North Texas
Cochlear Implant Program
Cochlear Implant Information Index -
General Information -
Internal Components -
External Components
Speech Processing -
Warranty/Loss Information -
Other Resources
Speech Processing
A. Speech Coding Strategies (or
Algorithms)
Speech coding strategies in speech
processors are driven by algorithm(s) used to decode and transmit speech to the
cochlea. There is still a great deal of
debate as to the best algorithm for improving speech recognition. It may even be true that one type of speech
processing strategy works well for one person but another strategy is optimal
for someone else.
Advanced Bionics Corp. - Clarion
• Three types of
stimulation available: simultaneous (SAS) stimulation, sequential stimulation
(CIS) or paired pulsatile (PPS)
Cochlear Corp. - Nucleus 24
• The ESPrit (ear-level processor) uses the SPEAK strategy. Cochlear Corp. anticipates that ACE will be
able to used in this processor soon.
• The SPrint speech processor (body worn) currently uses three
speech coding strategies: SPEAK, CIS and ACE (Advanced Combination
Encoders)
MED-EL - Combi 40+
• This speech processor uses sequential stimulation (CIS) and high
rate spectral peak extraction strategies (High-Rate N-of-M)
B. Processing Speed
Fast stimulation allows for detailed
representation of speech sounds to be delivered to the hearing nerve. The ideal situation is to send the most
information to the array by increasing the number of impulses (pitch pulses)
without causing the nerve to adapt. It
may be true that a high rate of stimulation will work very well for one person
and a slower rate of stimulation will work better for another.
Advanced Bionics Corp.- Clarion
• can transmit up to 91,000 pieces of information per second for
the SAS algorithm, 6,500 per second for the CIS algorithm
Cochlear Corp. - Nucleus 24
•
can transmit up to 14,400 pulses per second
MED-EL - Combi 40+
•
can transmit up to 18,180 pulses per second with their version of the CIS
strategy and with the N-of-M strategy (depending on the number of channels
activated by the waveform peaks)
C. Multiple Programs
Multiple programs (MAPS) are valuable
for storing different processing strategies to accommodate different listening
environments (e.g. speech in a noisy room, and speech in quiet). Different
programs store different information related to loudness levels, listening
thresholds or stimulating waveforms. It
is helpful to have at least two or more hearing programs in the speech
processor for this flexibility.
Advanced Bionics Corp.- Clarion
•
can store 3 programs
Cochlear Corp. - Nucleus 24
•
the body-worn speech processor can store 4 programs
•
the ear-level speech processor can store 2 programs
MED-EL - Combi 40+
•
the body worn speech processor can store 3 programs
• the ear-level speech processor is able to store 3 programs with 3
different volume settings for each of the 3 programs
D. Programmable Features
Advanced Bionics Corp.- Clarion
• built-in system quality check:
has a public alarm which indicates battery level and system status
Cochlear Corp. - Nucleus 24
• built-in system quality check:
has a public and/or private alarm which indicates battery level and LCD
display for system status
•
lockable controls for children
•MED-EL - Combi 40+
• built-in system quality check:
has a public and private alarm which indicates battery level and a
public alarm for the system status
•
has tamper-resistant controls