FingerFlex: Inferring Finger Trajectories from ECoG signals
This work addresses the development of high-precision motor brain-computer interfaces for potential applications in assistive technology, though it appears incremental as it adapts existing deep learning architectures to a specific dataset.
The paper tackled the problem of decoding finger movements from ECoG brain signals using a convolutional encoder-decoder model, achieving a correlation coefficient of up to 0.74 on a benchmark dataset.
Motor brain-computer interface (BCI) development relies critically on neural time series decoding algorithms. Recent advances in deep learning architectures allow for automatic feature selection to approximate higher-order dependencies in data. This article presents the FingerFlex model - a convolutional encoder-decoder architecture adapted for finger movement regression on electrocorticographic (ECoG) brain data. State-of-the-art performance was achieved on a publicly available BCI competition IV dataset 4 with a correlation coefficient between true and predicted trajectories up to 0.74. The presented method provides the opportunity for developing fully-functional high-precision cortical motor brain-computer interfaces.