Page 395 - Handbook of Biomechatronics

Current Advances in the Design of Retinal and Cortical Visual Prostheses 389

task force to develop substantive recommendations for the assessment base-
line (preimplant) visual status of potential patients (including specification of
the disease diagnosis and impact on visual functioning, as well as postoper-
ative visual function) (Rizzo III and Ayton, 2014).
The complexity of the phosphene percepts delivered via neuromodulation
has thus far been a key limitation to the advancement of the visual prosthesis
field that significantly influences the efficacy of these early-stage devices.
While one would hope that a grid of electrodes would produce a
corresponding perception of a grid of similarly spaced, punctate phosphenes,
the reported experiences of implant recipients have been significantly different
to this aspirational goal. While the electric fields that emanate from the
stimulating electrodes may indeed be shaped by way of charge containment,
altering the return path of the stimulation circuit, or manipulation of the
stimulus pulse profile, it remains that without an electrode-neuron interface
that approaches one-to-one, the phosphenes elicited will remain more com-
plex than in natural vision. Phosphene shapes are likely influenced significantly
by a combination of concomitant activation of multiple neurons, varying
degrees of recruitment of retinal network elements vs direct activation of
RGCs, or extraneous stimulation of passing axons from distal RGCs.
Stimulation of columns of neurons in the primary visual cortex may produce
geographic maps of phosphenes but this does not necessarily translate into syn-
thesis of images. Greater points of stimulation in the retina, geniculate nuclei,
or cortex do not necessarily mean there will be a higher resolution of visual
perception as a result. This may well be an intractable problem to do with the
fundamental approach of electrical stimulation and until efficacious steps are
taken to address this issue, this field of research may stagnate at a point close to
where we are now. However, as has been demonstrated in the cochlear
implant experience, very meaningful sensory input can be delivered in spite
of the fact that single electrodes elicit responses from multiple neurons. In the
example of the cochlear implant, the lack of a one-to-one correspondence
between electrodes and neurons has been managed rather than solved by
way of accepting that it occurs, and implementing novel stimulation strategies
that deliver the meaningful elements that describe sound. Perhaps, this is
the most plausible approach to be taken with visual prosthesis—that is, to
accept that phosphene complexity exists and focus on the delivery of the most
meaningful elements that describe the visual scene.
To conclude, the future of new visual bionic devices is directly linked to
expanding our understanding of the mechanisms and key barriers underlying
the generation of and transduction of electrical current across the biohybrid