Decoding muscle activity with local field potentials

Brain-machine interfaces (BMIs) have the potential to restore movement to people with paralyzed limbs by controlling functional electrical stimulation (FES) of paralyzed muscles. This concept recently was put into practice in monkeys using spikes to control electrical stimulation of a set of wrist muscles [1-2]. However, the viability of spike-based BMIs is limited by the inability to record spikes for more than a few years with current electrode technology. Therefore, field potentials, which may have better longevity, are being considered as an alternative. Recent studies have demonstrated the ability to decode the arm movement direction [3], and kinematics [3-4] with good accuracy using intracortical local field potentials (LFPs). Here we show that LFPs also can be used to decode the electrical activity of individual muscles with very high accuracy, similar to that obtained using spikes. This suggests that LFPs could be used in a BMI to control FES.