Opportunities For Improving Prosthesis Function

1.6.1 Capturing the Acoustic Signal

The first stage in auditory prosthesis signal processing involves capturing the acoustic signal that the patient wants to hear (see Figure 1.1). Most commonly, this is done with a small microphone that the subject wears near the ear. One of the most important problems that can be aided by the acoustic pick-up system is recognition of speech signals in a noisy environment. Noisy environments represent a particularly difficult situation for cochlear implant users, as well as for most individuals with any type of hearing impairment. The problem can be lessened by acoustic pick-up systems that focus on the source of the desired signal. A simple though not always practical approach is to put the pick-up close to the source of the signal. Thus, there are acoustic pick-ups that can be attached directly to a telephone, a television or a radio. In addition, where there is one-to-one contact between the speaker and the listener, the speaker can wear a microphone that transmits the signal to the subject's speech processor. Directional microphones worn by the listener can be advantageous. The best performance can be achieved by a beam former, which uses an array of microphones,48,49 but this is not always practical for the wearer. Further development in this area is warranted.

Single Channel

100 80 60 40 20 0

Early Multichannel

"S ioo

O 60

O 20 CD

Contemporary Multichannel

FIGURE 1.5 Speech recognition results for patients fitted with various types of cochlear prostheses. Each bar represents an individual patient. For each type of prosthesis, subjects are arranged in order based on their speech recognition scores. A-F: Results of a study by Gantz et al.32 The Iowa Sentence Test was used for all configurations tested in this study. The

FIGURE 1.5 Speech recognition results for patients fitted with various types of cochlear prostheses. Each bar represents an individual patient. For each type of prosthesis, subjects are arranged in order based on their speech recognition scores. A-F: Results of a study by Gantz et al.32 The Iowa Sentence Test was used for all configurations tested in this study. The

Another potential approach, still under development, is to extract the desired signal from background noise electronically. For example, signals with acoustic characteristics of speech might be extracted electronically from non-speech-like noise.50 Such feature-extraction approaches are also sometimes used for converting acoustic signals to electrical signals as described in Section 1.6.2.

Was this article helpful?

0 0

Post a comment