Does It Make Sense to Stimulate at Such High Rates?—
Since Auditory Nerve Fibers Cannot Discharge at Rates > 200 pps for any significant
period of time.
Quotes from the Literature:
-
There are two primary reasons for asking the question above:
-
Primary Dilema: If the nerve is stimulated at very high discharge
probabilities, severe distortions would almost certainly occur to the
temporal envelope information; i.e., to the very information we are
attempting to commmunicate using high-rate-pulsatile stimulation! An
early statement of this potential problem can be found in the following
excerpt
from page 63 of White,
1978.
-
Secondly, the long-term function and survival of fibers receiving
high-rate, high-amplitude stimulation is of very significant concern.
See the last item below for more information.
-
Kiang and Moxon, 1972 and Merzenich, et. al. 1973 experimentally observed
that Auditory Nerve fibers and Inferior Colliculus Neurons (respectively)
exhibited graded, stochastic input-output behavior: see excerpt
from p. 47 and pp. 52-53 White,
1978.
-
These observations were similar to those of Verveen (1959). At
relatively low stimulus intensities, stimulation will cause nerve fibers
to only discharge at low-to-moderate discharge rates, NOT at high discharge
rates. If some or all patients operate at these levels of response
then "there is 'no dilema.'" To incorporate Verveen's findings
into a modified version of Hill's nerve model White (1978, pp. 54-55) added
a stochastic source: excerpt.
-
A very simple "neural ensemble model of the cochlea" which incorporated
stochastic neurons was developed in an
attempt to connect cochlear neurophysiology with some "hard-to-understand"
cochlear implant psychophysics. In 1984, White used an early version
of this cochlear model. He found that relatively low model-neuron
discharge probabilities resulted in better qualitative predictions of the
small set of psychophysical data, see the
excerpt
on page 399 from
White,
1984b.
-
-
Again in 1987-1990 RTI, M White, UCSF (i.e., Pat Leake), and scientists
from other labs teamed-up to
obtain funding to better understand the basic neurophysiology,
psychophysics, and anatomy of the deaf auditory system. M White
authored the sections of the program project grant related to neural
and psychophysical modeling of the effects of
stochastic neural responses. All 5 projects in the proposal were eventually funded in
1990 after 2 submissions and a site visit -- $4,500,000 was requested over 5
years; however, it was initially funded for only 4 years.
See
excerpts and the complete project documents from this Program Project Grant
Proposal.
As with the 1987 publication, described in the paragraph above, this planning
and proposal-writing significantly improved RTI's understanding: RTI
realized that high-pulse-rate electrical stimulation would not necessarily
cause "neurons to be driven to 'anywhere near' saturation spike rates."
As previously noted, this hard-won understanding had a profound impact on
which processors and parameter values were tested by RTI:
See this email recommendation letter from Charlie Finley, written in
2003-2004.
-
Considerably later, these stochastic models were improved and extended.
In addition, a larger range of
psychophsical data were compared to model predictions. The data was qualitatively well-predicted by Bruce, et al's
model. They conducted several studies (Bruce,
et al_1999a,
Bruce, et al_1999b,
Bruce, et al_1999c)
that investigated the neurophysiological and psychophysical consequences of "operating
at different points on the Auditory Nerve's stochastic Input-Output
function." As before, the psychophysical data are most-often well
predicted when a large percentage of the model neurons are operating at low
to moderate discharge probabilities.
-
By stimulating all, or nearly all, nerve fibers at low-to-moderate discharge
probabilities we may be able to avoid these two potential problems. Related
to long-term function and nerve survival, Tykocinski
and Shepherd found that nerve damage was minimal
or non-existent IF the stimulus intensity was adjusted so that the experimental
animals were "clearly hearing the stimulus" but the stimulus was
judged "not too loud" based on the investigators' informal "stimulus-on/stimulus-off
behavioral observations." Two good
references are:
-
Xu, J, Shepherd, R.K., Millard, R.E. & Clark,
G.M. Chronic electrical stimulation of the auditory nerve at high stimulus
rates: A physiological and histopathological study. Hear. Res. 105:
1-29, 1997.
-
M Tykocinski, RK. Shepherd, GM Clark, “Reduction
in excitability of the auditory nerve following electrical stimulation at
high stimulus rates. II. Comparison of fixed amplitude with amplitude modulated stimuli.”
Hearing Research, 112: 1–2, October 1997, pp. 147-157.