Posters
Academic Level (Author 1)
Medical Student
Discipline Track
Biomedical ENGR/Technology/Computation
Abstract
Bimodal stimulation, the usage of both a cochlear implant and a hearing aid in the contralateral ear, is noted for improving speech perception in noisy backgrounds for some hearing-impaired individuals. However, the usage of both devices provides a range of variability in benefit that relate to difference in pitch fusion range. Hearing impaired populations are noted to have broad binaural fusion ranges compared to normal hearing population, effecting the ability to perceive certain syllables. In studies involving vowels classification and fusion, bimodal CI users showed significant difference in both CI-only conditions and HA-only conditions, compared to the bimodal conditions (Reiss et al 2016). Not much is known about the perception of consonants under bimodal conditions and how binaural fusion could affect perceptional cues. With this experiment, we hope to see the benefit in perceptional cues among consonants in bimodal stimulation in hearing impaired subjects.
For this experiment, natural and synthetic consonant perception is tested among 10 bimodal listeners under 3 conditions: 1) Bimodal, 2) CI only, and 3) HA only, to see if there is a benefit, averaging, or interference of the devices when it comes to speech perception. For the natural consonants testing, 16 consonants are presented: p t k f Ɵ s ʃ b d g v ð z ʒ m n, by 4 different speakers. A total of 64 stimuli is tested by the presence of |a| sound before the consonant. After presentation of stimulus, listeners indicate the consonant sound they thought they perceive. Subjects undergo 4 trial runs for each condition. To greater understand the difference in consonant perception, linguistic features such as manner, voicing, and place of each tested consonant are compared. Responses were recorded on a confusion matrix and percentage of correct responses per condition and feature were analyzed. For the synthetic consonant testing, consonant pairs with similar linguistic features, sa/sha and wa/ya were tested under the 3 conditions mentioned above. Each consonant pair are presented with 8 different fricative frequencies, encouraging the listeners to distinguish between the stimuli. 2 trial runs with 80 stimuli per trial for each condition are conducted per consonant pair. Listeners are instructed to select which consonant syllable they heard or take a guess if having difficulty distinguishing the sound. Responses are recorded and compared to the change in fricative frequency of the stimulus. In the natural consonants, the bimodal condition had the greatest mean at 70.015% ± 15.5, followed by CI only at 61.81% ± 16.1, and lastly, the HA only condition with an average mean of 55.48% ± 18.7. Confusion matrices patterns indicate some subjects benefit from the usage of bimodal stimulation, with a greater percentage correct of manner in all conditions. As for synthetic consonants, subjects’ responses to changes in fricative frequency of stimuli was graphed and a greater perceptional difference was indicated by a steep slope in percent consonant pair vs. stimulus. Bimodal conditions were more successful at distinguishing the fricative change, while CI only conditions varied in averaging the response. While there is some interference in bimodal conditions, analysis of fusion ranges and linguistic features can highlight difficulties and differences in perception. A greater understanding of consonant perception under bimodal conditions can determine if hearing impaired subjects actually benefit from the usage of both devices or could perhaps be better supported by changes in programming, or the usage of a unilateral CI.
Supported by NIH R01DC013307 and Everts-Smith Medical Student Scholarship
Presentation Type
Poster
Recommended Citation
Austin, Courtney J.; Dumont, Rose; Kain, Alexander; Sanders, Holden; and Reiss, Lina, "Does Bimodal Stimulation Improve or Interfere With Consonant Perception?" (2024). Research Colloquium. 1.
https://scholarworks.utrgv.edu/colloquium/2023/posters/1
Included in
Does Bimodal Stimulation Improve or Interfere With Consonant Perception?
Bimodal stimulation, the usage of both a cochlear implant and a hearing aid in the contralateral ear, is noted for improving speech perception in noisy backgrounds for some hearing-impaired individuals. However, the usage of both devices provides a range of variability in benefit that relate to difference in pitch fusion range. Hearing impaired populations are noted to have broad binaural fusion ranges compared to normal hearing population, effecting the ability to perceive certain syllables. In studies involving vowels classification and fusion, bimodal CI users showed significant difference in both CI-only conditions and HA-only conditions, compared to the bimodal conditions (Reiss et al 2016). Not much is known about the perception of consonants under bimodal conditions and how binaural fusion could affect perceptional cues. With this experiment, we hope to see the benefit in perceptional cues among consonants in bimodal stimulation in hearing impaired subjects.
For this experiment, natural and synthetic consonant perception is tested among 10 bimodal listeners under 3 conditions: 1) Bimodal, 2) CI only, and 3) HA only, to see if there is a benefit, averaging, or interference of the devices when it comes to speech perception. For the natural consonants testing, 16 consonants are presented: p t k f Ɵ s ʃ b d g v ð z ʒ m n, by 4 different speakers. A total of 64 stimuli is tested by the presence of |a| sound before the consonant. After presentation of stimulus, listeners indicate the consonant sound they thought they perceive. Subjects undergo 4 trial runs for each condition. To greater understand the difference in consonant perception, linguistic features such as manner, voicing, and place of each tested consonant are compared. Responses were recorded on a confusion matrix and percentage of correct responses per condition and feature were analyzed. For the synthetic consonant testing, consonant pairs with similar linguistic features, sa/sha and wa/ya were tested under the 3 conditions mentioned above. Each consonant pair are presented with 8 different fricative frequencies, encouraging the listeners to distinguish between the stimuli. 2 trial runs with 80 stimuli per trial for each condition are conducted per consonant pair. Listeners are instructed to select which consonant syllable they heard or take a guess if having difficulty distinguishing the sound. Responses are recorded and compared to the change in fricative frequency of the stimulus. In the natural consonants, the bimodal condition had the greatest mean at 70.015% ± 15.5, followed by CI only at 61.81% ± 16.1, and lastly, the HA only condition with an average mean of 55.48% ± 18.7. Confusion matrices patterns indicate some subjects benefit from the usage of bimodal stimulation, with a greater percentage correct of manner in all conditions. As for synthetic consonants, subjects’ responses to changes in fricative frequency of stimuli was graphed and a greater perceptional difference was indicated by a steep slope in percent consonant pair vs. stimulus. Bimodal conditions were more successful at distinguishing the fricative change, while CI only conditions varied in averaging the response. While there is some interference in bimodal conditions, analysis of fusion ranges and linguistic features can highlight difficulties and differences in perception. A greater understanding of consonant perception under bimodal conditions can determine if hearing impaired subjects actually benefit from the usage of both devices or could perhaps be better supported by changes in programming, or the usage of a unilateral CI.
Supported by NIH R01DC013307 and Everts-Smith Medical Student Scholarship