After the first CI (right), the patient participated in a longitudinal study examining monosyllabic word recognition over time (Skinner et al, 2007a
). shows her consonant-vowel nucleus-consonant (CNC) monosyllabic word scores (Peterson and Lehiste, 1962
), using the right CI only, at 12 test intervals from two weeks to two years post initial activation.  At two weeks, her word score was 44% and by six months, 87%. Her score, then, reached a plateau and was maintained at that level through the two year test interval. With a score of 87% on CNC words, the patient would be considered a “star performer”. She was scoring well above the average score of 62% for a group of 110 postlinguistically deafened adult CI users who participated in the same longitudinal study (Finley et al, 2010
). The patient reported that she was pleased with her communication abilities when wearing the CI and HA together; however, about a year after cochlear implantation, she stopped wearing the HA due to a decrease in hearing. When using the CI only, she was no longer satisfied with her ability to communicate in daily life. As can be seen in , her CI performance at one year was stable and above average for CNC words. Her difficulties understanding with just the CI were not related to a decrease in performance with the implant. The patient reported that she was comfortable with the CI during face-to-face communication in quiet environments, but when listening in background noise, she found herself struggling to understand. In fact, it was noted at her CI evaluations that her scores in noise were considerably lower than her scores in quiet. At her one year CI evaluation, her score on the TIMIT sentences (Lamil et al, 1986
) presented in the sound-field at 60 dB SPL in quiet using the implant only (no lip-reading) was 76%. When the TIMIT sentences were presented at the same level in four-talker babble (Auditec of St. Louis; +8 signal-to-noise [SNR]), she scored 34% using the CI only, a decrease of 42% from listening in quiet to listening in noise. The TIMIT sentences are difficult due to the composition of the talkers, using both male and female speakers, different regional dialects, and varied speaking rates. Furthermore, the language is complex. For example, “His name became synonymous with cold-blooded cruelty” and “Vietnamese cuisine is exquisite” are two TIMIT sentences. Because of the variability of the talkers and complexity of the sentences, the TIMIT sentences are thought to be more representative of real-life listening. It appeared that, indeed, the patient did fairly well with the CI listening in quiet (76%) but as reported, much poorer in noise (34%). Difficulty understanding speech in noise is a common report by CI users (Donaldson et al, 2009
). Additionally, research has shown that CI users have a significant decrease in sentence scores at SNRs that would not affect those with normal hearing (Spahr et al, 2007
). Because of the communication difficulties she encountered in daily life, the patient expressed a strong interest in obtaining a second CI.
CNC monosyllabic word scores obtained at twelve test intervals from two weeks to two years post-initial activation of the first CI.
As mentioned above, two years after obtaining the Nucleus Freedom CI for the right ear, the patient was implanted with the Nucleus 512 CI in the left ear. She was programmed using the Advanced Combination Encoder (ACE; Skinner et al, 2002
) speech processing strategy (pulse width = 25-μs/phase, monopolar stimulation) and 1800 pulses per second /channel (pps/ch) stimulation rate in the first implanted ear. At the initial activation of the second implant, a speech processor program or map was created using similar parameters to those used with the first CI (ACE strategy, 1800 pps/ch stimulation rate). The patient was familiar with the procedures used to program the speech processor. No programming issues were encountered and the maps created appeared typical. The patient scored 100% when simple sentences were presented live voice using the second CI only. Three days after initial activation of the second CI, the patient reported having a “full blown” vertiginous attack that lasted all day. She had not reported having an attack of vertigo since her left endolymphatic sac operation with shunt was performed several months prior to the second CI surgery. For the next month, the patient reported feeling unsteady to the point of having to remove the second CI speech processor every couple of days to recover.
The patient wore the first CI speech processor at a volume setting of 9 (range: 1–9) and sensitivity setting of 12 (range: 1–20), but in order to prevent feeling unsteady, the patient wore the second CI speech processor at a softer level, volume of 4 and sensitivity of 10. She reported that the first CI was louder than the second CI; however, she noted an improvement in speech understanding when the two were worn together. Furthermore, her CNC word score using the second CI only was 75% at her one month evaluation. Five weeks post-initial activation of the second CI, the patient could consistently wear both processors but still needed to keep the volume and sensitivity of the second CI speech processor at lower settings. In an attempt to equalize loudness between the two ears without causing further dizziness, the stimulation rate of the second CI was changed from 1800 pps/ch to 500 pps/ch. The stimulation rate used with the first CI remained at 1800 pps/ch. After using the 500 pps/ch map for two weeks, the patient reported that the unsteadiness was not occurring as frequently and that she was able to increase the volume to 5. She reported that the 500 pps/ch map sounded louder than the previous 1800 pps/ch map; moreover, she appreciated the longer battery life that the 500 pps/ch map provided. She stated that when she first used the 500 pps/ch map with the second CI and the 1800 pps/ch map with the first CI, there was a difference in sound quality between the ears. After several weeks of continued use, she no longer noticed a difference in sound quality when wearing both the right and left speech processors together. shows the speech processor parameters that are currently used by the patient for each ear. 
The patient’s CI device and speech processor program parameters for each ear.
shows the placement of the electrode array within the cochlea for the first and second CI.  CT scans were obtained both pre-operatively and post-operatively to create these images using a technique described by Skinner et al, 2007b
. The rectangular black and white marks within the cochlea represent the 22 electrodes of the Nucleus internal device (Nucleus Freedom with Contour Advance Electrode, first CI; Nucleus CI512 cochlear implant with Contour Advance Electrode, second CI). Ideally, all electrodes should be in scala tympani. The 3D reconstruction of the CT scans, which is based on a cochlear atlas model (Teymouri et al, 2010
), indicate that for the first CI (right), thirteen electrodes are located in scala tympani, and nine electrodes are involved in a transition from scala tympani through scala media to scala vestibuli with the two most apical electrodes in scala vestibuli. All electrodes remain in scala tympani for the second CI (left). Note that the electrodes are closer to the modiolar wall of the cochlea for the second CI compared to first CI; consequently, electrodes may be in closer proximity to spiral ganglion cells with the second CI compared to the first CI.
Figure 2 CT-based images of the placement of the Nucleus Contour Advance electrode array within the cochlea for the first and second cochlear implants. The black and white rectangular marks represent each of the 22 electrodes of the Nucleus device. Black marks (more ...)
When creating a CI speech processor program or map for the Nucleus device, two levels need to be set for each of the active electrodes, a minimum and a maximum electrical stimulation level. For the Nucleus device, the minimum electrical stimulation levels are commonly referred to as T-levels and maximum electrical stimulation levels are commonly referred to as C- levels. T-levels and C-levels are set based on the each patient’s perception of loudness. The technique used to set T- and C-levels can vary between CI centers, but in general, T-levels are set at a percept of “very soft”, and C-levels are set at a percept of “loud but comfortable”. Both levels can be adjusted to allow soft speech and sound to be audible and conversational speech comfortably loud in daily life. T- and C-levels are different for each patient and can vary between ears as well as across the array within a single ear for the same patient. The T- and C-levels used in the patient’s map for each ear (1800 pps/ch stimulation rate) were averaged across electrodes. For the first CI, averaged T- and C-levels were 178 and 209 (range: 1–255) current levels1
, respectively. For the second CI, averaged T- and C-levels were 85 and 128 current levels, respectively. This is a substantial difference in T- and C-levels between ears when using the same stimulation rate and may be related to the difference in the placement of the electrode arrays within the cochlea for the two implants.
Prior to the second CI surgery, the patient agreed to participate in a longitudinal study evaluating performance after sequential, bilateral cochlear implantation. Subjects in this study are tested before the second CI surgery (pre-operatively) and at one, three, six, nine, twelve, fifteen and eighteen months post-initial activation of the second CI (post-operatively). Pre-operative testing is carried out with the patient using a CI in one ear and a HA in the other ear. Post-operative testing is completed with the patient using an implant in each ear. Prior to speech recognition testing, each patient’s external equipment is checked and frequency modulated (FM) tone, sound-field threshold levels from 250–6000 Hz are obtained. If a patient reports changes in speech understanding or sound-field threshold levels are inconsistent with those previously obtained, a check of the map is performed with reprogramming as necessary. Testing through six months has been completed for this patient. FM tone, sound-field threshold levels obtained at the six-month test interval are shown in for each ear. The patient’s sound-field threshold levels with each CI are ≤ 20 dB HL across the frequency range ensuring the audibility of soft speech and sound (Holden et al, 2007
, Davidson et al, 2009
Frequency-modulated (FM) tone, sound-field threshold levels obtained from 250–6000 Hz at the six-month test interval with the first and second CI.
The test materials used to evaluate speech recognition performance for this study are difficult. Many of the tests are presented in noise or at soft presentation levels in order to replicate difficult listening situations that CI users encounter in daily life. Since the patient reported experiencing difficulty understanding speech in noise when using her first CI alone, the test results will predominately focus on her performance in noise. Two lists of each test were presented in a pseudo-randomized order at each test interval. All testing was performed in the sound-field. The majority of tests were performed with the patient seated at 0° azimuth approximately 1.5 meters from a loudspeaker. For the R-SPACE™
test environment (Revit et al, 2002
) and the localization test (Potts et al, 2009
), the loudspeaker configuration was different and will be described below.
shows CNC monosyllabic word scores for each ear separately at each study test interval (Pre-operatively: CI and HA; Post-operatively: first CI and second CI).  Pre-operatively, the patient scored 91% with the first CI and 0% with the HA. Post-operatively, scores with the second CI progress quickly and are higher than scores with the first CI at the three and six month test intervals.
CNC word scores obtained at four test intervals. Pre-operatively (prior to the second CI), the patient was tested using the first CI only and the HA only. Post-operatively (1 mo, 3 mos and 6 mos), the patient was tested using each implant separately.
The next several figures show results for each test interval in three conditions. Pre-operative scores are shown with the patient using the first CI only, the HA only, and then both together (bilateral condition). It is important to remember that at the pre-operative test interval, the patient had been using the first CI (right) for approximately two years. Post-operative scores are shown for the first and second CI separately and bilaterally. Scores for TIMIT sentences presented at 60 dB SPL in noise (four-talker babble, +8 dB SNR) are shown in .  Note the improvement in the bilateral scores from the pre-operative to the six-month test interval. In addition, the patient’s scores in noise are higher with the second CI than with the first CI at all test intervals. It is interesting to note an improvement in the noise score with the first CI at the six-month test interval compared to the pre-operative and three-month test intervals. At the six-month test interval, the patient had been using the first CI for approximately 2.5 years.
Figure 5 TIMIT sentences in noise scores obtained at three test intervals. Pre-operatively (prior to the second CI), the patient was tested with the first CI only, the HA only, and the CI and HA together. Post-operatively (3 mos and 6 mos), the patient was tested (more ...)
The Bamford-Kowal-Bench Speech in Noise Test (BKB-SIN; Killion et al, 2004
) is a sentence test in noise where the SNR varies throughout the test. The first sentence is presented at an SNR of +21 dB, and for each consecutive sentence, the SNR decreases by 3 dB. An SNR-50 score is obtained; that is, an SNR for which the subject can correctly repeat the key words within the sentences 50% of the time. The BKB-SIN was presented with the speech and noise (four-talker babble) coming from the same loudspeaker located in front of the patient. In addition, the noise was presented through a loudspeaker located 90° to the left of the patient as well as 90° to the right of patient with the speech always presented through the front speaker. Scores obtained when speech and noise were presented through the front speaker are shown in for the pre-operative, three and six-month test intervals.  The score is an SNR; therefore, a lower number represents a decrease in SNR and a better result. In other words, a low SNR means that the listener can understand speech in a greater amount of noise. Recall that pre-operatively, scores were obtained with the first CI, a HA, and both together (bilaterally), whereas post-operatively, scores were obtained with each CI separately and both together (bilaterally). The patient could not do the task when using the HA only. The scores improved considerably, especially in the bilateral condition, by the six-month test interval. At each postoperative test interval, the SNR is lower with the second CI than with the first CI. Again, note the improvement in SNR with the first CI at the three-month and six-month test interval following implantation of the second ear.
Figure 6 The average SNR obtained for the BKB-SIN test when sentences and noise were presented from a loudspeaker located in front of the patient. Pre-operatively, the patient was tested using the first CI only, the HA only, and the CI and HA together. Post-operatively, (more ...)
illustrates the benefits of bilateral cochlear implantation when noise was presented to the sides of the patient.  The left side of shows the SNR obtained when noise was presented 90° to the right of the patient. Pre-operatively, the patient was listening with the first CI (right) only and with the first CI and HA. Post-operatively, the patient was listening with the first CI only and with both implants. The right side of reveals the SNR obtained when noise was presented 90° to the left of the patient. Pre-operatively, the patient was listening with the HA in the left ear and then with the HA and first CI together. Post-operatively, the patient was listening with the second CI (left) only and with both implants. A substantial decrease in SNR was seen in the bilateral condition from the pre-operative to the six-month test interval regardless of the direction of the noise. When the noise was presented on the right side, an improvement in SNR of 12 dB was obtained by the addition of the second CI (left). When noise was presented to the left of the patient, an improvement in SNR of 9 dB was obtained when using two implants. When the noise was presented to the left, the SNR with the second implant (left) was lower, at each post-operative test interval, than the SNR with the first implant (right) when noise was presented to the right.
Figure 7 The average SNR obtained for the BKB-SIN test when sentences were presented from a loudspeaker located in front of the patient and noise was presented 90° to the right and 90° to the left of the patient. Pre-operatively, when noise was (more ...)
test environment is shown in .  The patient was seated in the center of eight loudspeakers through which restaurant noise was presented. The speech stimuli were paired HINT sentence (Nilsson et al, 1994
) lists presented from the loudspeaker directly in front of the patient. The noise was fixed at 60 dB SPL and the sentence presentation level was varied depending upon the patient’s response. An adaptive procedure in which the sentence presentation level was decreased for each correct response and increased for each incorrect response was utilized. For sentences 1–4, a 4 dB step size was used and then for sentences 5–20, a 2 dB step size was used. An SNR at which the patient could correctly repeat 50% of the sentences was determined by averaging sentence presentation levels for sentences 5–20 and a calculated presentation level for sentence 21. shows the considerable decrease (improvement) in SNR from the pre-operative to the six-month test interval for all three listening conditions.  A +6 dB SNR was obtained in the pre-operative bilateral test condition (HA and CI), whereas a -2 dB SNR was obtained in the bilateral test condition (both CIs) at six months. Note that for each post-operative test interval, a lower SNR is obtained with the second CI than with the first CI. Additionally, note the improvement in SNR, with the first CI at both the three-month and six-month test intervals compared to the pre-operative SNR.
Schematic of the R-SPACE™ test environment.
Figure 9 The average SNR obtained during R-SPACE™ testing. Pre-operatively, the patient was tested with the first CI only, with the HA only, and with both together. At all post-operative test intervals, the patient was tested with each CI separately and (more ...)
In addition to difficulties understanding speech in noise, many unilateral CI users report difficulty localizing sound. To evaluate localization skills, testing was performed pre-operatively using the CI and HA and post-operatively using both implants. The test environment consisted of a 140° loudspeaker array with 15 loudspeakers, 10° apart. is a schematic of the test configuration.  The localization task required the patient to state the perceived location of the stimulus (i.e. loudspeaker number 1–15). The patient faced the front (0° azimuth) prior to initiation of each trial, but was permitted to turn her head (i.e., turn toward the loudspeaker from which the word was perceived) during the trial. An equal number of words were presented from each of 10 selected positions; five of the visible loudspeakers were inactive (+/−60°, +/−40° and 0°), but the patient was not aware of this. The words were presented at a roving level of 60 dB SPL (±3 dB SPL). The scores are presented in terms of degree of RMS error. The RMS error is the mean deviation of the patient’s response from the actual location of the presentation, and it is not related to the direction of the deviation (Potts et al, 2009
). Pre-operatively, the patient obtained an RMS error of 59°, and at the six-month test interval, an RMS error of 26° degrees was obtained. An RMS error of 0° represents 100% accurate localization; therefore, the patient’s localization ability improved using two implants compared to using a HA and CI.
Schematic of the localization test environment. The boxes in gray represent the loudspeakers through which no sound was presented.
Even though this patient was scoring near 90% on the CNC word test with her first implant, she stated that she struggled to communicate in daily life when using only one CI. The results of the testing described above revealed that this patient obtains considerable benefit from bilateral cochlear implantation; however, these speech tests were all performed in a sound booth. No matter how difficult and varied the testing, the multitude of listening environments encountered in daily life can not be replicated in a sound booth. To better understand the benefits of bilateral implantation in different listening environments in daily life, the Speech, Spatial and Qualities of Hearing Scale (SSQ; Gatehouse and Noble, 2004
) was given at each test session. The test consists of three scales. The Speech scale questions the patient’s abilities to understand speech in a variety of difficult listening situations. A question from the Speech scale is a follows: “You are talking with one other person and there is a TV on in the same room. Without turning the TV down, can you follow what the person you’re talking to says?” The Spatial scale concerns spatial hearing or the ability to tell the direction and judge the distance of the sound. A question from the Spatial scale is as follows: “You are outside. A dog barks loudly. Can you tell immediately where it is without having to look?” The Sound Qualities scale of the SSQ questions a patient’s ability to segregate sounds. That is, can a patient recognize each of two sounds that are going on at once, for example water running and a radio playing? Additionally, this third scale examines the “naturalness” and “clarity” of sound and determines the amount of effort the patient exerts to listen (“Do you have to put in a lot of effort to hear what is being said in conversations with others?”). For each item on the SSQ, the patient provides a rating on a continuum from 0–10 with a higher number indicating greater ability. The patient was able to see her ratings from the previous test interval while completing each SSQ. indicates the average rating across items for each of the three scales.  An improvement from the pre-operative test interval, where the patient was using one CI, to the six-month test interval (bilateral CIs) was seen for all scales with the greatest improvement seen for the Spatial scale.
The average rating across test items for each of the three scales in the Speech, Spatial and Qualities of Hearing Scale (SSQ).