An investigation of the relationships between electrotactile stimulus parameters, primary afferent response, and perceived sensation

Sammanfattning: Sensory feedback possesses the possibility of adding a new dimension to many applications, including, but not limited to, prosthetics and surgical robots for improved control, virtual reality for incorporation of another sense, and phantom limb pain reduction for amputees. Electrotactile stimulation provides a compact, light-weight, energy efficient, highly responsive, and non-invasive option for sensory feedback; however, it has been found to commonly elicit unnatural or uncomfortable sensations for the user. To address this issue, this thesis was designed to test the impact of the different electrotactile stimulus parameters – current amplitude and polarity, pulse width, frequency, and waveform – on the user’s perceived sensation and afferent neural response. The relationship between sensation and neural response was also analysed. The aim of this thesis was to create guidelines to assist in the design and use of electrotactile stimulation. Neural data and matching psychophysical data from one healthy subject and purely psychophysical data from three others were gathered while applying electrotactile stimulations of different parameter combinations on the dorsal side of the hand or lower arm. Significant (p < 0.05) correlations and differences were found in all three relationships between electrotactile stimulus parameters, primary afferent response, and perceived sensation. Current (specifically negative) or pulse width control in monophasic waveforms were deemed most appropriate in applications that relay information through varying intensity. However, monophasic waveforms produced more discomfort, rendering biphasic waveforms more suitable when mild, local, and more natural sensations are of greater importance. Surprisingly, the data also suggested higher sensitivity to positive currents. While lower values of current amplitude and pulse width increased neural spike count, stimulus frequency could reliably control neural firing at all tested frequencies. Spike counts were moderately to strongly correlated with perceived intensity; however, practically identical neural responses could elicit different sensations. High current pulses at low frequencies induced spikes with the shortest latency – but with greater risk of discomfort. Due to limitations in sample size, generalisability is limited, and this thesis should be considered a pilot study to guide future investigations. The results suggest that recording from single and multiple afferent nerve fibres simultaneously would improve the understanding of the neural population response to electrotactile stimuli. Moreover, the one-to-one neural response to electrotactile stimuli raised the question of whether an electrotactile stimulation based on a natural spike pattern could replicate the original sensation. A future study testing this hypothesis may find a new approach to designing painless electrotactile stimulations for sensory feedback use.