Abstract
Real-time intramuscular electromyography (iEMG) decomposition, which is needed in biofeedback studies and interfacing applications, is a complex procedure that involves identifying the motor neuron spike trains from a streaming iEMG recording. We have previously proposed a sequential decomposition algorithm based on a Hidden Markov Model of EMG, which used Bayesian filter to estimate unknown parameters of motor unit (MU) spike trains, as well as their action potentials (MUAPs). Here, we present a modification of this original model in order to achieve a real-time performance of the algorithm as well as a parallel computation implementation of the algorithm on Graphics Processing Unit (GPU). Specifically, the Kalman filter previously used to estimate the MUAPs, is replaced by a least-mean-square filter. Additionally, we introduce a number of heuristics that help to omit the most improbable decomposition scenarios while searching for the best solution. Then, a GPU-implementation of the proposed algorithm is presented. Simulated iEMG signals containing up to 10 active MUs, as well as five experimental fine-wire iEMG signals acquired from the tibialis anterior muscle, were decomposed in real time. The accuracy of decompositions depended on the level of muscle activation, but in all cases exceeded 85 %. The proposed method and implementation provide an accurate, real-time interface with spinal motor neurons. The presented real time implementation of the decomposition algorithm substantially broadens the domain of its application.
Highlights
T HE electromyogram (EMG) is the recording of electrical activity of the muscle fibers, as generated during muscle contractions
We introduce Z[n] which represents the output of a pre-detection function z(Y [n + 1 : n + PD]), where PD denotes the length of pre-detection and is typically set to IR/2 where IR is the length of motor unit action potential (MUAP)
As an example, considering a sampling frequency of 10 kHz and ten active motor unit (MU) with mean inter-spike intervals (ISI) of 100 ms, the probability of having spikes occurring at a specific instant of time, given that one spike already occurs at the same instant, is 1 − (1 − 1/1000)(1 − 2/1000). . .(1 − 9/1000) = 0.044
Summary
T HE electromyogram (EMG) is the recording of electrical activity of the muscle fibers, as generated during muscle contractions. A real-time clustering and template matching algorithm for iEMG was presented in [9]. The approach presented in this work differs from its closest analogue [7] by utilization of single-channel iEMG instead of multichannel sEMG, which entails a fundamentally different model of EMG signal. Compared to another method of iEMG decomposition [9], which uses on-line clustering in order to classify MUAPs, the proposed approach resolves superpositions, providing more accurate decomposition and potentially scaling up to higher efforts at which MUAPs no longer occur isolated from each other.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
