Welcome to Cochlear Concepts, LLC.
Some of the links may not be active--my apology for this. Thank you for your interest in this site.
Eric L. Carmichel, owner, can be reached at
eric@elcaudio.com


Our mailing address:


1534 N Dorsey Ln
Tempe, AZ 85281


 

 

 

 

The link below goes to our parent site.

ELC Audio Engineering (click on image)

home | email  

Welcome to Cochlear Concepts™ LLC -- Home of the Veridical* Reality System (VRS)

News updates

I recently received a letter regarding a patent application that was initiated in 2007. Until some form of intellectual property protection was in place, I hesitated discussing new ideas regarding cochlear implants (CIs) to outside parties. I was ready to present the idea to the preeminent cochlear implant companies (Cochlear Corporation, Med-El, and Advanced Bionics) when word was received that another company was interested in purchasing the patent--this came after applying for a provisional patent. Interestingly, the company that purchased the design has no interest in CIs (at least not to my knowledge). But, as patents go, patent attorneys attempt to broaden the scope of a patent to cover as many features as possible. The downside to this is that using one feature of a patented item can end up in a legal battle if used in another application. Now, almost five years later, we can openly discuss aspects of the original patent application because any attempt to use the idea(s) in a commercial product requires licensing.

Before embarking on a doctoral program of study, I had the idea that conventional interleaving strategies and discrete electrodes wouldn't be necessary if we could send a virtual electrical pulse down an implanted device. Of course, a slow-traveling electric field is impossible, but an arrangement of discrete electrodes can be oriented to approximate a traveling wave. At first glance, this might appear to be identical to the conventional use of electrodes along an implanted array. Part of what makes our idea unique is implementation of a single, time-sweepable filter in lieu of contiguous filters. The filter's window has to remain open longer for low-frequencies sounds (or, more accurately, a sound's analogous electrical current), so the sweep is necessarily non-linear. This is akin to the slowing down of a traveling wave at the apical end of the basilar membrane: More time is needed for good low-frequency resolution (same goes for digital filtering, only more samples are needed). A control voltage (Vc) determines the time, and therefore place-frequency, that the filter is receptive to. The Vc is really little more than a modified sawtooth wave with a nonlinear ramp. An array of comparators, easily implemented via a programmable logic chip (at least what was used in the demo device), determines whether the input signal is of sufficient amplitude at both a specified TIME and frequency in order to allow output to a monostable multivibrator. In its simplest form, the multivibrator provides a fixed-width, fixed-amplitude pulse. Amplitude and pulse width need not be fixed, but the initial idea was to mimic the duration (including refractory period) of "typical" action potentials.

This explanation is simplified, but the filter and logic circuitry works in concert with a pseudo-active array. The inclusion of the active array broadens the scope of the patent/invention. A high electrode count is accomplished via multiplexing along the array (not ahead of it). This way, only a single wire (or circuit trace) carries the electrode current. The trace itself is a squiggly (or zig-zag shaped) buss. The opposite side of a thin film polymer has an overlapping and offset squiggly return buss; consequently, the electric field created along the buss's length effectively cancels. Minute currents are carried in parallel form to MOSFET switches situated along the array. Flexible, thin film semiconductor substrates make this physically realizable and relatively easy to implement. With five low-current carrying traces, 2**5 = 32 electrodes are controlled. Ten traces can control over 1000 electrodes. Basic multiplexing here. But remember, the cross-channel interaction (or bleedover) is minimal because the parallel bus carries such small currents. Also, inter-trace capacitance isn't too much of a problem because the frequency of operation is low. The preservation of a "clean" square wave is accomplished because the power buss carries a squared waveform instead of a steady-state dc voltage. Again, only a single power buss carries the stimulus current needed to power each electrode. Multiplexing such as this isn't new... until we add optical routing and optoelectronic switches. Phototransistors that are sensitive to specific wavelengths can be used, while a single "control" fiber optic (or substrate) can simultaneously carry multiple colors. The colors replace the need for parallel control lines. The downside to this is that optoelectronic switches generally require greater voltages to operate than MOS, so one has to be concerned with biological hazards associated with even minute voltages. Fortunately, lower-voltage devices are always surfacing. Note that this isn't at all akin to optical innervation of the auditory nerve; I had to explain this to my patent attorney because optical innervation has already been used.

A channel-picking strategy is intrinsic to the CI processor design described herein. Filtering is simplified by virtue of a single, time-sweepable filter. Loudness summation (or its electrical-current equivalent) occurs because the the number of "ON" electrodes increases as a center frequency's amplitude increases. Roughly speaking, there's more neural firing from the neurons flanking a center frequency when the traveling wave increases in amplitude. Because the above circuitry works in real time, localization ability may very well be improved. The initiation of the control voltage, Vc, begins with independent triggering at each ear (regardless of a sound's frequency or constituent frequencies). The initial prototype of the complete device (processor and electrode array) was built using National Instruments' Electronics Simulation Software (visit this LINK for info regarding NI's Electronics Workbench).

Original description (circa November 2011) follows:

Cochlear Concepts is currently developing a new cochlear implant (CI) processing strategy and active-circuit electrode array [see Footnote 1]. The array and its complementary processor allow for a high channel count but with minimal crosstalk between channels. The overall design results in a virtual “electrical transmission line” along the implanted membrane. The active array doesn't suffer from current smearing when a large number of electrodes are simultaneously energized. Furthermore, our strategy does not require superposition to create virtual channels made from widely-spaced, discrete electrodes.
The high channel count isn’t merely an attempt to achieve better frequency discrimination based on the place theory of frequency coding; instead, our implant design and multiplexing strategy follows many established principles of psychoacoustics--some of which are ignored in other CI designs. The high-density electrode will improve dynamic range and speech quality (as well as frequency discrimination) when one CI is used, and nearly-normal interaural time difference (ITD) coding when two implants are used. Additionally, there are theoretical benefits to our high channel density design when severe nerve ganglion damage is present. Because the implant is under development, studies to date have used only simulations based on software and hardware models. Because our design works on different principles than other implants, it is important to note that our acoustic simulations are considerably different from noise-band vocoder demonstrations used to simulate CI listening. Let’s just say that Cochlear Concepts isn’t simply “refining” established implant strategies: We’re defining on a new strategy!

To demonstrate improvements in any CI design (regardless of manufacturer), Cochlear Concepts is also focusing on controlled, but real-world, listening environments for researchers. The results of our research endeavors are intended to help EVERYONE who is involved with CI research.

Footnote 1: In our proposed design, much of the active circuitry lies on the implanted array. Newer technologies have made it feasible to fabricate flexible thin-film circuits while planar, and then folding or rolling the circuits to conform to different shapes or volumes.