A cochlear implant, by definition, is a small, complex device that provides sounds by way of electrical stimulation to an individual who is severely hard-of-hearing or deaf. While hearing aids amplify acoustic sounds so they may be detected and understood by damaged ears, a cochlear implant bypasses those damaged portions of the ear and sends direct stimulation to the auditory nerve using electrodes arrayed on the scala tympani (NIDCD,2013). While many recipients of this innovative device report various levels of success, there remain to be common reports of difficulty with speech processing, especially in the presence of noise. These patient concerns, while complex in each individual case, poses the research question of how many electrodes (or channels) in a cochlear implant are necessary for good speech recognition? This question will be analyzed further in this paper in various listening situations in order to understand speech recognition in cochlear implant processing as a function of the number of spectral channels, or electrodes.
Speech Recognition in Quiet
A review by Shannon et. al, 2007 attempts to describe how the cochlear implant functions to replace the function of the cochlea that is responsible for speech understanding. To complete this process, acoustic signal are analyzed into different frequency bands and the speech information from each band is presented to an electrode along the scala tympani that represents the corresponding frequency region (Shannon et. al, 2007). Dorman et al, 1997, used this processing strategy to measure speech recognition for vowels, consonants and sentences in normal hearing listeners are a function of the frequency bands in the representation. This research found that these normal hearing listeners required only about 4 to 6 bands of frequency information to obtain high levels of speech recognition in a quiet environment. The research project presented in this paper will attempt to replicate and add to earlier research by simulating a speech signal with cochlear implant processing in various channels in order to determine, the minimum amount of channels (or electrodes) necessary for understanding speech.
While cochlear implants deliver significant benefit for speech discrimination in quiet, its performance in music has not been ideal. According to a clinical review by Drennan et. al, 2008, implant users rank music as the second most important acoustic stimulus in their daily lives next to understanding conversational speech, and most of these users reported that music does not sound good when using the device. This lack of benefit in music can is explained by the fact that cochlear implant users typically only hear repetition rates up to about 300 Hz, and tend to lose much of the fine-structure characteristics that could be used to encode pitch. It has also been hypothesized that even though implanted patients get as many as 22 channels of stimulation in the last...