UniBCI: Towards a Unified Pretrained Model for Invasive Brain-Computer Interfaces

2026-04-30

AI Summary

This work addresses the challenges of heterogeneous few-shot data, cross-domain distribution shifts, and complex spatiotemporal dynamics in invasive brain–computer interfaces (BCIs) by proposing UniBCI—the first unified pretraining framework tailored for invasive neural signals. UniBCI integrates metadata with neural recordings through context-conditioned spatiotemporal tokenization (CST), a hierarchical interval–region attention mechanism (IAA), and a masked signal reconstruction objective for self-supervised learning. This enables effective representation learning across diverse, heterogeneous sources while capturing both local dependencies and global dynamics. Evaluated on multi-species, multi-region neural datasets, UniBCI achieves state-of-the-art performance across multiple downstream tasks with fewer parameters and lower latency, significantly improving accuracy, computational efficiency, and cross-domain generalization.

Abstract

Modeling invasive neural spike data is fundamental to advancing high-performance brain-computer interfaces (BCIs). However, existing approaches face critical challenges, including limited-scale heterogeneous data, cross-domain distribution shift, and the intrinsic spatiotemporal complexity of invasive neural signals. In this work, we propose UniBCI, a unified pretrained model for invasive Brain-Computer Interfaces. The model integrates three key components: (1) a context-conditioned spatio-temporal tokenization (CST) scheme that embeds neural signals together with metadata into a shared representation space; (2) a hierarchical Interval-Area Attention (IAA) mechanism that captures patterns of spike dynamics in slots via linear attention and locality dependencies via sliding-window attention; and (3) a scalable self-supervised masked signals reconstruction objective for learning generalizable neural representations from large-scale unlabeled data. We construct a pretraining corpus spanning multiple species, subjects, brain regions, and behavioral experiment paradigms. These heterogeneous recordings are standardize via our proposed unified normalization and tokenization. Comprehensive experiments demonstrate that UniBCI achieves SOTA performance across diverse downstream tasks while improving generalization. Moreover, the model achieves a strong balance between accuracy and efficiency, with fewer trainable parameters and lower inference latency. These results suggest that UniBCI provides a practical step toward general-purpose neural foundation models, enabling robust, scalable, and transferable representation learning for invasive neural data. The code for this paper is available at: https://anonymous.4open.science/r/UniBCI-C805.

Problem

• invasive brain-computer interfaces
• neural spike data
• distribution shift
• spatiotemporal complexity
• heterogeneous data

Innovation

• pretrained model
• spatio-temporal tokenization
• Interval-Area Attention
• self-supervised learning
• invasive BCI