Closed-Loop Design of Brain-Machine Interface Systems
Brain-machine interface (BMI) systems show great promise for restoring motor function to patients with motor disabilities, but significant improvements in performance are needed before they will be clinically viable. Moreover, BMIs must ultimately provide long-term performance that can be used in a variety of settings. One key challenge is to improve performance such that it can be maintained for long-term use in the varied activities of daily life. Leveraging the closed-loop, co-adaptive nature of BMI systems may be particularly beneficial for meeting these challenges. BMI creates an artificial, closed-loop control system, where the subject actively contributes to performance by volitional modulation of neural activity. In this work, we explore closed-loop design methods for BMI, which exploit the closed-loop and adaptive properties of BMI to improve performance and reliability.
We use a non-human primate model system, where subjects controlled 2-dimensional virtual cursors using spiking activity recorded from chronic electrode arrays implanted in motor cortex. We first explore closed-loop decoder adaptation (CLDA), which adapts the decoding algorithm as the user controls the BMI to improve performance. We present a CLDA algorithm that can rapidly and reliably improve performance regardless of the initial decoding algorithm, which may be particularly useful for clinical applications with paralyzed patients. We then demonstrate that CLDA can be combined with neural adaptation, and that leveraging both forms of adaptation may be useful for producing high-performance BMIs that can be maintained long-term. We also show that neural adaptation may be important for BMIs used in multiple contexts by exploring simultaneous motor and BMI control. We also show that both the selection of the neural signals for control will influence BMI operation. Finally, we explore neural representations of movement dynamics to explore alternative control signals for BMI. All together, this work shows the power of closed-loop engineering of BMI systems for motor neuroprostheses.