Meghan's PHY101 Project
Cochlear Implants
Background
Cochlear Implants are surgically implanted and excite certain regions
of the Basilar Membrane. The location of the stimulation determines
the pitch of the sound. Low frequencies excite the apex of the
Basilar Membrane, creating a low-pitch sound. Higher frequencies
stimulate the basal end of the membrane, creating a high pitch sound.
There are different types of Cochlear Implants, and they can contain
multiple different wires or channels. These wires will stimulate
different regions on the Basilar Membrane, causing patients to hear
certain frequencies.
Many people in the hearing world do not understand why it is so hard
to adjust to having a Cochlear Implant. Speech often sounds very
mechanical with an implant. The purpose of the experiment was to
educate the hearing world about what it may sound like to hear
using a Cochlear Implant. The experiment will also measure the
frequencies at which speech is most easily understood.
Procedure
The participant was informed that they would not be harmed during
the experiment and that they could cease participation at any time
if they became uncomfortable.

The participant was shown a list of 200 words, and informed that they
would hear 45 of them in a random order.

participant reading word list
The participant was then
played a series of sound files, and was instructed to write do
wn the
words that they heard. In order for the participant to understand what
they would be hearing, a clip from Comma Gets a Cure by Jill
McCullough & Barbara Somerville was read and filtered through C-Sound.
The clip began without any filter, and was repeated with a ban-pass filter
of 500 hz, then 1000hz, 2000hz, and 4,000 hz. The participant was not
required to write anything down at this point in the experiment, they were
simply instructed to listen.
The word list trials had several different filters. The first set contained
9 words unfiltered, 9 words filtered at 500hz, 9 words filtered at 1000hz,
9 words filtered at 2000hz, and 9 words filtered at 4,000hz. The next set contained
combinations of frequencies. The first nine words remained unfiltered, but the
following 9 words allowed frequencies of 500 and 1000hz, the next nine allowed
frequencies of 500 and 2000hz and the last set allowed frequencies of 500 and
4000 hz. The third and final set of words contained filters of 1000 and 2000,
1000 and 4000, and 2000 and 4000. The following words were recorded:
hid
feast
hunt
slip
fraud
bad
pan
cleanse
crash
plush
fern
pants
is
rise
fuss
no
such
creed
deed
are
smile
mange
grove
there
death
then
rub
bask
dike
pile
ford
folk
strife
ride
nook
bar
clove
hive
heap
wheat
not
toe
use
pest
box
cane
rag
end
dish
rat
The participant wrote down each word understood

The results were graphed by percent recognition per frequency.
See Page 2 for Results.