01 March 2010

New questions about the iPlant

A few days ago I received a set of unusually sharp thoughts and questions about the iPlant from an unknown person. My reply got rather lengthy, and I'm posting it here for future reference:


You're right that strong conditional rewarding brain stimulation, e.g. the kind that's been used to make rats exercise (Burgess et al., 1991; Garner et al., 1991), would probably create a powerful addiction in human patients and might disrupt normal interests and activities. Early on in the development of iPlants this might not be a huge concern, since patients undergoing the required surgery would already a) be in a desperate state and b) suffer from some form of addiction. I'm thinking of, for instance, morbidly obese patients in dire need of heavy physical exercise, or drug addicts who need to program themselves to stay clean.

In the context of widespread use of iPlants however surgeons will probably need to work with rewarding electrical current that is relatively weak. In my opinion, the current should if possible be just strong enough to motivate you through your two or so hours of very challenging behaviour; just enough to enable you to pull through, but no more. Rewarding brain stimulation is not an all-or-none phenomenon. For example, rats will get bored with pressing a button that delivers low intensity rewarding current after a certain number of trials, and will continue to prefer things like food, sex and play. It all depends on the intensity of the rewarding electrical current.

The problem, as far as widely used iPlants are concerned, would be getting the current intensity right, because although a lot may be learned from those first patients and from lab experiments, there would still be individual variability. The issue you raise at the end of your email may be a partial solution. That is, if we could monitor the amount of dopamine that's released in the brain in response to different intensities of electrical current, that might help surgeons find the current that works best for individual patients. However, although Paul Garris' and other labs have done some good work towards monitoring dopamine in the human brain, and an interesting study came out of MIT a few days ago on the subject, I still expect the growing experience of surgeons, combined with the reports and behaviour of the patient, to be the main way this problem is handled, just as is currently the case whenever deep brain stimulation is applied to treat psychiatric conditions. It's a hard and central problem, and my little novel-in-writing starts with an instance of this process going horribly, violently wrong.


Your second concern is equally central: how do we tie rewarding electrical current to specific activities, and ensure that the conditional rewarding brain stimulation remains conditional on the user performing the required behaviours? Again, this may not be a big problem in the early days, when patients can be required to perform work-outs, drug-tests, learning routines etc. at the hospital where they have their surgery. Patients would connect, e.g. to a rowing machine or an exercise bike, via the kind of transdermal communication equipment that's already used to re-program deep brain stimulation implants post-implantation (such a controller is shown next to an implant below). Only at the hospital would the iPlant be activated and pulses delivered, e.g. with each pull on the rowing machine or as long as the patient uses the exercise bike. Use of smoothly running equipment might not even require hospital staff supervision.



But again the question is how to this would work in the context of widespread public use. The risks of misuse and abuse are of course enormous and range from benign attempts to increase current intensity a little; through attempting to subvert the need to complete the required task to receive stimulation; to flat out mind control by external, malicious agents. Perhaps the least appealing solution would be to maintain the requirement that iPlants be activated and used only in certain supervised settings, such as hospitals and certified gyms. This would require hardware, software and policy routines that comprehensively prevent iPlants being used in other situations. Such access control is hard to ensure, even when the electronics are embedded under skin, and with widespread use some individuals bent on self-experimentation would probably sooner or later damage their implants or themselves: this is a very serious concern. It should also be noted that deep brain stimulation implants for certain psychiatric conditions already target the reward system (Greenberg et al., 2008, Schlaepfer et al., 2008, Malone et al., 2009, Bewernick et al., (2010)) and could thus in theory be hacked and used as iPlants, for good or ill.

If access control, encryption and intended use could be comprehensively ensured however, we can imagine a whole range of scenarios in which people might be able to use iPlants in the comfort of their own homes. The implant would deliver rewarding stimulation if activated by a computer command through the transdermal patch, and this command could be generated by all sorts of authorized hardware and software. An example I frequently use is a modified e-learning program designed to help people with learning difficulties by reinforcing correct answers typed into a software dialogue with rewarding current. More elaborate schemes include iPlant-driven research, where completion of some segment of a research protocol would give the user access to rewarding brain stimulation. See chapter two of the novel for more on this.


I hope this sheds at least some light on your questions, feel free to come back with more. Working on these problems in public and defining what iPlants would have to be like to function in our society is the whole point of the iPlant project. If you have a blog I hope you post your thoughts there too so other people can take part in the discussion, and with your permission I'd like to post your questions with my answers on my own blog at http://brainimplant.blogspot.com/2010/03/new-questions-about-iplant.html.

All good,
Chris
Post a Comment