Computational simulation and optimization of multipolar stimulation in cochlear implants

Abstract

Objective: This presentation proposes a method for multipolar stimulation that aims to achieve the narrowest possible pattern of current densities at target neurons in cochlear implants. Multipolar stimulation results in increased power consumption due to the simultaneous generation of intracochlear currents by multiple electrodes. Our objective is to identify those electrode current profiles that maximize focusing while minimizing power consumption. Methods: We have developed two objective functions: one for focusing and the other for power dissipation in the electrolyte. They are evaluated using a finite element-based volume conduction model that replicates a cochlear geometry. Several conductivities of the model are adjusted to match the transimpedance matrix measured in a patient, using an inverse optimization procedure. Once the model is calibrated, we apply multi-objective optimization using evolutionary algorithms. Results: Computational results reveal non-dominated solutions (optimal from a multi-objective optimization perspective) corresponding to a cochlear geometry with perimodiolar insertion of the electrode array, including: 1) a solution demonstrating that the modeled current density at target neurons can achieve tighter focus than a phased-array stimulus, albeit with higher power consumption; 2) solutions that simultaneously improve the focusing and power consumption of the phased-array stimulus. Conclusions: We have successfully obtained solutions that enhance the focusing of phased-array stimulation and diminish its power consumption, offering advantageous current profiles stimulation compared to phased-array.