A sample of EMG signals

This page contains the instructions to download a sample of EMG signals. The sample was obtained at Hospital Carlos III, Madrid, Spain. All the subjects gave written informed consent. The whole procedure was approved by the Ethics Committee of the  Hospital Carlos III. The database includes five files (each one of them comprise several segments of rest and voluntary activity) recorded from muscle right Tibialis Anterior.


The files are coded in EDF+ format (See also the EDF pages in Neurotraces).


The files included are (the size of each file is about 2 MB):

cemg01.edf.bz2

cemg02.edf.bz2

cemg03.edf.bz2

cemg04.edf.bz2

cemg05.edf.bz2

Additionally a file containing fibrillation and positive sharp waves is also included

possw.edf.bz2

The whole sample includes 15 normal records. I can not include them here because of limitations of space. If you are interested please send me a mail and I could upload the files to your server (jolivan@neurotraces.com). I hope that I will be able to include more pathological recordings as well as a tutorial describing the logic of the acquisition.

README:

You can decompress the files by using bzip2 (you have to use the command bunzip2). On Windows OS, bz2 format is supported by Power Archiver

The files were recorded from Right Tibialis Anterior using a concentric needle (TE/B64-408, Technomed Europe, 40 x 0.45 mm, recording area 0.068 mm²). The files contain a continuous acquisition of around 3.5 minutes sampled at 20 kHz. Each 10 seconds (more or less) the position of the needle is changed, so several segments of homogeneous rest/contraction relatively free of artifact are available for the same person, in the same muscle and in the same condition. Annotations describe segments of “rest” and “contraction” as well as a calibration segment. Each change in the position of the needle is indicated as “new_site”.

To access the contents of the files you can follow several ways:

- You can view the signals as well as the annotations using RASCHlab. This is a shot (you could set "speed" to 1000 mm/sec, equivalent to 10 ms/div, resolution  to 0.1 mm/uV, equivalent to 100 uV/div  and check "invert channel"  to get  "negativity up"  polarity).

- You can access the content of the files (signals and annotations) in Matlab, Octave, Scilab, Python, Perl, C or C++ using libRASCH.

- You can convert the signals into a file containing ASCII or float binary values using edfAsc and the annotations into an ASCII file using edfAnn. You can also access the header using edfEdit. These programs (edfEdit, edfAsc, edfAnn)  are available  at the EDF section of Neurotraces (documentation is included)

(You could also use many other utilities of EDF although most of them are designed to handle EEG signals)

   
Data contained here are accessible under the same Copying policy of Physionet.

Enjoy!!