Reduction of Crosstalk in the Electromyogram: Experimental Validation of the Optimal Spatio-Temporal Filter

Abstract

Objective: Crosstalk in surface electromyogram (EMG) is an important open problem and the common strategy of reducing it through spatial filters needs improvements. Methods: We evaluated experimentally the optimal spatio-temporal filter (OSTF), i.e., a recent approach that adapts to the subject, filtering different EMG channels both in time and space to emphasize the signal of a target muscle discarding that of adjacent ones. EMGs were recorded by a high-density recording system from pronator teres (target muscle) and flexor carpi radialis (crosstalk muscle) of 8 healthy subjects. OSTF was tested in different conditions, considering one channel per muscle (either single or double differential, SD and DD, respectively), changing the selectivity of detection (small electrodes close to each other, or large ones with higher inter-electrode distance), the force applied by the muscles (whose EMGs were summed to simulate different levels of crosstalk), and the duration of the signal to train the method. Results: OSTF was less affected by crosstalk than SD and DD filters. Statistically significant improvements were obtained in reducing the crosstalk-induced variations: for example, considering small electrodes, we obtained a percentage error of 157.30±57.11 % and 38.54±10.47 % (mean±std) in the estimation of the average rectified value (ARV), and an error of 23.57±3.92 % and 8.31±0.88 % in the estimation of the median frequency (MDF), for SD and OSTF, respectively. Conclusion: The OSTF can be applied in real-time, is easy to use, and is feasible even when using only few detection channels, as is customary in many applications.

Nonlinear spatio-temporal filter to reduce crosstalk in bipolar electromyogram

“…The experimental data used to test the algorithm are the same as in [36] (to which the reader can refer for details Monopolar EMGs were recorded with reference electrode on the elbow, using the amplifier Quattrocento (OT Bioelettronica, Turin, Italy) during selective contractions of FCR and PT. The two muscles are close to each other (making relevant the issue of crosstalk) and are devoted to different movements, i.e.…”

“…An alternative approach is the OSTF [35,36], which is adapted to the specific conditions (needing a preliminary training on selective data from target and crosstalk muscles) and can be applied in real-time. It requires simple recording systems (e.g.…”

“…It requires simple recording systems (e.g. a few EMG channels over the target and the crosstalk muscles) and showed lower sensitivity to crosstalk than traditional spatial filters [36] and advanced BSS approaches [37]. However, the output of the filter is not easy to be interpreted.…”