Imaging of the cochlea, cochlear nerve, brainstem, and auditory system is central to proper candidate selection and surgical planning in the setting of an auditory brainstem implantation. For the preoperative imaging of an auditory brainstem implant (ABI) candidate high-resolution computed tomography (CT) and magnetic resonance (MR) imaging provide complementary information. The imaging of an ABI candidate will not only help to fulfill the indication criteria but also help to assess the integrity of the auditory pathway from the brainstem up to the temporal cortex. Similarly imaging following implantation is again critical to confirm the appropriate electrode placement and, when needed, to evaluate possible complications.
Key wordsauditory brainstem implant – computed tomography – MR imaging – cochlear nerve – auditory pathway
3 Imaging of the Cochlea, Cochlear Nerve, Brainstem, and Auditory System
Imaging plays an important and indispensable role in the preoperative and postoperative assessment of cochlear and auditory brainstem implant (ABI) patients. The evaluation of the cochlea and cochlear nerve determines the eligibility of the patient for the cochlear versus auditory brainstem implantation. Additionally, the imaging of the posterior fossa as well as supratentorial structures is crucial for appropriate preoperative assessment of an ABI candidate. Following implantation, imaging is again required to confirm correct electrode placement as well as to evaluate possible complications; however, the implant itself becomes a source for artifact and potential hazard in a magnetic field. This chapter begins with a discussion of the radiological anatomy of the auditory pathway and will then address the important concepts related to preoperative and postoperative imaging of the ABIs.
3.2 Imaging Anatomy of the Inner Ear and Auditory Pathway
In modern-day practice the inner ear structures are evaluated with sectional imaging using computed tomography (CT) and magnetic resonance imaging (MRI). With either imaging technique, the cochlea appears as a spiral-shaped structure with 2.5 turns and a normal measured height of 5.1 mm (with a range of 4.4 to 5.9 mm). 1 The cochlear turns (basal, middle, and apical) are separated by interscalar septae, a bony plate radiating from the modiolus that forms the base of the cochlea (Fig. 3.1). The spiral lamina that also projects from the modiolus is a microanatomical osseous structure that separates the spiral of the cochlea into scala tympani (inferiorly), scala media, and scala vestibuli (superiorly). The spiral lamina can be seen with difficulty with conventional CT; however, it is easily appreciable with thin section MRI 2 (Fig. 3.2a). The scala tympani and scala vestibuli that are filled with fluid can be separated with high-resolution clinical 1.5 and 3 Tesla MRI; however, the scala media can only be visualized with high Tesla imaging. 3 , 4 The cochlear nerve passes from the internal auditory canal (IAC) to the modiolus through a bony canal called the cochlear aperture (or bony cochlear nerve canal) (Fig. 3.1). This “neck of the cochlea” has a normal measurable width of 1.9 mm (±0.24 mm). 2 , 5
3.2.2 IAC and the Vestibulocochlear Nerve
The size of IAC varies but the mean canal diameter is 4.21 ± 0.79 mm (with a range of 2–8 mm), and the two ears have almost symmetric size with a difference of up to 2 mm. 6 , 7 The vestibulocochlear and the facial nerve can only be well appreciated with high-resolution, high T2-weighted imaging (Fig. 3.2a). Although the axial slices can demonstrate the size of the IAC and help to evaluate the course of the vestibulocochlear nerve (VCN), the sagittal oblique images obtained perpendicular to the long axis of the IAC are best to distinguish each of the individual components of the VCN as the nerves are visualized in cross-section. 8 At the medial aspect of the IAC the VCN is seen as a crescent-shaped structure (Fig. 3.2b); however, more laterally, in the fundus, the three components of the VCN can be seen separately and the cochlear nerve lies in the anteroinferior aspect (Fig. 3.2c). 9 The normal size of the cochlear nerve on MRI measures 1.8 ± 0.2 mm at the porus acousticus and 1.2 ± 0.2 mm in the mid to distal IAC. 10 In the majority of cases, the cochlear nerve is larger than both the superior or inferior vestibular nerves. 11 The cochlear nerve is of similar size or larger than the facial nerve in more than half of the cases. 9
3.2.3 Cochlear Nucleus
MRI is better in delineating details of the brain anatomy but it is somewhat limited in the brainstem. 12 The difficulties in assessing the brainstem by using MRI arises not only from the small size of various brainstem structures, but also from the fact that those anatomical components do not exhibit enough contrast to enable their individual identification. 13 Therefore, when relaxation-based MR image contrast is used, despite high resolution, conspicuity of those structures such as cranial nerve nuclei cannot be achieved in clinical field strengths. 14 Nevertheless, the bulge of the medulla into the lateral recess of the fourth ventricle and to the foramen of Luschka caused by the cochlear nuclear complex can be easily identified by MRI (Fig. 3.2a). 15
3.2.4 Auditory Pathway
Morphologic Imaging Anatomy
Similar to the cochlear nuclear complex, the ascending fibers of the auditory pathway are not visible with normal visual inspection of routine MR sequences. The auditory radiation can only be demonstrated with dedicated fiber tracking obtained from the diffusion tensor imaging. 16 Functional MRI (fMRI) studies that allow noninvasive assessment of brain function can localize the auditory cortex (Fig. 3.3 ).
3.3 Preoperative Imaging of the Auditory Brainstem Implants
3.3.1 Imaging Techniques
For the preoperative imaging of ABI candidates, high-resolution CT and MRI provide complementary information.
CT remains the preferred examination for the initial evaluation of congenital sensorineural hearing loss (SNHL). CT of the temporal bone is able to delineate the detailed anatomy of the inner ear; but more importantly it can also help to evaluate the dimensions of the IAC and cochlear aperture to subsequently refer the patient for possible ABI placement. The two currently available and recommended CT scanners for the imaging of the temporal bone are multi-detector CT (MDCT) and cone-beam CT (CBCT).
MDCT has been and still is in most centers in the world the standard method to evaluate the temporal bone with CT. The image acquisition is performed in the axial plane; however, isotropic voxels allow reformatted images with high resolution in any additional plane. The imaging parameters are scanner specific but the collimation is usually 0.5 to 0.625 mm and has to be less than 1 mm. Images should always be processed with a bone algorithm and viewed with a window width of 4,000 HU and a window level of 200 to 500 HU. 17
Although MDCT is used worldwide, CBCT using flat panel detector technology is slowly taking over for detailed evaluation of the small temporal bone structures. 17 The CBCT uses a rotating gantry and a cone-shaped X-ray beam that generates three-dimensional volumetric dataset. 17 It results in higher resolution (0.15 mm thickness) with a lower dose. Additionally, it is less sensitive to metallic and beam hardening artifacts. It is however more sensitive to motion as the acquisition usually lasts for 40 seconds and anesthesia may be required for small children.
MRI is crucial in order to assess the cochlear nerve, the brain stem, and the integrity of the auditory pathways up to the temporal cortex. Again, the patient’s cooperation in limiting the motion is paramount and anesthesia is required for small children. The MRI should be performed with a 3.0 Tesla scanner whenever possible as higher field strength improves the signal to noise ratio (SNR) and increases the spatial resolution. 18 The MRI for the evaluation of an implant candidate should include high-resolution heavily T2-weighted sequence for a detailed evaluation of the membranous labyrinth, and especially for the assessment of the cochlear nerve. These sequences can be achieved with both gradient-echo (GRE) and fast spin-echo (FSE) T2-weighted techniques but the choice of which sequence to prefer is a heavily debated and published topic. 19 The most commonly used and widely available sequences include: constructive interference into steady state (CISS), fast imaging employing steady-state acquisition (FIESTA), driven equilibrium radio frequency reset pulse (DRIVE), 3D true-fast imaging with steady-state precession (FISP), and 3D T2 FSE or 3D T2 FSE with fast recovery (FRFSE) depending on the scanner vendor. The resolution of these heavily T2-weighted sequence should be increased with a slice thickness less than or equal to 1 mm. Although sagittal oblique reformatted images can be obtained from the axial dataset, bilateral direct sagittal oblique images, perpendicular to the IACs, with the same heavily T2-weighted sequence should always be acquired as the direct sagittal oblique images have a better resolution than the reformatted images. 8 The T2-weighted imaging of the entire brain is also required to assess the auditory pathway. 20 , 21 For neurofibromatosis type 2 (NF2) patients, postcontrast imaging of the entire brain is also required to look for additional intracranial extra-axial tumors.
3.3.2 Radiological Evaluation of the Auditory Brainstem Implant Candidates
Evaluation of Eligibility Criteria for Implantation in Congenital SNHL
The initial radiological assessment of an ABI candidate with congenital SNHL is indeed one of a cochlear implant (CI) candidate with ineligibility for the latter determining indication for the former. 22 , 23 The selected congenital inner ear anomalies that result in an ABI candidacy will be discussed in detail in another chapter, but from an imaging point of view not only the specific anomaly has to be detected and correctly labeled but also the size of the cochlear aperture and the presence of the cochlear nerve and its size should be evaluated. The structures that need to be carefully evaluated in a pre-implant imaging study are the following.
Cochlear hypoplasia represents a group of cochlear malformations which have dimensions less than normal cochlea with decreased number or height of the cochlear turns. 24 , 25 When the cochlea is hypoplastic the nerve may be also be hypoplastic or aplastic and needs to be assessed with MRI. 26 Similarly, common cavity anomaly, and type 1 and 2 incomplete partition anomalies have also been reported to have accompanying cochlear deficiencies including aplasia and need to be evaluated with MRI. 7 , 27 , 28
Atresia of the cochlear aperture is a strong indicator of underlying cochlear nerve anomaly. 7 , 29 Tahir et al reported that all 21 cases with cochlear aperture atresia in their series had accompanying cochlear nerve deficiency (either aplasia or hypoplasia). 7 The dimension of a patient’s cochlear aperture also needs to be assessed, as its diameter is a marker of the cochlear nerve status. 7 , 30 The cochlear aperture is considered stenotic when it is narrower than 1.4 mm (Fig. 3.4a). 5 , 31 , 32 , 33 , 34 In the series reported by Tahir et al, most of the stenotic cochlear aperture cases had accompanying hypoplasia/aplasia of the cochlear nerve with a normal size nerve in only 15% of the cases. 7 It is critical to realize that the aperture can be stenotic in the presence of a normal appearing and normal size cochlea; thus, a normal cochlear shape does not always indicate normal cochlear nerve structure and further imaging with MRI is required to assess the cochlear nerve status. 7
Internal Auditory Canal
The IAC is considered stenotic when the diameter at its midpoint is smaller than 2 mm. 35 The IAC stenosis or atresia may be easily demonstrated by CT. Although the exact measurement of the IAC diameter is more difficult with MRI, the high-resolution T2-weighted images can still demonstrate IAC hypoplasia or atresia with deceased or absence of high T2 signal of cerebrospinal fluid (CSF) within the IAC. Again the finding of a narrow or aplastic IAC raises concern for a deficiency of the cochlear nerve. 29 , 36 However, the IAC morphology is an unreliable surrogate marker of cochlear nerve integrity and as reported by Adunka et al, a normal IAC diameter can be seen in up to half of cochlear nerve aplasia patients. 7 , 37
The evaluation of the VCN and especially its cochlear branch is of extreme importance prior to cochlear implantation. In a normal-sized IAC the diagnosis of cochlear nerve aplasia is relatively straightforward with dedicated sagittal oblique high-resolution images (Fig. 3.4b). 38 However, in a very stenotic IAC, the diagnosis may be difficult because of the inability to separate the nerves. 35 , 37 Again the visualized inner ear may be normal or have subtle abnormality despite severe deficiency of the cochlear nerve. 38 , 39 Differentiation between hypoplasia and a normal size of the cochlear branch can also be challenging and require the highest possible resolution. 39 There isn’t a well-defined consensus regarding the definition cochlear nerve hypoplasia. Li et al defined the cochlear nerve hypoplasia as a cochlear nerve with a diameter smaller than that of the facial nerve, seen on the oblique sagittal images. 29 Similarly, Glastonbury et al designated the cochlear nerve as small when it appeared decreased in size compared with the other nerves of the IAC. 39 It is critical to recognize that there might be occasional discrepancy between the imaging and audiological findings regarding the presence/functionality of the cochlear nerve. 40 , 41 Several studies have shown that subsets of patients with cochlear nerve aplasia have positive audiological responses and might derive benefit from cochlear implantation. 40 , 41 , 42 Anatomical connections between the cochlear nerve and other branches of the vestibulocochlear complex that are below the resolution of the current MRI might be responsible for this radiological–audiological inconsistency. 43 Imaging with ultra-high field magnets with diffusion tensor imaging (DTI) fiber tractography might solve this problem in the future. 3 , 44
Brainstem and Supratentorial Brain
In every patient who is a candidate for an ABI placement, the imaging of the brainstem and supratentorial brain structures with MRI is crucial not only to verify the integrity of the auditory pathways up to the temporal cortex but also to determine possible underlying congenital or acquired malformations that might hinder post implant rehabilitation. 22 , 45 There is significant variability of the anatomy of the lateral recess in children with congenital deafness due to abnormalities of embryonic and fetal development. 46 It has been previously reported that congenital developmental abnormalities of the brain are more common in patients with auditory neuropathy spectrum disorder. 21 , 46 , 47 In patients with bilateral cochlear nerve deficiency, hindbrain anomalies such as pontine hypoplasia were reported to be the most common abnormal intracranial finding. 46 Additionally, there might be evidence of central pathologies such as chronic changes of hypoxic-ischemic injury, kernicterus, and chronic changes of congenital central nervous system (CNS) infections. 48 , 49 White matter lesions are also common findings in the preimplant imaging of the CI/ABI candidates. 21 , 50 These lesions are nonspecific, and more diffuse and prominent parenchymal changes were found to represent negative prognostic factors for speech and language development. 20 , 49 , 50 It is therefore critical to make a comprehensive evaluation of the brainstem and cerebrum in each ABI candidate. Additionally, with new developing technologies, MRI also has the potential to study the anatomical and functional organization of the auditory cortex through voxel-based morphometry and fMRI. 51 DTI metrics, such as fractional anisotropy, may prove important in selecting patients and predicting outcomes after the implantation. 52
Evaluation of Eligibility Criteria for Implantation in Acquired SNHL
The radiological evaluation of an ABI candidate with acquired SNHL is different from the cases with congenital SNHL as the imaging is more focused on the detection of possible surgical challenges and the following structures should be individually assessed.
The posterolateral medulla where the cochlear nuclear complex is located should have no appreciable signal changes. Post-traumatic or postoperative encephalomalacic changes as well as potential injury from prior radiation therapy are important considerations for this location. 53 , 54 Similarly, ischemic changes involving the brainstem and especially posterolateral medulla should be absent. 55 Although currently T2-weighted imaging is the recommended method of assessment for pontomedullary junction, in the future DTI may enable depiction of more subtle anomalies such as early detection of radiation-induced changes on auditory pathways in patients with vestibular schwannomas and might predict the success of the implantation preoperatively. 44 , 56
The Lateral Recess
The site of electrode placement should be normal in size without asymmetric widening as an enlargement at this location might contribute to the migration or rotation of the electrode array (Fig. 3.5 ). 57 Additionally, in NF2 patients, the presence or previous resection of tumors may result in distortion of the brainstem and the lateral recess may be difficult to identify. 58 In the setting of scar tissue from previous surgery, visible as obliteration of normal CSF signal, accompanying enhancement might be seen at the site of contemplated implantation and may result in distortion of the normal anatomy.
Anomalous lower cranial nerves and vascular anomalies around the lateral recess are potential variations that need to be looked for as those findings might prevent successful ABI placement or might cause unexpected surgical difficulties. 59 This assessment is best done with high-resolution heavily T2-weighted sequences. It has also been reported that previous meningitis might lead to excessive bleeding and increased surgical difficulties. It is thus important to evaluate postcontrast T1-weighted images for possible increased leptomeningeal enhancement. 60
In NF2 patients additional supratentorial masses such as meningiomas or meningiomatosis as well as parenchymal signal abnormalities should beinvestigated. 61 Similarly, possible accompanying traumatic encephalomalacia of the auditory cortex or sequel of shearing injury should be looked for in the setting of an ABI candidate with bilateral post-traumatic deafness. 60 , 62
3.4 Postoperative Imaging of the Implanted Patient
3.4.1 Imaging Issues after ABI
Following implantation, imaging is usually performed to determine the location of the electrode array to check for appropriate electrode array placement. 63 Additionally, a brain imaging might be required to investigate immediate postoperative complications or for the follow-up of an NF2 patient. 64
Plain X-ray films in anteroposterior (AP) and modified Stenvers’ views are the standard methods of imaging following ABI surgery and are sufficient in a majority of patients to depict the integrity and positioning of electrode array and to detect electrode kinking (Fig. 3.6 ). The evaluation on those plain films were standardized by Cerini et al on lateral and AP views, where angles rather than distances from specific landmarks were used to establish correct electrode positioning. 65 However, only limited information concerning the exact positioning of the electrode array can be extrapolated from plain films due to the innate planar representation. 66
CT evaluation is especially useful when postoperative radiographs fail to properly demonstrate the location of the electrode array or if a postoperative complication is suspected. 65 CT is suitable to assess the integrity of the device. Although the soft-tissue algorithm adequately demonstrates brainstem structures, it fails to distinguish the metallic electrodes and wires (Fig. 3.7a). Correspondingly, the bone algorithm delineates the electrodes well, but accompanying streak artifact limits the determination of the implant positioning with respect to surrounding soft tissues (Fig. 3.7b). 65 To overcome this limitation Lo et al suggested a superposition technique with inverted bone windows that were superimposed onto soft-tissue windows to better delineate the electrodes in relation to surrounding soft-tissue structures. 67 Advanced techniques such as View Angle Tilting (VAT) and Slice-Encoding for Metal Artefact Reduction (SEMAC) are also being developed to address the device artifacts with promising results. 66 CT is usually the adequate and the recommended imaging for the evaluation of potential early postoperative complications such as hemorrhage or CSF leak. 22