Imaging of Electrode Position after Cochlear Implantation with Flat Panel CT

Background. Postoperative imaging after cochlear implantation is usually performed by conventional cochlear view (X-ray) or by multislice computed tomography (MSCT). MSCT after cochlear implantation often provides multiple metal artefacts; thus, a more detailed view of the implant considering the given anatomy is desirable. A quite new method is flat panel volume computed tomography. The aim of the study was to evaluate the method's clinical use. Material and Methods. After cochlear implantation with different implant types, flat panel CT scan (Philips Allura) was performed in 31 adult patients. Anatomical details, positioning, and resolution of the different electrode types (MedEL, Advanced Bionics, and Cochlear) were evaluated interdisciplinary (ENT/Neuroradiology). Results. In all 31 patients cochlear implant electrode array and topographical position could be distinguished exactly. Spatial resolution and the high degree of accuracy were superior to reported results of MSCT. Differentiation of cochlear scalae by identification of the osseous spiral lamina was possible in some cases. Scanning artefacts were low. Conclusion. Flat panel CT scan allows exact imaging independent of implant type. This is mandatory for detailed information on cochlear electrode position. It enables us to perform optimal auditory nerve stimulation and allows feed back on surgical quality concerning the method of electrode insertion.


Introduction
Postoperative imaging after cochlear implantation usually is performed by conventional cochlear view (X-ray) or by multislice computed tomography (MSCT). Conventional cochlear view is routinely used mainly in children due to short investigation time and low radiation dose. This technique only gives projective information on the fact that insertion into the cochlea has been successful, but analysis of exact electrode position with regard to the topography of the cochlea is impossible [1,2]. MSCT after cochlear implantation allows for three-dimensional imaging; however, it unfortunately provides metal artefacts; thus, a more detailed view of the electrodes with regard to the given anatomical structures is desirable [3][4][5]. This is not only of major importance for quality control as far as surgical insertion methods are concerned but also with regard to special anatomic situations (e.g., mondini dysplasia, ossification of cochlea due to meningitis or otosclerosis) and fitting conditions dependent on electrode array position and results of neural response telemetry in rehabilitation. In order to gain more information on perfect design and function of electrodes, an exact way of imaging also is mandatory from researchers' and developers' point of view.
An additional imaging method is flat-panel computed tomography. The experimental use of flat-panel highresolution CT in temporal bone specimens was evaluated with regard to position of electrode array and with special concern on documentation of the highest possible accuracy in cranial base navigation [3,6,7].
The aim of our study was to evaluate the clinical use of flat-panel CT as far as feasibility, artefacts, position within     With patients temporal bone in system isocenter, the scan was performed with a propeller movement covering 207 • of the circular trajectory. 622 frames were exposed during the 20.7 s scan (30 frames/s), utilizing a detector format of 33 × 40 cm. Total examination time including bedding of the patient on the examination table demanded less than 2 minutes.

Materials and Methods
Source images were transferred to the workstation during and after the rotational acquisition, and a volume data set was created. The reconstruction appeared on the workstation monitor.
Multiplanar MIP reconstructions parallel to the cochlea were performed with a slice thickness of 1.5 mm and orthogonal to the cochlea with a slice thickness of 0.41 mm.

Evaluation/Analysis.
Images were analysed by two independent investigators (otorhinolaryngology 1, neuroradiology 1). Artefacts were characterised between very low (= 1) and very high (= 6). Identification of scala tympani, scala vestibule, and osseous spiral lamina as well as tip fold over, full insertion, identification of distance between electrode array and modiolus (next to (n) and far from (f)), and possibility of identification of single electrodes within the electrode array was of further interest.

Results
An overview on all subjects is illustrated in Table 1. In all cases, full insertion of all electrodes could be seen, which was in line with the surgeon's information. There was no tip fold over in any subject. The level of artefacts was calculated as 2.41 (1-6) in the mean for investigator I and 2.64 (1-6) for investigator II. Osseous spiral lamina could be identified in all subjects; scalae tympani and vestibuli were obviously visible in 6 of 31 cases (Figure 4). Differentiation of different types of implants was clearly feasible (Figures 2(a)-2(c)).
Additional identification of single electrodes was possible in all implants of Medel and Advanced Bionics Companies. Identification was not always possible in Cochlear Nucleus implants due to artefacts produced by narrower distances between electrodes. Dependening on implant type, electrode array insertion could apparently be identified as next to modiolus (Figures 3(a) and 3(b)) in 25 cases or far from modiolus (Figures 3(c) and 3(d)) in 6 cases.

Discussion
Radiological examinations after cochlea implantation are mostly performed by conventional X-ray (Figure 1(a)). This method offers information about the position of the device within the petrous bone. Positioning within the cochlea is approximately indirectly educible from the configuration of the electrode. Additionally, buckling of the electrode may be visible. In more complex cases, a more detailed view of the electrode is eligible. In these cases, conventional MSCT scans offer additional information of the electrode positioning within the cochlea (Figure 1(b)). Nevertheless, conventional CT scans are often affected by artefacts due to electrode and generator [3][4][5].
Several attempts have been made to improve imaging. The technique of cone beam computed tomography as a low-dose imaging technique for postoperative assessment of cochlear implantation, which was tested in postoperative patients, seems to be promising with fewer artefacts and higher resolution than multislice CT [8]. But cone beam CT devices are still rare. Therefore, we evaluated the clinical impact of flat-panel CT in a C-arm angiography unit.
Several publications deal with image quality of isolated temporal bone specimens [3] in flat-panel CT. They all revealed superior quality compared to multislice computed tomography.
Rotational computed tomography (RT) is based on three-dimensional digital subtraction angiography. Images are taken with a rotating C-arm in a single rotation [9]. After digital reconstruction, the intracochlear position of an electrode array can be identified [10]. The method offers new Blooming artefacts are also less comprehensive, and spatial resolution is superior in comparison to MSCT [11,12]. Due to higher spatial resolution, additional information of the position of the electrode in relation to the modiolus is offered. This has impact on the adjustment of the generator and the estimated life time of the battery.
Of course, tip fold over should be clearly visible. In our collective, no patient had one.
Electrode positioning within the scala tympani is clinically eligible. Otherwise, in case of positioning within the scala vestibuli, avoidable side effects like high impedances, reduced speech reception, and vertigo might occur due to damage of the intracochlear structures [13]. Exact positioning within cochlea was not assessable in our collective. As far as the literature is concerned, the assessment of positioning within the scala tympani most often could only be demonstrated in ex vivo specimens.
Aschendorff et al. [9,10] reported feasibility of assessing exact electrode position within the scala tympani or vestibuli in vivo by means of flat-panel CT. Scala media could not be visualized directly, and the position was concluded indirectly from curved reformatted images of the cochlea. In spite of similar examination techniques, the difference in the results might be based on different postprocessing. There is no possibility to perform curved reformations techniques which is impossible on our system. Some examinations have been performed to adjust radiation dose of FD CT in comparison to multislice CT. Radiation dose of FD CT is described to be lower than in MSCT.
The main weak point of our study is the lack of direct comparison of MSCT and flat-panel CT. Due to the high level of artefacts known from own experience, MSCT is not common in our house under this indication.
Our results indicate that flat-panel CT is a fast and accurate examination in the postoperative imaging of cochlear implants. It is of course superior to conventional X-ray, but it is also superior to MSCT mainly due to fewer artefacts. Additionally, radiation dose is lower than in MSCT.