unknown user

11-26-2007, 03:09 AM

Dear colleagues,

I am conducting a study in which I want to quantify agonist-antagonist

muscle co-contraction during gait. In this study, subjects walked at

several different slopes (including level walking). For each trial I

collected VO2 and EMG from eight muscles of one leg (Tibialis Anterior,

Medial Gastrocnemius, Lateral Gastrocnemius., Soleus, Vastus Medialis.,

Vastus Lateralis, Biceps Femoris, Gluteus Maximus) using a Noraxon Telemyo

system (sampling rate = 1000 Hz). I also collected foot switch data (both

feet). Prior to the walking trials, I collected EMG during quiet standing.

I am interested in determining if the reduced delta mechanical efficiency

observed in several of the subjects is related to an increase in

antagonist muscle co-contraction.

I would like to know what the best method for analyzing the EMG data is.

Here are some proposed steps for the initial EMG processing prior to

co-contraction analysis:

1) Filter EMG data (band pass: 10- 500 Hz; 2nd or 4th order Butterworth)

2) Rectify EMG data

3) Filter to get linear envelope (low pass: 6 Hz)

4) Normalize amplitude to mean integrated EMG of same muscle during level

walking trial (do not have MVC data)

5) Normalize time to step time using footswitch data (0% = heels strike,

100% = ipsilateral heel strike)

6) Determine when each muscle is OFF

Visually inspect and subjectively determine?

Or, Use prior research that gives the time (percentage of gait cycle)

that the muscle should not be activated?

Or, should I use some other method?

7) Determine when each muscle is ON

ON = When muscle activation is 3 standard deviations above the mean of

when muscle is OFF

Or, . Some other methods?

Steps to determine co-contraction (based on Falconer & Winter, 1985)

1) Determine what muscle is acting as the antagonist

Should I do this for each portion of step (weight acceptance, single

support, push off, swing) or for the whole step cycle?

The agonist will be determined as the muscle that has the smaller

integrated EMG envelope

Or, The agonist will be determined as the muscle that has the smaller

peak EMG.

Or is there some other method of determining which is the antagonist

muscle.

2) Get integral of linear envelope for both the agonist and antagonist

muscles.

Should we only compare one agonist muscle to one antagonist muscle at a

time (i.e. TA compared to Soleus)?

Or, should we compare all the agonist muscles about a joint (ankle) to

all antagonist muscle at the joint (i.e. TA compared to Soleus + Med. Gas

+ Lat. Gas)?

3) Get Total Integral of EMG (Itot)

Sum the two integrals from step six: Itot = Iant + Iagon

4) Calculate the co-contraction Index

CCI (%) = (2 x Iant)/Itot x 100

I worry this step could give me some misleading results. For example, lets

say I get a integral for the antagonist (Iant) equal to 1, and an integral

for the agonist (Iagon) equal to 2. Then the total integral of the EMG

(Itot) would be equal to 2 + 1 = 3. This would then give me a

co-contraction index of (2 x 1)/3 x 100 = 66%.

Then , lets say on the second trial I get a integral for the antagonist

(Iant) equal to 2, and a integral for the agonist (Iagon) equal to 4. Then

the total integral of the EMG (Itot) would be equal to 2 + 4 = 6 Vsec.

This would then give me a co-contraction index of (2 x 2)/6 x 100 = 66%.

Although the absolute magnitude of muscle activation increased for both

muscles but the co-contraction index stays the same. Does this seem

correct?

Other studies have quantified co-contraction using the duration of agonist

and antagonist muscle activation (Lamontagne et al, 2000; Dierick, et al,

2002). Is this a better method? What are its pros and cons?

Has anyone used the Noraxon Myoresearch Master Ed to analyze

co-contraction during gait? If so, is there an analysis protocol for this

software that is most appropriate.

Thank you for any help you may have to offer.

Respectfully,

Justus Ortega, Ph.D.

Biomechanics Lab

Department of Kinesiology

Humboldt State University

1 Harpst Street, Arcata CA 95521

(707) 826-4274

I am conducting a study in which I want to quantify agonist-antagonist

muscle co-contraction during gait. In this study, subjects walked at

several different slopes (including level walking). For each trial I

collected VO2 and EMG from eight muscles of one leg (Tibialis Anterior,

Medial Gastrocnemius, Lateral Gastrocnemius., Soleus, Vastus Medialis.,

Vastus Lateralis, Biceps Femoris, Gluteus Maximus) using a Noraxon Telemyo

system (sampling rate = 1000 Hz). I also collected foot switch data (both

feet). Prior to the walking trials, I collected EMG during quiet standing.

I am interested in determining if the reduced delta mechanical efficiency

observed in several of the subjects is related to an increase in

antagonist muscle co-contraction.

I would like to know what the best method for analyzing the EMG data is.

Here are some proposed steps for the initial EMG processing prior to

co-contraction analysis:

1) Filter EMG data (band pass: 10- 500 Hz; 2nd or 4th order Butterworth)

2) Rectify EMG data

3) Filter to get linear envelope (low pass: 6 Hz)

4) Normalize amplitude to mean integrated EMG of same muscle during level

walking trial (do not have MVC data)

5) Normalize time to step time using footswitch data (0% = heels strike,

100% = ipsilateral heel strike)

6) Determine when each muscle is OFF

Visually inspect and subjectively determine?

Or, Use prior research that gives the time (percentage of gait cycle)

that the muscle should not be activated?

Or, should I use some other method?

7) Determine when each muscle is ON

ON = When muscle activation is 3 standard deviations above the mean of

when muscle is OFF

Or, . Some other methods?

Steps to determine co-contraction (based on Falconer & Winter, 1985)

1) Determine what muscle is acting as the antagonist

Should I do this for each portion of step (weight acceptance, single

support, push off, swing) or for the whole step cycle?

The agonist will be determined as the muscle that has the smaller

integrated EMG envelope

Or, The agonist will be determined as the muscle that has the smaller

peak EMG.

Or is there some other method of determining which is the antagonist

muscle.

2) Get integral of linear envelope for both the agonist and antagonist

muscles.

Should we only compare one agonist muscle to one antagonist muscle at a

time (i.e. TA compared to Soleus)?

Or, should we compare all the agonist muscles about a joint (ankle) to

all antagonist muscle at the joint (i.e. TA compared to Soleus + Med. Gas

+ Lat. Gas)?

3) Get Total Integral of EMG (Itot)

Sum the two integrals from step six: Itot = Iant + Iagon

4) Calculate the co-contraction Index

CCI (%) = (2 x Iant)/Itot x 100

I worry this step could give me some misleading results. For example, lets

say I get a integral for the antagonist (Iant) equal to 1, and an integral

for the agonist (Iagon) equal to 2. Then the total integral of the EMG

(Itot) would be equal to 2 + 1 = 3. This would then give me a

co-contraction index of (2 x 1)/3 x 100 = 66%.

Then , lets say on the second trial I get a integral for the antagonist

(Iant) equal to 2, and a integral for the agonist (Iagon) equal to 4. Then

the total integral of the EMG (Itot) would be equal to 2 + 4 = 6 Vsec.

This would then give me a co-contraction index of (2 x 2)/6 x 100 = 66%.

Although the absolute magnitude of muscle activation increased for both

muscles but the co-contraction index stays the same. Does this seem

correct?

Other studies have quantified co-contraction using the duration of agonist

and antagonist muscle activation (Lamontagne et al, 2000; Dierick, et al,

2002). Is this a better method? What are its pros and cons?

Has anyone used the Noraxon Myoresearch Master Ed to analyze

co-contraction during gait? If so, is there an analysis protocol for this

software that is most appropriate.

Thank you for any help you may have to offer.

Respectfully,

Justus Ortega, Ph.D.

Biomechanics Lab

Department of Kinesiology

Humboldt State University

1 Harpst Street, Arcata CA 95521

(707) 826-4274