Comparison of a new method of equalization of electromyographic activity of knee muscles while walking with the equalization method based on maximum isometric contraction

Abstract:

In electromyography investigations, the normalization of signals is generally accepted in order to be able to compare and interpret. But due to the lack of a standard method for normalization, there are many differences in the normalization method.  These differences are related to the reference signal measurement method. The purpose of this study was to compare the reliability of three normalization methods: submaximal contraction, dynamic maximal contraction, and isometric maximal voluntary contraction for normalizing lower limb muscle activity. Method: Fourteen students of Bo Ali Sina University randomly participated in this study from among the available people. The average age, height and mass were (24.5 years), (165.2) and (55.9), respectively. To record muscle electromyographic activity, bipolar surface electrodes and 16-channel EMG device model MA-300-16 (with frequency 2500Hz, bandwidth 1250, low-pass filter 500-10 and notch filter 50Hz) and four Vicon cameras model MX T-series camera were used to separate movement phases with frequency 200Hz. Several different methods were used to obtain the reference signal, including, for the rectus femoris muscle, load carrying methods with 20% of body weight lying on the bed, Sargent jump and squat with 70% of one maximum repetition, latissimus dorsi and external latissimus dorsi, load carrying methods with 20% of body weight lying on the bed and squat with 70% of one maximum repetition, external and internal biceps, Sargent jump and load carrying methods with 20% of body weight In the standing position and for the tibialis anterior muscle carrying a load with 20% of the body weight while sitting on a chair and the conventional MVIC method. electrodes on the right thigh muscles, latissimus dorsi, latissimus dorsi, latissimus dorsi, quadriceps externus; Medial biceps and anterior tibia were located according to SENIAM guidelines. For statistical analysis, ANOVA with repeated measures, multivariate ANOVA and correlated T test were used. ICC and coefficient of variation (CV) were also used for data reliability. Results: For the right thigh muscle, the values ??of the coefficient of variation for the three methods of squatting, carrying a load, and Sargent's jump were 30.6, 40.6, and 20.4, respectively, and the ICC for these three methods were 0.97, 0.97, and 0.96, respectively. The correlation between two repetitions in the squat method was (0.73 to 0.93), for load carrying (0.97 to 0.98) and for Sargent's jump (0.72 to 0.95). The values ??of CV and ICC for the vastus medialis muscle were 29.6 and 39.2, 0.97 and 0.98 respectively in load carrying and squat methods. The correlation between the repetitions in the two methods of carrying a load and squat was also obtained (0.95 to 0.96) and (0.94 to 0.98), respectively. For the vastus externus, the CV and ICC values ??in the two methods of carrying a load and squat were 25.8, 27.2, 0.96 and 0.90, respectively. The correlation between two repetitions was (0.79 to 0.93) and for the squat method (0.90 to 0.95). For the external and internal biceps, in Sargent's carrying and jumping methods, the CV values ??were (42, 37) and (29, 38) respectively, and the ICC value was (0.96, 0.93) and (0.97, 0.96) respectively. The correlation between the repetitions for the external biceps muscle in the two methods of load carrying and Sargent's jump were (0.85 to 0.88) and (0.95 to 0.97), respectively, and for the internal biceps muscle these values ??were (0.75 to 0.94) and (0.78 to 0.86) respectively. In the tibialis anterior muscle, CV and ICC were 38 and 33, 0.97 and 0.96, respectively, in the two methods of load carrying and MVIC. The correlation between the repetitions for this muscle in the two load carrying methods and MVIC was (0.96 to 0.99) and (0.96 to 0.97) respectively. For the rectus femoris muscle, Sargent's jumping method was preferred because of the lower coefficient of variation. It is recommended for the latissimus dorsi due to its smaller coefficient of variation. In the external broad muscle, because the amount of all the factors were very close to each other, MVIC method is recommended to prevent more fatigue and reduce time. Considering all the factors for the external and internal twin muscles, Sargent jump is recommended due to the greater intensity of activity in Sargent's jump compared to carrying a load. For the tibialis anterior muscle, due to the smaller coefficient of variation in the MVIC method, using this method was preferred. Although the values ??in all factors were acceptable in other tests for all muscles.

Key words: normalization - electromyography - knee joint muscles - walking

1-1 Introduction

The electrical signal, in connection with muscle contraction, is called electromyography or EMG[1] (Winter [2] 2009).. In another definition, it can be said that electromyography is the study of electrical activity in muscles. The electromyography device can provide appropriate information about voluntary movements and muscle reflexes. In fact, the electromyography device measures the muscle action potential. EMG electrodes with relatively high sensitivity receive the action potential in the muscles and transfer it to the computer memory and printer (Peter Konrad; 2005). The value of the amplitude[3] and frequency[4] of the raw signal[5] is sensitive and variable to many factors.

DeLuca[6] (1997) believes that many internal and external factors affect these signals. External factors include the shape of the electrode, the distance between the electrodes, the distance of the electrodes from the motor-point and the outer edge of the muscles, as well as the origin of muscle fibers and skin preparation, apparent resistance [7], sweating and skin temperature. Internal factors include the physiological, anatomical and biomechanical characteristics of muscles, such as muscle activity intensity, muscle fiber type composition, blood flow in muscles, muscle fiber diameter [8], the distance between active fibers in the muscle with respect to the electrode and the amount of tissue between the surface of the muscle and the electrode.

Given that many factors affect EMG signals and voltages recorded from muscles, it is difficult to describe them; Therefore, it is problematic to describe the amplitude of the raw EMG signal, unless a normalization method [9] is performed. Normalization refers to converting the signal to a known scale and value.

It is reported that the normalized EMG signal was first introduced by Eberhart, Inman & Bresler in 1954. Since then, many normalization methods have been proposed, but there is no consensus on one method as the best. (Eberhart, H.D. Inman, V.T. Bresler, B. 1954) The data obtained by electromyography can be used to estimate muscle force and muscle fatigue. In order to achieve the above goals, we need to normalize the raw data. Normalization is done with the aim of making the data obtained from electromyography comparable and there are different methods for it. The most common normalization method is the maximum voluntary isometric contraction method [10] (MVIC), which considers the recorded maximum voluntary isometric contraction as a reference value. In this method, there are at least three repetitions and rest time between each repetition to reduce any fatigue. EMG signals from a maximal contraction are filtered after recording and their RMS is calculated [11]. The maximum value obtained from the processed signal in all repetitions is used as the reference value for the normalization of the EMG signals. This method is simple, however, researchers are trying to answer the question, what test should be used to produce maximum neuromuscular activity in the muscle? Unfortunately, there is no consensus about the test that can create maximum activity in the muscle. Another problem that exists in this method is whether the subject uses his maximum effort or if the subject group is a patient, can he make his maximum effort? In this regard, researchers use several other methods for normalization. One of these methods is maximum RMS during dynamic movement[12]. In this method, each activity point is divided by the recorded peak value. Another similar method is the average RMS during dynamic movement [13], and in that each point is divided by the recorded average RMS. Many researchers have used these two methods for normalization, but there are no reasons to prefer these methods over the MVIC method. Another method that can be mentioned for normalization is the maximum submaximal contraction method [14]. DeLuca stated that this method has close reliability with the maximum contraction method, in this method we use submaximal isometric contraction, which includes holding a member against a certain load or a percentage of the maximum load that the person can bear. There are more methods in the field of normalization, which are beyond the scope of the discussion. The ambiguous point is that the normalization method for muscle activity is still not accepted by everyone