Making brain-computer interfaces as reliable as muscles.

Area of Science:

• Neuroscience
• Neural Engineering
• Biomedical Engineering

Background:

• Brain-computer interfaces (BCIs) currently restore basic communication but lack the reliability of natural muscle-based actions.
• Existing BCI research primarily focuses on neural engineering to improve brain signal measurement and analysis.
• Neural engineering alone is insufficient to achieve the high reliability required for complex actions.

Purpose of the Study:

• To shift BCI research focus from solely neural engineering to include neuroscience principles.
• To develop BCI skills that emulate the reliability and adaptability of natural human skills.
• To explore the concept of 'heksors' in understanding and improving BCI functionality.

Main Methods:

• Conceptualizing natural skills as 'heksors'—neuronal networks that self-modify to maintain skill features.
• Introducing 'synthetic heksors' for BCIs, integrating neurons, synapses, and software.
• Analyzing the interaction and co-adaptation between synthetic heksors and the central nervous system.

Main Results:

• Natural skills are produced by adaptable neuronal networks ('heksors') that maintain key features through negotiated equilibrium.
• BCI-based skills can be produced by synthetic heksors, which require co-adaptation with the user's nervous system.
• Synthetic heksors can leverage multimodal sensory feedback and focus on maintaining essential skill features.

Conclusions:

• Integrating neuroscience principles with neural engineering is crucial for enhancing BCI reliability.
• Synthetic heksors offer a framework for developing BCI skills that mimic natural motor control.
• Future BCIs could achieve muscle-like reliability through a deeper understanding of skill acquisition and maintenance.