Uniform and Non-uniform Perturbations in Brain-Machine Interface Task Elicit Similar Neural Strategies

Area of Science:

• Neuroscience
• Computational Neuroscience
• Motor Control

Background:

• Brain-machine interfaces (BMIs) offer a direct method to study neural learning and adaptation.
• Understanding how neural populations adapt to changing sensorimotor relationships is crucial for advancing neuroscience.

Purpose of the Study:

• To investigate neural adaptation mechanisms during learning using a BMI paradigm.
• To compare neural responses to uniform (visuomotor rotation) versus non-uniform (decorrelation) perturbations.

Main Methods:

• Developed a BMI paradigm to enforce learning through imposed perturbations.
• Implemented a visuomotor rotation (VMR) task with a uniform 30° output rotation.
• Utilized a decorrelation task to decouple highly correlated neuronal activity by enforcing orthogonal preferred directions.

Main Results:

• Movement errors and trials to recover baseline performance were higher in the decorrelation task compared to VMR.
• Observed decreasing trends in preferred direction changes and cross-correlation coefficients during learning across both tasks.
• Final neural tuning adaptations were dependent on the type of perturbation controller (VMR or decorrelation).

Conclusions:

• Neural populations engage similar processes when adapting to new tasks, despite variations in perturbation type.
• The brain utilizes a global process to achieve individual solutions during motor learning and adaptation.
• BMI-controlled perturbations provide insights into the flexibility and specificity of neural adaptation.