Neural Prosthetics:A Review of Empirical vs. Systems Engineering Strategies

Abstract

Implantable electrical interfaces with the nervous system were first enabled by cardiac pacemaker technology over 50 years ago and have since diverged into almost all of the physiological functions controlled by the nervous system. There have been a few major clinical and commercial successes, many contentious claims, and some outright failures. These tend to be reviewed within each clinical subspecialty, obscuring the many commonalities of neural control, biophysics, interface materials, electronic technologies, and medical device regulation that they share. This review cites a selection of foundational and recent journal articles and reviews for all major applications of neural prosthetic interfaces in clinical use, trials, or development. The hard-won knowledge and experience across all of these fields can now be amalgamated and distilled into more systematic processes for development of clinical products instead of the often empirical (trial and error) approaches to date. These include a frank assessment of a specific clinical problem, the state of its underlying science, the identification of feasible targets, the availability of suitable technologies, and the path to regulatory and reimbursement approval. Increasing commercial interest and investment facilitates this systematic approach, but it also motivates projects and products whose claims are dubious.

Figure 1

Most currently approved and clinically successful neural prostheses use fairly bulky hermetic titanium cans with multiple feedthroughs and polymerically encased flexible leads to relatively large platinum-alloy electrodes, similar to cardiac pacemakers of the 1970s.

Figure 2

Epiretinal visual prosthesis consists of a highly flexible electrode array tacked onto the inner surface of the retina, which is connected to a multichannel stimulator that is strapped to the eyeball and receives power and command signals from an external video camera and video processing unit (VPU) via an externally generated radio-frequency magnetic field. Photos used with permission of the manufacturer, Second Sight Inc., Sylmar, CA 91342.

Figure 3

Functional electrical stimulation (FES) research system for paraplegic locomotion, adapted from [58].

Figure 4

Penetrating microelectrode arrays to stimulate and/or record activity in individual neurons in peripheral nerves (a) [114] and cerebral cortex (b) [115]. Command signals from peripheral motor axons could be used to control prosthetic limbs, and stimulation of somatosensory afferents could provide perceptual feedback. Command signals from motor cortex can be used to control prosthetic limbs for amputees or neuromuscular stimulation for quadriplegic patients. Illustration provided courtesy of The Johns Hopkins University Applied Physics Laboratory.

Figure 5

Two types of single-channel monolithic neuromuscular stimulators that can be implanted into individual muscles or near nerves, where they generate well-controlled stimulus pulses that are powered and commanded by an externally generated radio-frequency magnetic field.