WO2018209495A1

WO2018209495A1 - Muscle fatigue monitoring system

Info

Publication number: WO2018209495A1
Application number: PCT/CN2017/084343
Authority: WO
Inventors: 潘盈秀
Original assignee: 东莞市棒棒糖电子科技有限公司
Priority date: 2017-05-15
Filing date: 2017-05-15
Publication date: 2018-11-22
Also published as: CN111491559B; CN111491559A

Abstract

A muscle fatigue monitoring system, comprising: an sEMG amplifier module (101) for receiving sEMG signals and amplifying the received sEMG signals; a wave filter module (103) connected to the sEMG amplifier module (101); a bit stream converter (105) connected to the wave filter module (103) and used to digitize the sEMG signals and convert the sEMG signals to a discrete signal on the basis of a single threshold value but not to digitize the entire sEMG signals; and a bit stream cross-correlator (107) connected to the bit stream converter (105), the bit stream cross-correlator (107) comprising a plurality of correlation cascades (201) that are connected in series, a plurality of counters (303) respectively connected to the plurality of correlation cascades (201), and a maximum value selector (205) connected to the counters (303), the bit stream cross-correlator (107) being used to continuously correlate the sEMG signals during a given time window, calculate all time points with the same sEMG signal, repeatedly compare all the counters (303), and find, via the counter (303) having the maximum value, the distance between specific reference points on the sEMG signals.

Description

Muscle fatigue detection system

Technical field

This patent application relates generally to medical electronics, and more particularly to a muscle fatigue testing system.

Background technique

Muscle Fiber Conduction Velocity (hereinafter referred to as MFCV) is a measure of the conduction velocity of a motor unit's action potential in muscle tissue and is one of the most important items in response to muscle activity. Compared to monitoring only power spectral density, MFCV provides a more detailed understanding of muscle fatigue and muscle recovery. Compared to the median/average frequency analysis, MFCV monitoring can better track muscle fatigue.

One conventional MFCV tracking method is to extract information only from the detected surface myoelectric signal (hereinafter referred to as sEMG). In general, spectral analysis tools are needed. This method is sensitive to noise and introduces large deviations in the results. Another conventional method is to compare two or more detected sEMG signals along the direction of the muscle fibers. The electrodes are placed perpendicular to the underlying muscle fibers. The algorithm according to the method includes finding the distance between the reference points. This method assumes that the noise of the two measured signals is exactly the same. Therefore, when the delay is introduced, the two signals have the same shape. Thus, any particular reference point, such as a valley, peak, or zero, can be used to align the two signals and estimate the delay between them. Therefore, the phase difference between the detected signals can be calculated to estimate the MFCV. Cross-correlation Function) The maximum time delay can be used as an estimate of the delay.

Based on SoC (System-on-Chip) CMOS (Complementary Metal Oxide Semiconductor) solutions offer significant promise in small form factor, low power consumption and high accuracy wearable medical device solutions. Therefore, it is necessary to implement the cross-correlation method using low-power digital CMOS logic with low computational complexity, high efficiency, and good anti-interference.

Summary of the invention

This patent application provides a muscle fatigue monitoring system. In one embodiment, a muscle fatigue monitoring system includes: a sEMG amplifier module for receiving a sEMG signal and amplifying the received sEMG signal; a filtering module coupled to the sEMG amplifier module; and a bit stream coupled to the filtering module a converter for digitizing the sEMG signal and converting the sEMG signal into a discrete signal based on a single threshold without digitizing the entire sEMG signal; a bitstream cross correlator coupled to the bitstream converter, the bitstream cross correlator comprising a series a plurality of correlation stages, a plurality of counters respectively coupled to the plurality of correlation stages, and a maximum value selector coupled to the counter, the bit stream cross correlator for continuously correlating the sEMG signals within a given time window, Calculating all the time points of the same sEMG signal, cyclically comparing all the counters, and finding the distance between specific reference points on the sEMG signal by a counter having a maximum value; a bias generator; and a cross-correlator with the bit stream a connected timing control module; and a serial peripheral connection connected to the timing control module and the maximum value selector mouth. The sEMG amplifier module includes multiple dual channel measurement amplifiers and an external floating high pass filter. The filtering module includes two low pass filters for extracting a frequency band of 10 Hz–500 Hz signal properties. The bit stream converter includes two analog comparators. The maximum value selector includes a plurality of comparison modules for comparing the values of the counters in pairs and processing to evaluate the comparison results, each comparison module for comparing two 14-bit numbers. Each correlation stage includes a delay module, a counter, and a correlator. The delay module is a D-type flip-flop, and the delay time of the delay block is controlled by the sampling frequency of the system. The counter for each correlation stage is a 14-bit traveling wave counter whose size is selected by analyzing the previous sEMG data. The correlator includes an AND gate and an AND gate connected to the same OR gate.

The low pass filter can be Sallen with a cutoff frequency of 2.5 kHz Key low pass filter. The reference voltages of the two analog comparators can be kept separate to allow for offset mismatch compensation.

In another embodiment, a muscle fatigue monitoring system includes: a sEMG amplifier module for receiving a sEMG signal and amplifying the received sEMG signal; a filtering module coupled to the sEMG amplifier module; and a bit coupled to the filtering module A stream converter for digitizing the sEMG signal and converting the sEMG signal to a discrete signal based on a single threshold without digitizing the entire sEMG signal; and a bitstream cross correlator coupled to the bitstream converter. The bitstream cross correlator includes a plurality of correlation stages connected in series, a plurality of counters respectively coupled to the plurality of correlation stages, and a maximum value selector coupled to the counter. The bitstream cross correlator is configured to continuously correlate the sEMG signal within a given time window, calculate all time points of the same sEMG signal, cyclically compare all the counters, and then find out the sEMG signal by using a counter having a maximum value The distance between specific reference points. The sEMG amplifier module includes multiple dual channel measurement amplifiers and an external floating high pass filter. The filtering module includes two low pass filters. The bit stream converter includes two analog comparators. Each correlation stage includes a delay block, a counter, and a correlator.

The muscle fatigue monitoring system can also include a bias generator, a timing control module coupled to the bitstream correlator, and a serial peripheral interface coupled to the timing control module and the maximum value selector. The filtering module can be used to extract the frequency band to 10 Hz–500 Hz signal properties. The maximum value selector can include a plurality of comparison modules for comparing the values of the counters in pairs and then evaluating the comparison results, each comparison module for comparing two 14-bit numbers. The delay module can be a D-type flip-flop, and the delay time of the delay block can be controlled by the sampling frequency of the system. The counter of each correlation stage can be a 14-bit traveling wave counter whose size can be selected by analyzing the previous sEMG data. The correlator can include an AND gate and an AND gate connected to the same OR gate.

DRAWINGS

1 is a block diagram of a muscle fatigue monitoring system in accordance with an embodiment of the present patent application.

2 is a bitstream cross correlator of the muscle fatigue monitoring system of FIG. 1.

Figure 3 shows the correlation level in Figure 2.

Figure 4 illustrates the timing logic used by the maximum selector of Figure 2.

detailed description

The muscle fatigue monitoring system of the present patent application will be described in detail below with reference to the accompanying drawings and embodiments.

1 is a block diagram of a muscle fatigue monitoring system in accordance with an embodiment of the present patent application. Referring to FIG. 1, the muscle fatigue monitoring system includes a sEMG amplifier module 101, a filtering module 103 connected to the sEMG amplifier module 101, a bit stream converter 105 connected to the filtering module 103, and a bit stream cross-connected with the bit stream converter 105. Device (bit-stream cross correlator) 107.

The sEMG amplifier module 101 includes a plurality of dual channel measurement amplifiers for receiving sEMG signals and amplifying the sEMG signals it receives. The sEMG amplifier module 101 is capable of filtering 300 out of the biopotential electrode mV DC polarization voltage.

The sEMG amplifier module 101 also includes an external floating high pass filter. The external floating high-pass filter does not require a grounding resistor compared to the traditional passive high-pass filter, resulting in a very large common-mode input impedance.

The filtering module 103 includes two low pass filters for extracting the frequency band from 10 Hz to 500 Hz signal properties. Preferably, the low pass filter employs a Sallen Key low pass filter with a cutoff frequency of 2.5 kHz.

The bitstream converter 105 includes two analog comparators for digitizing the sEMG signal. The reference voltages of the two comparators remain separated to allow for offset mismatch compensation.

The bitstream cross correlator 107 is configured to continuously correlate sEMGs within a given time window, calculate all time points of the same sEMG signal, cyclically compare all counters, and then find out the specifics on the sEMG signal by a counter having a maximum value The distance between the reference points.

Figure 2 shows the bitstream cross correlator 107 of the muscle fatigue monitoring system of Figure 1. Referring to Figure 2, the bitstream cross correlator 107 includes a plurality of correlation stages 201 in series. The bitstream cross correlator 107 also includes a plurality of counters 203 that are respectively coupled to a plurality of correlation stages 201. At the end of the relevant time window, all counters 203 of the system are read. The correlation level (i.e., delay) of the counter with the maximum value represents the time lag between the two input signals.

Referring to FIG. 2, the bitstream cross correlator 107 further includes a maximum value selector 205 coupled to the counter 203 for cyclically comparing all of the counters 203. The maximum value selector 205 includes a number of comparison modules, each of which is used to compare two 14-bit numbers. The maximum selector compares all results (such as the value of the counter) in pairs, and the post-processing evaluates the results of the above comparison.

Referring to FIGS. 1 and 2, the muscle fatigue monitoring system further includes a bias generator 202, a timing control module 204 coupled to the bitstream cross correlator 107, and a string coupled to the timing control module 204 and the maximum value selector 205. Line peripheral interface.

Figure 3 shows the correlation level in Figure 2. Referring to FIGS. 2 and 3, each correlation stage 201 includes a delay module 301, a counter 303, and a correlator 305. In this embodiment, the delay module 301 is a D-type flip-flop. The delay time is controlled by the sampling frequency of the system. The counter is a 14-bit traveling wave counter. The size of the counter is selected by analyzing the previous sEMG data to allow operation of the associated time window and high sampling frequency for more than 1 second. The correlator 305 includes an AND gate 3051 and an AND gate 3053 connected to the same OR gate 3051. The same OR gate 3051 is used as a bit correlator that improves the traditional AND gate design by considering all possible digital cases.

FIG. 4 shows the timing logic used by the maximum value selector 205 of FIG. 2. Referring to Figure 4, each maximum operation returns a binary flag that is passed to the next comparison and indicates which of the two comparison values is the maximum. A binary one indicates that the first of the two values is larger. When the operation is completed, the counter position value (delay value) is returned instead of the counter value.

In this embodiment, the bitstream cross correlator 107 is used to perform cross correlation. The algorithm calculates the time delay between the sEMG signals. The algorithm is applied to find distances between specific reference points, such as valleys, peaks, or zeros, thereby simplifying the process of cross-correlation. The sEMG signal is converted to a discrete signal by the bitstream converter based on a single threshold without the need to digitize the entire sEMG signal while preserving the necessary information for cross-correlation and delay estimation. This eliminates the need to cross-correlate the entire sEMG signal, but only requires a single digit approximation of the cross-correlated sEMG signals, and thus the architecture of the cross-correlator is greatly simplified.

In this embodiment, the bitstream buffer window is eliminated by continuously intersecting the two sEMG signals within a given time window. This is achieved by calculating all the same time points for the two sEMG signals. For an x(n), the cross-correlation time window replaces the buffer window.

In this embodiment, the discrete time lag of the cross-correlation output is obtained by continuously delaying the input signal to obtain cross-correlation results for each discrete time lag. The time lag between the two signals is returned by a counter having a larger value, and therefore, the number of required transistors is greatly reduced.

The above description is only a preferred embodiment of the present patent application, and is not intended to limit the scope of the present application. Although the present application is disclosed in the preferred embodiments, the present application is not intended to limit the patent application. A person skilled in the art can make a slight change or modify the equivalent embodiment by using the technical content disclosed above, without departing from the technical solutions of the present patent application, without departing from the technical scope of the present patent application. Any simple modifications, equivalent changes and modifications made to the above embodiments in accordance with the teachings of the present application are within the scope of the technical solutions of the present patent application.

Claims

A muscle fatigue monitoring system comprising:

a sEMG amplifier module for receiving the sEMG signal and amplifying the received sEMG signal;

a filter module connected to the sEMG amplifier module;

a bitstream converter coupled to the filtering module for digitizing the sEMG signal and converting the sEMG signal to a discrete signal based on a single threshold without digitizing the entire sEMG signal;

a bitstream cross-correlator coupled to the bitstream converter, the bitstream cross-correlator comprising a plurality of correlation stages connected in series, a plurality of counters respectively coupled to the plurality of correlation stages, and a maximum value selection connected to the counter The bitstream cross correlator is configured to continuously cross-correlate the sEMG signal within a given time window, calculate all time points of the same sEMG signal, cyclically compare all the counters, and find the sEMG through a counter having a maximum value The distance between specific reference points on the signal;

Bias generator

a timing control module coupled to the bitstream cross correlator;

a serial peripheral interface coupled to the timing control module and the maximum value selector; wherein:

The sEMG amplifier module includes a plurality of dual channel measurement amplifiers and an external floating high pass filter;

The filtering module includes two low pass filters for extracting signal properties in a frequency band of 10 Hz - 500 Hz;

The bit stream converter includes two analog comparators;

The maximum value selector includes a plurality of comparison modules for comparing the values of the counters in pairs and then processing to evaluate the comparison results, each comparison module for comparing two 14-bit numbers;

Each correlation stage includes a delay module, a counter, and a correlator;

The delay module is a D-type flip-flop, and the delay time of the delay block is controlled by the sampling frequency of the system;

The counter of each correlation stage is a 14-bit traveling wave counter whose size is selected by analyzing the past sEMG data;

The correlator includes an AND gate and an AND gate connected to the same OR gate.

2. The muscle fatigue monitoring system according to claim 1, wherein the low pass filter is Sallen with a cutoff frequency of 2.5 kHz. Key low pass filter.

3. The muscle fatigue monitoring system of claim 1 wherein the reference voltages of the two analog comparators remain separated to allow offset mismatch compensation.

4. A muscle fatigue monitoring system comprising:

a filter module connected to the sEMG amplifier module;

a bitstream cross-correlator coupled to the bitstream converter, the bitstream cross-correlator comprising a plurality of correlation stages connected in series, a plurality of counters respectively coupled to the plurality of correlation stages, and a maximum value selection connected to the counter The bitstream cross correlator is configured to continuously correlate the sEMG signal within a given time window, calculate all time points of the same sEMG signal, cyclically compare all the counters, and then find the sEMG signal through a counter having a maximum value The distance between specific reference points; where:

The filtering module includes two low pass filters;

The bitstream converter includes two analog comparators;

Each correlation stage includes a delay block, a counter, and a correlator.

5. The muscle fatigue monitoring system according to claim 4, wherein the muscle fatigue monitoring system further comprises a bias generator, a timing control module coupled to the bitstream correlator, and the timing control module and the maximum value The serial peripheral interface to which the selector is connected.

6. The muscle fatigue monitoring system according to claim 4, wherein the filtering module is configured to extract a frequency band of 10 Hz - 500 Hz signal properties.

7. A muscle fatigue monitoring system according to claim 4, wherein the maximum value selector comprises a plurality of comparison modules for comparing the values of the counters in pairs and then evaluating the comparison results, each comparison module being used for Compare two 14-digit numbers.

8. The muscle fatigue monitoring system according to claim 4, wherein the delay module is a D-type flip-flop, and the delay time of the delay block is controlled by a sampling frequency of the system.

9. The muscle fatigue monitoring system according to claim 4, wherein the counter of each correlation stage is a 14-bit traveling wave counter, and the size of the counter is selected by analyzing the past sEMG data.

10. The muscle fatigue monitoring system of claim 4, wherein the correlator comprises an AND gate and an AND gate connected to the same gate.