NeurDuo-EEG: A Long-Sequence EEG Foundation Model with Persistent State and Explicit Memory

2026-09-29

AI Summary

This study addresses the limitation of existing models in capturing long-range temporal EEG dynamics by proposing a causal EEG foundation model equipped with persistent states and explicit memory. Methodologically, it introduces multi-timescale memory management and selective retrieval mechanisms to efficiently model continuous long sequences, alongside a pretraining strategy based on discrete spectral codes for multi-channel autoregressive prediction. Experimental results demonstrate that the proposed model achieves state-of-the-art performance on four out of five benchmarks, yields significant improvements in AUC for seizure detection, and supports low-latency streaming inference. Collectively, this work establishes an efficient new paradigm for long-horizon EEG signal analysis.

Abstract

Electroencephalography (EEG) is recorded continuously over hours, with relevant dynamics spanning timescales from milliseconds to hours. Most EEG foundation models nevertheless process fixed windows independently, limiting their ability to capture information encoded in long-timescale dynamics. State-space architectures enable persistent recurrent processing, but long-range information remains implicitly compressed in recurrent states. We present NeurDuo-EEG, a causal EEG foundation model with channel-resolved persistent memory. NeurDuo-EEG introduces multi-timescale memory management with learned consolidation and selective retrieval, enabling persistent modelling of continuous EEG with fixed-size state. It is pre-trained on 3,955 hours of EEG from 17 public datasets using multichannel autoregressive prediction of discrete spectral codes. Across three short-window and two long-sequence downstream tasks, NeurDuo-EEG achieves the best performance on four of five benchmarks, including all three short-window tasks and seizure detection, where AUC-PR improves from $0.285$ to $0.471$ over the strongest non-NeurDuo baseline. NeurDuo-EEG also remains competitive on sleep staging and supports efficient streaming inference, with nearly constant per-chunk latency as the available history grows to one hour. Notably, the Small variant achieves this with only 4.7M backbone parameters. These results demonstrate the value of persistent, multi-timescale modelling for both long-sequence and short-window EEG analysis. Our code is available at https://github.com/YifaNNW/NeurDuo-EEG.

Problem

• EEG foundation model
• long-sequence modeling
• multi-timescale dynamics
• persistent memory

Innovation

• EEG foundation model
• multi-timescale memory management
• persistent state
• autoregressive prediction
• streaming inference