Current Focusing to Reduce Channel Interaction for Distant Electrodes in Cochlear Implant Programs

Abstract

Speech understanding abilities are highly variable among cochlear implant (CI) listeners. Poor electrode-neuron interfaces (ENIs) caused by sparse neural survival or distant electrode placement may lead to increased channel interaction and reduced speech perception. Currently, it is not possible to directly measure neural survival in CI listeners; therefore, obtaining information about electrode position is an alternative approach to assessing ENIs. This information can be estimated with computerized tomography (CT) imaging; however, postoperative CT imaging is not often available. A reliable method to assess channel interaction, such as the psychophysical tuning curve (PTC), offers an alternative way to identify poor ENIs. This study aimed to determine (a) the within-subject relationship between CT-estimated electrode distance and PTC bandwidths, and (b) whether using focused stimulation on channels with suspected poor ENI improves vowel identification and sentence recognition. In 13 CI listeners, CT estimates of electrode-to-modiolus distance and PTCs bandwidths were measured for all available electrodes. Two test programs were created, wherein a subset of electrodes used focused stimulation based on (a) broad PTC bandwidth (Tuning) and (b) far electrode-to-modiolus distance (Distance). Two control programs were also created: (a) Those channels not focused in the Distance program (Inverse-Control), and (b) an all-channel monopolar program (Monopolar-Control). Across subjects, scores on the Distance and Tuning programs were significantly higher than the Inverse-Control program, and similar to the Monopolar-Control program. Subjective ratings were similar for all programs. These findings suggest that focusing channels suspected to have a high degree of channel interaction result in quite different outcomes, acutely.

Keywords: channel selection; imaging; psychophysics; speech perception.

Figure 1.

CT view of cochlea and electrode array along the midmodiolar axis (red and yellow dashed line), for all subjects, organized by electrode array type. The evenly spaced red dots represent electrodes; the outermost dot represents the insertion depth marker. The white line represents the 0° reference point from which insertion depth is measured, extending from the midmodiolar axis. Row 1: 1 J; Row 2: 1 J Helix and 1 J with positioner; Row 3: MidScala electrode array. CT = computerized tomography; ERB = equivalent rectangular bandwidth. Source: Reproduced with permission from DeVries and Arenberg (2018).

Figure 2.

PTCs across the electrode array for an example subject (S53). The x-axis denotes the masker electrode (apical to basal). The y-axis represents the masker threshold in percentage dynamic range. The numbered lines inside the plot are probe electrodes, labeled at the tip of the PTC. The mean PTC ERB_DR in millimeters is marked inside the plot. ERB_DR = equivalent rectangular bandwidth; PTC = psychophysical tuning curve. Source: Reproduced with permission from DeVries and Arenberg (2018).

Figure 3.

(a) Tuning curve ERB_DR as a function of electrode-to-modiolus distance (mm) for S43. The numbers inside the figure indicate channels. The green dashed line is set at the median electrode-to-modiolus distance and the purple dashed line is set to the median PTC ERB_DR. (b) Electrode configuration schemes for “Distance,” “Tuning,” “Inverse-Control,” and “Monopolar-Control” programs. Gray boxes are channels that use pTP stimulation, and black boxes are channels that use monopolar stimulation for each program. Channels 1 and 16 are not used because of the nature of sQP stimulation. ERB_DR = equivalent rectangular bandwidth; PTC = psychophysical tuning curve; pTP = partial tripolar; sQP = steered quadrupolar.

Figure 4.

Tuning curve ERB_DR (mm) as a function of electrode-to-modiolus distance (mm) for 13 subjects, arranged by Pearson’s r from highest to lowest. The numbers inside the figure indicate channels. The green dashed line is set at the median electrode-to-modiolus distance and the purple dashed line is set to the median PTC ERB_DR. The red line is the best fit line for each subject. ERB_DR = equivalent rectangular bandwidth; PTC = psychophysical tuning curve.

Figure 5.

(a) Raw scores (percentage correct) on the medial vowel identification task, within (left of dashed line) and across subjects (right of dashed line). Scores are presented for Monopolar-Control (blue), Inverse-Control (red), Distance (green), and Tuning (purple) programs. Brackets with asterisks indicate significance differences between programs. (b) Raw scores (percentage correct) on the sentence recognition task. All subjects used IEEE sentences except S40, S46, and S49, which are marked with an asterisk. All other conventions are the same as the top panel. ERB = equivalent rectangular bandwidth.

Figure 6.

(a) Performance relative to the Monopolar-Control program (percentage correct) on the medial vowel identification task, within (left of dashed line) and across subjects (right of dashed line). Scores are presented for Inverse-Control (red), Distance (green), and Tuning (purple) programs. (b) Performance relative to the Monopolar-Control program (percentage correct) on the sentence recognition task. All subjects used IEEE sentences except S40, S46, and S49, which are marked with an asterisk. All other conventions are the same as the top panel. MP = monopolar.

Figure 7.

(a) Raw scores (percentage correct) on the sentence recognition task for ERB_DR–Distance correlation (left) and no correlation groups (right). All subjects used IEEE sentences except S40, S46, and S49. All subjects were combined for ease of viewing. Scores are presented for Monopolar-Control (blue), Inverse-Control (red), Distance (green), and Tuning (purple) programs. Brackets with asterisks indicate significance differences between programs. (b) Raw scores (percentage correct) on the medial vowel identification task for ERB_DR–Distance correlation (left) and no correlation groups (right). ERB_DR = equivalent rectangular bandwidth.

Figure 8.

Boxplots showing the distribution of subjects CNC word scores with subject’s everyday program (y-axis) for the ERB_DR–Distance no correlation (left) and correlation groups (right). ERB_DR = equivalent rectangular bandwidth.

Figure 9.

Scatterplots showing individual (black symbols) and mean (red symbol) sound quality ratings (y-axis) for clarity in quiet (top left), clarity in noise (top right), ease of listening (bottom left), and naturalness (bottom right) for each program tested (x-axis). Note that S49, S46, and S49 were not tested in noise and therefore did not assess clarity in noise. MP = monopolar.