Transcript Reproducibility of surface EMG variables and peak torque during

Can you directly compare EMG amp between subjects?
Trial 2-1
Difference
scores of
raw data
Raw Data
Change in mean
127.83
Lower conf limit
67.23
Upper conf limit
188.43
Trial 1
Trial 2
Trial 2-1
Dave
439.4
550.5
111.1
Megan
271.8
415
143.2
Typical error
52.09
Jeremiah
199.1
333.8
134.7
Lower conf limit
35.01
Dianna
785.2
1021
235.8
Upper conf limit
108.83
Ogechi
215.2
351.4
136.2
Ambar
238.3
244.3
6
Total error
102.14
Limits of agreement
144.29
Pearson r
0.981
Intraclass r
0.958
Reproducibility of surface EMG
variables and peak torque during
three sets of ten dynamic
contractions
Barbro Larsson , Bjarne Månsson ,
Christian Karlberg , Peter Syvertsson ,
Jessica Elert and Björn Gerdle
Introduction
Isokinetic dynamometers are commonly
used for assessment of dynamic muscle
strength, endurance and fatigue.
 For measurement of reproducibility, intraclass correlation (ICC) is preferred.

Fatigue and EMG
Peripheral muscle fatigue during sustained
static contractions is generally
characterized by increases in signal
energy (RMS or iEMG) and shifts in the
EMG spectrum towards lower frequencies
(spectral shift).
 Fatigue increases amplitude
 Fatigue decreases frequency

Problems with Dynamic
Contractions and EMG





The interpretation of the EMG from dynamic
contractions might be difficult—especially for
frequency spectrum variables—because the
movement per se introduces additional factors
that might affect its characteristics
changes in force throughout the range of motion
changes in fiber and muscle length
movement of the neuromuscular junction with
relation to the electrodes position
problems with non-stationary of the signal
(recruitment and de-recruitment of MUs
Methods -- Isokinetic
3 sets of 10 isokinetic contractions at 90 d/s
 One hour between sets
 The electrodes were NOT REMOVED
 ROM was constrained to 90 – 15 deg ext.
 Subjects relaxed during knee flexion and
the immediately performed extension.

Methods -- EMG
Surface EMG from VL, VM & RF
 20 mm interelectrode distance on center of
muscle in line with muscle fibers.
 Sampled at 2KHz with 12 bit A/D
 EMG band pass filtered at 16-500 Hz
 Torque & Position low pass filtered at 40 Hz

EMG Processing & Statistics
FFT with Hamming window (2 Hz resolution)
 RMS used for amplitude
 Signal-amplitude ratio (SAR) of relaxation
(flexion phase) to contraction (extension
phase).
 ICC (3,1) Shrout & Fleiss
 One-way ANOVA was used to test for
differences between sets.

Results
No significant differences between the
three sets.
 Peak torque had high reproducibility (0.99)
 Rectus femoris generally had lower ICC
than the two other muscles
 RMS generally had somewhat higher ICC
than the MNF

Discussion
High ICC may be due to limited ROM
which may control movement effects.
 Previous studies have reported good
reproducibility for during and between day
static contractions.
 Not REMOVING ELECTRODES may have
contributed to high ICC

Discussion -- Cont
We have reported that the MNF
correspond to physiological properties
during dynamic contractions
 Positive significant correlations have been
reported between the proportion of Type-II
muscle fibers and MNF during single
dynamic (non-fatiguing) contractions
