Hi Hannes, 

There are two great papers about EMG normalization during cycling: 

In addition to those, there is a very good review (Nick Ball and Joanna Scurr, Journal of applied biomechanics, 2013) about EMG normalization during high-velocity muscle actions.

The normalization method that you choose in the end, will depend on your research question... But is important to have in mind that this method should be both repeatable (reliable) and sensitive. Both issues are very well discussed in the Albertus-Kajee paper, so maybe you would like to use that article as a reference. 

In my opinion, in case that you are analyzing high-velocity cycling, sub-maximal dynamic reference contractions are better for normalization in comparison to an isometric MVC.

Hope that this helped and good Luck with your work,

Kind Regards,

Eduardo 

How about to take the parentage of average level to maximum peak activity during the dynamic contraction.  (avrag/peak*100)

One could first measure the MVIC or you could measure the EMG during a 30-second full-out Wingate or sprint test to measure maximum EMG for normalisation.

Hi Hannes,

Fernandez-Pena et al. (2008) verified a good reliability of 6-s isokinetic sprints (80-rpm) in order to normalize EMG during cycling. I think it is a good method if you use to compare with submaximal workload with similar cadences. I think it is a preferable method than IMVC for cycling tests.

Depends on the workload which will use you would select a specific cadence to normalize EMG by 6-s sprints.

Two different processing for MPF or MDF calculation during intermittent isometric exercise

Ph. Germain O. Buttelli

Motility Performance Laboratory, University of sport sciences of Orleans, France.

Mean Power Frequency (MPF) and Median Frequency (MDF) are usually used to investigate neuromuscular activation (Duchêne and Goubel, 1993). When such frequency parameters are determined over a finite length of signal, a random error and a systematic error or bias are introduced (Hof, 1991). Unfortunately, sometimes contraction time duration of intermittent isometric exercise are too short, inducing restricted data number. It’s likely that our experiment investigate possibilities to calculate MPF and MDF through the whole intermittent isometric exercise including activity and rest. We investigated maximal activation of biceps brachii by surface electromyography at a recording frequency of 1024 Hz. Several intermittent isometric contractions were performed with different combination of 1 to 6 seconds of muscle action time and 1 to 20 seconds of rest period between contractions. Activation time duration were long enough which could permit us to compare mean of 6 successive MPF and MDF computed without rest period (A processing) to MPF and MDF of the whole 6 activations  computed with rest period (B processing). Eight voluntary subjects for this experiment performed 6 trials for each condition. MPF and MDF were normalised in reference to the maximal value of individual trials.

MPF and MDF values with A processing compared to MPF and MDF with B processing are closed to 0,1% and 1,2%, respectively. Muscle action time duration induce a variability of MPF and MDF comprise between 0,05 and 0,7% and between 0,7 and 1,4%, respectively. Rest time duration induce MPF and MDF variability from 0,1 to 1,4% and 0,9 to 1, 8%, respectively. No statistical differences appeared between each MPF and MDF values instead of calculation procedure. This results support the hypothesis of an alternative possibility for MPF and MDF calculation if activation time during intermittent contraction is not long enough.

Hof At. L. (1991) I.E.E.E. 38(11), 1077-1088.

Duchêne J. Goubel F. (1993) CRC Crit. Rev. Biomed. Eng. 21, 313-397

V.O2 and EMG Activity Kinetics During Moderate and Severe Constant Work Rate Exercise in Trained Cyclists

Christophe Cleuziou1, Stéphane Perrey2, Fabio Borrani3, Anne Marie Lecoq4, Daniel Courteix1, Philippe Germain1, and Philippe Obert5

(2004). VO2 and EMG activity kinetics during moderate and severe constant work rate exer- cise in trained cyclists. Can. J. Appl. Physiol. 29(6): 758-772. © 2004 Canadian Society for Exercise Physiology.

. Key words: median power frequency, O2 uptake, recruitment of fast-twitch fibers, VO2slow

component

Mots-clés: fréquence moyenne du spectre, prélèvement d’oxygène, recrutement des fibres .

rapides, composante lente de VO2

Abstract/Résumé

The purpose of this study was to compare O2 uptake (V.O2) and muscle electromyography activity kinetics during moderate and severe exercise to test. the hypothesis of progressive recruitment of fast-twitch fibers in the explanation of the VO2 slow component. After an incremental test to exhaustion, 7 trained cyclists (mean ± SD, 61.4 ± 4.2 ml·min–1·kg–1) performed several square-wave transitions for 6 min at moderate and severe intensities on

. a bicycle ergometer. The VO2 response and the electrical activity (i.e., median power fre-

quency, MDF) of the quadriceps vastus lateralis and vastus medialis of both lower limbs were measured continuously during exercise. After ~2 to 3 min of exercise onset, MDF values increased similarly during moderate and severe exercise for almost all muscles whereas a V.O2 slow component occurred during severe exercise..There was no relationship between the increase of MDF values and the magnitude of the VO2slow component during the severe exercise. These results suggest that the origin of the slow component may not be due to the progressive recruitment of fast-twitch fibers.

1Lab. de Physiologie de l’Exercice Musculaire, Univ. d’Orléans, 45100 Orléans la Source, France; 2EA 2991, UFR STAPS, 700 av du pic saint loup, 34090 Montpellier, France; 3Institut des Sciences du Sport et de l’EP, Univ. de Lausanne, 1015 Lausanne, Switzerland; 4Dépt. de Physiologie Respiratoire, CHR, 45000 Orléans, France; 5Lab. des Adaptations Cardiovasculaires à l’Exercice, Univ. d’Avignon, 84000 Avignon, France.

Constant-rate-high-intensity exercise and EMG

P. Germain, C. Cleuziou, R. Halin, O. Buttelli, Z. Taoutaou, P. Obert

Abstract

Introduction – In one hand, during a fatiguing exercise the analysis of the myoelectric signal support the hypothesis of an increase of proportional number of active slow motor units. In the other hand, the slow component of oxygen consumption during a constant-rate-high- intensity exercise, seems to dependent on a progressive recruitment of fast-twitch fibers.

Methods and results – In front of this contradiction, the purpose of this study is to investigate the EMG activity of quadriceps femoris muscle during a 90% of PMA cycling exercise.

Two main results were observed: 1) Integrated EMG decreased all exercise long (from 0,99 to 0,86 v); 2) MPF and MDF decreased in the first part of the exercise (respectively from 82,5 to 81,25 Hz and from 108,75 to 107 Hz), followed by a progressive increase (respectively from 81,25 to 84,5 Hz and from 107 to 109,25 Hz).

Conclusion – Our results suggest a progression in the average frequency of the motor unit discharge towards the high frequencies, which coheres with the hypothesis of the progressive recruitment of fast-twitch fibers during the V~ O2 slow component.

© 2003 Éditions scientifiques et médicales Elsevier SAS. All rights reserved. Mots clés : Exercice intensif ; Consommation d’oxygène ; Electromyographie Keywords: Intensive exercise; ElectroMyoGraphy; Oxygen conscemption

Hi Hannes,

It depends on your research question. If you want to compare between muscles, then a IMVC is more appropriate as you will be relating the muscle activity to the maximum activity of the muscle (provided that the IMVCs produce maximums). We have used the following reference for IMVCs to produce maximum in 7 leg muscles: Rutherford DJ, Hubley-Kozey CL, Stanish WD (2011) Maximal voluntary isometric contraction exercises: a methodological investigation in moderate knee osteoarthritis. J Electromyogr Kinesiol. 21(1):154-60.

Torque-Velocity method may not produce the maximum activity in all muscles and as such, you will be relating the activity to the maximum produced during that specific task. As a results, comparisons between muscles would not be appropriate as different muscles would be active at different levels during the normalisation task. However, if you are wanting to look at the activity levels compared of individual muscles to what they could produce in that task then Torque-Velocity method may be better.

Regards

Mark

Dear Hannes,

I think it very depends on your application. but have an experience with EMG signals during hand movements. when I wanted to normalize my Muscle Activation signals (EMG Envelopes), some of my channels loose their amplitude as their MVC is too high in comparison with their usual activation.

so as [Andrea d'Avella et al 2006] mentioned, I normalized each channel to its maximum amplitude across conditions, so it kept the normalized signal activation pattern comparable with other channels.

Regards

Payman

Hi all,

thanks a lot for your interesting and original Input! 

hi!. you can have some information after reading this document

Hi all, for all of you that are interested in EMG during cycling. Check this article that we recently published. It might help in your research.