Continuous Force Decoding from Deep Brain Local Field Potentials for Brain Computer Interfacing

Syed A Shah, Huiling Tan, Peter Brown

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract / Description of output

Current Brain Computer Interface (BCI) systems are limited by relying on neuronal spikes and decoding limited to kinematics only. For a BCI system to be practically useful, it should be able to decode brain information on a continuous basis with low latency. This study investigates if force can be decoded from local field potentials (LFP) recorded with deep brain electrodes located at the Subthalamic nucleus (STN) using data from 5 patients with Parkinson's disease, on a continuous basis with low latency. A Wiener-Cascade (WC) model based decoder was proposed using both time-domain and frequency-domain features. The results suggest that high gamma band (300-500Hz) activity, in addition to the beta (13-30Hz) and gamma band (55-90Hz) activity is the most informative for force prediction but combining all features led to better decoding performance. Furthermore, LFP signals preceding the force output by up to 1256 milliseconds were found to be predictive of the force output.

Original languageEnglish
Title of host publication2017 8th International IEEE/EMBS Conference on Neural Engineering (NER)
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Pages371-374
Number of pages4
Volume2017
ISBN (Electronic)978-1-5090-4603-4
ISBN (Print)978-1-5090-4604-1
DOIs
Publication statusPublished - 15 Aug 2017

Fingerprint

Dive into the research topics of 'Continuous Force Decoding from Deep Brain Local Field Potentials for Brain Computer Interfacing'. Together they form a unique fingerprint.

Cite this