Hardware Acceleration of EEG-Based Emotion Classification Systems: A Comprehensive Survey

Area of Science:

• Neuroscience
• Biomedical Engineering
• Computer Science

Background:

• Growing interest in electroencephalography (EEG)-based wearable emotion classifiers for real-time patient monitoring.
• Potential applications in neurological disorders such as Amyotrophic Lateral Sclerosis (ALS), Autism Spectrum Disorder (ASD), and Alzheimer's disease.
• Need for improved healthcare outcomes and social integration for patients through emotion classification technology.

Purpose of the Study:

• To present the first hardware-focused critical review of EEG-based wearable emotion classifiers.
• To survey implementation perspectives, algorithmic foundations, and feature extraction methodologies.
• To provide a neuroscience-based analysis of current hardware accelerators for emotion classifiers.

Main Methods:

• Critical review of existing hardware platforms for EEG-based emotion classification.
• Survey of implementation perspectives, algorithms, and feature extraction techniques.
• Neuroscience-based analysis of hardware accelerators and identification of research gaps.

Main Results:

• Identified gaps in current hardware platforms for emotion classification in healthcare.
• Surveyed diverse approaches to implementation, algorithms, and feature extraction.
• Provided a neuroscience-based analysis highlighting areas for hardware accelerator development.

Conclusions:

• Several research opportunities exist for advancing EEG-based wearable emotion classifiers.
• Future directions include multi-modal hardware platforms and robust accelerators.
• Development of pre-processing libraries for universal EEG datasets is crucial.