Advertisement
Published in:
2019 | OriginalPaper | Chapter
Hybrid Gaussian Point-Process Model for Finer Control of Myoelectric Robotic Hands
Authors : Sohail Siadatnejad, Francesco Negro, Luca Citi
Published in: Converging Clinical and Engineering Research on Neurorehabilitation III
Publisher: Springer International Publishing
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
Abstract
This paper presents a novel mathematical approach to decode information from neural spike trains recorded using high-density surface EMG for the control of a robotic hand. During precision grasps characterised by low firing rates, the algorithm uses a point-process approach to extract the information encoded in the spike times. When performing power grasps, the increased firing rate makes the point-process less reliable and the algorithm automatically and gradually switches to an approach based on the EMG power. Unlike current point-process decoding algorithm relying on a Gaussian approximations of the observation model, this model adopts a fully Gaussian state-space representation thanks to a parametrization of the rate of an inverse Gaussian distribution and by modelling the inter-spike interval directly.