Download files
Abstract
Retinal implants aim to provide artificial vision to those profoundly blind by stimulating the residual network to elicit visual percepts. While human clinical trials have demonstrated encouraging results including the presence of visual percepts as well as partial visual restoration, the vision quality provided remains limited. One potential cause of this poor performance has been attributed to the indiscriminate activation of functionally-different retinal ganglion cell (RGC) types. To combat this problem, a promising strategy has been to design stimulation strategies that are capable of selectively, or preferentially, activating different cell types. One such approach to realise this goal has been through the use of high frequency stimulation (HFS) which was shown to be effective in preferentially activating two major retinal ganglion cell types– ON and OFF. While encouraging, the utility of the technique to target a broader range of cell types, and under different stimulation conditions and environments was still unclear.
The studies presented in this thesis were designed to improve the understanding of HFS based preferential activation. Using in vitro whole-cell patch clamp of RGCs in mice (C57BL/6J and rd1), an investigation into whether HFS could be used to preferentially activate four major RGC types namely, ON-sustained (ONS), ON-transient (ONT), OFF-sustained (OFFS), and OFF-transient (OFFT), was undertaken. Results suggested that three of the four targeted cell types could be preferentially activated against the remaining population. A subsequent study documented the responses and the preferential activation capabilities of the aforementioned cell types when the high frequencies were modulated with short stimulation bursts, varying sequence orders and in a continuous waveform. It was shown that the ON (sustained and transient) RGCs typically exhibited more consistent responses and preferential activation regions irrespective of the frequency order, or when presented as a continuous waveform. A final study examined the responses of rd1 ON and OFF RGCs to HFS both with and without the presynaptic degenerate network. The network did appear to have an effect on the HFS evoked responses, and particularly increased the variability of the responses which in turn affected the preferential activation of the cell types. Additionally, a comparison into the specific intrinsic properties between the rd1 and healthy RGCs found that these properties may differ between the cell groups. Overall, this thesis investigated the usefulness of HFS to preferentially activate different cell types and across various stimulation conditions and environments and found that HFS remains a viable stimulation technique to reduce indiscriminate activation of functionally-distinct cell types.
