ETD

• Computational modeling
• Intracortical microstimulation
• Local Field Potential
• Phase-dependent excitability
• Visual prostheses

Intracortical microstimulation of primary visual cortex (V1) is a promising strategy to restore vision in blind individuals. However, clinical translation is limited by inter-trial response variability, whereby identical stimulation pulses can evoke inconsistent neural responses. Experimental studies have suggested that this variability may partly depend on cortical excitability, which may be reflected by the phase of ongoing local field potential (LFP) oscillations. Computational modelling offers a controlled framework to investigate this hypothesis and clarify the mechanisms linking background cortical activity to stimulation efficacy.
This thesis uses multi-compartment neuronal modelling to study phase-dependent excitability in cortical neurons under intracortical microstimulation. A morphologically detailed layer 5 pyramidal neuron was implemented in NEURON and driven by synaptic inputs coherent with resting-state LFP activity recorded from a blind human volunteer implanted with a Utah array. Single-pulse microstimulation was delivered across a set of delta-band phase timings, current amplitudes, and electrode locations, allowing the relationship between LFP phase and stimulation-evoked firing to be quantified.
Results show clear phase-dependent modulation of neuronal excitability, consistent across electrode locations but strongly dependent on stimulation amplitude. At low, near-threshold currents, the probability of activation peaks near the LFP trough, where the membrane potential is more depolarized and closer to firing threshold. In contrast, at high, supra-threshold currents, stimulation efficacy decreases near the trough, likely because increased baseline spiking induces refractory channel states or masks stimulus-evoked spikes with spontaneous activity. Overall, these findings suggest that timing stimulation pulses shortly after the delta-band LFP trough may improve stimulation efficiency. This work provides mechanistic insight into state-dependent variability in cortical microstimulation and offers design principles for future closed-loop visual prostheses.