Optimisation of and mechanisms behind neural distortion responses obtained with cochlear-implant stimulation by interleaved pulse trains

Purpose: We investigated the mechanisms behind the generation of Neural Distortion Responses (NDRs) measured with electroencephalography when interleaving amplitude-modulated (AM) pulse trains in cochlear implant (CI) users. Methods: Previous experiments measured NDRs using interleaved pulse trains in which odd- and even-numbered pulses were respectively modulated at F1 and F2 Hz, where F1=2*F2/3 and F1≈80Hz. Experiment 1 separated quadratic and cubic distortion products using a relationship of F1 and F2 such that F1 = (2*F2/3)+2. Experiment 2 retained the original F1=2*F2/3 relationship but shifted the starting AM phase of F2 relative to that of F1. Experiment 3 simultaneously separated the quadratic and cubic distortion products and varied the relative phase of F1 and F2. Experiment 4 flipped the leading polarity of the second pulse in each pulse-pair such that the central phase of the resulting pseudo-triphasic pulses was anodic. All experiments were conducted with Cochlear CI users and interpreted using two computational models of the electrically stimulated auditory nerve. Results: Both quadratic (QDR) and cubic distortion products (CDR) were observed. Varying relative AM phase had an effect on NDR amplitudes, with the largest amplitude when maxima of the 2 AM patterns were aligned. Stimulating with an anodic central phase within the AM pulse pairs resulted in stronger responses for 4 of 9 CI users. Computational modelling suggested a stronger role for refractoriness than for facilitation in generating the NDR, but failed to fully account for the effect of flipping the polarity. Conclusion: NDRs consist of both quadratic and cubic components, and are greatest when maxima of the F1 and F2 pulse trains align. Refractoriness is likely more important than facilitation for generating the NDR.