One way for quadriplegic patients to feed themselves and perform other motor activities is to connect an interface to neural populations, and by using a complicated algorithm decode task-oriented activity and determine the parameters necessary for these external devices to work. In a fascinating experiment from the Washington National Primate Research Center published in Nature, Moritz et al. have shown that it is possible to create artificial connections between haphazardly chosen motor cortex cells to multiple muscles and in order to restore movement to a paralyzed arm.
The researchers used operant conditioning to reward monkeys for specified motions in the wrist. Within 10 minutes of the first practice session the monkeys had volitional control of the firing of most neuron cells involved, which is incredible. The researchers then increased the complexity of the experiment so that the monkeys were rewarded for maintaining a certain amount of torque in the wrist, which requires that the monkeys be able to both control the extension of the muscles and relax them on cue. The monkeys were able to do this as well, as each individual cell increased in control with practice.
This technique may prove to be a revolutionary strategy for recreating motor control in paralyzed patients, with signal pathways made directly from cells to muscles instead of the typical approach of attempting to derive neural signals from a neuron population at large. Their method piggybacks on the natural plasticity of individual cells to tremendous effect. The optimal strategy now probably would combine this new technique with the old method somehow. The paper is short and fascinating, required reading for anybody interested in brain computer interfacing at all.
Moritz TC, Perlmutter SI, Fetz EE 2008 Direct control of paralysed muscles by cortical neurons. Nature advanced online publication October 15. doi:10.1038/nature07418.