EMGONOFF

Finds the "on" and "off" indicies of a single raw surface EMG burst (muscle contraction).

Description

This function takes as input the raw EMG data of a single muscle contraction. Output is a 1x2 array, the first number is the EMG onset, the second number is the EMG "offset" index.

There needs to be enough data before the "on" burst to establish an average resting EMG level.

The resting levels (before and after burst) are defined by the user with MATLAB's ginput function.

One option is available by editing the .m file at line 65. By setting "xi" to x(2), the search for onset begins with the end of the ginput-defined resting EMG zone, and thus ignores any muscle bursts within the user-defined resting zone.

Algorythm for EMG onset & offset taken from Hodges, P.W. and B.H. Bui, _A comparison of computer-based methods for the determination of onset of muscle contraction using electromyography._ Electroencephalography & Clinical Neurophysiology, 1996. 101(6): p. 511-9

Created by: Kieran A. Coghlan, BSME, MSES
SUNY at Buffalo, New York
Last modified: 1 May, 2006

• MATLAB release: MATLAB 7.1.0 (R14SP3)

[roiohan robin]

no

[Joe Chang]

 

[J B]

For loops need vectorizing

[Armindo]

Dear Kieran A. Coghlan

After read with very interest the article of Hodges “A comparison of computer-based methods for the determination of onset of muscle contraction using electromyography” I saw your matlab function what is very nice indeed (thank you for sharing) however I saw that the combination of parameters used in the article also account for the cutoff frequency used in the low pass filtering. According to Hodges “The criteria that most accurately identified the onset of EMG (25 ms/3 SD/50 Hz, 50 ms/1 SD/50 Hz, 10 ms/1 SD/ 500 Hz) resulted from the most appropriate combinations of the positive and negative influences of each of the parameters.” But in this function it seems that you don’t use a low pass filter cutoff frequency (is not an argument). Perhaps this information about low pass filter should be in the function or an option (low pass filter) could be added to the function so the raw data can be filtered before calculating the EMG onset and offset.

Kind regards,
Armindo Marques