In the early days of performing CI, the main imaging modality for CI candidates was temporal bone computed tomography (TBCT), which focused on assessing the condition of the inner ear, including labyrinthine ossification or congenital malformation, and the experience with magnetic resonance imaging (MRI) was limited [11, 12]. With the improvement of MRI resolution, four nerves in the IAC, the cochlear nerve, facial nerve, and superior and inferior vestibular nerves, can be clearly identified, and CND has been identified as an important cause of congenital SNHL [13].

Most children with CND have severe-to-profound hearing loss that presents as either SNHL or auditory neuropathy spectrum disorder (ANSD) [14, 15]. CND may be present in up to 18 % of children with SNHL [14] and in 30 % of those with ANSD [15]. A higher incidence of CND has been reported in children with IEMs [1, 3], stenosis of the IAC or cochlear nerve canal (CNC) [16, 17], and ANSD [15, 18]. A study including 59 children with IEMs showed that 19.6 % of them had CND and that the severe forms of IEMs such as common cavity and cochlear aplasia were associated with a much higher incidence of CND [1]. Stenosis of the CNC or IAC is also a strong indicator of CND [16, 17, 19]. A study including 54 ears with CND showed that 36 (66.7 %) had CNC stenosis (<1.5 mm) and 25 (46.3 %) had IAC stenosis (<3.0 mm) [16]. In another study including 42 ears with SNHL, none of the 32 ears with normal CNC had CND but 8 of 10 ears with CNC stenosis (<1.5 mm) had CND [17]. About one-third of children with ANSD have been reported to have CND [15, 18]. A study including 14 children with ANSD showed that five children (35.7 %) had CND [18], and another study reported that 15 of 54 children (28 %) with ANSD had CND [15]. The incidence of CND is low in hearing-impaired children with normal inner ear morphology [3, 17], but on rare occasions, CND can be present in cases where the structure of the inner ear, CNC, and IAC are normal [20].

CND is challenging for the decision-making process about whether to proceed with CI, because this condition may result in very poor speech perception or even anacusis after CI [14, 20].

The range of CND includes cochlear nerve hypoplasia and cochlear nerve aplasia. CND is diagnosed using MRI of the IAC [21]. The cochlear nerve is considered to be normal if on the MR image it is the same size or larger than the other nerves in the IAC, including the facial nerve and superior and inferior vestibular nerves. The cochlear nerve is considered to be hypoplastic (small) if it is smaller than the other nerves in the IAC and to be aplastic (absent) if it is not seen on MRI [19]. Heavily T2-weighted gradient echo or turbo-spin echo sequences should be used, and images of the axial and parasagittal planes must be acquired with submillimetric slice thickness to assess accurately the size of the cochlear nerve [22].

Children with CND meet the audiometric criteria for CI because most of them have severe-to-profound hearing loss [20]. Although there has been a perception that pathology of the cochlear nerve precludes deaf children from receiving CI, the surgery has been performed for children with CND based on the rationale that children with cochlear nerve hypoplasia can benefit from CI [23] and that some children with cochlear nerve aplasia may have cochlear nerve fibers that are below the resolution of current imaging modalities [14].

However, the outcomes of CI performed for children with CND are unreliable and extremely variable [23–27]. A study that reported the outcome of CI performed in six children with CND showed that five of the children could detect only the presence of sounds and one child could understand some common phrases [24]. All the children achieved very poor speech production at grade 2 speech intelligibility rating (SIR). Many studies of CI outcomes in children with CND have shown that the majority of children with CND obtained very limited benefit after CI, such as improved access to environmental sounds, and that only sporadic cases could acquire open-set speech perception and spoken language [23–27]. In children with CND, therefore, CI should be considered carefully after patients and family members have been fully informed about the uncertain speech and language outcomes.

It is very difficult to predict the outcome of CI prior to surgery in children with CND, but high-resolution MRI and electrical auditory brainstem response (EABR) testing can help. A study including 139 implanted children with CND showed that children with cochlear nerve hypoplasia had a good outcome after CI [23]. In that study, the cochlear nerve aplasia group (one-fifth of the children) showed significantly worse scores on categories of auditory performance (CAP) and SIR after CI than the non-CND group. However, no significant difference was noted between the cochlear nerve hypoplasia group (four-fifths of the children) and the non-CND group. Although MRI is the best imaging tool to visualize the cochlear nerve, it may for various reasons not always differentiate cochlear nerve hypoplasia from aplasia [22]. If images are obtained using a thick section, the resolution can be poor, movement artifacts may obscure the images, or a concomitant narrow IAC or CNC may mask the cochlear nerve. To depict accurately the cochlear nerve, submillimetric thin-section images of the axial and parasagittal planes must be acquired using a 3-tesla MRI scanner rather than a 1.5-tesla MRI scanner [28]. In addition to MRI, EABR using intracochlear electrical stimulation can help predict the outcome of CI in children with CND [15, 29]. The ideal method is transtympanic EABR using promontory or round window stimulation performed before surgery, but studies have failed to show a predictive value of preoperative transtympanic EABR on CI outcomes and the status of the cochlear nerve. One study showed that there were no significant differences in speech perception and production between deaf children with a clear promontory EABR wave and those with no wave [30]. The other study reported that promontory EABR was absent in four children with a narrow IAC who were subsequently found during surgery for auditory brainstem implant (ABI) to have thin vestibulocochlear nerves [31]. However, EABR performed during CI surgery using intracochlear electrical stimulation was reported to have predictive value for the speech perception outcomes after CI in children with CND [15, 29] and children with narrow IAC [32]. The information about the prognosis obtained from EABR performed during or very soon after CI surgery in children with CND can assist in the choice of habilitation method and facilitate decision-making about whether to convert to ABI [32].

An ABI should also be considered for children with CND [14, 22, 31]. A report showed that the children with CND who had a poor performance with a CAP score of 2 or less after CI achieved CAP scores of 2–7 after receiving an ABI for the ipsilateral ear [14]. Another report showed that children with CND who received an ABI outperformed those with CND who received CI [33]. However, the surgical risks related to ABI and the much more difficult programming procedure for the ABI must be considered [22, 33]. Furthermore, ABI is recommended for appropriate candidates aged over 18 months [34]. Therefore, for children with CND, it is better to perform CI as early as possible to minimize the period of auditory deprivation. If a child with CI has an aplastic cochlear nerve on MRI and shows no EABR and no auditory progress, a contralateral ABI should be considered [33].

The incidence of IEMs in children with congenital SNHL has been reported to range from 20 to 40 % [35–37]. Most IEMs are considered to be the result of arrested and/or aberrant development during the embryogenesis of the inner ear [35]. Single-gene mutations can also cause IEMs, including the SLC26A4 gene mutation that causes an enlarged vestibular aqueduct [38] and the POU3F4 gene mutation that causes incomplete partition type 3 [39].

When CI surgery was first performed, children with IEMs were regarded as poor candidates for CI because of their abnormal tonotopic organization and the increased risks of surgical complications including CSF gusher, facial nerve injury, or electrode misinsertion into the IAC in patients with a defect of the fundus [2–9]. Currently, however, the presence of IEMs is no longer an obstacle to performing CI and achieving favorable speech perception abilities [1, 40]. There have been numerous reports of CI performed in children with IEMs and promising outcomes have been reported [1, 4, 36, 37, 40, 41]. However, children with some severe forms of IEM, including common cavity or cochlear aplasia, are still difficult candidates for CI, and these types of malformation should be carefully assessed [42–44].

The best modality to depict IEMs is high-resolution TBCT. Most IEMs can be diagnosed without difficulty using axial and coronal TBCT images that are obtained with submillimetric slice thickness. In cases of an extremely malformed inner ear structure, a three-dimensional volume-rendering technique using MR images can help depict the morphology of the inner ear [37].