Device for detecting the location of action potential generation and method for detecting the same

Abstract

There is a need for technology that can acquire information about the activity of living organisms or nerves connected to living organisms, as well as information transmitted from living organisms to nerves, from various action potentials, including those of muscles. [Solution] The disclosed technology may be an action potential generation location detection device. The disclosed device measures biopotentials with multiple electrodes, obtains multiple digital signals by sampling each of the multiple biopotential signals obtained from the measurement at a sampling frequency higher than the frequency of the action potential, and can determine the generation location of the action potential from the multiple digital signals.

Description

Technical Field

[0001] The present disclosure relates to an apparatus for detecting the generation position of action potentials and a method for detecting the same.

Background Art

[0002] Conventionally, electromyography has been known as a method for measuring muscle activity. Electromyography acquires an electrical signal based on the action potential of a muscle by an invasive or non-invasive method. From an electromyogram composed of the obtained electrical signal, for example, the intensity and fatigue of muscle activity and the evaluation of muscle function can be obtained.

Prior Art Documents

Non-Patent Documents

[0003]

Non-Patent Document 1

Summary of the Invention

[0004] An electromyogram obtained by conventional electromyography can evaluate various muscle activities as described above. However, the conventional electromyography method has limitations in capturing finer changes in muscle activity.

[0005] Therefore, a technique for acquiring information on the activity of a living body or a nerve connected to the living body and information transmitted from the living body to the nerve from various action potentials including muscles is desired.

[0006] The disclosed technique can be an apparatus for detecting the generation position of action potentials. The disclosed apparatus measures a bioelectric potential with a plurality of electrodes, acquires a plurality of digital signals obtained by sampling each of the signals of the plurality of bioelectric potentials measured at a sampling frequency higher than the frequency of the action potential, and can obtain the generation position of the action potential from the plurality of digital signals.

[0007] One aspect of this disclosure may be a method for detecting the location of an action potential. The disclosed method involves measuring biopotentials with multiple electrodes, obtaining multiple digital signals by sampling each of the multiple biopotential signals obtained at a sampling frequency higher than the frequency of the action potential, and determining the location of the action potential from the multiple digital signals.

[0008] Further details will be described in the embodiments below. [Brief explanation of the drawing]

[0009] [Figure 1] Figure 1 is a schematic diagram of the action potential generation location detection device according to an embodiment. [Figure 2] Figure 2 is a graph showing the time evolution of the potential of the electromyographic signal (EMG signal) input to the comparator. [Figure 3] Figure 3 is a graph showing an example of the digital signals of synchronized electromyography signals acquired from two surface electrodes. [Figure 4] Figure 4 is a conceptual diagram illustrating the time evolution of the potential of electromyographic signals measured at three points. [Figure 5] Figure 5 is a flowchart illustrating the process of determining the time difference of action potentials in order to find the location of the action potential generation performed by the analyzer. [Figure 6] Figure 6 is a conceptual diagram illustrating the process shown in Figure 5. [Figure 7] Figure 7 is a schematic diagram showing an electromyography measurement device as a comparative example. [Figure 8] Figure 8 is a graph showing the envelope of the digital signal of the electromyographic signal input to the analyzer. [Figure 9] Figure 9 is a graph showing the time evolution of the High / Low ratio of the normalized electromyographic signal. [Figure 10] Figure 10 is a graph showing the time evolution of the High / Low ratio of the electromyographic signal after applying a moving average filter to the electromyographic signal in Figure 9. [Modes for carrying out the invention]

[0010] <1. Overview of the device for detecting the location of action potentials and its detection method> (1) The action potential generation location detection device according to the embodiment measures biopotentials with multiple electrodes, obtains multiple digital signals by sampling each of the multiple biopotential signals obtained from the measurement at a sampling frequency higher than the frequency of the action potential, and can determine the generation location of the action potential from the multiple digital signals. In this case, fine changes in the action potential can be sufficiently observed and evaluated. Therefore, the action potential generation location detection device according to the embodiment can estimate the generation location of an action potential caused by muscle activity.

[0011] (2) The location of the generation of the action potential can be determined from the time difference of the action potential signals in the plurality of digital signals.

[0012] (3) The time difference of the action potential signals can be determined by the cross-correlation function of the action potential signals.

[0013] (4) The sampling may be performed synchronously between the signals of the plurality of biopotentials. In this case, since the synchronized signals of the plurality of biopotentials can be acquired as digital signals, the action potential generation location detection device according to the embodiment can accurately detect the generation location of the action potential with less computational load.

[0014] (5) The sampling may be performed on a single FPGA (Field-Programmable Gate Array) to which the signals of the multiple biopotentials are input. This makes it possible to easily achieve synchronization between the signals of the multiple biopotentials in the action potential generation position detection device according to the embodiment.

[0015] (6) The sampling frequency may be 100 kHz or higher.

[0016] (7) The action potential is an action potential generated in muscle fibers, and based on pre-stored muscle-nerve correspondence data associating the muscle fibers with the types of nerves connected to the muscle fibers, and the determined generation position of the action potential, the type of nerve connected to the muscle fibers is determined, and the activity of the determined nerve can be evaluated.

[0017] (8) An amplifier that amplifies each of the plurality of bioelectric potential signals measured by the electrode, a band-pass filter that removes signals outside a predetermined frequency band corresponding to the effective frequency band of the bioelectric potential signal from each of the plurality of bioelectric potential signals output from the amplifier, a quantizer that quantizes each of the plurality of bioelectric potential signals output from the band-pass filter, a signal processing device that synchronously samples the quantized plurality of bioelectric potential signals output from the quantizer, and an analysis device that determines the generation position of the action potential based on the digital signal output from the signal processing device may be provided.

[0018] (9) The signal processing device may be a single FPGA (Field-Programmable Gate Array) that synchronously samples the quantized plurality of bioelectric potential signals.

[0019] (10) The bioelectric potential may be a surface electromyogram.

[0020] (11) The method for detecting the generation position of an action potential according to the embodiment measures a bioelectric potential with a plurality of electrodes, obtains a plurality of digital signals obtained by sampling each of the plurality of bioelectric potential signals measured at a sampling frequency higher than the frequency of the action potential, and determines the generation position of the action potential from the plurality of digital signals.

[0021] <2. Example of an apparatus for detecting the generation position of an action potential and its detection method> The embodiments will be described in more detail below with reference to the drawings. In the embodiments, the biopotential is described as surface electromyography (EMG) measured by surface electrodes placed on the skin surface, and the action potential is the action potential of muscle fibers. However, the biopotential disclosed may be based on other electrical activity such as that of the heart or brain. Furthermore, the EMG is not limited to surface EMG, but may be EMG measured by needle electrodes.

[0022] Figure 1 is a schematic diagram of the action potential generation location detection device 10 according to the embodiment.

[0023] The action potential generation location detection device 10 (hereinafter simply referred to as "detection device 10") includes a plurality of surface electrodes 11a, 11b, and 11c, a preamplifier 12, a bandpass filter 13, a comparator 14, an FPGA (Field-Programmable Gate Array) 15, and an analyzer 21. The preamplifier 12, bandpass filter 13, and comparator 14 are actually arranged according to the number of surface electrodes (three in Figure 1), but for explanatory purposes, only one preamplifier 12, bandpass filter 13, and comparator 14 are shown in Figure 1.

[0024] Surface electrodes 11a, 11b, and 11c are multiple electrodes placed on the skin surface 1 to measure electromyography (EMG). Surface electrodes 11a, 11b, and 11c, for example, have three surface electrodes and can measure EMG at multiple points. Surface electrodes 11a, 11b, and 11c are positioned on the skin surface 1 along the direction of the muscle fibers 3 of the muscle 2. Excitation of motor nerve cells 6 in the spinal cord 5, etc., propagates along the nerve axon 7 and reaches the neuromuscular junction 8. EMG is based on action potentials generated when this excitation reaches the muscle fibers 3, and consists of complex action potentials of multiple muscle fibers 3.

[0025] In the following explanation, when surface electrodes 11a, 11b, and 11c are not distinguished, they will simply be referred to as surface electrode 11.

[0026] The preamplifier 12 is an example of an amplifier that amplifies each electromyographic signal based on the electromyographic potential measured by the surface electrode 11 by a predetermined magnification (e.g., 200 times).

[0027] The bandpass filter 13 (BPF13) allows signals within a predetermined frequency band to pass through the electromyographic signal amplified by the preamplifier 12, while removing signals of other frequencies. The predetermined frequency band corresponds to the effective frequency band of the electromyographic signal of the muscle fiber 3, which is from 5 Hz to 500 Hz, for example, signals from 20 Hz to 498 Hz.

[0028] Comparator 14 is an example of a quantizer, which compares each electromyographic signal output from the bandpass filter 13 against a predetermined threshold voltage and quantizes each electromyographic signal. Comparator 14 consists of a number of comparators corresponding to the desired number of quantization bits; for example, there is one comparator for 1 bit and seven comparators for 3 bits.

[0029] Here, Figure 2 is a graph showing the time evolution of the potential of the electromyographic (EMG) signal input to the comparator 14. For example, the comparator 14 binarizes each EMG signal with respect to a threshold voltage of 2.04V.

[0030] FPGA 15 is an example of a signal processing device and is an integrated circuit that can be programmed as required. FPGA 15 has a number of sampling circuits 16 corresponding to the number of surface electrodes 11 (number of channels). FPGA 15 samples the multiple quantized electromyographic signals output from each comparator 14 in synchronization with the electromyographic signals. Specifically, FPGA 15 samples three electromyographic signals in a clock-synchronized state by sampling circuits 16 that operate with a common sampling clock from a clock generator 17. FPGA 15 samples at a sampling frequency higher than the frequency of the action potential of the muscle fiber 3. It is preferable that the sampling is sufficiently fast. In this embodiment, the sampling frequency is, for example, 100 kHz or higher (100 samples per 1 msec), which is sufficiently higher than the effective frequency of the electromyographic signal (5 Hz to 500 Hz). The sampling frequency is more preferably 1 MHz or higher (1,000 samples per 1 msec), more preferably 10 MHz or higher (10,000 samples per 1 msec), and even more preferably 100 MHz or higher (100,000 samples per 1 msec). Since the duration of a single action potential (muscle pulse) is approximately 1 msec, increasing the sampling frequency to the above levels allows for observation of a single action potential (muscle pulse) at a high frequency, making it easier to individually observe individual action potentials included in the complex action potential of multiple muscle fibers 3. The upper limit of the sampling frequency is not particularly limited, but 10 GHz is used as an example. Here, Figure 3 is a graph showing an example of the digital signals of synchronized electromyographic signals acquired by two surface electrodes 11a and 11b.

[0031] The detection device 10 uses an FPGA 15 with processing capabilities corresponding to the number of electromyographic signals for sampling, thereby enabling it to obtain multiple electromyographic signals sampled in a synchronized manner. In other words, the FPGA 15 allows the subsequent analysis device 21 to suitably perform processing on multiple electromyographic signals measured simultaneously.

[0032] The analysis device 21 may be a personal computer such as a desktop PC, notebook PC, or mobile PC. The analysis device 21 has a processor and memory. The processor is, for example, a CPU (Central Processing Unit). The memory includes flash memory, EEPROM (Electrically Erasable and Programmable Read Only Memory), ROM (Read Only Memory), RAM (Random Access Memory), etc. The memory may be a primary storage device or a secondary storage device. The analysis device 21 also has a display unit for displaying content and an input unit for receiving instructions.

[0033] The analyzer 21 can determine the location of an action potential from multiple synchronized electromyographic signals, which are multiple digital signals. The location of the action potential may be, for example, the neuromuscular junction 8, which is the starting point of the propagation of the electromyographic potential, or the location of the innervation zone, which is a collection of neuromuscular junctions 8. The analyzer 21 can determine the location of an action potential from, for example, the time difference of the action potentials (included) in multiple electromyographic signals. The analyzer 21 can determine the time difference using the cross-correlation function of the action potential signals.

[0034] Here, Figure 4 is a conceptual diagram illustrating the time evolution of the potential of electromyographic signals measured at three points.

[0035] For example, if the distance (position) between surface electrodes 11a and 11b is known, the time delay between the electromyographic signal measured at surface electrode 11a and the electromyographic signal measured at surface electrode 11b is proportional to the difference in distance from surface electrodes 11a and 11b to the location where the action potential is generated. Therefore, the distance from the skin surface 1 to the location where the action potential is generated (depth from the skin surface 1) can be estimated by triangulation from the electromyographic signals measured at the two points of surface electrodes 11a and 11b.

[0036] Furthermore, the location (three-dimensional position) of the action potential can be estimated from the electromyographic signals measured at three surface electrodes 11a, 11b, and 11c.

[0037] Figure 5 is a flowchart illustrating the process of determining the time difference of action potentials in order to determine the generation location of the action potentials performed by the analyzer 21. Figure 6 is a conceptual diagram illustrating the process in Figure 5. In Figures 5 and 6, an example is used to explain how to determine the time difference of electromyographic signals measured at two points where two surface electrodes 11a and 11b are placed, using a cross-correlation function.

[0038] First, the analyzer 21 takes multiple electromyographic signals measured by surface electrodes 11a and 11b and acquired from FPGA 15, and extracts signals within a certain interval from t to (t+a), as shown in Figure 3, and defines them as signals F1(t) and F2(t) (step S1 in Figure 5). The certain interval is determined according to the frequency of the signal to be detected, for example, 25ms.

[0039] Next, the analyzer 21 generates a waveform (signal) by shifting the signal F2(t) in the time axis direction by n × dt (where n is between -N and +N (N is an integer)) (step S2). As shown in Figure 6, the analyzer 21 generates a signal F2(t + n × dt) for a predetermined time while shifting it by a small amount of time dt.

[0040] Next, the analyzer 21 calculates the correlation coefficient F3(n) between F1 and F2, as shown in Figure 5, for each generated signal F2(t+n×dt) (step S3).

[0041] Next, the analyzer 21 finds the value of n that maximizes the correlation coefficient F3(n) (step S4). Based on this, the analyzer 21 finds the time difference, which is the time delay n between the signal F1(t) and the signal F2(t) (step S5).

[0042] Based on the time delay n obtained between each surface electrode 11 in this manner, the analyzer 21 determines the location of the action potential generation (distance from the surface electrode 11 or the three-dimensional position of the action potential generation location) as described above.

[0043] The analyzer 21 can determine the type of nerve connected to the muscle fiber 3 based on the location of the action potential generation, evaluate the activity of the determined nerve, and output the results to the user. For this purpose, the analyzer 21 has muscle-nerve correspondence data 22 (correspondence data 22). Correspondence data 22 is data pre-stored in the analyzer 21 that associates the muscle fiber 3 with the type of nerve connected to this muscle fiber 3. Based on the correspondence data 22 and the determined location of the action potential generation, the analyzer 21 determines the type of nerve connected to the muscle fiber 3 and evaluates the activity of the determined nerve.

[0044] For example, the analyzer 21 can evaluate, based on the location of the action potential generation, whether the connection between the muscle fiber 3 from which the action potential originated and the nerve is good, and whether the nerve signal is properly reaching the muscle fiber 3. Furthermore, if a subject to be measured with surface electrodes 11 attached intends to move but the movement does not occur, the analyzer can evaluate nerve activity and the state of the muscle fiber 3 based on the location of the action potential generation, the type of nerve involved, and whether or not an action potential was generated. For example, if no action potential is observed despite the intention to move, it can be inferred that there is some kind of abnormality in the nerve (such as fatigue). Also, if action potentials are only observed in a limited area, it can be inferred that there is some kind of abnormality in the muscle fiber 3 (such as fatigue).

[0045] Furthermore, the detection device 10 can acquire the location of action potential generation and evaluate muscle activity not only for efferent signal transmission from nerve to muscle 2, but also for mesial signal transmission from muscle 2 to nerve.

[0046] Furthermore, the detection device 10 can also be used, in the same way as known electromyography measurement devices, for detecting the envelope of electromyographic signals, which is used as a time-domain analysis method for muscle activity.

[0047] Here, Figure 7 is a schematic diagram showing an electromyography measurement device 30 as a comparative example.

[0048] The electromyography measurement device 30 includes a surface electrode 11a, a preamplifier 12, a bandpass filter 13, a rectifier 24, a lowpass filter 25, an analog-to-digital converter 26, and an analyzer 21. The preamplifier 12, bandpass filter 13, and analyzer 21 are substantially the same as those of the detection device 10 shown in Figure 1, so their explanation is omitted.

[0049] The surface electrode 11a is one of the electrodes 11 of the detection device 10 shown in Figure 1.

[0050] The rectifier 24 full-wave rectifies the electromyography signal, which includes the signal in the effective frequency band of the electromyography signal output from the bandpass filter 13, and generates a DC signal.

[0051] The low-pass filter 25 (LPF25) smooths the DC signal output from the rectifier 24 by passing only the electromyography signals below a predetermined value (e.g., 200 Hz or less). The analog-to-digital converter 26 (ADC26) converts the electromyography signal into a multi-bit digital signal at a sampling frequency of 1 kHz, taking into account the typical frequency of electromyography. The analysis device 21 performs the necessary processing on the digital signal acquired from the analog-to-digital converter 26.

[0052] Figure 8 is a graph showing the envelope of the digital signal of the electromyographic signal input to the analyzer 21. Muscle activity corresponding to voltage can be read from this envelope.

[0053] To confirm that the detection device 10 of this embodiment has similar performance to the electromyography measurement device 30, the envelope was determined from the same electromyography signal using the following method. Note that the detection device 10 uses only one surface electrode 11a, and does not use surface electrodes 11b and 11c.

[0054] The surface electrodes 11a of the electromyography (EMG) measurement device 30 and the detection device 10 are common to the bandpass filter 13. The EMG signal output from the bandpass filter 13 is branched to the rectifier 24 of the EMG measurement device 30 and the comparator 14 of the detection device 10. Furthermore, in order to synchronize the branched signals in the FPGA 15 and the configuration from the rectifier 24 onward of the EMG measurement device 30, a switch connected to these components is mounted on the FPGA 15. Measurements were performed so that the subject exerted muscle force as quickly as possible after pressing the switch. In Figures 8 to 10, the timing of the switch being pressed (Switch Rising Edge) was set to 0.0 seconds.

[0055] As shown in Figure 1, FPGA 15 may further have a normalization circuit 18. Note that the normalization circuit 18 is provided for the purpose of explaining the comparative example, and the normalization circuit 18 may be omitted. The normalization circuit 18 normalizes the digital signal of the sampled electromyography signal. Specifically, FPGA 15 normalizes the proportion of time the electromyography signal was High (High / Low ratio) over a 1ms period to 7 bits. Here, Figure 9 is a graph showing the time change of the High / Low ratio of the normalized electromyography signal.

[0056] The analyzer 21 applied a moving average filter with a window width of 100 ms to the electromyographic signal output from FPGA 15, smoothing the electromyographic signal as shown by the Higha / Low ratio in Figure 9. Figure 10 is a graph showing the time change of the High / Low ratio of the electromyographic signal after applying the moving average filter to the electromyographic signal in Figure 9.

[0057] Comparing the time evolution of the electromyographic signal shown in Figure 8, obtained by the electromyographic measurement device 30 as a comparative example, with the time evolution of the electromyographic signal shown in Figure 10, obtained by the detection device 10, it was observed that the rise time of the electromyographic signal around 0.6 seconds after the switch was pressed, and the subsequent time evolution of the electromyographic signal, were almost identical.

[0058] From this, it can be said that the detection device 10 of this embodiment is capable of time-domain electromyography (EMG) measurement in much the same way as the EMG measurement device 30, which is an example of a known EMG measurement device. Furthermore, the detection device 10 can also analyze the frequency components of the signal by performing a Fast Fourier Transform on the time change of the High / Low ratio of the normalized EMG signal using the analysis device 21, and converting the time-domain signal to the frequency domain.

[0059] The present invention is not limited to the above embodiments, and various modifications are possible.

[0060] For example, the electromyographic signal measured by the surface electrode 11 may be subjected to signal correction processing according to the characteristics (transfer function) of the signal transmission path. This makes it possible to correct signal distortion in the time domain or amplitude (frequency domain) of the signal.

[0061] The configuration of the detection device 10 is just one example; it is not limited to this configuration, as long as it can acquire multiple digital signals sampled at a sampling frequency higher than the frequency of the action potential, and the location of the action potential can be obtained from the multiple digital signals. [Explanation of Symbols]

[0062] 1:Skin surface 2: Muscle 3: Muscle fibers 5: Spinal cord 6: Motor neurons 7: Nerve axons 8: Neuromuscular junction 10: Action potential generation location detection device (detection device) 11a: Surface electrode 11b: Surface electrode 11c: Surface electrode 12: Preamplifier 13: Bandpass filter (BPF) 14: Comparator 15: FPGA 16: Sampling Circuit 17: Clock generator 18: Normalization circuit 21: Analyzer 22: Muscle-nerve correspondence data (correspondence data) 24: Rectifier 25: Low-pass filter (LPF) 26: Analog-to-Digital Converter (ADC) 30: Electromyography Measurement Device

Claims

1. Bioelectric potential is measured using multiple electrodes. Each of the multiple biopotential signals obtained by measurement is sampled at a sampling frequency higher than the action potential frequency to acquire multiple digital signals. From the aforementioned multiple digital signals, the location where the action potential is generated is determined. A device for detecting the location of action potential generation.

2. The location of the generation of the action potential is determined from the time difference of the action potential signals in the plurality of digital signals. The device for detecting the location of an action potential generation according to claim 1.

3. The time difference of the action potential signals is determined by the cross-correlation function of the action potential signals. The device for detecting the location of an action potential generation according to claim 2.

4. The sampling is performed in synchronization with the signals of the plurality of biopotentials. The device for detecting the location of an action potential generation according to claim 1.

5. The sampling is performed on a single FPGA (Field-Programmable Gate Array) to which the signals of the multiple biopotentials are input. The device for detecting the location of an action potential generation according to claim 4.

6. The sampling frequency is 100 kHz or higher. The device for detecting the location of an action potential generation according to claim 1.

7. The aforementioned action potential is an action potential generated in a muscle fiber. Based on pre-stored muscle-nerve correspondence data that associates the muscle fibers with the types of nerves connected to them, and the determined location of the action potential, the type of nerve connected to the muscle fiber is determined, and the activity of the determined nerve is evaluated. The device for detecting the location of an action potential generation according to claim 1.

8. An amplifier that amplifies the signals of each of the plurality of biopotentials measured by the electrodes, A bandpass filter removes signals from each of the plurality of biopotential signals output from the amplifier from a predetermined frequency band corresponding to the effective frequency band of the biopotential signals, A quantizer that quantizes each of the plurality of biopotential signals output from the bandpass filter, A signal processing device that synchronously samples the multiple quantized biopotential signals output from the quantizer, An analysis device that determines the location of the action potential based on the digital signal output from the signal processing device, The action potential generation location detection device according to claim 1, comprising:

9. The signal processing device is a single FPGA (Field-Programmable Gate Array) that synchronously samples the quantized signals of the plurality of biopotentials. The device for detecting the location of an action potential generation according to claim 8.

10. The bioelectric potential mentioned above is the surface electromyographic potential. The device for detecting the location of an action potential generation according to claim 1.

11. Bioelectric potential is measured using multiple electrodes. Each of the multiple biopotential signals obtained by measurement is sampled at a sampling frequency higher than the action potential frequency to acquire multiple digital signals. From the aforementioned multiple digital signals, the location where the action potential is generated is determined. A method for detecting the location of action potential generation.

Images