CLINICAL MOTIVATIONS FOR MRI IMAGING OF THE OPTIC NERVE
The optic nerve can be affected by numerous pathologies including primary neoplasms, metastases, inflammation, intracranial hypertension, and glaucoma. Although the retina and optic nerve head can be assessed with ophthalmoscopy and optical coherence tomography, large portions of the optic nerve are “hidden” from clinical examination by the globe and bones of the orbit and skull base. Computed tomography (CT) provides rapid acquisition in the emergency setting and excellent bone detail. MRI provides superior visualization of the orbital soft tissues and optic pathways without ionizing radiation (1,2). MRI better characterizes postchiasmatic portions of the visual pathways that also may be affected by different disorders.
Clinical MRI of the Optic Nerve
The optic nerve is not well characterized on conventional brain MRI because of its relatively small size and surrounding orbital fat. Optimized and targeted orbital sequences that include fat saturation ideally should be obtained before advanced MRI sequences described below are considered. A orbital MRI protocol is distinct from a routine brain MRI protocol—combining the 2 complete protocols may result in an excessively long examination for the patient and may not be reimbursed by insurance. The following is a brief synopsis of a typical orbital protocol, common optic nerve disease imaging findings, and when to clinically consider CT instead of MRI.
Precontrast axial and coronal T1 (without fat saturation), axial T2, and coronal T2 images with fat suppression are common components of a clinical MRI orbit protocol. High signal from fat on precontrast T1 images without fat suppression helps to assess marrow in the bony orbit walls, evaluate the size of the extraocular muscles, and detect intraconal masses. After intravenous gadolinium injection, axial, and coronal T1 images with fat suppression are acquired. The orbital apex and optic canal are best evaluated on axial sequences. The optic nerve, chiasm and tracts are best assessed in the coronal plane. Postcontrast fat-suppression pulse sequences are an essential feature of the orbital protocol since the orbital portion of the optic nerve is surrounded by fat (3). Fat-suppression pulses are sensitive to magnetic field inhomogeneities and become suboptimal because of local geometric distortions in patients with dental or craniofacial hardware. A short tau inversion recovery (STIR) pulse sequence can be substituted to provide superior image quality (4). However, this longer sequence is prone to artifacts from eye motion and vessel pulsation.
Common Disease Findings and Interpretation
Because the optic nerve is a central nervous system structure, it is surrounded by all 3 meningeal layers, surrounded by cerebrospinal fluid (CSF), and its axons are myelinated by oligodendrocytes. The optic nerve, meninges, and CSF together constitute the radiologic optic nerve sheath complex (ONSC).
The most common ONSC tumors are meningioma and optic pathway glioma (5). Both tumors demonstrate similar fusiform enlargement with enhancement. Meningiomas arise from arachnoid cells within the meninges so the centrally located optic nerve becomes compressed. Circumferential meningiomas have a typical “track–track” pattern of enhancement (6), whereas other meningiomas may have a more eccentric location. In contrast, gliomas expand the nerve centrifugally and there are often additional findings including nerve kinking, mucinous cyst formation, and skip lesions (7–9). MRI scans also can detect glioma spread to postchiasmatic structures (10).
Optic nerve inflammation may mimic tumor on MRI in the acute setting with nerve expansion, edema, and enhancement. Typically, findings are most apparent in the orbital segment of the nerve, which will be swollen with increased T2 signal (11). Contrast enhancement is best seen with fat-suppression techniques in the coronal plane. Enhancement will be seen in greater than 90% of cases if the MRI is obtained within 3 weeks of symptom onset (12). Of note, MRI of the optic nerve for a patient with clinically-established MS presenting with new symptoms of optic neuritis is unnecessary since there are rarely alternative causes (13).
CSF surrounding the optic nerve provides excellent contrast when evaluating the ONSC. Increased intracranial pressure in idiopathic intracranial hypertension can enlarge the CSF space of the ONSC. This is best seen on fat-suppressed T2 images as vertical or horizontal buckling of the nerve. Papilledema appears as abnormal anterior convexity of the optic disc at the posterior globe on axial T2-weighted images (14). Alternatively, with intracranial hypotension there may be almost no CSF surrounding the optic nerve on MRI. Optic nerve atrophy from any cause will become obvious because of contrast between the nerve and surrounding CSF signal on T2 sequences.
Reasons to Choose CT Over MRI
Although MRI offers superior visualization of orbital soft tissues, the evaluation of structures with a low concentration of mobile protons (e.g., cortical bone) is limited. Hence, CT is recommended to evaluate orbital trauma or bony lesions (15,16). MRI also has longer scan times compared to CT that can lead to motion degradation in uncooperative, pediatric, elderly, or claustrophobic patients. CT also should be performed when MRI is not clinically feasible or contraindicated, including patients with implanted medical devices and in patients with CT-proven intraorbital ferromagnetic foreign bodies (e.g., from previous trauma or gunshot wound) (17). There are other causes of artifact for the orbital MRI such as cosmetic makeup and tattoos, dental braces, and asymmetric fat suppression. Some institutions now perform orbital MRI in patients with pacemakers when directly supervised by electrophysiology or cardiology.
CHALLENGES TO MRI OF THE OPTIC NERVE
There are several inherent challenges to imaging the optic nerve. The optic nerve is small in caliber varying between 1.5 and 4 mm in diameter and being most narrow in the optic canal (18). Especially for noise-limited MRI sequences like diffusion, there can be a stark trade-off between signal-to-noise ratio (SNR) and sufficient spatial resolution to minimize partial volume effects with CSF and/or orbital fat.
The adjacent air-filled paranasal sinuses and skull base create large magnetic field inhomogeneities that result in geometric distortions and signal drop-out that can mask optic nerve abnormalities. This problem is exaggerated at higher magnetic field strengths or in echoplanar techniques where long echo-train lengths are required to generate sufficient spatial resolution for the optic nerve. These magnetic field inhomogeneities also degrade fat saturation sequences.
Artifacts from patient head motion or, more commonly, involuntary eye movements blur the optic nerve and confound quantitative approaches. The range of movement for the orbital optic nerve is highest, whereas the nerve is relatively tethered at the optic canal and there is minimal motion for its precanalicular portion (19). The orbital optic nerve has been documented to move a mean total distance of 11.8 mm over a mean area of 5.2 mm2 during a 3-minute diffusion acquisition in healthy volunteers asked to fixate on a stationary target placed on the inner surface of the MRI scanner bore (20). The relative position of the orbital nerve also can drift during the scan even in highly-motivated healthy subjects and conjugate gaze is not always maintained (especially if the eyes are “closed”). Asking the patient to fix their eyes on a single point for 5 minutes did not significantly reduce movement (21). Therefore, rapid MRI acquisitions to reduce motion degradation are a key emphasis of current research. Conversely, using real-time MRI to characterize motion of the optic nerve and other orbital structures has been relatively unexplored (22,23). Motion of the optic nerve at 3 different positions is demonstrated in the attached video using a coronal single-shot fast spin echo sequence (see Supplemental Digital Content, Video, http://links.lww.com/WNO/A227). Optic nerve motion limits the utility of conventional postprocessing realignment techniques.
Optic nerve orientation changes over its course and is consistently oblique to the standard axial, sagittal, and coronal planes of conventional MRI. The nerve is not simply a cylinder extending from globe to chiasm but is composed of 2 obliquely-oriented parts that taper and form a pivot at the orbital apex. The right and left optic nerves do not run parallel but rather diverge from each other anterolaterally toward their respective orbital apices. This limits MRI slice geometries to sample both optic nerves simultaneously. Most groups have focused on coronal orientations, but because of the small optic nerve size relative to 2D geometry slice thickness and diverging oblique orientations, there are still significant partial volume effects.
Volumetric 3D isotropic MRI sequences minimize partial volume effects by facilitating multiplanar or curviplanar reformats. The optic nerve is better evaluated with 3D T1 and T2 sequences when compared with respective 2D sequences (24), However, volumetric 3D sequence acquisition times require 5–6 minutes. A further limitation is that volumetric T2 sequences are “CSF-weighted” such that pathological T2 signal intensity changes within the optic nerve can be hard to appreciate. The oblique optic nerve orientation also creates potential for artifactual hyperintensity in short echo time MRI sequences (e.g., proton density images) due to the magic angle effect, which is a complex phenomenon resulting from the quantum mechanics of spin interactions relative to their orientation to the main magnetic field (25,26).
RECENT INNOVATIONS FOR MRI OF THE OPTIC NERVE
Discussion of MRI innovations should begin with the use of high-field clinical scanners, that is, 3-Tesla which have theoretically-doubled SNR compared with 1.5-T. However, it is not always so simple or straightforward—the MRI protocols must be adjusted because at 3-T more radiofrequency energy is deposited into the patient, T1 contrast decreases, T2 relaxation times shorten, whereas chemical shift and susceptibility artifacts may be increased. Several new research sequence innovations and technological advances described below have the potential for improved imaging the optic nerve (Table 1).
Traditional fat-suppressed postcontrast T1 turbo spin echo imaging is often motion-degraded because of respiration, vascular pulsation, and eye movements. The fat-suppressed T1-weighted 3D radial gradient-recalled echo-volumetric interpolated breath hold examination (Radial VIBE) samples the x and y planes of k-space in a radial spoke-wheel fashion (27,28). Note, “k-space” is a graphical representation of the data directly obtained during MRI scanning and represents an array of spatial frequencies present—the Fourier transformation of k-space creates the anatomical MRI images. Compared with conventional Cartesian sampling technique, radial sampling has reduced sensitivity to motion because of the varying sample directions and the oversampling of the center of the k-space (29). The radial-VIBE sequence has demonstrated superior image quality when evaluating the orbits compared with conventional postcontrast images (30) (Fig. 1). This sequence also lends itself to dynamic temporal resolution of contrast enhancement in the optic nerve or associated mass lesions (i.e., for potential future modeling characterization of perfusion and permeability).
Half-Fourier Acquisition Single-shot Turbo Spin Echo
Half-Fourier acquisition single-shot turbo spin echo (HASTE) imaging is a single-shot echo-planar fast spin echo sequence for high-resolution T2 images (31,32) specific to Siemens (but other MRI vendors provide similar versions). HASTE uses phase-conjugate symmetry to take advantage of “mirror-image” properties in k-space. Thus, just over half the data needs to be acquired while the remaining lines of k-space are estimated. HASTE images (<1 second) can virtually eliminate motion artifacts. HASTE 2D slices also are concatenated making it further resistant to motion degradation as only the portion of the study acquired during motion is imperfect. The precision of HASTE MRI exceeds ultrasonographic studies for determining optic nerve diameter (33) and HASTE measures of decreased caliber correlated with glaucoma severity (34). We routinely use orbital HASTE to supplement a diagnostic head MRI in patients with suspected clinically-isolated syndrome or MS (Fig. 2). The HASTE acquisition of 43 seconds gives us a relatively “free” look at the optic nerves. The acquisition can be acquired with or without fat saturation.
Diffusion-weighted MRI (DWI) characterizes how the random Brownian motion of water is altered in normal and pathologic nervous tissue environments (35). Reduced water diffusion has proven sensitive, but not specific, to many pathologies in the brain including ischemia, infection, and inflammation. There are reports demonstrating reduced diffusion for acute optic nerve ischemia (36,37) (Fig. 3) and acute optic neuritis (38,39). Diffusion also can increase the sensitivity for detecting highly-cellular malignancies that infiltrate the ONSC such as lymphoma or primitive neuroectodermal tumor. However, the typical axial DWI sequence for the brain is a poor choice for the orbits because of suboptimal resolution of the obliquely-oriented optic nerves and potential anisotropy effects.
Myelin creates a largely impermeable barrier to water diffusion orthogonal to the long axis of the axon. With insufficient angular resolution, routine diffusion trace MRI signal intensity can appear artificially increased (and called “pathologic”) because one of the limited contributing diffusion MRI sequence gradients is oriented orthogonal to the long axis of the normal, healthy optic nerve. Diffusion-tensor imaging (DTI) is the simplest mathematical extension of DWI to derive the principal orientations of water diffusion, quantify and eliminate these anisotropy artifacts, and generate tractography that correlates with specific white matter pathways (35). Fractional anisotropy is a scalar measure of the degree of diffusion coherence in one direction—values vary between 0 and 1, where CSF fractional anisotropy approaches “0” ,whereas for highly coherent white matter, like the corpus callosum, fractional anisotropy approaches “0.8–1.” A useful benchmark for diffusion MRI quality then is the reported fractional anisotropy value for the normal optic nerve—low values generally imply partial volume effects from orbital fat or motion degradation due to long scan times.
This sensitivity to white matter structure and organization makes DTI and/or diffusion tractography useful surrogate markers for axon and myelin integrity. DTI parameters are altered for the optic nerves and radiations by glaucoma (40) and correlate with glaucoma severity (41). DTI values correlate well with visual dysfunction on visual evoked potentials in optic neuritis with MS (42). Unlike other clinical MRI sequences for the optic nerve, diffusion MRI provides a quantitative, objective marker of optic nerve pathology and seems sensitive to subtle microscopic tissue changes, such as early selective axonal loss, that are not easily recognized by radiologists, referring physicians and sometimes patients. Most clinical applications thus far have focused on DTI characterization of the optic nerves since the diffusion data required for this analytical model is feasible in patients. There are newer models of nervous tissue complexity (43) that may provide more specific measures of axon and myelin integrity, but these currently require long acquisitions that may not be practical in patients.
Reduced Field-of-View or Multishot Methods
DWI or DTI of the optic nerve remains challenging because you need high spatial resolution, yet, doubling spatial resolution reduces SNR by 87% and diffusion is already a “signal-starved” MRI sequence. It also is only clinically feasible to acquire DWI with echoplanar imaging where the relatively high spatial resolution needed requires a long echo train that becomes more vulnerable to magnetic susceptibility artifacts (e.g., from the adjacent sphenoid sinus). Reducing the field-of-view to exclude the unwanted tissues can reduce long echo-train lengths, but will lead to aliasing or wrap-around artifacts unless the MRI signal from the unwanted tissue volume is suppressed or never excited to resonance in the first place (19). Using this technique, diffusion values correlated well with clinical and electrophysiological findings of optic neuritis (44). More recently, this method was used for DTI of the optic nerve in normal subjects and those with optic neuritis (21,45). Alternatively the echo-train length also can be reduced by segmenting it into separate components (“multi-shot”) to reduce geometric image distortion (Fig. 4). This approach applied to DTI of the optic nerve provides a higher SNR (since overall phase-encode steps are not reduced) and improved resolution with fewer artifacts compared with single-shot techniques (46). The trade-off is increased scan time. Scan time may be reduced with parallel imaging, partial Fourier acquisitions and/or recently-developed simultaneous multislice (SMS) acquisitions (47).
A series of spin echo sequences with increasing echo time can characterize exponential MRI signal decay to estimate T2 values for nervous tissue. However the long scan times required are not clinically feasible (e.g., 30 minute scans). Multiecho fast spin echo pulse sequences, which acquire multiple measures of signal decay per 90° pulse, can reduce times for T2 mapping to 5–6 minutes but overestimate T2 values with wide variance (48). We have implemented an echo-modulation curve T2 mapping package that models both stimulated echoes and radiofrequency inhomogeneities to generate images independent of the scanner and parameter values and at clinically-feasible scan times (49). This method can be used to increase anatomical contrast and allow the accurate quantification of optic nerve T2 values. We have begun to implement this novel method in patients with optic neuritis and glaucoma, where subtle changes would not be seen on conventional imaging (Fig. 5). Like diffusion, T2 maps may detect optic nerve gliosis, demyelination, and axonal loss not visually-obvious on conventional MRI, but unlike diffusion there is ample SNR, the acquisition is fast, more resistant to susceptibility artifacts, and can be implemented even with older 1.5-T MRI scanners.
New MRI Acceleration Techniques
A major limitation to MRI of the optic nerve is motion—patient cooperation and acquisition speed are the key factors to successful imaging. A recently-developed SMS approach can provide 2- to 4-fold acceleration of diffusion acquisitions without degrading image quality (47). This SMS method theoretically can be applied to any 2D acquisition such as turbo spin echo. We recently applied SMS diffusion acquisitions to generate 500-μm super-resolution diffusion-based parameter maps of the in vivo human brainstem using Track Density Imaging (50,51) (Fig. 6). Compared with conventional 3-T MRI approaches, this approach provides direct visualization of optic tracts, oculomotor nuclear complex, and medial longitudinal fasciculus. Another promising method is compressed sensing reconstructions that take advantage of intrinsic image sparsity to undersample k-space potentially without meaningful compromise to image quality (52). With incoherent sampling and iterative nonlinear reconstructions, compressed sensing images can be generated with only 10%–20% k-space sampling and thus substantially reduced scan times while still providing images of sufficient diagnostic quality. This method has been combined with parallel imaging (53) and could serve as a rapid screening tool in the outpatient setting.
New Coils for Better Visualization
The sensitivity of surface coils falls off dramatically with increasing depth away from the patient. The move toward small multichannel surface coils often leads to dramatic improvements in SNR particularly for peripheral structures near the head surface. The disadvantage is nonuniformity of the image where the periphery is accentuated and signal from the center is attenuated. This can be corrected with postprocessing filters. The 32- and 64-channel coils now available commercially easily outperform coils that were state of the art a decade ago. We have explored the creation of new targeted orbit surface coils that do not overtly cover the eyes (Fig. 7). Specific, flexible multiple coil geometries targeted to the optic nerve and orbit can dramatically improve image quality (54,55). Coils can be designed for highly-specific indications and regions of the optic nerve if there is sufficient clinical justification.
Characterization of the optic nerve is a common indication for MRI in routine clinical practice. Orbital MRI is often used to exclude compressive lesions, inflammation or optic nerve tumors. The small size and oblique orientation of the optic nerve, motion and distortion from nearby air-filled structures pose inherent challenges to imaging the optic nerve well. It should be emphasized that high-quality conventional MRI of the optic nerve should first be optimized and put into clinical use. There also are promising new approaches using high-resolution sequences and advanced quantitative techniques that may prove helpful in the near-future for clinical diagnosis and treatment in individual patients. Emerging MRI technologies that emphasize rapid acquisition should improve visualization of the optic nerve and facilitate accurate quantification of MRI properties that can detect visually-occult pathology. As advanced MRI becomes more adept at characterizing the optic nerve, its role in clinical management will increase. There remains a strong need for collaborative, multidisciplinary teams of radiologists, imaging scientists, neurologists, and ophthalmologists to further explore and develop MRI of the optic nerve to improve clinical management in patients.
STATEMENT OF AUTHORSHIP
Category 1: a. Conception and design: T. M. Shepherd, M. J. Hoch, and M. T. Bruno; b. Acquisition of data: T. M. Shepherd, M. J. Hoch, and M. T. Bruno; c. Analysis and interpretation of data: T. M. Shepherd, M. J. Hoch, and M. T. Bruno. Category 2: a. Drafting the manuscript: T. M. Shepherd, M. J. Hoch, and M. T. Bruno; b. Revising it for intellectual content: T. M. Shepherd, M. J. Hoch, and M. T. Bruno. Category 3: a. Final approval of the completed manuscript: T. M. Shepherd, M. J. Hoch, and M. T. Bruno.
The authors thank Graham Wiggins, Noam Ben-Eliezer, Sohae Chung, Max Sale, and Jennifer Barger for their help with preliminary data used in this review. The authors also thank Eugene Hagiwara for providing a second case of ischemic optic neuritis.
1. Wu AY, Jebodhsingh K, Le T, Law C, Tucker NA, DeAngelis D, Oestreicher JH, Harvey JT. Indications for orbital imaging by the oculoplastic surgeon. Ophthal Plast Reconstr Surg. 2011;27:260–262.
2. Lee AG, Johnson MC, Policeni BA, Smoker WR. Imaging for neuro-ophthalmic and orbital disease—a review. Clin Exp Ophthalmol. 2009;37:30–53.
3. Simon J, Szumowski J, Totterman S, Kido D, Ekholm S, Wicks A, Plewes D. Fat-suppression MR imaging of the orbit. AJNR Am J Neuroradiol. 1988;9:961–968.
4. Atlas SW, Grossman RI, Hackney DB, Goldberg HI, Bilaniuk LT, Zimmerman RA. STIR MR imaging of the orbit. AJNR Am J Neuroradiol. 1988;9:969–974.
5. Shields JA, Shields CL, Scartozzi R. Survey of 1264 patients with orbital tumors and simulating lesions: the 2002 Montgomery Lecture, part 1. Ophthalmology. 2004;111:997–1008.
6. Johns TT, Citrin CM, Black J, Sherman JL. CT evaluation of perineural orbital lesions: evaluation of the “tram-track sign”. AJNR Am J Neuroradiol. 1984;5:587–590.
7. Aoki S, Barkovich AJ, Nishimura K, Kjos BO, Machida T, Cogen P, Edwards M, Norman D. Neurofibromatosis type 1 and 2: cranial MRI findings. Radiology. 1989;172:527–534.
8. Menor F, Marti-Bonmati L, Mulas F, Cortina H, Olague R. Imaging considerations of central nervous system manifestations in pediatric patients with neurofibromatosis type 1. Pediatr Radiol. 1991;21:389–394.
9. Azar-Kia B, Naheedy MH, Elias DA, Mafee MF, Fine M. Optic nerve tumors: role of magnetic resonance imaging and computed tomography. Radiol Clin North Am. 1987;25:561–581.
10. Taylor T, Jaspan T, Milano G, Gregson R, Parker T, Ritzmann T, Benson C, Walker D. Radiological classification of optic pathway gliomas: experience of a modified functional classification system. Br J Radiol. 2008;81:761–766.
11. Lebedis CA, Sakai O. Nontraumatic orbital conditions: diagnosis with CT and MR imaging in the emergent setting. Radiographics. 2008;28:1741–1753.
12. Kupersmith MJ, Alban T, Zeiffer B, Lefton D. Contrast-enhanced MRI in acute optic neuritis: relationship to visual performance. Brain. 2002;125:812–822.
13. Optic Neuritis Study Group. The clinical profile of optic neuritis: experience of the Optic Neuritis Treatment Trial. Arch Ophthalmol. 1991;109:1673–1678.
14. Suzuki H, Takanashi J, Kobayashi K, Nagasawa K, Tashima K, Kohno Y. MR. Imaging of idiopathic intracranial hypertension. AJNR Am J Neuroradiol. 2001;22:196–199.
15. Wippold FJ, for the Expert Panel on Neurologic Imaging. Orbits, vision, and visual loss: ACR appropriateness criteria. AJNR Am J Neuroradiol. 2010;31:196–198.
16. Lee AG, Brazis PW, Garrity JA, White M. Imaging for neuro-ophthalmic and orbital disease. Am J Ophthalmol. 2004;138:852–862.
17. Conneely MF, Hacein-Bey L, Jay WM. Magnetic resonance imaging of the orbit. Semin Ophthalmol. 2008;23:179–189.
18. Dodds NI, Atcha AW, Birchall D, Jackson A. Use of high-resolution MRI of the optic nerve in Graves' ophthalmopathy. Br J Radiol. 2009;82:541–544.
19. Wheeler-Kingshott CA, Parker GJ, Symms MR, Hickman SJ, Tofts PS, Miller DH, Barker GJ. ADC mapping of the human optic nerve: increased resolution, coverage, and reliability with CSF-suppressed ZOOM-EPI. Magn Reson Med. 2002;47:24–31.
20. Moodley A, Rae WID, Brijmohan Y, Miranda D, Connolly C, Michowicz A, Bhigjee A. The impact of optic nerve movement on optic nerve magnetic resonance diffusion parameters. S Afr J Rad. 2014;18. Art. #596, 6 pages.
21. Wheeler-Kingshott CA, Trip SA, Symms MR, Parker GJ, Barker GJ, Miller DH. In vivo diffusion tensor imaging of the human optic nerve: pilot study in normal controls. Magn Reson Med. 2006;56:446–451.
22. Abramoff MD, Van Gils APG, Jansen GH, Mourits MP. MRI dynamic color mapping: a new quantitative technique for imaging soft tissue motion in the orbit. Invest Ophthalmol Vis Sci. 2000;41:3256–3260.
23. Piccirelli M, Luechinger R, Rutz AK, Boesiger P, Bergamin O. Extraocular muscle deformation assessed by motion-encoded MRI during eye movement in health subjects. J Vis. 2007;7:5.1–10.
24. Held P, Nitz W, Seitz J, Frund R, Muller HM, Haffke T, Hees H. Comparison of 2D and 3D MRI of the optic and oculomotor nerve anatomy. Clin Imaging. 2000;24:337–343.
25. Chan KC, Tse Z, Jan N, Schuman JS, Kim S, Sigal IA. Magic angle enhanced MR microscopy of fibrous structures in the eye. Proc Intl Soc Mag Reson Med 21st Annual Scientific Meeting & Exhibition; April 20–26, 2013; Salt Lake City, UT, USA.
26. Chappell KE, Robson MD, Stonebridge-Foster A, Glover A, Allsop JM, Williams AD, Herlihy AH, Moss J, Gishen P, Bydder GM. Magic angle effects in MR neurography. AJNR Am J Neuroradiol. 2004;25:431–440.
27. Lin W, Guo J, Rosen MA, Song HK. Respiratory motion-compensated radial dynamic contrast-enhanced (DCE)-MRI of chest and abdominal lesions. Magn Reson Med. 2008;60:1135–1146.
28. Song HK, Dougherty L. Dynamic MRI with projection reconstruction and KWIC processing for simultaneous high spatial and temporal resolution. Magn Reson Med. 2004;52:815–824.
29. Chandarana H, Block TK, Rosenkrantz AB, Lim RP, Kim D, Mossa DJ, Babb JS, Kiefer B, Lee VS. Free-breathing radial 3D fat-suppressed T1-weighted gradient echo sequence: a viable alternative for contrast- enhanced liver imaging in patients unable to suspend respiration. Invest Radiol. 2011;46:648–653.
30. Bangiyev L, Raz E, Block TK, Hagiwara M, Wu X, Yu E, Fatterpekar GM. Evaluation of the orbit using contrast-enhanced radial 3D fat-suppressed T1 weighted gradient echo (Radial-VIBE) sequence. Br J Radiol. 2015;88:20140863.
31. Semelka RC, Kelekis NL, Thomasson D, Brown MA, Laub GA. HASTE MR imaging: description of technique and preliminary results in the abdomen. J Magn Reson Imaging. 1996;6:698–699.
32. Patel MR, Klufas RA, Alberico RA, Edelman RR. Half-Fourier acquisition single-shot turbo spin-echo (HASTE) MR: comparison with fast spin-echo MR in diseases of the brain. AJNR Am J Neuroradiol. 1997;18:1635–1640.
33. Lagrèze WA, Lazzaro A, Weigel M, Hansen HC, Hennig J, Bley TA. Morphometry of the retrobulbar human optic nerve: comparison between conventional sonography and ultrafast magnetic resonance sequences. Invest Ophthalmol Vis Sci. 2007;48:1913–1917.
34. Bley TA, Weigel M, Gaggl M, Munk R, Langer M, Hennig J, Lagrèze W. Quantification of glaucomatous optic atrophy utilizing high resolution MRI of the optic nerve. Proc. Intl. Soc. Mag. Reson. Med 16th Annual Scientific Meeting & Exhibition; May 3–9, 2008; Toronto, Ontario, Canada.
35. Mukherjee P, Berman JI, Chung SW, Hess CP, Henry RG. Diffusion tensor MR imaging and fiber tractography: theoretic underpinnings. AJNR Am J Neuroradiol. 2008;29:632–641.
36. Mathur S, Karimi A, Mafee MF. Acute optic nerve infarction demonstrated by diffusion-weighted imaging in a case of rhinocerebral mucormycosis. AJNR Am J Neuroradiol. 2007;28:489–490.
37. Chen JS, Mukherjee P, Dillon WP, Wintermark M. Restricted diffusion in bilateral optic nerves and retinas as an indicator of venous ischemia caused by cavernous sinus thrombophlebitis. AJNR Am J Neuroradiol. 2006;27:1815–1816.
38. Bender B, Heine C, Danz S, Bischof F, Reimann K, Bender M, Nagele T, Ernemann U, Korn A. Diffusion restriction of the optic nerve in patients with acute visual deficit. J Magn Reson Imaging. 2014;40:334–340.
39. Fatima Z, Motosuqi U, Muhi A, Hori M, Ishigame K, Araki T. Diffusion-weighted imaging in optic neuritis. Can Assoc Radiol J. 2013;64:51–55.
40. Garaci FG, Bolacchi F, Cerulli A, Melis M, Spano A, Cedrone C, Floris R, Simonetti G, Nucci C. Optic nerve and optic radiation neurodegeneration in patients with glaucoma: in vivo analysis with 3-T diffusion-tensor MR Imaging. Radiology. 2009;252:496–501.
41. Chang ST, Xu J, Trinkaus K, Pekmezci M, Arthur SN, Song SK, Barnett EM. Optic nerve diffusion tensor imaging parameters and their correlation with optic disc topography and disease severity in adult glaucoma patients and controls. J Glaucoma. 2014;23:513–520.
42. Kolbe S, Chapman C, Nguyen T, Bajraszewski C, Johnston L, Kean M, Mitchell P, Paine M, Butzkueven H, Kilpatrick T, Egan G. Optic nerve diffusion changes and atrophy jointly predict visual dysfunction after optic neuritis. NeuroImage. 2009;45:679–686.
43. Fieremans E, Benitez A, Jensen JH, Falangola MF, Tabesh A, Deardorff AT, Spampinato MVS, Babb JS, Novikov DS, Ferris SH, Helpern JA. Novel white matter tract integrity metrics sensitive to Alzheimer Disease Progression. AJNR Am J Neuroradiol. 2013;34:2105–2112.
44. Hickman SJ, Wheeler-Kingshott CA, Jones SJ, Miszkiel KA, Barker GJ, Plant GT, Miller DH. Optic nerve diffusion measurement from diffusion-weighted imaging in optic neuritis. AJNR Am J Neuroradiol. 2005;26:951–956.
45. Trip SA, Wheeler-Kingshott CA, Jones SJ, Li W, Barker GJ, Thompson AJ, Plant GT, Miller DH. Optic nerve diffusion tensor imaging in optic neuritis. NeuroImage. 2006;30:498–505.
46. Jeong H, Dewey BE, Hirtle JAT, Lavin P, Sriram S, Pawate S, Gore JC, Anderson AW, Kang H, Smith SA. Improved diffusion tensor imaging of the optic nerve using multishot two-dimensional navigated acquisitions. Magn Reson Med. 2015;74:953–963.
47. Setsompop K, Gagoski BA, Polimeni JR, Witzel T, Wedeen VJ, Wald LL. Blipped-controlled aliasing in parallel imaging for simultaneous multislice echo planar imaging with reduced g-factor penalty. Magn Reson Med. 2012;67:1210–1224.
48. Ben-Eliezer N, Sodickson DK, Block KT. Rapid and accurate T2 mapping from multi-spin-echo data using Bloch-simulation-based reconstruction. Magn Reson Med. 2015;73:809–817.
49. Ben-Eliezer N, Sodickson DK, Shepherd T, Wiggins G, Block KT. Accelerated and motion-robust in vivo T2 mapping from radially undersampled data using Bloch-simulation-based iterative reconstruction. Magn Reson Med. 2016;75:1346–1354.
50. Calamante F, Tournier JD, Jackson GD, Connelly A. Track-density imaging (TDI): super-resolution white matter imaging using whole-brain track-density mapping. Neuroimage. 2010;53:1233–1243.
51. Hoch MJ, Chung S, Ben-Eliezer N, Bruno MT, Fatterpekar GM, Shepherd TM. New-clinically feasible 3T MRI protocol to discriminate internal brain stem anatomy. AJNR Am J Neuroradiol. 2016;37:1058–1065.
52. Lustig M, Donoho D, Pauly JM. Sparse MRI: The application of compressed sensing for rapid MR imaging. Magn Reson Med. 2007;58:1182–1195.
53. Liang D, Liu B, Wang J, Ying L. Accelerating SENSE using compressed sensing. Magn Reson Med. 2009;62:1574–1584.
54. Georgouli T, James T, Tanner S, Shelley D, Nelson M, Chang B, Backhouse O, McGonagle D. High-resolution microscopy coil MR-eye. Eye (Lond). 2008;22:994–996.
55. Sirin S, Schlamann M, Metz KA, Bornfeld N, Schweiger B, Holdt M, Temming P, Schuendein MM, Goericke SL. High-resolution MRI using orbit surface coils for the evaluation of metastatic risk factors in 143 children with retinoblastoma: part 2: new vs. old imaging concept. Neuroradiology. 2015;57:815–824.