Looking at the stimulation level data as a function of pulse duration, there is a nonconstant amount of energy required to evoke a CAP of the same magnitude, as we vary the pulse duration. It has been suggested that optical stimulation of neural activity occurs via a thermal mechanism [5
]. With all other parameters being equal (e.g., wavelength, tissue hydration), the amount of energy absorbed into the tissue governs the temperature increase subsequent to the light absorption. Our results then indicate that it is not the total energy and, therefore, total thermal rise, that governs optical stimulation in the gerbil cochlea. It could be the time over which this energy is deposited that governs optical stimulation. When the stimulation level data  are calculated in terms of peak optical power, we see another interesting trend of data (). In general, the peak power is constant across all pulse durations, except for 35 μ
s, in which it increases. This trend could indicate that the time over which the energy is deposited is more important for laser stimulation. However, it is also possible that the rise time of the optical pulse is significant in optical stimulation of the cochlea. At present, we are limited by the fixed pulse rise time of our current laser, but we will investigate this parameter in future experiments.
Fig. 6 Peak power at stimulation level as a function pulse duration. The peak power measured for a 50-μ CAP is constant for pulse durations 100 μs−1 ms, but increases at a pulse duration of 35 μs. These are the same experiments (more ...)
When examining the CAPs, it is clear that the CAPs evoked from laser pulses at 35, 100, and 300 μs are all primarily composed of N1/P1 peaks. However, at 600μs and 1-ms pulse durations, there is a large secondary peak that is evident in the CAPs. The large secondary peak could account for the variability in the error estimates seen in the stimulation measurements at these pulse durations. The second maximum could be explained by a second action potential from some of the nerve fibers that contribute to the CAP. There’s also the possibility that a different subpopulation of nerve fibers that do not contribute to the CAP are responsible for the second maximum. In depth studies of single auditory nerve fibers are underway and will help determine the sources contributing to the CAP.
Other data that were acquired during these experiments examined the influence of wavelength on the evoked response, at a pulse width of 35 μ
s. Starting at the shortest possible penetration depth and increasing, the CAP amplitude increased significantly for approximately 300 μ
m, then almost plateaued for the remaining increase in OPD (). (OPD is defined as the distance over which the incident light is reduced in magnitude by 1/e
and is a function of the absorption coefficient of the tissue.) The steady increase in CAP amplitude is most likely due to an increasing number of neurons in the optical path that receive suprathreshold irradiation as the penetration depth increases. Notice that the OPD distance over which the CAP increases, 300 μ
m, corresponds well with the approximate length of the optical path through the spiral ganglion cell population in the upper basal turn of the gerbil cochlea, ~ 250 μ
m. Further into the optical path from the spiral ganglion cells in the upper base, there is supporting tissue that is nonneural. Therefore, as the penetration depth is increased beyond the initial spiral ganglion cell population, there are no further neurons to receive suprathreshold irradiation, which likely contributes to the plateauing of the data. For use in human cochlear implants, it is possible that a different wavelength, corresponding to a longer OPD, will be better suited to stimulating the auditory neurons as the human cochlea is larger than that of a gerbil [26
]. However, this parameter will ultimately be determined by the location for the optical implant array in the scala tympani.
The continual stimulation experiments conducted for 6 h at 13 Hz, the upper limit of this laser, indicate that we can evoke a very stable CAP. The CAP is a very sensitive marker for the physiological state of the cochlea, such as cochlear damage and changes in temperature [27
]. Acoustic tones activate a segment of neurons along the length of the cochlea and the CAP thresholds reflect the synchronous activation of neurons in this small cochlear segment. In contrast to stimulating with acoustic stimuli, for optical radiation, neural stimulation occurs over the depth of the optical path and the corresponding evoked CAP amplitudes are governed by the number of cells that receive suprathreshold stimulus levels. In case optical irradiation damages the cells, the number of neurons that respond to optical radiation should decrease and consequently the peak-to-peak amplitude of the CAP. Variability in CAP amplitude towards the end of the experiments may be attributed to changes in cochlear function caused by the surgery, by inadvertent cooling of the cochlea, or by effects on the animal from anesthesia. The same type of effect was seen in the extended acoustic stimulation previously [3
In order to clinically implement an optical cochlear implant, certain design criteria must be met first. The most important objectives are 1) improved spatial selectivity of stimulation, 2) ability to optically stimulate at rates that mimic the native firing rates of the auditory system, and 3) safe optical stimulation for the aforementioned parameters. In terms of the spatial objective, mid-infrared light does not spread laterally upon incidence into tissue as does electric current. Only tissue that is directly in the optical path absorbs the light [29
]. We have demonstrated selective optical stimulation of the auditory system in comparison to electric stimulation [4
Single auditory neurons have maximum firing rates of ~300–400 Hz [31
]. To adequately restore a sense of hearing, an optical cochlear implant will need to operate up to this stimulation rate without causing damage. To make a conservative estimate of the likelihood of heat accumulation following optical stimulation, we can use an equation for the tissue thermal relaxation time, τtherm
, which describes the time-dependent heat transport out of the tissue by diffusion following laser irradiation (this term is independent of the incident laser energy)
is the temperature conductivity, which has a value of ~ 1.4*10−7
/s in most tissues [33
], and α
is the wavelength-dependent absorption coefficient of the material. Based on the wavelengths used here, we would arrive at a thermal relaxation time of ~0.5 s, which corresponds to 2 Hz. By these conservative calculations, 2 Hz would be the upper limit to the safe region in which we could stimulate without causing a build-up of heat and, therefore, tissue damage. However, we have shown that it is possible to optically stimulate the cochlea at much higher rate of stimulation with no apparent thermal effects. The ability to stimulate at high rates with no damage is likely due to the very low energies that we are using. The types of laser-tissue interactions for which the thermal relaxation equation is typically used are higher powered irradiation, for applications such as coagulation and ablation. It is also possible that the perfusion of the cochlea, which is not incorporated into the equation, increases the ability to dissipate thermal energy, thereby decreasing the time parameter above. We have recently acquired a pulsed diode laser that can operate up to 1-kHz repetition rate and experiments are underway to examine high rate optical stimulation of the auditory system. While we have measured a steady CAP in response to optical stimulation at 400 Hz over several hours, we are conducting further studies to validate these data using single fiber population studies, which will be the topic of a future manuscript. We will also explore the safety of an optical cochlear implant with chronic animal studies, in which the cochlea is optically stimulated over a period of months. Electrophysiology and histology data will reveal the safe parameters for optical stimulation in a cochlear implant.
To apply optical stimulation for use in other neuroprostheses, two general goals must be met: 1) improvement of optical stimulation versus electrical stimulation for the application; 2) safely optically stimulate to achieve desired function. Other likely candidates for optical stimulation in a sensory system include the dorsal root ganglia, which carries sensory information such as tactile, pain, and temperature sensations from the periphery to the central nervous system; and the retinal ganglion cells, which transmit the signals from the photoreceptors to the brain. Each of these applications would require unique design criteria, for instance the physical size of an implant, the method of delivering the light, and the as-yet unknown optical parameters required to evoke a response from the neurons. Introducing optical stimulation into a neuroprosthesis for the tactile sensory system could allow information feedback for individuals with motor prostheses. A retinal prosthesis incorporating optical stimulation could improve the resolution of vision that is restored to the blind individual. These are just a few of the many applications of optical stimulation of neural tissue.