KR20000072178A - Method and system for monitoring muscle fatigue using of EMG median frequency - Google Patents

Abstract

PURPOSE: A muscle fatigue measuring method is provided to enable a user to maintain the strength of an appropriate exercise by monitoring a muscle fatigue state in real time. CONSTITUTION: A method of measuring muscle fatigue comprises measuring a surface electromyogram(EMG), obtaining median frequency(MDF) data through a power spectrum analysis from a stored signal in a constant epoch, performing the power spectrum analysis using the obtained power spectrum analysis(PSA) data as the subject to confirm a main signal frequency band, passing a main signal frequency band through a digital filter to obtain new median frequency data, and outputting a reduction aspect of the new median frequency data according to muscle fatigue generation and an index for judging the muscle fatigue.

Description

Method and system for monitoring muscle fatigue using EMG median frequency change

The present invention relates to a method and system for measuring muscle fatigue of muscles in motion, comprising: measuring surface EMG, storing EMG signals, and analyzing frequency spectrum from stored signals : Obtaining the median frequency (MDF) data at a certain period through the PSA), confirming the main signal frequency band by performing PSA on the obtained MDF data, the main signal frequency of the MDF data Obtaining a new MDF data by applying a digital filter to pass only the band, the new MDF data obtained is reduced with the occurrence of muscle fatigue, and the method comprising the step of outputting the index for determining the fatigue fatigue calculated from this data And the system to objectively measure and monitor muscle fatigue. Since the noise reduction effect during the dynamic contraction exercise is superior to the conventional method, muscle fatigue occurs while the user or the patient is easily carrying it for all kinds of exercises in sports related institutions or physical therapy rooms. By monitoring the degree in real time, it is possible to accurately maintain the intensity of the exercise suitable for the user and to check the training status of the muscle.

In muscle strengthening exercise, according to the DeLorme axiom and the overload principle, in order to induce maximum muscle activity, maximum resistance or maximum repetition is performed until muscle fatigue occurs. The principles apply similarly to endurance training.

Therefore, the muscle fatigue is not only necessary to determine the appropriate level of resistance and repetition frequency required for muscle strengthening exercise and endurance training, but also, especially in various neuromuscular lesions and myopathy where cardiovascular disease or muscle fatigue occurs early. It is necessary to closely monitor the level of muscle fatigue so as not to weaken due to excessive use.

However, the fatigue of the human body has an inherently behavioral definition of "a decrease in performance as a result of repeating a specific task", and its causes may occur at various levels of the human motor system, and whether or not muscle fatigue occurs. When trying to judge, subjective dispositions of both exercisers and examiners could be affected, so there was a limit in objectively evaluating them.

Since the discovery of the characteristic that the power spectrum analysis (PSA) of EMG signals shifts to the low frequency band as local muscle fatigue progresses due to continuous movement, it has undergone various mechanisms and significance tests, It began to be used.

When the PSA obtained through the Fast Fourier transformation is S _m (f), the median frequency (MDF) is higher than the mean frequency (MNF) as in Equation 1 below. Since it is known to be a little noisy, it is widely used as a representative parameter for measuring muscle fatigue. It was found to have intra-subject reliability.

Based on this fact, it was thought that it would be useful in various clinical fields if practical muscle fatigue measuring instrument could be used, so in 1982, Stulen and DeLuca introduced analog hardware. Since its development, equipment using microprocessors or digital signal processors and software has been developed occasionally, but all of them have been used only for isometric contraction, and they are not portable because they are not portable.

On the other hand, when performing muscle strengthening training in a physical therapy room or a gym, except for a special case, an isotonic exercise is performed as a dynamic exercise similar to real life muscle activity. Therefore, if you use a muscle fatigue meter in a real situation, you will usually target this kind of exercise.

However, EMG signals obtained from dynamic movements include random noise due to changes in muscle movement under electrodes, electrode swings, periodic muscle length, muscle contraction mode, muscle torque, and changes in blood flow. Since there are so many, it was difficult to obtain reliable MDF data and fatigue monitering variables from this signal.

In the meantime, there are not many attempts to solve this problem, such as applying a FFT after applying a moving average or the like to an EMG signal, or using MDF data using successively overlapped FFT epochs. In some cases, the root mean square (RMS) processing was performed on the obtained MDF data. In addition, because EMG signals from isotonic motion are irregular, FFT is sometimes performed by selecting only the static signal parts to make it suitable for FFT, but none of them sufficiently removes noise from MDF data.

Therefore, the object of the present invention is that the effect of reducing the noise generated during the dynamic contraction (dynamic contraction) exercise is superior to the conventional method, the user or patient to target all kinds of exercise in sports-related institutions or physical therapy room The present invention provides a method and system for measuring muscle fatigue using EMG intermediate frequency changes, which allows the user to accurately maintain the strength of exercise suitable for the user by monitoring the degree of fatigue of the muscle in real time while being easily carried. .

In order to achieve the above object as well as other objects that can be easily expressed, the present invention provides a method for measuring surface electromyography, storing an EMG signal, and performing a constant period through power spectrum analysis from a stored signal. Obtaining median frequency (MDF) data, confirming the main signal frequency band by performing PSA on the obtained MDF data, and passing only the main signal frequency band of the MDF data. ) To obtain new MDF data, reduce the new MDF data according to the occurrence of muscle fatigue, and output the index for determining muscle fatigue calculated from the data. Objectively measured and monitored.

1 is a schematic flowchart of a system of the present invention,

2 is a schematic hardware diagram of a system of the present invention;

3 to 5 are graphs showing muscle fatigue diagram measured by the method of the present invention in three subjects and a graph showing muscle fatigue diagram measured by another method for control.

* Description of the main parts of the drawing

1. Muscle 2. EMG measurement and amplification means

3. Bandpass filter 4. RMS converter

5. A / D Converter 6. Microprocessor

7. FFT analysis (MDF) 8. Low pass filter

9. RAM 10. Indication means

The present invention is described in more detail as follows.

The muscle fatigue measurement method using the change in the EMG intermediate frequency according to the present invention comprises measuring a surface EMG, storing an EMG signal, and performing a constant cycle through a power spectrum analysis (PAS) from the stored signal. epoch) to obtain median frequency (MDF) data, to determine the main signal frequency band by performing PSA on the obtained MDF data, and to pass only the main signal frequency band of MDF data. obtaining new MDF data by applying a filter), and obtaining new MDF data as the fatigue fatigue occurs and outputting the index for determining the fatigue fatigue calculated from the data.

The electrode used in the step of measuring EMG is not particularly limited, and in the present invention, the AE-131 circular surface EMG disposable electrode (NeuroDyne Medical Corp., MA., USA) is used. Three metal disks with a diameter of 12 mm are fixed in a triangular array with a distance of 20 mm between the centers of the disks. The active electrode and the reference electrode are placed in the direction of muscle fibers and the ground electrode is ground. It is located in the side or the inside more than these.

Compression sponge disks underlying the metal electrodes are flexible and fit on the curved surface of the muscle, are disposable with adhesive material on the floor, and the part connected to the cable is a quick snap button.

Therefore, the electrode was convenient to use for isotonic muscle contraction, and it is judged to be suitable for use in the general public when applied to the muscle fatigue measuring instrument.

The electrode attachment position was chosen as the lower third of the abdominal muscle of the biceps brachi, which is promoted at 90 ° bend, taking into account that the position of the muscle belly changes with muscle contraction. The skin was always disinfected with alcohol swab.

EMG equipment was used by connecting EMG100B, an EMG module, to MP100WSW (Biopac Systems Inc., CA., USA), sampling rate 1024Hz, low pass filter. One channel of EMG signal was collected at 500 Hz and high pass filter 30 Hz. Setting of measurement parameters and computer storage of EMG signals were done using the Acknowledge 3.53 (Acqknowledge, Biopac Systems Inc., CA., USA) program.

In the present invention, the elbow flexor maximum voluntary contraction (MVC) of the subject was measured, and in order to measure the elbow flexor maximum muscle strength, a digital tensiometer module TSD121C was connected to the MP100WSW. 1: 1 Chest pulley weights (Preston, MI., USA) were used to impose a load on the subject's elbow flexion contractions. The size of the load was adjusted, and a metronome (Wittner, Germany) was used to control the speed of the isotonic movement.

Subject was standing up next to the chest pulley weights with the back and upper arm back against the wall, and the wrist cuffs with leather cuffs on top of the wrist with towel cuffs firmly worn. At the elbow flexion, the upper forearm muscles were exercised with the lower arm completely supinated to avoid the effect of supination of the lower arm. First, to check the maximum strength of the elbow flexion, install a digital tanometer between the leather cuff and the pulley rope fixed to the floor with the elbow joint bent at 90 °. The test was repeated several times to select the maximum tension among them.

After strength measurement, the tanciometer and the rope fixed to the floor were removed, and the leather cuff end was connected to the pulley and the weight corresponding to 25% of the maximum strength, and the EMG electrode was attached.

And, in the same posture, bend and stretch the elbows in 25 minutes a minute (2.4 seconds per repetition). Endurance time Subjects reported verbally to the examiner and measured the EMG signal of the forearm biceps.

Completing the step of storing the surface EMG signal by storing the measured EMG signal in a computer using the Acknowledge 3.53 (Biopac Systems Inc., CA., USA) program. .

Next, frequency spectrum analysis is performed from the stored EMG signal, and frequency spectrum analysis is performed at an interval of 0.5 seconds to obtain MDF data, and PSA is performed on the obtained intermediate frequency data to identify the main signal frequency band. A step of obtaining new MDF data by applying a digital filter to pass only the main signal frequency details.

At this time, MDF data was obtained from EMG signal at regular time interval using “Romeo” software developed by Yonsei University's Medical Engineering Department. FFT (Frequency Spectrum Analysis) and digital filtering to check the main signal frequency details Silver 3.53 (Acqknowledge, Biopac Systems Inc., CA., USA) program was used.

The obtained data is output by using various output devices such as a display, a voice, an alarm, and the like, so as to prevent excessive exercise of the user or the patient and to exercise an appropriate amount of muscle.

The following examples and comparative examples illustrate the invention in more detail, but do not limit the scope of the invention.

Example 1

Three 20-year-old males stood up next to the chest pulley weights with their backs and upper arms attached to the wall, maintaining a standing posture, and wearing a leather cuff with metal pulley connectors on top of the wrist cuffs. At the time of bending, the upper forearm biceps exercised with the lower arm completely behind to avoid the effect of lagging the lower arm.

First, to check the maximum strength of the elbow flexion, install a digital tanometer between the leather cuff and the pulley rope fixed to the floor with the elbow joint bent at 90 °. The test was repeated several times to select the maximum tension among them.

After strength measurement, the tanciometer and the rope fixed to the floor were removed, and the leather cuff end was connected to the pulley and the weight corresponding to 25% of the maximum strength, and the EMG electrode was attached.

And, in the same posture, bend and stretch the elbows in 25 minutes a minute (2.4 seconds per repetition). Endurance time Subjects reported orally to the examiner, and measured and stored the EMG signal of the forearm biceps.

Then, frequency spectrum analysis is performed from the stored EMG signals, and frequency spectrum analysis is performed at intervals of 0.5 seconds to obtain MDF data, and when PSA is performed on the obtained MDF data, the main signal frequency band is 0.023 Hz or 0.031 Hz in all subjects. It was below (average 0.027Hz).

Next, new MDF data was obtained by applying a digital low pass filter of 0.023 Hz or 0.031 Hz, and this is output as a graph over time, and is represented by FIGS. 3 to 5 (d).

Comparative Example 1

MDF data was obtained in the same manner as in Example 1 until only PSA was performed at 0.5 second intervals from the stored EMG signal, and this is output as a graph as shown in FIGS. 3 to 5 (a).

Comparative Example 2

Perform frequency spectrum analysis, but obtain MDF data in the same manner as in Comparative Example 1, except that the PSA epoch is set to 2.4 seconds, and output it as a graph over time as shown in FIGS. 3 to 5 (b). It was.

Comparative Example 3

MDF data was obtained in the same manner as in Comparative Example 1, and this was calculated as a moving average for each of 13 consecutive data, and the graphs were output as graphs and are shown as (c) of FIGS. 3 to 5.

As can be seen from Example 1 and Comparative Examples 1 to 3, noises outside the regression line were the most severe in Comparative Example 1, but the noise decreased in order of Comparative Example 2, Comparative Example 3, and Example 1 to match the regression line. The year was showing.

In Comparative Example 1, since the first MDF value appeared 2.4 seconds after the start of the measurement, there was a blank section.In Comparative Example 2, since the moving average calculation value was set to be displayed at point 7, which is the center of 13 data points, the average value was calculated. In order to collect data only for the first and last 6 points, that is, the first three seconds and the last three seconds, there is no calculation value for the blank section, while in Example 1 there was no blank section of the data.

As a result of comparing the correlation coefficient of the regression line and the improvement effect of the coefficient of determination due to the reduction of random noise after the signal processing by Kendall analysis, the correlation in the order of Comparative Example 1 <Comparative Example 2 <Comparative Example 3 <Example 1 The coefficient and the coefficient of determination were increased so that the average correlation coefficient in Example 1 was 0.91, which is about 2.5 times that of Comparative Example 1, and the average coefficient of determination was 0.83, which is about 5.5 times that of Comparative Example 1. Was 0.86, indicating a small individual difference.

In addition, the F-test p value, which is a significance test of the regression line, also decreased in the order of Comparative Example 1> Comparative Example 2> Comparative Example 3 = Example 1, and averaged in Comparative Example 3 and Example 1 0.000, and the degree of agreement among subjects was 0.46.

Therefore, although the comparative examples 2 and 3 of Example 1 as well as the random noise reduction effect, it was found that Example 1 has the strongest noise reduction effect in most subjects.

On the other hand, as a result of comparing the effects of the various signal processing methods on the slope or intercept of the regression line by Kendall analysis, the slope of the regression line was shown in Comparative Example 2 <Comparative Example 1 <Comparative Example 3 <Example 1 However, since the correlations were increased in the order of, the slope of Example 1 was -0.10 Hz / sec. With the gentlest mean, and the degree of agreement among the subjects of this sequence was 0.74.

The Y-axis intercept obtained from the regression equation, that is, the initial MDF value, decreased in the order of Comparative Example 2> Comparative Example 1> Comparative Example 3> Example 1, and it was no longer possible to maintain the time signature of each subject's motion in the regression equation. The final MDF value obtained by inputting the endurance time decreased in the order of Comparative Example 3> Comparative Example 1> Comparative Example 2> Example 1.

Therefore, in Example 1, it was found that the effect of lowering the initial MDF value and the last MDF value was the greatest among the four methods described above, and the degree of agreement among the subjects of these sequences was 0.73 and 0.74, respectively, but there was not much individual difference.

When calculating the rate of change of MDF ((first MDF-final MDF) / final MDF), the order of decrease in the order of Comparative Example 2> Comparative Example 1> Example 1> Comparative Example 3 is only 0.35. There were a lot of individual differences.

Therefore, in Example 1, it was found that the effect of lowering the initial MDF and the final MDF and making the slope of the regression line somewhat smoother in most subjects was affected by the digital low pass filter. Seems crazy. When using MDF over time, the reliability of the test itself can be greatly increased and the reduction pattern is still obvious, which is not a serious problem.

As described above, in the present invention, the measurement of the surface EMG, the step of storing the EMG signal, and the intermediate frequency (Median) at a constant period through a frequency spectrum analysis (Power spectrum analysis: PSA) from the stored signal (Median) frequency: MDF) to obtain the main signal frequency band by performing PSA on the obtained MDF data, applying a digital filter to pass only the main signal frequency band of the MDF data new MDF Obtaining data, outputting new MDF data as the fatigue fatigue occurs, and outputting the index for determining muscle fatigue calculated from the data, to objectively measure and monitor muscle fatigue. Especially during dynamic contraction Since the noise reduction effect is superior to the conventional method, the user or the patient can easily carry the exercise for all kinds of exercise in sports related institutions or physical therapy rooms and monitor the degree of muscle fatigue in real time to monitor the intensity of the exercise suitable for the user. You can keep it correctly and check the training status of the muscles.

Claims (5)

Obtaining median frequency (MDF) data at a constant epoch through measuring the surface EMG, storing the EMG signal, and power spectrum analysis (PSA) from the stored signal. Step, checking the main signal frequency band by performing PSA on the obtained MDF data, obtaining a new MDF data by applying a digital filter to pass only the main signal frequency band of the MDF data, obtained new A method of measuring muscle fatigue using EMG intermediate frequency change, characterized in that the MDF data is reduced with the occurrence of muscle fatigue and the step of outputting the index for determining muscle fatigue calculated from the data. The method of claim 1, wherein the measuring of the EMG comprises sampling a MG signal of one channel at a sampling rate of 1024 Hz, a low pass filter of 500 Hz, and a high pass filter of 30 Hz. Muscle fatigue measurement method using the EMG intermediate frequency change, characterized in that performed by. The method of claim 1, wherein the digital filter passing only the main signal frequency band of the MDF data is a low pass filter. The method of claim 3, wherein the frequency of the filter is 0.05 Hz or less. The method of claim 1, wherein the method of outputting MDF data is a display, a voice, or an alarm.