A computational model for multiobjective optimization of multipolar stimulation in cochlear implants: An enhanced focusing approach

Multipolar stimulation has been demonstrated to improve auditory perception in individuals with cochlear implants by generating more focused electric fields through simultaneous activation of multiple electrodes. In this study, we propose a novel approach to multipolar stimulation that aims to achieve the narrowest possible pattern of current densities at target neurons. Our goal is to find the optimal profile of currents delivered by the electrodes that maximizes the focusing for a specific power consumption, or alternatively, which minimizes the power for a given focusing. To this end, we have designed two objective functions which are optimized through multiobjective evolutionary algorithms. These objective functions are evaluated using a patient-specific finite element volume conduction model that replicates the cochlear geometry and electrical behavior of the implant. Experimental results demonstrate that this approach achieves tighter current density focusing compared to phased-array stimulation, albeit with higher power consumption. Additionally, it is possible to reach non-dominated solutions that simultaneously improve the focusing and power consumption of both monopolar and phased-array stimulation. • Computational procedure combining FEM-EAs for multiobjective optimization is handled. • FEM model fitted to patient’s transimpedance matrix using surrogate based EAs. • Definition of objective functions for focusing and power consumption are provided. • Injected corner high quality solutions in EAs population improve non-dominated front. • Design principles to improve focusing of standard multipolar stimulation are proposed.