Accurate Emotion Decoding Based on Emotion-Induced Scalp EEG Datasets and Large-Scale Models: Progress, Theoretical Dilemmas, Methodological Pitfalls, and Practically Feasible Pathways

This study provides a critical examination of the scientific foundations and application prospects of affective and emotional brain–computer interfaces (BCIs), with a particular focus on the feasibility of deploying online affective BCIs in clinical and practical contexts. First, the conceptual boundaries between emotion and affect are delineated, clarifying their interrelations and distinctions. The paper then reviews recent advances in multimodal signal acquisition, emotion induction paradigms, and decoding algorithms, while addressing key challenges such as signal–label alignment, the trade-off between real-time performance and robustness, and the ethical as well as academic boundaries of the field. The discussion emphasizes that the inherent subjectivity of emotions and the instability of their recognition constitute major theoretical and technical obstacles to large-scale implementation. The study concludes by advocating for a cautious and scientifically rigorous trajectory of future development, ensuring that progress in this emerging domain remains both responsible and sustainable.