Flow Matching for Few-Trial Neural Adaptation with Stable Latent Dynamics

Puli Wang, Yu Qi, Yueming Wang, Gang Pan

TL;DR: A novel neural alignment approach based on flow matching.

Abstract: The primary goal of brain-computer interfaces (BCIs) is to establish a direct linkage between neural activities and behavioral actions via neural decoders. Due to the nonstationary property of neural signals, BCIs trained on one day usually obtain degraded performance on other days, hindering the user experience. Existing studies attempted to address this problem by aligning neural signals across different days. However, these neural adaptation methods may exhibit instability and poor performance when only a few trials are available for alignment, limiting their practicality in real-world BCI deployment. To achieve efficient and stable neural adaptation with few trials, we propose Flow-Based Distribution Alignment (FDA), a novel framework that utilizes flow matching to learn flexible neural representations with stable latent dynamics, thereby facilitating source-free domain alignment through likelihood maximization. The latent dynamics of FDA framework is theoretically proven to be stable using Lyapunov exponents, allowing for robust adaptation. Further experiments across multiple motor cortex datasets demonstrate the superior performance of FDA, achieving reliable results with fewer than five trials. Our FDA approach offers a novel and efficient solution for few-trial neural data adaptation, offering significant potential for improving the long-term viability of real-world BCI applications.

Lay Summary:

Brain-computer interfaces (BCIs) aim to translate neural activity into intended behavior, but their performance often degrades over time due to the nonstationary nature of neural signals. This makes it difficult to maintain reliable decoding across days, especially when only limited target data is available.

We introduce Flow-Based Distribution Alignment (FDA), a novel method that leverages flow matching to align neural activity across sessions with minimal data. FDA does not require access to source data during adaptation and performs well with as few as five target trials.

Evaluations on multiple motor cortex datasets show that FDA consistently delivers more stable and accurate decoding than existing approaches. Our method addresses a core challenge in real-world BCI applications and provides a practical solution for efficient neural alignment.

Application-Driven Machine Learning: This submission is on Application-Driven Machine Learning.

Link To Code: https://github.com/wangpuli/FDA

Primary Area: Applications->Neuroscience, Cognitive Science

Keywords: Brain-Computer Interface, Neural Decoding, Flow Matching, Neural Alignment

Submission Number: 8350