- Email: [email protected]
Computed tomography and magnetic resonance imaging in the preoperative work-up for cochlear implantation
European Journal of Radiology, 15 (1992) 89-92 0
1992 Elsevier Science Publishers
B.V. All rights reserved. 0720-048X/92/$05.00
89
EURRAD 00287
Computed tomography and magnetic resonance imaging in the preoperative work-up for cochlear implantation Hans-Martin
Klein”, K. Bohndorf”,
H. Hermesb, W.F. Schiitzb, R.W. Gunther”
and G. SchldndorEb
aDepartment of Radiology and b Departmellt of Otolaryngology. Technical University Aachen, Aachen, Germany (Received 21 November
Key words: Computed tomography,
comparative
1991; accepted after revision 15 February 1992)
study; Magnetic resonance, comparative study; Magnetic resonance, technology; Ear, CT; Ear, MRI; Cochlear implant
Abstract
The role of CT and MRI in the evaluation of patients for possible insertion of a multichannel intracochlear hearing device was appraised. The study included 52 patients who underwent both CT and MRI examinations, 40 of whom were later operated on. Coronal and axial T,-weighted spin-echo sequences were performed in 25 volunteers with normal hearing and in 47 adult patients. In 5 patients, instead of a T,-weighted spin-echo sequence, a Tz*-weighted gradient echo 3D sequence with axial presaturation was used. In 39 patients with normal appearances on CT and MRI, the implant device was successfully inserted. One patient who underwent surgery had a reduced cochlear signal on MRI but a normal CT scan; however, at surgery, the implant device could only be inserted into the first turn of the cochlea, due to fibrous obliteration. In 3 of 12 patients who were not operated upon, the results of diagnostic imaging indicated that the insertion of an intracochlear hearing device was not useful. Our experience indicates that, with reduced cochlear fluid signal intensities on MRI, fibrous obliteration of the cochlear turns is likely to be present. MRI proved to be a useful adjunct to CT, but the latter was necessary for the evaluation of bony abnormalities. Gradient echo sequences can successfully replace time-consuming T,-weighted spin-echo sequences.
Introduction A detailed assessment of bony and soft tissue structures is mandatory before inner ear surgery; the surgeon needs information on possible obstructions of the cochlear duct prior to insertion of an intracochlear hearing device [ 11. Bony structures are best demonstrated by highresolution CT, which provides reliable information on disorders of the temporal bone [2-41. However, CT cannot satisfactorily demonstrate fibrous obliterations, as there is only a slight X-ray absorption difference between liquid and soft-tissue substances. This has led to the demand for a more sensitive imaging method which is able to differentiate between endocochlear fluid Correspondence to: Dr. Hans-Martin Klein, Klinik fur Radiologisthe Diagnostik, Klinikum der RWTH, Aachen 5100, Aachen, Germany.
and fibrotic occlusions. Limited experience is available to indicate whether MRI may be a reliable solution to this problem [6,7]. Patients and methods In a preliminary study, 25 volunteers with normal hearing were examined by MRI to determine the detectibility of fluid signals in the endolymphatic system. Following this, 52 patients (age range 1 l-69 years, mean 41 years) with profound bilateral hearing loss were examined. Hearing threshold was worse than 95 dB at 1000 and 2000 Hz. All patients underwent a high resolution CT examination and MR studies of the temporal bone on the same day. Forty of these patients underwent cochlear implantation; only the operated ears of these patients are evaluated in this study. High resolution CT was performed with a Siemens Somatom DRH. Slice thickness was 1 mm and there
90
was no interslice distance. The gantry was angled 20” craniocaudal to the orbitomeatal line (Fig.1). MRI was performed with a superconducting 1.5 T system (Magnetom, Siemens) and a circularly polarized head coil (FOV: 300 mm). Two MRI protocols were used: 1. T,-weighted (SE TR = 2300 ms/TE = 90 ms) axial and coronal oriented spin-echo sequences. Slice thickness was 3 mm without interslice gaps. One acquisition was used. 2. T,*-weighted gradient echo sequence (FISP 3D, TR = 50 ms/TE = 12 ms/Flip angle = 40”) with perpendicular presaturation to minimize inflow artifacts, mainly of the internal carotid artery. The thickness of the axial image volume was 32 mm. With a zoom factor of 1.2 and a slice thickness of 1 mm isotropic image voxels were provided. This enabled us to perform additional 3D reformatting in arbitrarily oriented planes. All CT and MR images were reviewed retrospectively by two experienced investigators who had no
Fig. 1. High resolution computed tomography of the temporal bone, axial orientation. Basal run of the cochlear duct (B), secondary and helical turn of the cochlear duct (S), vestibulum (V) and internal auditory canal (arrow).
knowledge of the surgical and clinical outcome. There was unanimous agreement in all cases after discussion. CT examinations were assessed for the demonstration of: (1) the internal auditory canal (2) the labyrinthine structures (3) the tympanum and (4) adjacent regions of the mastoid process. On MRI (1) the fluid-filled compartments of the endolymphatic system (2) the structures in the internal auditory canal and (3) the tympanum were reviewed. Results Using MRI protocol (1): 144 ears were examined (25 normal volunteers and 47 patients). Two patients interrupted the examination because of claustrophobia. Total imaging time was 35 min. Demonstration of the cochlea in both axial and coronal orientation proved to be of particular significance; the basal cochlear turn is presented more clearly in the axial planes and the secondary and helical turns in the coronal planes. Using both planes, the cochlear duct showed a high signal intensity, indicating fluid, in the endolymphatic system of all ears (n = 50) of volunteers with the normal hearing (Fig.2).
Fig. 2. Fig 2. MRI of a normal cochlea (arrow) with T,-weighted spin-echo sequences (SE 2300/90), slice thickness: 3 mm.(a) Coronal orientation: The helical turn of the cochlea is well demonstrated (arrow). (b) Axial orientation: Signal intense fluid (arrows) is shown in the basal run of the cochlea.
Fig. 3. Gradient echo sequence (FISP 3D 50/12/40) of the temporal bone in axial orientation (with presaturation). Slice thickness 1 mm. Internal auditory canal (I), basal turn of the cochlea (B), secondary turn (arrow) and vestibulum (V) show a high signal intensity. The internal carotid artery(C) shows a low signal intensity since axial presaturation is applied.
MRI protocol (2): In five patients we produced axially oriented planes of 1 mm thickness, with a sufficient signal-to-noise ratio and a very high contrast-to-noise ratio between intracochlear fluid and bone. All examinations could be completed and were of good quality. The flow-dependent distortion caused by the internal carotid artery, particularly in hypertensive patients, could be reduced by axial presaturation (Fig.3). Inplane and z-axis resolution were equal, facilitating coronary oriented secondary reconstructions (Fig.4). Imaging time was 17 min. Of the 40 candidates who received a cochlear implant, 6 patients suffered from congenital, 20 patients from acquired and 14 patients from deafness of unknown origin. In one of the congenitally deaf patients a unilateral bony obstruction was found on CT, with no high signal intensity in the suspected cochlear region in MRI; the implant was successfully inserted into the opposite (normal) cochlea. All other operated congenitally non-hearing patients (5/6) had normal findings on both CT and MRI. In 20 patients with acquired deafness, 19 cases showed normal appearance on both modalities. One patient showed normal findings on CT, but had reduced signal intensity on MRI, indicating abnormal cochlear content. During operation, fibrous occlusion was found and the implant could be only partially in-
Fig. 4. Four subsequent slices alter secondary reformatting of the cochlea in coronal orientation from a FISP 3D volume. Slice thickness 1 mm. The cochlea is filled with a high signal intensity fluid (arrow).
serted. Nevertheless, 8 of 16 electrodes of the device could be used to stimulate the cochlear nerve. Two patients suffered from otosclerosis; CT demonstrated the typical “double ring sign”, whereas the MR signal in the cochlear duct was of normal intensity. No false negative results occured on MRI. On CT, results in one patient with a history of inflammatory disease were judged false negative. This case was judged true positive on MRI. No false positive results were found on either CT or MRI (Table 1).
TABLE
1
CT and MRI-results in deaf patients correlated with surgical findings (n = number of patients)
MR spin-echo (n = 35) MRI gradient echo (?I= 5) CT (n = 40)
True negative
True negative
False negative
False negative
34
1
0
0
5
0
0
0
39
0
1
0
92
Twelve of the examined patients did not undergo cochlear implantation. In nine cases, imaging findings were normal, and in two patients the MR examination was terminated prematurely: these patients refused operation. In another two patients, complete bony obstruction of the cochlear duct was found on CT; MRI showed no signal in the cochlear duct. One patient showed a reduced signal intensity in both ears on MRI whereas CT was normal; this patient suffered from acquired deafness caused by meningitis. Based on experiences in a similar examination with the same combination of “suspicious” MRI, history of inflammatory and fibrous obliteration found at surgery, this patient did not undergo implantation.
Discussion Modern diagnostic imaging provides valuable information in the pre-operative assessment of cochlear implant candidates [ 1,5]. Substantial data on the condition of the temporal bone can be provided by CT and MRI, which assists surgical planning and the selection of the implant device. For CT, the investigation protocol is largely standardized [ 2-51. High-resolution back-projection algorithms are used. The imaging plane lies para-axial to follow the long axis of the temporal. With a 5122 matrix and a field of view of 250 mm (head mode) the pixel size is 0.5 mm. High-resolution CT has a high sensitivity for bony occlusions [ 31, but assessment of the intracochlear fluid is difficult, due to the very small difference in attenuation between fluid (O-10 HU) and soft-tissue (40-50 HU). Furthermore, CT resolution is limited by scattered radiation. In the temporal bone region, beam hardening artifacts are significant, caused by the preferential absorption of low energy X-rays by the pericochlear bone. Jackler et al. [ 51 reported a rate of 46 y0 false negative CT diagnoses; they proposed to perform an explorative cochleostomy for preoperative investigation of the cochlear duct. Intracochlear fluid has a high signal intensity on MRI and can easily be differentiated from the intermediate signal of fibrous tissue in T,-weighted sequences. MRI is hampered by partial volume effects if a slice thickness of 3 mm has to be used but, in our experience, the large difference in signal intensity between fluid and soft-tissue provides sufficient compensation. However, if 3 mm slices are performed, two imaging planes (coronal and axial) are necessary to facilitate a more detailed evaluation of the basal, secondary and apical runs of
the cochlea. This procedure is relatively time consuming, since the total examination time is 35 min. To our knowledge for the first time a gradient echo volume sequence (FISP 3D) was successfully used for the evaluation of the cochlea. This technique enables to produce slices of less than 1 mm thickness, with a good signal-to-noise ratio. Coronal reformatting is possible due to the isometric voxel dimensions. Another advantage is the minimal interslice cross talk between contiguous thin slices compared to 2D images. Distortion arising from flow artifacts can be diminished by presaturation. The imaging time is reduced to 17 min. In one of the surgically proved 40 cases, different findings were demonstrated on CT and MRI. In this case, at the beginning of the study, despite bilateral abnormalities revealed on MRI, a cochlear implantation was performed. Insertion of the device was complicated by fibrotic occlusion; only a partial implantation was possible. This corresponds well with the results of Harnsberger et al. and Laszig et al., who found a significantly reduced signal of the cochlear duct in the presence of fibrous occlusion [ 6,7]. High-resolution CT remains mandatory for the assessment of bony conditions in cochlear implant candidates. For example, a diagnosis of otosclerosis has to be known before surgery, but does not prevent insertion of an intracochlear device. However, we consider MRI to be a helpful adjunct in the pre-operative work-up of these patients, providing additional information on possible fibrous obliterations of the labyrinthine system. References 1 Mueller DP, Dolan KD, Gantz BO. Temporal bone computed tomography in the preoperative evaluation for cochlear implantation. Ann Otol Rhino1 Laryngol 1989; 98: 346-349. 2 Swartz JD. High-resolution computed tomography of the middle ear and mastoid. Radiology 1983; 148: 449-454. 3 Virapongse C, Rothman SLG, Kier EL, Sarwar M. Computed tomographic anatomy of the temporal bone. AJR 1982; 139: 739749. Rosenberg RA, Cohen NL, Reede DL. Radiographic imaging for the cochlear implant. Ann Otol Rhino1 Laryngol 1987; 96: 300304. Jackler RK, Luxford WM, Schindler RA, McKerrow WS. Cochlear patency problems in cochlear implantation. Laryngoscope 1987: 97: 801-805. Harnsberger HR, Dart DJ, Parkin JL, Smoker WRK, Osbom AG. Cochlear implant candidates: assessment with CT and MR imaging. Radiology 1987; 164: 53-57. Laszig R, Terwey B, Battmer RD, Hesse G. Magnetic resonance imaging (MRI) and high resolution computed tomography (HRCT) in intracochlear implant candidates. Stand Audio1 Suppl 1988; 30: 197-200.
1992 Elsevier Science Publishers
B.V. All rights reserved. 0720-048X/92/$05.00
89
EURRAD 00287
Computed tomography and magnetic resonance imaging in the preoperative work-up for cochlear implantation Hans-Martin
Klein”, K. Bohndorf”,
H. Hermesb, W.F. Schiitzb, R.W. Gunther”
and G. SchldndorEb
aDepartment of Radiology and b Departmellt of Otolaryngology. Technical University Aachen, Aachen, Germany (Received 21 November
Key words: Computed tomography,
comparative
1991; accepted after revision 15 February 1992)
study; Magnetic resonance, comparative study; Magnetic resonance, technology; Ear, CT; Ear, MRI; Cochlear implant
Abstract
The role of CT and MRI in the evaluation of patients for possible insertion of a multichannel intracochlear hearing device was appraised. The study included 52 patients who underwent both CT and MRI examinations, 40 of whom were later operated on. Coronal and axial T,-weighted spin-echo sequences were performed in 25 volunteers with normal hearing and in 47 adult patients. In 5 patients, instead of a T,-weighted spin-echo sequence, a Tz*-weighted gradient echo 3D sequence with axial presaturation was used. In 39 patients with normal appearances on CT and MRI, the implant device was successfully inserted. One patient who underwent surgery had a reduced cochlear signal on MRI but a normal CT scan; however, at surgery, the implant device could only be inserted into the first turn of the cochlea, due to fibrous obliteration. In 3 of 12 patients who were not operated upon, the results of diagnostic imaging indicated that the insertion of an intracochlear hearing device was not useful. Our experience indicates that, with reduced cochlear fluid signal intensities on MRI, fibrous obliteration of the cochlear turns is likely to be present. MRI proved to be a useful adjunct to CT, but the latter was necessary for the evaluation of bony abnormalities. Gradient echo sequences can successfully replace time-consuming T,-weighted spin-echo sequences.
Introduction A detailed assessment of bony and soft tissue structures is mandatory before inner ear surgery; the surgeon needs information on possible obstructions of the cochlear duct prior to insertion of an intracochlear hearing device [ 11. Bony structures are best demonstrated by highresolution CT, which provides reliable information on disorders of the temporal bone [2-41. However, CT cannot satisfactorily demonstrate fibrous obliterations, as there is only a slight X-ray absorption difference between liquid and soft-tissue substances. This has led to the demand for a more sensitive imaging method which is able to differentiate between endocochlear fluid Correspondence to: Dr. Hans-Martin Klein, Klinik fur Radiologisthe Diagnostik, Klinikum der RWTH, Aachen 5100, Aachen, Germany.
and fibrotic occlusions. Limited experience is available to indicate whether MRI may be a reliable solution to this problem [6,7]. Patients and methods In a preliminary study, 25 volunteers with normal hearing were examined by MRI to determine the detectibility of fluid signals in the endolymphatic system. Following this, 52 patients (age range 1 l-69 years, mean 41 years) with profound bilateral hearing loss were examined. Hearing threshold was worse than 95 dB at 1000 and 2000 Hz. All patients underwent a high resolution CT examination and MR studies of the temporal bone on the same day. Forty of these patients underwent cochlear implantation; only the operated ears of these patients are evaluated in this study. High resolution CT was performed with a Siemens Somatom DRH. Slice thickness was 1 mm and there
90
was no interslice distance. The gantry was angled 20” craniocaudal to the orbitomeatal line (Fig.1). MRI was performed with a superconducting 1.5 T system (Magnetom, Siemens) and a circularly polarized head coil (FOV: 300 mm). Two MRI protocols were used: 1. T,-weighted (SE TR = 2300 ms/TE = 90 ms) axial and coronal oriented spin-echo sequences. Slice thickness was 3 mm without interslice gaps. One acquisition was used. 2. T,*-weighted gradient echo sequence (FISP 3D, TR = 50 ms/TE = 12 ms/Flip angle = 40”) with perpendicular presaturation to minimize inflow artifacts, mainly of the internal carotid artery. The thickness of the axial image volume was 32 mm. With a zoom factor of 1.2 and a slice thickness of 1 mm isotropic image voxels were provided. This enabled us to perform additional 3D reformatting in arbitrarily oriented planes. All CT and MR images were reviewed retrospectively by two experienced investigators who had no
Fig. 1. High resolution computed tomography of the temporal bone, axial orientation. Basal run of the cochlear duct (B), secondary and helical turn of the cochlear duct (S), vestibulum (V) and internal auditory canal (arrow).
knowledge of the surgical and clinical outcome. There was unanimous agreement in all cases after discussion. CT examinations were assessed for the demonstration of: (1) the internal auditory canal (2) the labyrinthine structures (3) the tympanum and (4) adjacent regions of the mastoid process. On MRI (1) the fluid-filled compartments of the endolymphatic system (2) the structures in the internal auditory canal and (3) the tympanum were reviewed. Results Using MRI protocol (1): 144 ears were examined (25 normal volunteers and 47 patients). Two patients interrupted the examination because of claustrophobia. Total imaging time was 35 min. Demonstration of the cochlea in both axial and coronal orientation proved to be of particular significance; the basal cochlear turn is presented more clearly in the axial planes and the secondary and helical turns in the coronal planes. Using both planes, the cochlear duct showed a high signal intensity, indicating fluid, in the endolymphatic system of all ears (n = 50) of volunteers with the normal hearing (Fig.2).
Fig. 2. Fig 2. MRI of a normal cochlea (arrow) with T,-weighted spin-echo sequences (SE 2300/90), slice thickness: 3 mm.(a) Coronal orientation: The helical turn of the cochlea is well demonstrated (arrow). (b) Axial orientation: Signal intense fluid (arrows) is shown in the basal run of the cochlea.
Fig. 3. Gradient echo sequence (FISP 3D 50/12/40) of the temporal bone in axial orientation (with presaturation). Slice thickness 1 mm. Internal auditory canal (I), basal turn of the cochlea (B), secondary turn (arrow) and vestibulum (V) show a high signal intensity. The internal carotid artery(C) shows a low signal intensity since axial presaturation is applied.
MRI protocol (2): In five patients we produced axially oriented planes of 1 mm thickness, with a sufficient signal-to-noise ratio and a very high contrast-to-noise ratio between intracochlear fluid and bone. All examinations could be completed and were of good quality. The flow-dependent distortion caused by the internal carotid artery, particularly in hypertensive patients, could be reduced by axial presaturation (Fig.3). Inplane and z-axis resolution were equal, facilitating coronary oriented secondary reconstructions (Fig.4). Imaging time was 17 min. Of the 40 candidates who received a cochlear implant, 6 patients suffered from congenital, 20 patients from acquired and 14 patients from deafness of unknown origin. In one of the congenitally deaf patients a unilateral bony obstruction was found on CT, with no high signal intensity in the suspected cochlear region in MRI; the implant was successfully inserted into the opposite (normal) cochlea. All other operated congenitally non-hearing patients (5/6) had normal findings on both CT and MRI. In 20 patients with acquired deafness, 19 cases showed normal appearance on both modalities. One patient showed normal findings on CT, but had reduced signal intensity on MRI, indicating abnormal cochlear content. During operation, fibrous occlusion was found and the implant could be only partially in-
Fig. 4. Four subsequent slices alter secondary reformatting of the cochlea in coronal orientation from a FISP 3D volume. Slice thickness 1 mm. The cochlea is filled with a high signal intensity fluid (arrow).
serted. Nevertheless, 8 of 16 electrodes of the device could be used to stimulate the cochlear nerve. Two patients suffered from otosclerosis; CT demonstrated the typical “double ring sign”, whereas the MR signal in the cochlear duct was of normal intensity. No false negative results occured on MRI. On CT, results in one patient with a history of inflammatory disease were judged false negative. This case was judged true positive on MRI. No false positive results were found on either CT or MRI (Table 1).
TABLE
1
CT and MRI-results in deaf patients correlated with surgical findings (n = number of patients)
MR spin-echo (n = 35) MRI gradient echo (?I= 5) CT (n = 40)
True negative
True negative
False negative
False negative
34
1
0
0
5
0
0
0
39
0
1
0
92
Twelve of the examined patients did not undergo cochlear implantation. In nine cases, imaging findings were normal, and in two patients the MR examination was terminated prematurely: these patients refused operation. In another two patients, complete bony obstruction of the cochlear duct was found on CT; MRI showed no signal in the cochlear duct. One patient showed a reduced signal intensity in both ears on MRI whereas CT was normal; this patient suffered from acquired deafness caused by meningitis. Based on experiences in a similar examination with the same combination of “suspicious” MRI, history of inflammatory and fibrous obliteration found at surgery, this patient did not undergo implantation.
Discussion Modern diagnostic imaging provides valuable information in the pre-operative assessment of cochlear implant candidates [ 1,5]. Substantial data on the condition of the temporal bone can be provided by CT and MRI, which assists surgical planning and the selection of the implant device. For CT, the investigation protocol is largely standardized [ 2-51. High-resolution back-projection algorithms are used. The imaging plane lies para-axial to follow the long axis of the temporal. With a 5122 matrix and a field of view of 250 mm (head mode) the pixel size is 0.5 mm. High-resolution CT has a high sensitivity for bony occlusions [ 31, but assessment of the intracochlear fluid is difficult, due to the very small difference in attenuation between fluid (O-10 HU) and soft-tissue (40-50 HU). Furthermore, CT resolution is limited by scattered radiation. In the temporal bone region, beam hardening artifacts are significant, caused by the preferential absorption of low energy X-rays by the pericochlear bone. Jackler et al. [ 51 reported a rate of 46 y0 false negative CT diagnoses; they proposed to perform an explorative cochleostomy for preoperative investigation of the cochlear duct. Intracochlear fluid has a high signal intensity on MRI and can easily be differentiated from the intermediate signal of fibrous tissue in T,-weighted sequences. MRI is hampered by partial volume effects if a slice thickness of 3 mm has to be used but, in our experience, the large difference in signal intensity between fluid and soft-tissue provides sufficient compensation. However, if 3 mm slices are performed, two imaging planes (coronal and axial) are necessary to facilitate a more detailed evaluation of the basal, secondary and apical runs of
the cochlea. This procedure is relatively time consuming, since the total examination time is 35 min. To our knowledge for the first time a gradient echo volume sequence (FISP 3D) was successfully used for the evaluation of the cochlea. This technique enables to produce slices of less than 1 mm thickness, with a good signal-to-noise ratio. Coronal reformatting is possible due to the isometric voxel dimensions. Another advantage is the minimal interslice cross talk between contiguous thin slices compared to 2D images. Distortion arising from flow artifacts can be diminished by presaturation. The imaging time is reduced to 17 min. In one of the surgically proved 40 cases, different findings were demonstrated on CT and MRI. In this case, at the beginning of the study, despite bilateral abnormalities revealed on MRI, a cochlear implantation was performed. Insertion of the device was complicated by fibrotic occlusion; only a partial implantation was possible. This corresponds well with the results of Harnsberger et al. and Laszig et al., who found a significantly reduced signal of the cochlear duct in the presence of fibrous occlusion [ 6,7]. High-resolution CT remains mandatory for the assessment of bony conditions in cochlear implant candidates. For example, a diagnosis of otosclerosis has to be known before surgery, but does not prevent insertion of an intracochlear device. However, we consider MRI to be a helpful adjunct in the pre-operative work-up of these patients, providing additional information on possible fibrous obliterations of the labyrinthine system. References 1 Mueller DP, Dolan KD, Gantz BO. Temporal bone computed tomography in the preoperative evaluation for cochlear implantation. Ann Otol Rhino1 Laryngol 1989; 98: 346-349. 2 Swartz JD. High-resolution computed tomography of the middle ear and mastoid. Radiology 1983; 148: 449-454. 3 Virapongse C, Rothman SLG, Kier EL, Sarwar M. Computed tomographic anatomy of the temporal bone. AJR 1982; 139: 739749. Rosenberg RA, Cohen NL, Reede DL. Radiographic imaging for the cochlear implant. Ann Otol Rhino1 Laryngol 1987; 96: 300304. Jackler RK, Luxford WM, Schindler RA, McKerrow WS. Cochlear patency problems in cochlear implantation. Laryngoscope 1987: 97: 801-805. Harnsberger HR, Dart DJ, Parkin JL, Smoker WRK, Osbom AG. Cochlear implant candidates: assessment with CT and MR imaging. Radiology 1987; 164: 53-57. Laszig R, Terwey B, Battmer RD, Hesse G. Magnetic resonance imaging (MRI) and high resolution computed tomography (HRCT) in intracochlear implant candidates. Stand Audio1 Suppl 1988; 30: 197-200.