8. Lee MJ, Kim S, Lee SA et al. Overcoming artifacts from metallic
orthopedic implants at high-field-strength MR imaging and
multi-detector CT. Radiographics 2007; 27: 791–803.
9. Gross B. Findings of magnetic susceptibility artifacts of the
vertebral column. Veterinary Radiology and Ultrasound 2008; 49:
317–318.10. Matsuura H, Inoue T, Ogasawara K et al. Quantitative analysis
of magnetic resonance imaging susceptibility artifacts caused by
neurosurgical biomaterials: comparison of 0.5, 1.5, and 3.0 Tesla
magnetic fields. Neurologia Medico-Chirurgica (Tokyo) 2005; 45:395–399.
11. Stradiotti P, Curti A, Castellazi G et al. Metal-related artifacts in
instrumented spine. Techniques for reducing artifacts in CT and
MRI: state of the art. European Spine Journal 2009; 18(Suppl 1):102–108.
12. Olsen RV, Munk PL, Lee MJ et al. Metal artifact reduction
sequence: early clinical applications. RadioGraphics 2000; 20: 699–
712.
13. Zhang Y, Cheng J, Bai J et al. Tiny ferromagnetic intraocular
foreign bodies detected by magnetic resonance imaging: a report
of two cases. Journal of Magnetic Resonance Imaging 2009; 29:
704–707.
14. Hecht S, Adams WH, Narak J et al. Magnetic resonance imaging
susceptibility artifacts due to metallic foreign bodies. Veterinary
Radiology and Ultrasound 2011; 52: 409–414.15. Harris CA, White LM. Metal artifact reduction in
musculoskeletal magnetic resonance imaging. Orthopedic Clinics ofNorth America 2006; 37: 349–359.
16. Hargreaves BA, Worters PW, Pauly KB et al. Metal-induced
artifacts in MRI. American Journal of Roentgenology 2011; 197:
547–555.17. Farrelly C, Davarpanah A, Brennan S et al. Imaging of soft
tissues adjacent to orthopedic hardware: comparison of 3-T and
1.5-T MRI. American Journal of Roentgenology 2010; 194: 60–64.18. Dees DD, Knollinger AM, Simmons JP et al. Magnetic resonance
imaging susceptibility artifact due to pigmented intraorbital
silicone prosthesis. Veterinary Ophthalmology 2012; 15: 386–390.
19. Armour MD, Broome MR, Dell’Anna G et al. A review of orbital
and intracranial magnetic resonance imaging in 79 canine and 13
feline patients (2004–2010). Veterinary Ophthalmology 2011; 14:215–226.
20. Cherubini GB, Platt SR, Howson S et al. Comparison of
magnetic resonance imaging sequences in dogs with multi-focal
intracranial disease. Journal of Small Animal Practice 2008; 49:634–640.
21. Cho J. Surgery of the globe and orbit. Topics in CompanionAnimal Medicine 2008; 23: 23–37.
22. Spiess BM. Diseases and Surgery of the Canine Orbit. In:
Veterinary Ophthalmology, 4th edn. (ed. Gelatt KN) Iowa,
Blackwell, 2007; 551–553.23. Gelatt KN, Gelatt JP. Surgery of the orbit. In: Small Animal
Ophthalmic Surgery: Practical Techniques for the Veterinarian. (edsGelatt KN, Gelatt JP) Oxford, Butterworth-Heinemann, 2001;
52–63.24. Lin CT, Hu CK, Liu CH et al. Surgical outcome and ocular
complications of evisceration and intraocular prosthesis
implantation in dogs with end stage glaucoma: a review of 20
cases. Journal of Veterinary Medical Science 2007; 69: 847–850.25. Hamor RE, Roberts SM, Severin GA. Use of orbital implants
after enucleation in dogs, horses, and cats: 161 cases (1980–1990). Journal of the American Veterinary Medical Association 1993;
203: 701–706.26. Thelen A, Bauknecht HC, Asbach P et al. Behavior of metal
implants used in ENT surgery in 7 Tesla magnetic resonance
imaging. European Archives of Otorhinolaryngology 2006; 10: 900–905.
27. Arena L, Morehouse HT, Safir J. MR imaging artifacts that
simulate disease: how to recognize and eliminate them.
Radiographics 1995; 15: 1373–1394.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 17, 184–189
e v a l u at i on o f p ro s t h e t i c p i gm ent a t i on u s i n g mr i a nd c t 189