Xupeng Chen

The paper uses graph-theoretical connectivity from ECoG to predict eloquent cortex during presurgical mapping, finding varying accuracy across functional domains.

The paper identifies the source and timing of a corollary discharge signal in human speech using ECoG recordings and Granger-causality analysis.

This paper introduces a deep learning framework for decoding speech from neural signals, achieving high correlation with original speech spectrograms and providing open-source code.

This paper introduces a transformer-based model for speech decoding from neural signals using surface and depth electrodes, achieving high performance across multiple participants and electrode types.

This study uses a deep learning model to decode speech from neural signals, revealing that frontal and temporal cortex process both feedforward and feedback information during speech production.