A method for removing cochlear implant artifact

When cortical auditory evoked potentials (CAEPs) are recorded in individuals with a cochlear implant (CI), electrical artifact can make the CAEP difficult or impossible to measure. Since increasing the interstimulus interval (ISI) increases the amplitude of physiological responses without changing the artifact, subtracting CAEPs recorded with a short ISI from those recorded with a longer ISI should show the physiological response without any artifact. In the first experiment, N1-P2 responses were recorded using a speech syllable and tone, paired with ISIs that changed randomly between 0.5 and 4 s. In the second experiment, the same stimuli, at ISIs of either 500 or 3000 ms, were presented in blocks that were homogeneous or random with respect to the ISI or stimulus. In the third experiment, N1-P2 responses were recorded using pulse trains with 500 and 3000 ms ISIs in 4 CI listeners. The results demonstrated: (1) N1-P2 response amplitudes generally increased with increasing ISI. (2) Difference waveforms were largest for the homogeneous and random-stimulus blocks than for the random-ISI block. (3) The subtraction technique almost completely eliminated the electrical artifact in individuals with cochlear implants. Therefore, the subtraction technique is a feasible method of removing from the N1-P2 response the electrical artifact generated by the cochlear implant.