EEG-based hybrid QWERTY mental speller with high information transfer rate

Area of Science:

• Neuroscience
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Brain-computer interfaces (BCIs) offer communication pathways for individuals with severe motor impairments.
• Traditional BCIs often rely on single paradigms like P300 or SSVEP, which can have limitations in accuracy and speed.
• Developing hybrid BCIs that leverage the strengths of multiple paradigms is crucial for improving assistive communication technologies.

Purpose of the Study:

• To introduce and evaluate a novel hybrid Brain-Computer Interface (BCI) QWERTY speller.
• To combine P300 and Steady-State Visually Evoked Potentials (SSVEP) features within a single BCI system.
• To assess the performance improvements of the hybrid speller compared to traditional P300 and SSVEP spellers.

Main Methods:

• A hybrid BCI paradigm was developed, integrating P300 and SSVEP detection.
• P300 signals were utilized as a time-division multiplexing index to reduce the number of required SSVEP frequencies.
• Each flickering frequency was assigned a unique color to further enhance system accuracy and user experience.

Main Results:

• The hybrid QWERTY speller achieved an average classification accuracy of 96.42% across 20 subjects.
• A mean information transfer rate (ITR) of 131.0 bits per minute was recorded during free spelling trials.
• Statistical analysis confirmed significantly superior performance (accuracy and ITR) compared to traditional P300 and QWERTY SSVEP spellers.

Conclusions:

• The developed hybrid QWERTY speller demonstrates superior performance over conventional P300 and QWERTY SSVEP spellers.
• This hybrid approach offers a more efficient and accurate communication method for BCI users.
• The findings suggest a promising direction for the advancement of assistive communication technologies through hybrid BCI design.