Synthetic Data Generation for Brain-Computer Interfaces: Overview, Benchmarking, and Future Directions

2026-03-11

AI Summary

This study addresses the critical challenges in brain–computer interface (BCI) research—namely, the scarcity, heterogeneity, and privacy sensitivity of neural signal data—that severely hinder the development of deep learning models. The work presents the first systematic review and categorization of existing brain signal generation methods, proposing a unified taxonomy comprising four generative paradigms: knowledge-based, feature-based, model-based, and translation-based. A comprehensive benchmark is established across four representative BCI tasks, with cross-paradigm performance evaluated through multidimensional metrics. The analysis not only elucidates the strengths and limitations of each approach but also introduces the first open-source benchmark codebase for synthetic brain signals. This resource lays a foundational framework for advancing data-efficient and privacy-preserving BCI systems, while outlining future directions prioritizing accuracy, computational efficiency, and enhanced privacy protection.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Humans and AI: Brain-Sensing and Analysis
• Natural Language Processing: Generation

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Web Mining and Content Analysis: Web data generation and simulation
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Deep learning has achieved transformative performance across diverse domains, largely driven by the large-scale, high-quality training data. In contrast, the development of brain-computer interfaces (BCIs) is fundamentally constrained by the limited, heterogeneous, and privacy-sensitive neural recordings. Generating synthetic yet physiologically plausible brain signals has therefore emerged as a compelling way to mitigate data scarcity and enhance model capacity. This survey provides a comprehensive review of brain signal generation for BCIs, covering methodological taxonomies, benchmark experiments, evaluation metrics, and key applications. We systematically categorize existing generative algorithms into four types: knowledge-based, feature-based, model-based, and translation-based approaches. Furthermore, we benchmark existing brain signal generation approaches across four representative BCI paradigms to provide an objective performance comparison. Finally, we discuss the potentials and challenges of current generation approaches and prospect future research on accurate, data-efficient, and privacy-aware BCI systems. The benchmark codebase is publicized at https://github.com/wzwvv/DG4BCI.

Problem

• brain-computer interfaces
• synthetic data generation
• neural recordings
• data scarcity
• privacy-sensitive data

Innovation

• synthetic data generation
• brain-computer interfaces
• generative models
• benchmarking
• neural signal synthesis