Operant conditioning reveals task-specific responses of single neurons in a brain-machine interface

Area of Science:

• Neuroscience
• Biomedical Engineering
• Computational Neuroscience

Background:

• Volitional modulation of single cortical neurons is crucial for developing brain-machine interfaces (BMIs).
• Single-neuron control in BMIs allows for rapid learning of associations between neural activity and machine control.
• Understanding the limits of single-neuron control is essential for advancing BMI technology.

Purpose of the Study:

• To compare the effectiveness of two different brain-machine interface (BMI) tasks for single-neuron control.
• To investigate the neural mechanisms underlying performance in thresholded versus windowed activity BMI tasks.
• To determine the optimal BMI task design for leveraging cortical circuit adaptability in skill acquisition.

Main Methods:

• Rats were trained to control a one-dimensional actuator using single neurons in layer V of the motor cortex.
• Two BMI tasks were employed: a thresholded activity task and a windowed activity task, differing in reward contingency.
• Neural activity of controlled and neighboring neurons was recorded during task performance.

Main Results:

• Single neurons in motor cortex improved performance in both thresholded and windowed activity BMI tasks.
• Windowed activity task performance involved activation of neighboring, non-controlled neurons.
• Thresholded activity task performance at the time of reward involved only neurons directly controlling the BMI.

Conclusions:

• Thresholded activity single-neuron BMI implementations are superior to windowed activity tasks for exploiting cortical circuit adaptability.
• Thresholded tasks promote learning of arbitrary skills by engaging only the directly controlled neural pathways.
• These findings have significant implications for the design of more effective and efficient brain-machine interfaces.