On this page are the major files related to the papers:

(1) "Speed dependence of averaged EMG profiles in walking." 
Gait & Posture 16: 78-86 (2002)

and, THIS NEW PAPER:

(2) "Detection of non-standard EMG profiles in walking."
Gait & Posture, in press.

By Hof, Elzinga, Grimmius and Halbertsma

There are two subdirectories: ASCII and MFILES.

The ASCII directory contains in ascii format, TAB delimited:
allmeas.txt    containing the measured average profiles for 14 muscles at 5 speeds
fdata.txt       with the processed data: F0, F1,F2,f0,f1
ddata           D0,D1,D2
newlim          H0, H1, L0, L1

These files can be downloaded into a spreadsheed program like EXCEL, etc.
No processing program have been supplied. In principle it is possible to do the 
required matrix multiplications in EXCEL.

The MFILES directory contains the data and some example programs in MatLab format,
 version 5.2 and higher. 

allmeas.mat    containing the measured average profiles for 14 muscles at 5 speeds
fdata.mat       with the processed data: F0, F1,F2,D0,D1,D2,f0,f1
newlim       (NEW) with data from (2) to calculate upper and lower limits: L0 L1 L2, H0 H1 H2

In order to use the following programs, these three mat files should be loaded into the Matlab
workspace:
>>> load allmeas ; load fdata ; load newlim

fitplot.m         a program to verify the claims made in  paper (1). Select a muscle
  and a speed and the result is a figure that shows the measured profile, 
  the estimate with f0/f1 and the estimate with F0F1F2. 
  In the command screen the r.m.s. errors are shown.

Plotmusc.m  plots the five profiles for the selected muscle. 
  This is a Matlab function that requires an array from allmeas as an  argument, e.g.
>>> plotmusc(TA)       shows a plot of Tibialis Anterior, something like Figure 3a.
winterfig.m  does a similar thing, but now the EMGs are filterd 
  with a 3 Hz critically damped filter, as recommended by DA Winter (see Discussion)
>>> winterfig(TA)    show the filtered plot of Tibialis Anterior, something like Figure 3b.

NEW
plotlimgem.m    plots a recorded EMG together with the estimated profile 'est' and the upper and
lower limits 
    'high' and 'low', respectively. In the command screen the gain g and the deviation D^2 are
presented. 
    You are prompted for the muscle (1-14 from the string array 'allmusc') and the speed (1-5 = [
0.75 1.00 1.25 1.50 1.75 ] m/s).
    The real attractive thing is to adapt the program in order to analyse your own measured average
EMG profiles.
    I have software available to compute step-averaged EMG profiles, for who is interested.

In general, by far the most convincing show is to execute the Matlab programs. 
If you don't have the program, maybe a colleague is willing to run it. 

Good luck,
At Hof