Published online Dec 31, 2021.
https://doi.org/10.13104/imri.2021.25.4.281
Cranial Nerve Disorders: Clinical Application of High-Resolution Magnetic Resonance Imaging Techniques
Abstract
Cranial-nerve disorders can be caused by a wide spectrum of diseases, including congenital, inflammatory, and tumorous diseases, and are often encountered in practice. However, the imaging of cranial-nerve disorders is challenging, and understanding the anatomical differences of each region is essential for conducting the best protocols and for detecting subtle changes in cranial nerves during magnetic resonance imaging (MRI) examinations. In this review we discuss which MRI techniques are best for observing normal and pathologic appearance, according to the different regions of the cranial nerves.
INTRODUCTION
Cranial-nerve disorders are frequently encountered in practice and can be caused by a wide spectrum of diseases, including congenital and inflammatory diseases and tumors of the cranial nerves, or by disease propagation from the adjacent structures of the brain, skull base, or other craniofacial structures. Given its excellent soft-tissue contrast, MRI is the best imaging modality for investigating cranial-nerve disorders (1, 2). Although various methods have been used to identify cranial-nerve disorders, their imaging remains challenging because of the small caliber of cranial nerves and the complex regional anatomy (3).
With recent developments in MRI techniques, visualization of the normal anatomy of cranial nerves or their pathologic changes became feasible because of the high spatial resolution, isotropic three-dimensional imaging, and fast image acquisition of these techniques. Guided by medical history and the results of neurological examination, radiologists should set the coverage range appropriately and select optimal sequences when conducting MRI examinations for patients with suspected cranial-nerve abnormalities.
Cranial nerves comprise 12 pairs of symmetric nerves that originate from the cerebrum and the brainstem (4). The peripheral cranial nerves consist of ten cranial nerves, from the oculomotor nerve to the hypoglossal nerve (2). Olfactory and optic nerves are direct parenchymal extensions from the telencephalon and diencephalon, respectively (5). These ten peripheral cranial nerves pass through similar anatomical structures, from the nuclei (located in the brainstem), through the skull base, and downwards to extracranial structures (2). When doing cranial-nerve-dedicated MRI examinations, one must understand the anatomical characteristics of each region in order to use the best protocols for detecting subtle changes in the cranial nerves (1, 2). In this review, we will discuss appropriate MRI techniques, focusing on various regions of the peripheral cranial nerves.
Voxel Size and Coverage
Although large cranial nerves can be partly seen even in routine MRI sequences (6, 7), few of them are usually visible with conventional MRI, because of their small caliber (4). These small structures pass through the cisternal space of the posterior fossa to the skull base, and therefore tend to be more susceptible to artifacts related to pulsation from blood vessels and cerebrospinal fluid (CSF) spaces and from the air-bone interface. Dedicated high-resolution MRI is essential to overcome these challenges (2, 4). High-resolution, thin-section imaging enables full-length tracing of the small cranial nerves. The visualization of the trochlear nerve (CN IV) is especially challenging because of its small caliber and proximity to multiple vascular structures (8). Previous studies have shown that carefully conducted high-resolution imaging (0.3 mm × 0.3 mm × 0.25 mm), with a scan plane set to an oblique axial direction perpendicular to the long axis of the cerebral aqueduct, can make the cisternal segment of the CN IV visible better than can conventional high-resolution images (0.67 mm × 0.45 mm × 1.4 mm) (9). In practice, increased spatial resolution is better for visualizing small cranial nerves, although it involves a trade-off with reduced scan coverage, decreased signal-to-noise ratio, and image degradation associated with long acquisition times (3). Therefore, image acquisition and reconstruction techniques, including parallel imaging and compressed sensing, can be applied to reduce scan time while preserving image quality (10, 11).
Multiple sub-millimetric three-dimensional isotropic images must be acquired with 0.6-mm voxels for cranial-nerve MRI (2, 4). Moreover, the scan range should be focused to cover the areas of interest chosen in terms of the patients’ history and neurological signs (2). In our institution, routine high-resolution cranial-nerve imaging consists of three-dimensional heavily T2-weighted (T2W) imaging; three-dimensional contrast-enhanced T1-weighted (T1W) imaging; and non-enhanced and enhanced three-dimensional T2 fluid-attenuated inversion recovery (FLAIR) with axial and coronal reconstruction. Multiplanar reconstruction is additionally done when indicated. In selected cases, CN IV sequence or special sequences for extracranial segments of cranial nerves are added to conventional cranial-nerve protocols (9, 12).
Visualization of Normal Cranial Nerves
Cisternal and Dural Segments
Because the cisternal segment of the cranial nerve is surrounded by CSF, sub-millimetric fluid-sensitive sequences are well suited for the visualization of the cisternal positions of the cranial nerves (1, 2). High-resolution three-dimensional T2W sequences show dark cranial nerves against the background of the bright CSF and offer high resolution with a section thickness of less than 1 mm and voxel volume of less than 1 mm3 (12, 13, 14, 15) (Fig. 1). These sequences could be obtained by using fast gradient-echo or fast spin-echo (FSE) techniques. Fast gradient-echo sequences (true fast imaging with steady-state free precession [FISP] fast imaging employing steady-state acquisition [FIESTA], or balanced fast-field echo [b-FFE]) are typically susceptible to dephasing caused by small field inhomogeneities, resulting in banding artifacts. These artifacts could be reduced by doing two acquisitions with and without phase frequency alternations in the constructive interference into steady-state and fast imaging, employing steady-state acquisition with phase cycling sequences (constructive interference in steady state [CISS], FIESTA-C) (13). However, they typically involve longer scanning times and can be susceptible to blurring (16). One can use three-dimensional isotropic T2 fast spin-echo techniques to visualize the cisternal segment of cranial nerves, e.g., CUBE, volumetric isotropic turbo spin-echo acquisition (VISTA), or sampling perfection with application-optimized contrasts by using different flip-angle evolutions (SPACE). These approaches were developed with short, non-spatially selective radiofrequency pulses, very long echo-train length, and variable flip angles for refocusing radiofrequency pulses (suppressing blurring while reducing flow and chemical shift artifacts). These techniques enable high-resolution isotropic imaging (up to 0.3 mm) with a high contrast-to-noise ratio (17, 18). These three-dimensional T2W images have become routine in otologic imaging for demonstrating labyrinthine anomalies (19, 20), detection or differentiation of labyrinthitis ossificans (21), intralabyrinthine schwannomas (22), or the preoperative work-up of cochlear implantation (to evaluate the size of the cochlear nerve and cochlea as well as the patency of the cochlear aperture) (23, 24). High-resolution cisternographic images are the modality of choice in neurovascular compression, enabling vascular contact in the transition zone. Three-dimensional data offer benefits for multiplanar reconstruction in visualizing the culprit vessel by showing the full course of the transition zones of various cranial nerves (25, 26, 27, 28). Because three-dimensional T2W images provide the highest resolution among three-dimensional imaging sequences, this sequence is the modality of choice for patients with congenital trochlear nerve palsy in order to detect congenital aplasia or hypoplasia of the trochlear nerve (29, 30).
Fig. 1
Cranial nerves in the cisternal segment on high-resolution three-dimensional T2 weighted sequences. (a) Oculomotor nerves (white arrows). (b) Trochlear nerves (white arrows). (c) Trigeminal nerve (white arrow), a meningioma in the left Meckel's cave (asterisk). (d) Abducens nerves entering the Dorello's canal (white arrows). (e) Facial nerve (white arrow). (f) Vestibulocochlear nerve on axial (left) and oblique sagittal reconstruction image (right), cochlear nerve (white arrows), vestibular nerves (black arrowheads), facial nerve (asterisk). (g) Glossopharyngeal nerve (white arrow). (h) Vagus and spinal accessory nerves (left, axial; right, oblique sagittal), vagus nerve-spinal accessory nerve complex (white arrow on axial), glossopharyngeal nerve (black arrow), vagus nerve (white arrowhead), and spinal accessory nerve (black arrowhead). (i) Hypoglossal nerve (white arrow).
Interdural and Foraminal Segments
The dura mater consists of two layers, the outer periosteal layer and the inner meningeal or cerebral layer. The outer layer adheres to the inner table of the skull and is continuous with the outer periosteum through sutures and the foramina. The inner layer is usually fused with the outer layer; however, there are interdural spaces in which these two layers are not fused, such as in regions with dural venous sinuses (1, 2). The cavernous sinus and petroclival venous plexus are examples of the interdural spaces.
Except for the vestibulocochlear nerve, most of the cranial nerves exit the cranium and run into extracranial structures via the skull-base foramina. The foraminal segment of a cranial nerve is defined from the inner margin of the internal orifice of the skull-base foramen to the outer margin of the outer table at the external orifice. Because the cranial nerve and abundant circumneural venous plexus exist in the foraminal segment of cranial nerves, normal nerves appear as non-enhanced filling-defect-like structures, surrounded by the prominently enhancing venous plexus on MRI (31) (Fig. 2). Therefore, three-dimensional high-resolution contrast-enhanced T1W images could be used to visualize normal cranial nerves in the interdural or foraminal segment. The mainstay of three-dimensional sequences consists of ultrafast spoiled GE sequences with small flip angles and short repetition times (1). Isotropic data offer various benefits for multiplanar reconstruction in the complex skull base (32). The sequences for acquiring these data typically have pre-pulses (magnetization preparation) to provide T1W and T2W or fat suppression and can be combined with partial k-space sampling to further reduce acquisition times. Examples include volumetric interpolated breath-hold examination (VIBE) from Siemens and T1W as well as high-resolution isotropic volume examination from Philips (33). Special FSE sequences, such as VISTA and SPACE, can be used as alternatives to fast GE sequences to produce high-resolution fat-suppressed T1W imaging (34). They are also resistant to susceptibility artifacts, which is a useful characteristic at the skull base. Another advantage of these sequences arises from their inherent flow suppression caused by the dephasing of flowing spins, which can prevent vascular time-of-flight effects from simulating lesions at the skull base (35).
Fig. 2
Cranial nerves in the foraminal and interdural segments on high-resolution contrast-enhanced T1 weighted sequences. (a) Cranial nerves within the cavernous sinus, oculomotor nerve (asterisk), ophthalmic nerve (white arrowhead), maxillary nerve (black arrowhead), mandibular nerve (curved white arrow), and abducens nerve (black arrow). (b) Lower cranial nerves at the jugular foramen, glossopharyngeal nerve (white arrowhead), vagus nerve (black arrowhead), spinal accessory nerve (curved white arrow), jugular spine (asterisk).
Mixed-weighting steady-state free precession (SSFP) or reversed fast imaging with steady-state precession (PSIF) and contrast administration can be also used to visualize nerves in the foraminal segments with the advent of high spatial resolution and T1 shortening effect after contrast administration. Normal cranial nerves can be well visualized inside the cavernous sinus, surrounded by bright contrast-enhanced venous blood (36). However, gadolinium-enhanced T1-weighted sequences have limitations for distinguishing pathologic from normal enhancement in some cranial nerves in the interdural or foraminal segments, including the trigeminal ganglion within the Meckel’s cave (2, 37) and the geniculate, tympanic, or mastoid segment of the facial nerve (38, 39).
Extraforaminal Segment
The extraforaminal segment includes parts starting from the outer cortex of the outer margin of the skull-base foramen (1, 2). The cranial nerve passes through different types of soft tissue, including fat, vessels, glands, muscles, and viscera, until it reaches its final destination (1). The normal extraforaminal cranial nerve could appear as a hypointense tubular structure within hyperintense fatty tissue on high-resolution T1WI without fat suppression (2). However, for detecting pathologic signal changes of enhancement of the cranial nerves, fat suppression is mandatory to maximize the tissue contrast between the suppressed fatty tissue and the hyperintense nerve (2). Sequences such as CISS and FIESTA, which are fully refocused (balanced) steady-state sequences that demonstrate both T2 and T1 components, could help the differentiation of nerves with high resolution from tissues having various properties (1). Specifically, several reports have shown the value of special high-resolution sequences for extracranial cranial-nerve imaging, such as three-dimensional PSIF with diffusion-weighted imaging (DWI) or three-dimensional double echo steady-state with water excitation (3D DESS-WE). The DESS sequence involves the acquisition of two different echoes: the gradient echo used in the FISP sequence and spin echo used in the PSIF sequence. The PSIF signal accentuates the signal intensities of components with long T2, whereas the FISP signal intensity provides more anatomic details with a contrast dominated by the T1/T2 ratio. In the WE technique, only water is excited by using section-selective composite pulses, thus avoiding disruptions of the steady state and ensuring uniform fat suppression. When DESS is combined with WE, the facial nerve is clearly shown without relying on the fatty-tissue background. The DESS-WE sequences also demonstrate flow sensitivity, because of which stationary fluids (such as saliva in parotid ducts) show high signal intensity with suppressed carotid artery flow. This technique works well for visualizing the intraparotid facial nerve and extracranial trigeminal nerve (40, 41, 42) (Figs. 3, 4).
Fig. 3
Extraforaminal segment of facial nerve on double-echo steady-state sequence with water excitation (DESS-WE) sequence (left, sagittal; right, axial, arrows).
Fig. 4
A 35-year-old patient with a mucoepidermoid carcinoma in the left parotid gland. A double-echo steady-state sequence with water excitation (DESS-WE) sequence with coronal reconstruction shows the relationship between the main trunk of the facial nerve (thin arrows) and the tumor.
The advantages of three-dimensional PSIF-DWI arise from its dominant T2 contrast of reversed FISP and the accentuation of anisotropic diffusion characteristics of peripheral nerves in combination with DWI (43, 44). The use of three-dimensional PSIF-DWI could achieve excellent nerve-muscle contrast with blood suppression and simultaneously display the intraparotid facial nerve and parotid duct with excellent soft-tissue contrast.
Imaging of Pathologic Signals of Cranial Nerves
Traditionally, gadolinium-enhanced T1W spin-echo and three-dimensional gradient echo sequences were used to evaluate the pathologic enhancement in cranial nerves. However, because of the presence of the circumneural venous plexus, these techniques have shown significant limitations by inhibiting the evaluation of the real pathologic enhancement of the nerve caused by inflammatory or neoplastic breakdown of the blood-nerve barrier (38, 39, 45).
In the last decade, isotopic three-dimensional T2W FLAIR, using three-dimensional turbo spin-echo sequences with variable flip angles (VISTA, SPACE, or CUBE), was introduced on the 3T MR system (34). This sequence uses variable refocusing flip angles to constrain T2 decay over a long echo train with minimal blurring and can acquire isotropic resolutions in a clinically acceptable scan time. This ability of the three-dimensional high-resolution FLAIR sequence to find minute abnormalities mainly results from FLAIR imaging’s high sensitivity to subtle changes in longitudinal magnetization in the fluid space. This sensitivity declines with increasing concentration of the contrast agent in a manner that is different from that of other T1W images (46, 47). The three-dimensional-FLAIR sequence also suppresses the signal intensity from the fluid flowing at a velocity exceeding 1.0 cm/s (48). Three-dimensional FLAIR sequences provide exquisite CSF suppression with fewer CSF-related flow artifacts in the posterior fossa, while enabling visualization of the T2 prolongation of the affected lesion (49, 50) (Figs. 5, 6, 7), . Previous studies have shown the advantage of three-dimensional FLAIR sequences for detecting pathologic signal alterations and abnormal enhancement of the nerve itself, which could be beneficial for the diagnosis of various cranial-nerve pathologies, including facial neuritis (51), vestibular neuritis (52, 53), Ramsay-Hunt syndrome (54), and antiGQ1b antibody syndrome (55). The role of high-resolution three-dimensional FLAIR imaging has also been highlighted in otologic imaging to detect abnormal labyrinthine signal changes in the onset of sudden sensorineural hearing loss (56, 57, 58, 59) and in vivo imaging of endolymphatic hydrops (60, 61, 62). Table 1 summarizes the indications and representative sequences of cranial-nerve MRI.
Fig. 5
A 50-year-old patient with chronic inflammatory demyelinating polyneuropathy. Bilateral facial nerves show contrast enhancement, which is more visible on contrast-enhanced fluid attenuated inversion recovery (FLAIR) (arrows in the upper row) than on a contrast-enhanced T1 weighted image (arrows in the lower row). Thickened and enhanced trigeminal nerves (black asterisks).
Fig. 6
A 57-year-old woman with anti-GQ1b antibody syndrome, presenting with diplopia, areflexia, and ataxia. Precontrast three-dimensional fluid attenuated inversion recovery (FLAIR) image (a) shows mild hyperintensity of both oculomotor nerves (arrows). Post three-dimensional FLAIR image (b) shows multifocal enhancement of the left oculomotor nerve (arrow) including the neuromuscular junction (arrowhead). Three-dimensional CE T1-turbo field echo (TFE) (c) shows long segmental enhancement of both oculomotor nerve (arrows).
Fig. 7
A 84-year-old woman with invasive fungal sinusitis, with headache, fever, and diplopia. Postcontrast three-dimensional fluid attenuated inversion recovery (FLAIR) shows prominent enhancement of the right abducens nerve (thin arrows) when it enters the Dorello’s canal, whereas three-dimensional CE-T1 turbo field echo (TFE) and precontrast FLAIR show equivocal findings (upper row). High-resolution images show the thickening and enhancement of the right oculomotor (black arrowhead) and maxillary nerves (white arrowheads) (asterisks, fungal ball in the left sphenoid sinus, lower row).
Table 1
Recommended Pulse Sequences for Cranial Nerve MRI according to Clinical Situation and Cranial Nerve Segments
Recent Advances in Cranial-Nerve Imaging
Cranial-nerve imaging requires a relatively long scan time to render small cranial nerves visible, and these are often accompanied by motion artifacts associated with long scan time (3). Recent imaging techniques, including parallel imaging and compressed sensing (CS), have been introduced to reduce the scan time while preserving acceptable image quality (11). Conventional parallel-imaging techniques, such as sensitivity encoding (SENSE) and generalized autocalibrating partial-parallel acquisition (GRAPPA), have been widely used in cranial-nerve imaging. However, these have had limitations in reducing scan time and having a good image quality/scan range (63, 64). CS is a recently introduced technique in MR imaging, characterized by non-uniform sampling of k-space, iterative reconstruction, and reduced scan time (65, 66, 67). A recently published study has shown the benefit of combining parallel imaging and CS (such as in SENSE-CS) in patients with suspected neurovascular compression to reduce scan time yet have acceptable image quality (10).
In conclusion, recent advances in MRI technology have enabled anatomical imaging of small cranial nerves that could not have been seen on conventional brain MR imaging. In this review, we have described the best MRI techniques for observing normal and pathologic changes in the cranial nerves for different regions of the cranial nerves. Radiologists should be familiar with the normal anatomy and basic sequences for cranial-nerve imaging in order to guide optimized imaging, which will eventually lead to improved patient management.
References
-
Kim HJ, Seong M, Kim Y. Normal anatomy of cranial nerves III-XII on magnetic resonance imaging. J Korean Soc Radiol 2020;81:501–529.
-
-
Kim TK, Kim HY, Yu IK, Son HJ, Chang DS, Jang YD. Schwannoma of the tongue base with imaging features and differential diagnosis: a rare case report and literature review. Investig Magn Reson Imaging 2019;23:385–389.
-
-
Seo M, Choi Y, Lee S, et al. Diagnostic value of susceptibility-weighted MRI in differentiating cerebellopontine angle schwannoma from meningioma. Investig Magn Reson Imaging 2020;24:38–45.
-
-
Cho SJ, Choi YJ, Chung SR, Lee JH, Baek JH. High-resolution MRI using compressed sensing-sensitivity encoding (CS-SENSE) for patients with suspected neurovascular compression syndrome: comparison with the conventional SENSE parallel acquisition technique. Clin Radiol 2019;74:817.e9–817.e14.
-
-
Kim JH, Hwang JM. Usefulness of MR imaging in children without characteristic clinical findings of Duane's retraction syndrome. AJNR Am J Neuroradiol 2005;26:702–705.
-
-
Williams LS, Schmalfuss IM, Sistrom CL, et al. MR imaging of the trigeminal ganglion, nerve, and the perineural vascular plexus: normal appearance and variants with correlation to cadaver specimens. AJNR Am J Neuroradiol 2003;24:1317–1323.
-
-
Kato Y, Higano S, Tamura H, et al. Usefulness of contrast-enhanced T1-weighted sampling perfection with application-optimized contrasts by using different flip angle evolutions in detection of small brain metastasis at 3T MR imaging: comparison with magnetization-prepared rapid acquisition of gradient echo imaging. AJNR Am J Neuroradiol 2009;30:923–929.
-
-
Yagi A, Sato N, Taketomi A, et al. Normal cranial nerves in the cavernous sinuses: contrast-enhanced three-dimensional constructive interference in the steady state MR imaging. AJNR Am J Neuroradiol 2005;26:946–950.
-
-
Maeda M, Tsuchida C. "Ivy sign" on fluid-attenuated inversion-recovery images in childhood moyamoya disease. AJNR Am J Neuroradiol 1999;20:1836–1838.
-
-
Fukuoka H, Hirai T, Okuda T, et al. Comparison of the added value of contrast-enhanced 3D fluid-attenuated inversion recovery and magnetization-prepared rapid acquisition of gradient echo sequences in relation to conventional postcontrast T1-weighted images for the evaluation of leptomeningeal diseases at 3T. AJNR Am J Neuroradiol 2010;31:868–873.
-
-
Naganawa S, Koshikawa T, Nakamura T, et al. Comparison of flow artifacts between 2D-FLAIR and 3D-FLAIR sequences at 3 T. Eur Radiol 2004;14:1901–1908.
-