MEG-GPT: A transformer-based foundation model for magnetoencephalography data
This work addresses the challenge of analyzing large-scale brain dynamics for neuroscience, offering a foundation model that could benefit computational neuroscience and neural decoding applications, though it is incremental as it adapts existing transformer methods to a new domain.
The paper tackled the problem of modeling complex spatiotemporal patterns in magnetoencephalography (MEG) data by introducing MEG-GPT, a transformer-based foundation model with a novel tokenizer, which improved zero-shot generalization accuracy from 0.54 to 0.59 across sessions and from 0.41 to 0.49 across subjects compared to baselines.
Modelling the complex spatiotemporal patterns of large-scale brain dynamics is crucial for neuroscience, but traditional methods fail to capture the rich structure in modalities such as magnetoencephalography (MEG). Recent advances in deep learning have enabled significant progress in other domains, such as language and vision, by using foundation models at scale. Here, we introduce MEG-GPT, a transformer based foundation model that uses time-attention and next time-point prediction. To facilitate this, we also introduce a novel data-driven tokeniser for continuous MEG data, which preserves the high temporal resolution of continuous MEG signals without lossy transformations. We trained MEG-GPT on tokenised brain region time-courses extracted from a large-scale MEG dataset (N=612, eyes-closed rest, Cam-CAN data), and show that the learnt model can generate data with realistic spatio-spectral properties, including transient events and population variability. Critically, it performs well in downstream decoding tasks, improving downstream supervised prediction task, showing improved zero-shot generalisation across sessions (improving accuracy from 0.54 to 0.59) and subjects (improving accuracy from 0.41 to 0.49) compared to a baseline methods. Furthermore, we show the model can be efficiently fine-tuned on a smaller labelled dataset to boost performance in cross-subject decoding scenarios. This work establishes a powerful foundation model for electrophysiological data, paving the way for applications in computational neuroscience and neural decoding.