Optimizing Stimulus Energy for Cochlear Implants with a Machine Learning Model of the Auditory Nerve
Authors:
Jacob de Nobel,
Anna V. Kononova,
Jeroen Briaire,
Johan Frijns,
Thomas Bäck
Abstract:
Performing simulations with a realistic biophysical auditory nerve fiber model can be very time consuming, due to the complex nature of the calculations involved. Here, a surrogate (approximate) model of such an auditory nerve fiber model was developed using machine learning methods, to perform simulations more efficiently. Several machine learning models were compared, of which a Convolutional Ne…
▽ More
Performing simulations with a realistic biophysical auditory nerve fiber model can be very time consuming, due to the complex nature of the calculations involved. Here, a surrogate (approximate) model of such an auditory nerve fiber model was developed using machine learning methods, to perform simulations more efficiently. Several machine learning models were compared, of which a Convolutional Neural Network showed the best performance. In fact, the Convolutional Neural Network was able to emulate the behavior of the auditory nerve fiber model with extremely high similarity ($R^2 > 0.99$), tested under a wide range of experimental conditions, whilst reducing the simulation time by five orders of magnitude. In addition, we introduce a method for randomly generating charge-balanced waveforms using hyperplane projection. In the second part of this paper, the Convolutional Neural Network surrogate model was used by an Evolutionary Algorithm to optimize the shape of the stimulus waveform in terms energy efficiency. The resulting waveforms resemble a positive Gaussian-like peak, preceded by an elongated negative phase. When comparing the energy of the waveforms generated by the Evolutionary Algorithm with the commonly used square wave, energy decreases of 8% - 45% were observed for different pulse durations. These results were validated with the original auditory nerve fiber model, which demonstrates that our proposed surrogate model can be used as its accurate and efficient replacement.
△ Less
Submitted 14 November, 2022;
originally announced November 2022.
A 0.042 mm^2 programmable biphasic stimulator for cochlear implants suitable for a large number of channels
Authors:
W. Ngamkham,
M. N. van Dongen,
W. A. Serdijn,
C. J. Bes,
J. J. Briaire,
J. H. M. Frijns
Abstract:
This paper presents a compact programmable biphasic stimulator for cochlear implants. By employing double-loop negative feedback, the output impedance of the current generator is increased, while maximizing the voltage compliance of the output transistor. To make the stimulator circuit compact, the stimulation current is set by scaling a reference current using a two stage binary-weighted transist…
▽ More
This paper presents a compact programmable biphasic stimulator for cochlear implants. By employing double-loop negative feedback, the output impedance of the current generator is increased, while maximizing the voltage compliance of the output transistor. To make the stimulator circuit compact, the stimulation current is set by scaling a reference current using a two stage binary-weighted transistor DAC (comprising a 3 bit high-voltage transistor DAC and a 4 bit low-voltage transistor DAC). With this structure the power consumption and the area of the circuit can be minimized. The proposed circuit has been implemented in AMS 0.18um high-voltage CMOS IC technology, using an active chip area of about 0.042mm^2. Measurement results show that proper charge balance of the anodic and cathodic stimulation phases is achieved and a dc blocking capacitor can be omitted. The resulting reduction in the required area makes the proposed system suitable for a large number of channels.
△ Less
Submitted 29 January, 2015;
originally announced February 2015.
