Assessing vibrotactile feedback strategies by controlling a cursor with unstable dynamics

Area of Science:

• Neuroscience
• Human-Computer Interaction
• Rehabilitation Engineering

Background:

• Brain-computer interface (BCI) control primarily relies on visual feedback, limiting its application in tasks requiring diverted visual attention or lacking visual cues.
• Real-world motor control utilizes diverse feedback, unlike the predominantly visual feedback in current BCI systems.
• This reliance on visual feedback poses challenges for BCI usability in tasks like grasping or when visual attention is compromised.

Purpose of the Study:

• To investigate the efficacy of non-visual feedback modalities, specifically vibrotactile feedback, for BCI control.
• To compare different types of vibrotactile feedback (proportional vs. on-off) against visual feedback in a BCI control task.
• To introduce a novel task, the Critical Stability Task (CST), to assess BCI control under varying system instability.

Main Methods:

• Developed and utilized the Critical Stability Task (CST), a modified version of the Critical Tracking Task, to evaluate BCI control.
• Recruited 9 human subjects to control an unstable virtual system using free arm movements or pinch force.
• Provided subjects with visual, proportional vibrotactile, or on-off vibrotactile feedback regarding the system's state.

Main Results:

• Subjects demonstrated the ability to control the unstable system using solely vibrotactile feedback, albeit with lower performance compared to visual feedback.
• Proportional vibrotactile feedback yielded marginally better control performance than on-off vibrotactile feedback.
• Significant intra-subject variability was observed, suggesting that optimal input and feedback methods are subject-specific.

Conclusions:

• Vibrotactile feedback is a viable alternative to visual feedback for BCI control, particularly in scenarios where visual attention is limited.
• The Critical Stability Task (CST) offers a valuable complement to existing BCI assessment paradigms, enhancing the real-world applicability of BCI research.
• Tailoring feedback and input methods to individual users is crucial for achieving high-level BCI control and improving user experience.