Decomponi


decomp

General Description: Decompni is a tool included in the free software OT BioLab for the processing of the multi-channel EMG signals recorded with non-invasive two-dimensional arrays of electrodes. It is based on blind source separation techniques and identifies the contributions of motor units to the surface EMG.

Details: Decomponi plug-in result is a series of motor unit action potential trains that represents the neural drive sent to the muscles. In the literature, this approach is also called "EMG decomposition". EMG decomposition has been classically possible with invasive (needle) electrodes and time-consuming manual editing. Decomponi allows decomposition from fully non-invasive recordings in a fully automatic way. Decomponi has been developed by OT Bioelettronica based on research methods and results published by the groups of Prof. Dario Farina, Dr. Francesco Negro and Prof. Ales Holobar.

 Decomponi plug-in technical specifications:

- It needs as input two-dimensional EMG signals recorded by matrices >=32 electrodes per muscle.

- It provides as output both the series of discharges of the identified motor units as well as the respective action potentials for each location of the recording grid.

- It tests the validity of the extracted results by indexes of accuracy that have been tested in the scientific literature. Each extracted action potential train is provided with an estimated confidence of its correctness.

- It works with signals recorded in both monopolar and differential modalities.

- Is limited to isometric contractions.

- In addition to the series of discharge timings, it also provides basic statistical characteristics of the extracted motor units, such as the discharge rate and interspike interval variability.

References:

- Negro F., Muceli S., Castronovo A.M., Farina D., (2016) Multi-channel intramuscular and surface EMG decomposition by convolutive blind source separation. Journal of neural engineering, 13(2), 026027.
- Holobar A., Zazula D. (2007). Multichannel blind source separation using convolution kernel compensation. IEEE Trans. Signal Proc., vol 55. pp, 4487-4496.