SPICED: A Synaptic Homeostasis-Inspired Framework for Unsupervised Continual EEG Decoding

2025-09-22

AI Summary

Addressing the challenge of balancing stability and plasticity in unsupervised, cross-subject continual brain signal decoding, this paper proposes a biologically inspired, dynamically expandable continual learning framework motivated by synaptic homeostasis. The framework integrates three complementary mechanisms: (i) discriminative memory replay for selective reactivation of task-relevant representations; (ii) synaptic consolidation to mitigate catastrophic forgetting; and (iii) synaptic renormalization to preserve network dynamical equilibrium. Evaluated on three EEG datasets, our method significantly alleviates catastrophic forgetting and enables robust, unsupervised adaptation to newly introduced subjects. To the best of our knowledge, this is the first work to systematically incorporate synaptic homeostasis principles into unsupervised brain–computer interface continual learning. The proposed paradigm offers a scalable, biologically plausible solution for real-world scenarios involving dynamic subject expansion—advancing both neuroscientific interpretability and practical deployability of adaptive neural decoding systems.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Life-Long and Continual Learning
• Humans and AI: Brain-Sensing and Analysis

Application Category

• Economics, Online Markets and Human Computation: Social networks and social learning
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Human brain achieves dynamic stability-plasticity balance through synaptic homeostasis. Inspired by this biological principle, we propose SPICED: a neuromorphic framework that integrates the synaptic homeostasis mechanism for unsupervised continual EEG decoding, particularly addressing practical scenarios where new individuals with inter-individual variability emerge continually. SPICED comprises a novel synaptic network that enables dynamic expansion during continual adaptation through three bio-inspired neural mechanisms: (1) critical memory reactivation; (2) synaptic consolidation and (3) synaptic renormalization. The interplay within synaptic homeostasis dynamically strengthens task-discriminative memory traces and weakens detrimental memories. By integrating these mechanisms with continual learning system, SPICED preferentially replays task-discriminative memory traces that exhibit strong associations with newly emerging individuals, thereby achieving robust adaptations. Meanwhile, SPICED effectively mitigates catastrophic forgetting by suppressing the replay prioritization of detrimental memories during long-term continual learning. Validated on three EEG datasets, SPICED show its effectiveness.

Problem

• Addresses unsupervised continual EEG decoding with inter-individual variability
• Mitigates catastrophic forgetting during long-term continual learning adaptation
• Achieves dynamic stability-plasticity balance through bio-inspired neural mechanisms

Innovation

• Synaptic homeostasis-inspired neuromorphic framework for EEG
• Dynamic expansion through critical memory reactivation mechanism
• Synaptic consolidation and renormalization prevent catastrophic forgetting