The promise and threat of modern cybernetics.

Cybernetics has historically been defined as "Communication and control in human and machine." In modern terms, this clearly encompasses the increasingly important area of research in which human and technology are merged into an overall system. It is worth emphasizing that from a cybernetics viewpoint, it is the functioning of the overall system that is of prime importance. In this paper, some of the realistic opportunities and dangers of this concept are considered.

[ILLUSTRATION OMITTED]

From a medical and ethical standpoint, it is apparent that the reasons for a specific human-technological merger could be broadly classified by some as being for either therapeutic purposes or for enhancement. However, to anyone with even a basic understanding of the field, this is an extremely naive viewpoint, in that there is no clear and present line which divides the two. For example, what might be a therapeutic connection for one person may well be an enhancement for another.

The focus of attention here is placed on the use of technology to alter the capabilities of a human. While external devices such as eye glasses clearly achieve this, these are distinctly temporary measures. Rather, we wish to look at the use of invasive implant techniques to improve the range of abilities of an individual.

Medical Opportunities

Implants can take on a variety of forms. Perhaps best well known would be heart pacemakers and artificial hips. Here, however, I wish to consider the class of implant which can potentially have the most profound effect, to the extent of completely altering the characteristics and abilities of an individual. One example in widespread use is the cochlear implant, employed to assist those with hearing loss. Artificial retinas are also under research, (1) to replace a malfunctioning visual system.

In general terms, signals can be monitored, by means of implant technology, from the human brain and nervous system to provide an extra communication channel. In chronically paralyzed individuals, this can mean allowing them direct control of a switch or cursor which is not otherwise possible, thereby potentially increasing their quality of life. (2) For an amputee, this could permit direct neural control of a replacement limb. (3) Alternatively, neurons can be stimulated by the same method to allow for sensory replacement or feedback from external influences.

Electronic neural stimulation has been extremely successful in a range of areas, including the treatment of Parkinson disease symptoms. With Parkinson disease, diminished levels of the neurotransmitter dopamine cause over-activation in the ventral posterior nucleus and the subthalamic nucleus, resulting in slowness, stiffness, gait difficulties and hand tremors. By implanting electrodes into the subthalamic nucleus to provide a constant stimulation pulse, this over-activity can be inhibited, allowing the patient, to all external intents and purposes, to function normally. (4)

It is still not clear exactly how high frequency deep brain stimulation actually alleviates movement disorders. The basal ganglia circuitry has a tendency to develop spontaneous oscillations due to the membrane properties of many of the excitatory neurons. It is felt that in Parkinson disease, these rhythms become larger, slower and uncontrollable because of the lack of dopamine. With electrical stimulation, the neurons become locked to the stimulation frequency and lose any tendency for spontaneous oscillation. Therefore, this type of stimulation is probably highly effective because it takes command of these networks, preventing them from relapsing into the slow synchronous cycles that appear to cause motor symptoms.

Some of the most impressive human research to date has been carried out using the microelectrode array (Fig.). The individual electrodes are only 1.5 mm long and taper to a tip diameter of less than 90 [micro]m. Although a number of nonhuman trials have occurred, (5) human testing is currently limited to two studies. In the second of these studies, the array has been employed in a recording-only role (6) with outgoing signals being decoded to direct cursor movement. This has enabled an individual to position a cursor on a computer screen, using neural signals for control combined with visual feedback. The first use of the microelectrode array is discussed in the following section.

[FIGURE OMITTED]

Therapy Plus

The microelectrode array has been implanted into the median nerve fibers of a healthy human individual (myself) to test bidirectional functionality in a series of experiments. (7,8) As well as monitoring neural signals via the implant to neurally control remote technology, a stimulation current was directed onto the nervous system to allow information to be transmitted to the user (me).

The microelectrode array employed in this work contains a total of 100 electrodes which, when implanted, become distributed within the nerve fascicle. In this way, it is possible to gain direct access to muscle spindles, motor neural signals to particular motor units and sensory receptors. Essentially, such a device allows a bidirectional link between the human nervous system and a computer.

The array was pneumatically inserted into the median nerve such that the body of the array sat adjacent to the nerve fibers. With the median nerve fascicle estimated to be 4 mm in diameter, the electrodes penetrated well into the fascicle.

After extensive experimentation, over a 6-week period, it was found that injecting currents below 80 [micro]A onto the median nerve fibers had little perceivable effect. Between 80 [micro]A and 100 [micro]A, all the functional electrodes were able to produce a recognizable stimulation, with an applied voltage of 40 to 50 V, dependent on the series electrode impedance. Increasing the current above 100 [micro]A had no apparent additional effect; the stimulation switching mechanisms in the median nerve fascicle exhibited a nonlinear thresholding. In all subsequent successful trials, the current was applied as a biphasic signal with pulse duration of 200 microseconds and an interphase delay of 100 microseconds.

By this means, a number of experimental trials have been successfully concluded (9,10):

Specifically:

1. Extrasensory (ultrasonic) input was successfully implemented and utilized.

2. Extended control of a robotic hand across the Internet was achieved, with feedback from the robotic fingertips being sent back as neural stimulation to give a sense of force being applied to an object (this was achieved between New York, USA and Reading, UK).

3. A primitive form of telegraphic communication directly between the nervous systems of two humans was performed.

4. A wheelchair was successfully driven around by means of neural signals.

5. The color of jewelry was changed as a result of neural signals--as indeed was the behavior of a collection of small robots.

In each of these cases, a clear therapy/enhancement dichotomy exists. For example, in Experiment 1 the technology could be used to provide an alternative sense for an individual who is blind; for others, however, it would clearly be seen as an extra sense. Meanwhile, in Experiment 4, the neural control of a mobile platform differs only trivially, whether the vehicle is a wheelchair for an individual who is paralyzed or an automobile for a healthy subject. In both cases, the technology is the same, and from a cybernetic viewpoint, the end application is of minor technical concern. Essentially, if the technology has been developed for one purpose, it is a relatively trivial exercise for anyone to use it for an alternate reason.

One further aspect of the research was to investigate the human body's acceptance or rejection of such an implant. At the end of the 3-month long trial, no infection whatsoever was witnessed during the course of the experiment and during extraction, it was observed that body tissue had grown around the array holding it in its original place.

Discussion

It is clear that the interaction of electronic signals with the human brain can cause the brain to operate in a distinctly different manner. Such is the situation with the stimulator implants that are successfully used to electronically counteract the tremor effects associated with Parkinson disease. Such technology can also potentially be employed to modify the normal functioning of the human brain and nervous system in a number of different ways. Perhaps understandably, invasive implants are presently far less well developed than their external electrode counterparts. However, numerous animal trials have been carried out as well as a small number of human trials, the more pertinent of which have been indicated here.

The potential for implant technology applications in individuals with paralysis or other motor dysfunction is enormous, as cerebral functioning can generate command signals, despite impairment of the motor neural pathways. The major role of the implant is to either relay a signal of intention to the appropriate actuator muscles, or to reinterpret the neural signals to operate connected technology, thereby acting as an enabler. In these situations, it may be that no other medical "cure" is available, which presents a huge motivation for an invasive implant solution for the millions of affected individuals. Clearly, though, bidirectional signaling is important, not only to monitor and enact an individual's intent but also to provide feedback on that individual's resultant interaction with the real world. For grasping, walking and even as a defensive safety stimulant, feedback is vital.

Where invasive interfaces are employed in human trials, it is in most cases the situation that a purely therapeutic scenario exists. In a small number of instances, such as use of the microelectrode array as an interface, an individual has been given different abilities, something which opens up the possibilities of an extended view of therapy as described above. These latter cases, however, raise more topical ethical questions with regard to the need and use of implant technology. What might be seen as a new means of communication for a paralyzed individual or new sensory input for the blind, may also be seen by some as an unnecessary extra for another individual, even though it may provide novel commercial opportunities.

While there are still many technical problems to be overcome in the development of implants, significant recent experimental results have indicated that a sufficient technological infrastructure now exists for further major advances to be made. Although a more detailed understanding of the underlying neural processes will be needed in the years ahead, it is not felt that this will present a major hold up over the next few years; rather, it will provide an avenue of research in which many new results will appear through trials and experimentation, possibly initially through animal studies, although it is only through human studies that a full analysis can be made and all encompassing conclusions can be drawn. Nevertheless, the topic opens up various ethical questions that need to be addressed and as such, research in this area should only proceed in light of a pervasive ethical consensus.

By means of the implant technology described here, the opportunities are many. Assistance appears to be possible, in one way or another, for those with a spinal injury, perhaps rerouting the nervous system or as an aide to regeneration, for those with multiple sclerosis and even motor neuron disease. But it is the extension of human abilities that is felt to be particularly exciting and challenging. It took 6 weeks for my brain to learn to reliably recognize new signals input via the peripheral nervous system--thereby allowing for novel methods of feedback, sensing and information input. Also, the direct use of neural signals to control all manner of technology, over vast distances, must necessarily change our views on the capabilities of an implanted human--if the nervous system is linked directly with a computer and the internet, where does an individual body stop?

Merging humans with machines is the ultimate technological development in cybernetics. Such a step holds tremendous promise in that through the use of implant technology, the possibility exists to assist those with a disability. However, the same technology also offers the possibility of upgrading humans, endowing them with enhanced abilities. This can also be seen as a promising move in that it forms an exciting evolutionary direction and opens up many new opportunities. For others, however, the concept of upgrading humans could be seen as a threat.

Already, extra senses have been shown to be operable and experiments have been performed with a new form of 'telepathic' communication. Adding to memory and long distance control appear to be relatively straightforward next steps. Clearly, those who are upgraded may well regard ordinary humans with sympathy, but it will be obvious that the abilities of such humans are relatively limited. Hence, while the enhancement of humans holds tremendous promise for those who would like to be part of the new wave, it poses a considerable threat for those humans who wish to remain just as they are.

References

1. Rizzo J, Wyatt J, Humayun M, et al. Retinal prosthesis: an encouraging first decade with major challenges ahead. Opthalmology 2001;108:13-14.

2. Kennedy P, Andreasen D, Ehirim P, et al. Using human extra-cortical local field potentials to control a switch. J Neural Eng 2004;1:72-77.

3. Warwick K, Gasson M, Hutt B, et al. The application of implant technology for cybernetic systems. Arch Neurol 2003;60:1369-1373.

4. Pinter M, Murg M, Alesch F, et al. Does deep brain stimulation of the nucleus ventralis intermedius affect postural control and locomotion in Parkinson's disease? Mov Disord 1999;14:958-963.

5. Branner A, Normann R. A multielectrode array for intrafascicular recording and stimulation in sciatic nerve of cats. Brain Res Bull 2000;51:293-306.

6. Donoghue J, Nurmikko A, Friehs G, et al. Development of neuromotor prostheses for humans. Suppl Clin Neurophysiol 2004;57:592-606.

7. Warwick K, Gasson M, Hutt B, et al. Thought communication and control: a first step using radiotelegraphy. IEE Proceedings-Communications 2004;151:185-189.

8. Gasson M, Hutt B, Goodhew I, et al. "Invasive neural prosthesis for neural signal detection and nerve stimulation", Proc. International Journal of Adaptive Control and Signal Processing 2005;19(5): pp 365-375.

9. Warwick K, Gasson M, Hutt B, et al. An attempt to extend human sensory capabilities by means of implant technology. Presented at: Proceedings IEEE International Conference on Systems, Man and Cybernetics, October 2005; Hawaii, pp 1663-1668.

10. Warwick K, Gasson M. Practical interface experiments with implant technology. In: Sebe N, Lew MS, Huang TS. (eds). Computer Vision in Human-Computer Interaction: ECCV 2004 Workshop on HCI. Prague, Czech Republic, May 16, 2004, Proceedings (Lecture Notes in Computer Science). New York, Springer, 2004, Vol. 3058, pp 7-16.

Kevin Warwick, PhD

From the Department of Cybernetics, School of Systems Engineering, The University of Reading, Reading, United Kingdom.

Reprint requests to Professor Kevin Warwick, PhD, Cybernetics at Reading, School of Systems Engineering, The University of Reading, Reading, United Kingdom. Email: [e-mail]