JP6857573B2 - EMG measuring device, method and program - Google Patents

Description

The present invention relates to electromyographic measuring devices, methods and programs for measuring surface electromyograms.

In recent years, surface electromyography (sEMG) has been used to observe muscle activity in the fields of medical research and clinical rehabilitation, as well as in the fields of human optics such as sports science and occupational medicine. The surface electromyogram can be measured relatively easily by simply placing a disposable electrode (disposable electrode) or a dry electrode on the skin close to the muscle to be observed.

In the analysis of electromyograms, it is common to convert the measurement signal into features such as the root mean square (RMS). However, in the measurement of surface EMG, when the electrode floats or shifts due to body movement or the like and the contact state between the skin and the electrode changes, the change becomes a noise component and is included in the measured value, and the RMS value changes. To do. Therefore, when analyzing the muscle activity of the observation target, it is necessary to analyze the potential change due to the muscle activity and the potential change due to the noise component separately.

Therefore, conventionally, a method of removing a noise component from a measured value of a surface electromyogram by a filter processing technique generally used as a signal processing technique has been proposed. For example, Non-Patent Document 1 proposes a method of removing a body motion noise component by using a bandpass filter having a passing frequency band of 20 to 450 Hz.

Carlo J. De Luca et al. “Filtering the surface EMG signal Movement artifact and baseline noise contamination”, Journal of Biomechanics, Volume 43, Issue 8, Pages 1573-1579, (2010)

However, even if a bandpass filter is used, the noise component contained in the 20 to 450 Hz band cannot be removed, so that further improvement in analysis accuracy cannot be expected.

The present invention has been made by paying attention to the above circumstances, and the purpose of the present invention is to make it possible to remove noise components that cannot be removed by the filtering technique, thereby further improving the analysis accuracy. To provide methods and programs.

In order to achieve the above object, the first aspect of the electromyographic measuring device or method according to the present invention is from each of the electrodes installed on the surface of the interlocking first and second muscle parts, respectively. The first and second measurement signals showing the surface EMGs by the second muscle parts are acquired, and the feature amounts of the surface EMGs are calculated based on the first and second measurement signals, respectively, and the above-mentioned A correlation value indicating the degree of correlation between the calculated feature amounts is calculated, and each feature amount is corrected based on the calculated correlation value. In the correction, the calculated correlation value is used. When the value is equal to or less than a preset threshold value, the calculated feature amount is corrected based on the correction information stored in the storage medium in advance, and the feature amount when the correlation value reaches the maximum value. Is obtained, and each feature amount is stored in the storage medium as the correction information .

According to the first aspect of the present invention, the following effects can be achieved. That is, since the feature amounts of the surface electromyograms measured at the interlocking first and second muscle parts are substantially the same when only the components due to muscle activity are included, the correlation value takes the maximum value. On the other hand, for example, when skin movement or electrode misalignment that does not depend on muscle activity occurs, a noise component is mixed in each of the above feature amounts, and the correlation value decreases. Therefore, by correcting the feature amount based on the correlation value between the above feature amounts, it is possible to remove the noise component that cannot be removed only by the filter processing technique, which can further improve the analysis accuracy. ..

FIG. 1 is an overall configuration diagram of a system including an electromyographic measuring device according to an embodiment of the present invention. FIG. 2 is a block diagram showing a functional configuration of an electromyographic measuring device according to an embodiment of the present invention. FIG. 3 is a flowchart showing a procedure and processing contents of myoelectric measurement control by the myoelectric measurement device shown in FIG. It is a figure which shows an example of RMS which shows the muscle activity of the temporalis muscle and the correlation coefficient R at the time of a masticatory movement. It is a figure which shows an example of RMS which shows the muscle activity of the temporalis muscle when the tooth is clenched, and the correlation coefficient R. It is a figure which shows an example of RMS which shows the muscle activity of the temporalis muscle at the time of blinking, and the correlation coefficient R. It is a figure which shows an example of RMS which shows the muscle activity of the temporalis muscle when the brim of a hat is moved left and right, and the correlation coefficient R.

Hereinafter, embodiments according to the present invention will be described with reference to the drawings.
[One Embodiment]
(Constitution)
FIG. 1 is an overall configuration diagram of a system including an electromyographic measuring device according to an embodiment of the present invention.
As shown in FIG. 1, electrodes 2L and 2R are attached to the skin surfaces on the left and right head muscles of the user, respectively. The left and right head muscles are, for example, one of the muscles for performing masticatory movements and are located near the temples.

The electrodes 2L and 2R consist of, for example, cloth electrodes having an electrode size of 30 mm × 30 mm, and _{output surface electromyograms (sEMG) showing muscle activity of the left and right head muscles as measurement signals IS L} and IS _R , respectively. To do. These measurement signals IS _L and IS _R are sent to the myoelectric measurement device 1 via, for example, a signal cable.

The electrode 2L, the 2R, for example Bluetooth and attaching a radio unit which adopts a low-power radio communication standard (registered trademark) or the like, thereby wirelessly transmitting the measurement signal IS _L, the IS _R to electromyographic measuring device 1 It may be configured to do so. Further, the attachment positions of the electrodes 2L and 2R can be arbitrarily selected as long as they are on the left and right temporalis muscles, and the size, shape, material, attachment means, etc. of the electrodes 2L and 2R are also surface electromyograms. Anything can be used as long as it can measure.

The myoelectric measurement device 1 uses, for example, a personal computer, and is configured as follows. FIG. 2 is a block diagram showing a functional configuration of the hardware and software. The myoelectric measurement device 1 includes a control unit 10, a storage unit 20, and an input / output interface unit 30.

_{The input / output interface unit 30 has a function of converting measurement signals IS L} and IS _R representing surface electromyograms (sEMG) output from the electrodes 2L and 2R into digital data that can be handled by the control unit 10. Be prepared. The input / output interface unit 30 also has a function of outputting the corrected measurement data output from the control unit 10 to an external device. As the external device, for example, a user terminal equipped with a display, and an information processing device such as a personal computer or a server that determines a user's masticatory movement based on measurement data are assumed.

The storage unit 20 uses, for example, a non-volatile memory such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive) that can be written and read at any time and a volatile memory such as a RAM as a storage medium. In addition to the program storage unit, the area is provided with a measurement data storage unit 21 and a baseline value storage unit 22.

The measurement data storage unit 21 is used to store measurement data representing the surface electromyograms of the left and right temporalis muscles detected by the electrodes 2L and 2R. The baseline value storage unit 22 is used to store the baseline value used to correct the measurement data when the measurement data representing the surface electromyogram (sEMG) contains a noise component. Will be done.

The control unit 10 has at least one processor and a working memory, and has sEMG measurement data acquisition control unit 11, BPF processing unit 12, and RMS calculation unit as processing functions necessary for implementing the present embodiment. A base line value calculation unit 14, a correlation coefficient calculation unit 15, and a correction processing unit 16 are provided. All of these processing functions are realized by causing the processor to execute the program stored in the storage unit 20.

The sEMG measurement data acquisition control unit 11 takes in measurement data corresponding to the surface electromyogram (sEMG) representing the movement of each of the left and right temporal muscles from the input / output interface unit 30 for a predetermined time, and these measurement data. Is controlled to be stored in the measurement data storage unit 21 in chronological order.

The BPF processing unit 12 has a bandpass filter processing function in which a frequency pass band is preset in order to remove noise components due to body movement or the like. Then, by performing the bandpass filter processing on the measurement data read from the measurement data storage unit 21, noise components due to body movement or the like are removed from the measurement data.

The RMS calculation unit 13 sets a window having a predetermined time width for the measurement data filtered by the BPF processing unit 12, and performs a process of calculating the feature amount of the measurement data for each window. As the feature amount, for example, the root mean square (RMS) is used.

In the calibration phase, the baseline value calculation unit 14 calculates the average value of the feature amount (RMS) of the measurement data calculated for each window by the RMS calculation unit 13 for a predetermined time. Then, a process of storing the calculated average value of RMS as a baseline value in the baseline value storage unit 22 is performed.

In the normal measurement mode after the baseline value is set, the correlation coefficient calculation unit 15 is calculated between the RMSs of the surface electromyograms (sEMGs) of the left and right temporal muscles calculated for each window by the RMS calculation unit 13. Performs a process of calculating the correlation coefficient (R) of.

The correction processing unit 16 compares the correlation coefficient (R) calculated by the correlation coefficient calculation unit 15 with a preset threshold value, and as a result of this comparison, the correlation coefficient (R) is equal to or less than the threshold value. In this case, it is determined that a large amount of noise component still remains in the RMS at this time. Then, a process of correcting the RMS value based on the baseline value stored in the baseline value storage unit 22 is performed.

(motion)
Next, the measurement operation of the surface EMG by the EMG measuring device 1 configured as described above will be described.
(1) Setting of Baseline Value Prior to the actual measurement of the surface EMG, the EMG measuring device 1 executes a calibration process for determining the baseline value used for RMS correction as follows.

First, as illustrated in FIG. 1, the electrodes 2L and 2R are attached to the skin on the left and right temporalis muscles of the user. For example, cloth electrodes 2L and 2R having an electrode size of 30 mm × 30 mm are positioned and attached to the skin on the left and right temporal muscles of the user so that the center distance between the electrodes is 40 mm.

Then, in the resting state, the control unit 10 under the control of the sEMG measurement data acquisition control unit 11 inputs the measurement data representing the surface EMG measured by the electrodes 2L and 2R to the input / output interface unit 30 for 5 seconds. Is taken in from, and stored in the measurement data storage unit 21. In this case input and output interface unit 30, the sampling rate 1000 Hz, the gain 1,000-fold magnification, measuring signal IS _L, is performed sampled IS _R, it is converted into measurement data consisting of the digital signal.

Next, the myoelectric measurement device 1 reads out each measurement data from the measurement data storage unit 21 by the BPF processing unit 12, and performs bandpass filter processing on the measurement data. At this time, the pass band of the bandpass filter processing is set to 20-450 Hz. As a result, the BPF processing unit 12 can obtain measurement data in which the noise component included in the band other than 20-450 Hz is removed from the pass band.

Subsequently, under the control of the RMS calculation unit 13, windows are set every 100 msec for each measurement data from which noise components have been removed by the BPF processing unit 12, and the root mean square root of each measurement data is set for each of these windows. (RMS) is calculated. Then, for each of the above measurement data, the average value of the RMS calculated for each of the above windows for the above 5 seconds is calculated, and the average value of the calculated RMS is used as the baseline value of each measurement data as the baseline value. It is stored in the storage unit 22. Thus, a baseline value corresponding to each of the left and right temporalis muscles is obtained.

(2) Measurement of surface EMG When the determination of the baseline value is completed, the EMG measuring device 1 executes, for example, a measurement process of a user's masticatory movement as follows. FIG. 3 is a flowchart showing the processing procedure and the processing content.

Similar to the above baseline value setting process, the myoelectric measurement device 1 first controls the left and right temporal muscles detected by the electrodes 2L and 2R in step S11 under the control of the sEMG measurement data acquisition control unit 11. The measurement data representing the surface EMG is sequentially fetched from the input / output interface unit 30 in chronological order and stored in the measurement data storage unit 21.

Next, under the control of the BPF processing unit 12, the myoelectric measurement device 1 reads out the measurement data corresponding to the left and right temporal muscles from the measurement data storage unit 21 in step S12, and 20 for these measurement data. Performs filtering processing that allows only the −450Hz band to pass and attenuates other bands. This filtering process removes most of the noise components unrelated to the masticatory movement.

Then, under the control of the RMS calculation unit 13, the myoelectric measurement device 1 sets a window with a time width of 100 msec for the measurement data corresponding to the left and right temporal muscles after the filtering process in step S13. Then, for each of these windows, the root mean square (RMS) representing the feature amount of each of the above measurement data is calculated.

Next, under the control of the correlation coefficient calculation unit 15, the myoelectric measurement device 1 sets a window for 5 seconds for the RMS of the measurement data corresponding to each of the left and right temporal muscles in step S14. The correlation coefficient (R) between the RMSs of each of the above measurement data during the period of this window is calculated.

Then, under the control of the correction processing unit 16, the myoelectric measurement device 1 first compares the calculated correlation coefficient (R) with the threshold value in step S15, and determines the correlation coefficient (R). Determine if it is less than or equal to the value. The threshold is set to, for example, 0.9.

As a result of the above comparison, if the correlation coefficient (R) is larger than the threshold value, the similarity of RMS calculated from the measurement data of the left and right temporal muscles is high, and the noise component that imbalances the left and right RMS values. Is not included. Then, in step S16, the process proceeds to step S18 without correcting the RMS value, and the measured RMS value is used as the data OS representing the measurement result as it is from the input / output interface unit 30 to the user terminal or management server (not shown). To output to etc.

On the other hand, it is assumed that the correlation coefficient (R) is equal to or less than the threshold value as a result of the above comparison. In this case, the correction processing unit 16 determines that the RMS contains a large amount of noise components and needs to be corrected. Then, in step S17, the baseline value is read from the baseline value storage unit 22, and the RMS value is replaced with the baseline value. Then, the process proceeds to step S18, and the corrected RMS value is output from the input / output interface unit 30 to a user terminal or a management server (not shown) as a data OS representing the measurement result.

(3) Specific examples of measurement results (3-1) Masticatory movement When a masticatory movement is performed, the left and right temporalis muscles move with the same contractile force at almost the same time in the closing movement, so the RMS value is the same on the left and right. As a result, the correlation coefficient (R) becomes a high value, that is, a value close to "1".
FIG. 4 shows an example of the change in the left and right RMS values and the correlation coefficient (R) when the masticatory movement is performed for 5 seconds. As shown in the figure, the RMS value is about the same on the left and right, and as a result, the correlation coefficient (R) is as high as “0.97”.

(3-2) When clenching teeth When clenching teeth, the left and right temporalis muscles are linked, so the RMS value is about the same on the left and right, as in the case of the above masticatory movement, and the correlation coefficient. (R) is also a high value.
FIG. 5 shows an example of the change in the left and right RMS values and the correlation coefficient (R) when the tooth clenched state is held for 5 seconds. As shown in the figure, the RMS value is similar on the left and right, and as a result, the correlation coefficient (R) is as high as “0.95”.

(3-3) When blinking On the other hand, when blinking is performed, the skin on the surface of the temporalis muscle is deformed even though the temporalis muscle is not active. Therefore, a noise component that cannot be removed by the bandpass filter processing is mixed in the RMS. In this case, the degree of mixing of the noise component into the measured value due to blinking differs between the left and right, and the relationship is not constant, so that the correlation coefficient becomes low.
FIG. 6 shows an example of the change in the left and right RMS values when blinking is performed for 5 seconds and the calculation result of the correlation coefficient (R). As shown in the figure, the RMS value is less related to the left and right, and as a result, the correlation coefficient (R) is "0.83", which is lower than that in the case of masticatory movement or the like.

(3-4) When the brim of the hat is moved left and right When the user holds the brim of the hat and moves it left and right, the noise component is caused by the electrode shifting even though the temporalis muscle is not active in this case as well. Will be mixed. In this case, the degree of mixing of noise components differs between the left and right and is not constant, so that the correlation coefficient becomes low.
FIG. 7 shows an example of the change in the left and right RMS values when the hat is moved left and right for 5 seconds, and the calculation result of the correlation coefficient (R). As shown in the figure, the RMS value is less related to the left and right, and as a result, the correlation coefficient (R) is "0.8", which is lower than that in the case of masticatory movement.

In the present embodiment, the threshold value is set to "0.9" based on the above measurement results, and the necessity of correction is determined. Therefore, it is possible to distinguish the masticatory movement of the user from other movements that cause noise and make an accurate determination.

(effect)
As described in detail above, in one embodiment, measurement data of the surface electromyogram (sEMG) of each of the left and right temporalis muscles is acquired, and only 20-450 Hz is passed through these measurement data to pass other frequency components. After performing the bandpass filter processing to be removed, the RMS indicating the feature amount is calculated. Then, the correlation coefficient (R) between the RMSs corresponding to the left and right temporal muscles is calculated, and when the correlation coefficient (R) is equal to or less than the threshold value, the RMS indicates the movement of the temporal muscles. Considering that a large amount of noise components that do not approach each other remain, the RMS value is replaced with a preset baseline value.

Therefore, it is possible to remove noise components that cannot be removed only by the filter processing technique, which can further improve the analysis accuracy. In addition, since noise removal processing using a bandpass filter is also used, it is possible to eliminate the effects of body movement components other than masticatory movements in advance, which makes it possible to more accurately calculate the RMS and determine the necessity of correction. It becomes possible to do. Further, since the baseline value for correction is stored in the baseline value storage unit 22 in advance, the correction process can be performed by a simple process simply by reading the baseline value. Moreover, since the threshold value used for determining the necessity for correction is set based on the data measured in advance for each user, an appropriate threshold value can be set for each user.

[Other Embodiments]
In the above embodiment, as a correction processing method, when the calculated correlation coefficient is equal to or less than the threshold value, the calculated RMS is replaced with the baseline value. However, the present invention is not limited to this, and a correction amount for the baseline value may be calculated according to the difference between the correlation coefficient and the threshold value, and the calculated RMS value may be corrected based on this correction amount.

In the above embodiment, the baseline value is set based on the actual measurement result, but the calibration process may be omitted and the standard baseline value prepared in advance may be stored. .. Further, the threshold value may be updated regularly or at any time based on the analysis accuracy of the masticatory movement or the like, in addition to using a fixed value set in advance.

Further, in the above-described embodiment, a case where the movements of the left and right temporalis muscles are measured in order to analyze the masticatory movement has been described as an example. However, it is not limited to this, and in addition to the temporalis muscle, the masseter muscle located near the back teeth is also linked to the left and right during the masticatory movement, so the noise component is removed by the same processing procedure and the surface electromyogram (sEMG) is performed. Can be measured.

In addition to the masticatory movement, the present invention is also applicable to other movements such as bending and stretching movements in which the muscles move in tandem with left-right symmetry. In addition, the configuration of the EMG measuring device, the procedure and processing content of the surface EMG measurement process, the threshold value, and the like can be variously modified and implemented without departing from the gist of the present invention.

In short, the present invention is not limited to the above-described embodiment as it is, and at the implementation stage, the components can be modified and embodied within a range that does not deviate from the gist thereof. In addition, various inventions can be formed by an appropriate combination of the plurality of components disclosed in the above-described embodiment. For example, some components may be removed from all the components shown in the embodiments. In addition, components from different embodiments may be combined as appropriate.

1 ... EMG measuring device, 2R, 2L ... Electrode, 10 ... Control unit, 11 ... sEMG measurement data acquisition control unit, 12 ... BPF processing unit, 13 ... RMS calculation unit, 14 ... Baseline value calculation unit, 15 ... Phase Relationship number calculation unit, 20 ... storage unit, 21 ... measurement data storage unit, 22 ... baseline value storage unit, 30 ... input / output interface unit.

Claims (4)

An acquisition unit that acquires first and second measurement signals showing surface electromyograms of the first and second muscle parts, respectively, from electrodes installed on the surfaces of the first and second muscle parts that are interlocked with each other. When,
A feature amount calculation unit that calculates the feature amount of the surface EMG based on the first and second measurement signals, respectively.
A correlation value calculation unit that calculates a correlation value indicating the degree of correlation between each of the calculated feature quantities,
It is provided with a correction unit that corrects each feature amount based on the calculated correlation value.
The correction unit
A storage medium for storing correction information for correcting each feature amount, and
When the calculated correlation value is equal to or less than a preset threshold value, a correction processing unit that corrects each of the calculated feature amounts based on the stored correction information, and a correction processing unit.
A correction information generation unit that obtains each of the feature amounts when the correlation value takes the maximum value and stores each of the feature amounts in the storage medium as the correction information.
Electromyographic measuring device comprising a. The feature amount calculation unit
A filter unit that extracts signals in a preset frequency band from the first and second measurement signals, respectively.
The myoelectric measurement device according to claim 1, further comprising a calculation unit for calculating the feature amount from each signal extracted by the filter unit. It is an EMG measurement method executed by an EMG measuring device.
The process of acquiring the first and second measurement signals showing the surface electromyograms of the first and second muscle parts, respectively, from the electrodes installed on the surfaces of the first and second muscle parts that are interlocked with each other. ,
The process of calculating the feature amount of the surface EMG based on the first and second measurement signals, respectively, and
The process of calculating the correlation value indicating the degree of correlation between the calculated feature quantities, and
It includes a process of correcting each feature amount based on the calculated correlation value .
The correction process is
When the calculated correlation value is equal to or less than a preset threshold value, the process of correcting each of the calculated feature amounts based on the correction information stored in the storage medium in advance, and
A process of obtaining each feature amount when the correlation value takes a maximum value and storing each feature amount as the correction information in the storage medium.
EMG measurement method with a. A program that causes a processor to function as each of the parts included in the myoelectric measurement device according to claim 1 or 2.

Images