Clustering Neural Patterns in Kernel Reinforcement Learning Assists Fast Brain Control in Brain-Machine Interfaces
Neuroprosthesis enables the brain control on the external devices purely using neural activity for paralyzed people. Supervised learning decoders recalibrate or re-fit the discrepancy between the desired target and decoder’s output, where the correction may over-dominate the user’s intention.
Tracking Neural Modulation Depth by Dual Sequential Monte Carlo Estimation on Point Processes for Brain Machine Interfaces
We propose a dual sequential Monte Carlo adaptive point process method, which models and decodes the gradually changing modulation depth of individual neuron over the course of a movement. We use multi-channel neural spike trains from the primary motor cortex of a monkey performing a target pursuit task, and test our computational approach with successful tracking on the neural modulation depth over time and improvement on kinematic reconstruction.
Learning Recurrent Waveforms within EEGs
We explore a modeling approach that automatically learns the reoccurring waveforms within EEG traces. To summarize waveforms learned across electrodes and subjects we propose a cluster analysis protocol using shift-invariant k-means. The spatial amplitude patterns associated with a subset of the learned waveforms are shown to discriminate between motor imagery modalities.