JP2023114575A - Biological signal processing device - Google Patents

Abstract

To provide a biological signal processing device capable of acquiring a period of a living body motion having periodicity with a high degree of reliability without taking much processing time.SOLUTION: A biological signal processing device includes: a plurality of biological signal generation units (13 and 14) for generating biological signals that change according to motions of a living body including periodical motions (cardiac motions and respiratory motions) through different processes respectively; period detection units (15 and 16) for detecting a period of a signal component that changes according to periodical motions of the living body in the biological signal; reliability determination units (15 and 16) for determining a degree of reliability of the period detected by the period detection units; and a living body motion period determination unit (20) for determining the most probable period as the living body motion period on the basis of the degree of reliability acquired by the reliability determination units for each of the plurality of periods from the plurality of periods each acquired by the period detection units.SELECTED DRAWING: Figure 3

Description

TECHNICAL FIELD The present invention relates to a biological signal processing apparatus that processes biological signals that change according to the movement of a living body including periodic movements such as heartbeat and respiratory movement of the living body.

2. Description of the Related Art There has been proposed an apparatus for detecting the respiration rate and heart rate of a living body such as a person from a living body signal that changes according to the movement of the living body (for example, see Patent Document 1). This device has a sensor that outputs a detection signal that changes according to the load placed on the bed, and frequency analysis (for example, FFT method) of the detection signal output from the sensor while a person is lying on the bed. I do. Based on the result of frequency analysis of the detection signal, biological physiological information such as a frequency component corresponding to respiration and a frequency component corresponding to pulse (heartbeat) is extracted. According to such a device, it is possible to know biological physiological information such as respiration and heartbeat of a person lying on the bed without touching the person, and also to confirm whether the person is alive or not. can do.

JP-A-2003-552

However, in the apparatus using the frequency analysis (for example, FFT method) of the detection signal as described above, it takes a long time to obtain a plurality of obtained data and increase the reliability by statistics. This makes it unusable for systems that need to display real-time respiration or heart rate, or alerts based on that respiration or heart rate.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a biological signal processing apparatus capable of obtaining a period of periodic biological motion with high reliability without taking much processing time. is.

A biosignal processing apparatus according to the present invention includes a plurality of biosignal generation units that generate biosignals that change according to the motion of a living body including periodic motion (heartbeat, respiratory motion) through different processes; a cycle detection unit for detecting a cycle of a signal component in the biosignal generated by each of the biosignal generation units that changes according to the periodic movement of the living body; and a reliability determination unit for determining a degree of reliability, based on the degrees of reliability obtained by the reliability determination unit for each of the plurality of cycles from the plurality of cycles each obtained by the cycle detection unit. , and a body motion cycle determination unit that determines the most probable cycle as the body motion cycle.

With such a configuration, each of the plurality of biomedical signal generators generates biomedical signals that change according to the movement of the living body through different processes. A cycle of a signal component in the biosignals generated by each of the plurality of biosignal generators, which changes according to the periodic movement of the living body, is detected, and the degree of reliability of the detected cycle is determined. Then, from the plurality of cycles obtained from the plurality of biosignals generated by the plurality of biosignal generation units, based on the degree of reliability obtained for each of the plurality of cycles, the most probable cycle is the biodynamic cycle. is determined as

In the biological signal processing apparatus according to the present invention, each of the plurality of biological signal generators outputs a detection signal according to the movement of the living body obtained through one of the plurality of different parts of the living body. and a biosignal extraction unit for extracting, as the biosignal, a signal containing a signal component that changes according to the periodic motion of the living body from the detection signal output from the detector. .

With such a configuration, in the plurality of biomedical signal generation units, signals containing signal components that change according to the periodic motion of the biomedical body are generated from detection signals from different detectors corresponding to a plurality of parts of the biomedical body. is extracted as Then, by using different detectors, the cycles of the signal components that change according to the periodic motion of each of the plurality of biological signals generated through different processes are detected, and the detected A degree of confidence in the period is determined.

In the biological signal processing apparatus according to the present invention, the plurality of biological signal generators have a common detector that outputs a detection signal corresponding to the movement of the living body obtained through a predetermined part of the living body, and the plurality of each of the biological signal generators includes a signal converter that converts a detection signal output from the detector according to one conversion characteristic among a plurality of conversion characteristics assigned to the plurality of biological signal generators; and a biological signal extraction unit for extracting, as the biological signal, a signal containing a signal component that changes according to the periodic motion of the living body from the signal obtained by the signal conversion unit.

With such a configuration, in the plurality of biological signal generation units, the detection signal from the common detector is converted according to the plurality of conversion characteristics, and each of the plurality of converted signals changes according to the periodic movement of the living body. A signal containing a signal component to be extracted as a biosignal. Then, the cycles of the signal components that change according to the periodic movement of the living body are detected for each of the plurality of biological signals that have been generated through different processes by being converted with different conversion characteristics, and the A degree of confidence in the detected period is determined.

In the biosignal processing apparatus according to the present invention, the biosignal extraction unit includes a heartbeat signal extraction unit that extracts, as the biosignal, a heartbeat signal containing a signal component that changes according to the heartbeat of the living body. be able to.

With such a configuration, each of the plurality of biosignal generators extracts, as a biosignal, a heartbeat signal containing a signal component that changes according to the heartbeat of the living body. Then, the period of the signal component that changes according to the heartbeat of the heartbeat signal is detected, and the degree of reliability of the detected period is determined.

In the biomedical signal processing apparatus according to the present invention, the biomedical signal extractor includes a respiratory signal extractor that extracts, as the biomedical signal, a respiratory signal containing a signal component that changes according to the respiratory motion of the living body. be able to.

With such a configuration, each of the plurality of biosignal generators extracts, as a biosignal, a respiratory signal containing a signal component that changes according to the respiratory motion of the living body. Then, the period of the signal component that changes according to the respiratory motion of the respiratory signal is detected, and the degree of reliability of the detected period is determined.

In the biomedical signal processing apparatus according to the present invention, the period detection unit includes a signal extraction unit that extracts a signal portion for a predetermined time from the heartbeat signal obtained by the heartbeat signal extraction unit and holds it as an extraction waveform portion; The clipped waveform portion obtained by the clipping unit is moved by a certain time width, and each time the clipped waveform portion is moved, the clipped waveform portion and the waveform portion of the heartbeat signal are compared to obtain a degree of similarity representing the degree of matching between them. a degree of similarity calculation unit that calculates the heart rate based on the timing at which the degree of matching between the clipped waveform portion represented by the similarity obtained by the similarity calculation unit and the waveform portion of the heartbeat signal peaks and a heartbeat cycle determination unit that determines the cycle of the signal.

With such a configuration, a signal portion for a predetermined time is cut out as a cutout signal waveform portion from the heartbeat signal obtained by each of the plurality of biosignal generation units, and each time the cutout waveform portion is moved by a predetermined time width, The clipped waveform portion and the waveform portion of the heartbeat signal are compared to calculate a degree of similarity representing the degree of matching between them. Then, the period of the heartbeat signal is determined based on the timing at which the degree of matching between the clipped waveform portion represented by the degree of similarity and the waveform portion of the heartbeat signal peaks.

In the biomedical signal processing device according to the present invention, the reliability determination unit is determined by the heartbeat cycle determination unit based on the similarity used when the heartbeat cycle determination unit determines the cycle of the heartbeat signal. determining a degree of confidence in the period of the heartbeat signal obtained.

With such a configuration, the reliability of the period determined for the heartbeat signal based on the similarity used when determining the period of the heartbeat signal, that is, the similarity of the peak obtained in the process of moving the cutout waveform portion. is determined.

In the biomedical signal processing apparatus according to the present invention, the signal clipping section can be configured to shift the clipping timing of the clipped waveform portion by a time shorter than the time width of the clipped waveform portion.

With such a configuration, the timing for cutting out the cutout waveform portion from the heartbeat signal is shifted by a shorter time than the time portion of the cutout waveform portion. Then, each time a cutout waveform portion is cut out from the heartbeat signal, the cutout waveform portion is moved to determine the period of the heartbeat signal.

In the biological signal processing apparatus according to the present invention, the period detection unit includes a peak detection unit for detecting a signal value as a peak value of the respiratory signal obtained by the respiratory signal extraction unit, and a peak value as a bottom value of the respiratory signal obtained by the respiratory signal extraction unit. a bottom detection unit for detecting a signal value; and when the signal value detected as the peak value by the peak detection unit is maintained as the peak value for a predetermined time, the signal value is determined as the peak value, and the bottom detection is performed. a control unit for determining the signal value as the bottom value when the signal value detected as the bottom value by the unit is maintained as the bottom value for a predetermined time; and the peak value and the bottom value determined by the control unit. a respiratory cycle determination unit configured to determine the cycle of the respiratory signal based on the detection timing of at least one of the values.

With such a configuration, when a certain signal value is detected as a peak value in the respiratory signals obtained by each of the plurality of biological signal generators, there is no signal value exceeding that signal value, and the signal value reaches the peak. When the value is maintained for a predetermined period of time, the signal value is determined as the peak value. Further, in the respiratory signal, when a certain signal value is detected as the bottom value, there is no signal value lower than that signal value, and when the signal value is maintained as the bottom value for a predetermined period of time, the signal value becomes the bottom value. is determined as Then, the cycle of the respiratory signal is determined based on the determined detection timing of at least one of the peak value and the bottom value.

In the biomedical signal processing apparatus according to the present invention, the reliability determination unit determines the respiration signal determined by the respiratory signal determination unit based on the difference between the peak value and the bottom value determined by the control unit. Arrangements can be made to determine a degree of confidence in the period of the respiratory signal.

With such a configuration, if a pseudo peak not based on the respiratory motion of the living body occurs in the respiratory signal, the difference between the value of the pseudo peak and the bottom value of the respiratory signal is the normal peak value and the bottom value. values, so that the reliability of the period determined using the spurious peak value is different from the reliability of the period determined using the normal peak value, e.g. can do.

According to the present invention, from a plurality of cycles obtained from a plurality of biological signals generated through different processes, based on the degree of reliability obtained for each of the plurality of cycles, the most probable cycle is the biological movement. Since it is determined as a period, it is possible to obtain the periodic period of living body motion with high reliability without taking much processing time.

FIG. 1 is a diagram showing an arrangement example of a plurality of sensor pads on a bed. FIG. 2 is a diagram showing the positional relationship between a plurality of sensor pads on the bed and a living body (human) on the bed. FIG. 3 is a block diagram showing the configuration of the biological signal processing device according to the first embodiment of the present invention. 4 is a block diagram showing a functional configuration of a selector in the biological signal processing apparatus shown in FIG. 3. FIG. FIG. 5 is a diagram for explaining the extraction of the cutout waveform portion g(x) from the heartbeat signal f(x) and the movement (shift) of the cutout waveform portion g(x) in the SSF processing section. FIG. 6 is a diagram showing another example of extraction of the waveform portion g(x) extracted from the heartbeat signal f(x) in the SSF processing section. FIG. 7 is a diagram for explaining the respiratory signal processed by the PBD processing unit and the procedure for determining the period of the respiratory signal. It is a block diagram showing a biological signal processing device according to a second embodiment of the present invention.

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings.

In the biological signal processing device according to the first embodiment of the present invention, a plurality of (for example, four) sensor pads arranged as shown in FIGS. 1 and 2 are used.

1 and 2, the bed 100 is provided with four sensor pads 10a, 10b, 10c and 10d. Each of these sensor pads 10a-10d is configured similarly to the "sensor pads" described in Japanese Patent No. 44423481, for example, and includes piezoelectric elements 11a-11d (detectors). The four sensor pads 10a to 10d are placed at certain intervals, for example, on the shoulder region, the chest region, the abdomen region, and the waist region of the person H (living body) lying on his back on the bed 100. are arranged so as to correspond to each other. Each of the piezoelectric elements 11a-11d outputs a detection signal whose amplitude changes according to vibration transmitted from the person H to the corresponding sensor pads 10a-10d. In the description of the first embodiment, the reference number 10 is used for the sensor pad as a generic term, and the reference number 11 is used for the piezoelectric element as a generic term.

A biological signal processing apparatus according to a first embodiment of the present invention is configured as shown in FIG. This biological signal processing device has four processing channels CH1 to CH4 that perform different processes by using the four piezoelectric elements 11a (sensor pads 10a) to 11d (sensor pads 10d) described above. there is The four processing channels CH1 to CH4 have the same configuration, and each includes a preamplifier 12, a high-pass filter (HPF) 13, a low-pass filter (LPF) 14, an SSF (Slide Shift Fitting) processing section 15, and a PBD (Peak Bottom Detection) processing. It has a portion 16 .

The piezoelectric element 11 of the sensor pad 10 outputs a detection signal according to the movement (vibration) of the person H (living body) on the bed 100 transmitted to the piezoelectric element 11 through the sensor pad 10 . A detection signal output from the piezoelectric element 11 undergoes impedance conversion in the preamplifier 12 and is provided in parallel to the high-pass filter 13 and the low-pass filter 14 . The high-pass filter 13 extracts a signal of relatively high frequency components from the detection signal as a heartbeat signal containing a signal component that changes according to the heartbeat of the person H (heartbeat signal extractor: biological signal extractor). The low-pass filter 14 extracts a signal of relatively low frequency components from the detection signal as a respiratory signal containing a signal component that changes according to the respiratory motion of the person H (respiratory signal extractor: biological signal extractor). Here, in each of the processing channels CH1 to CH4, the piezoelectric element 11, the preamplifier 12, and the high-pass filter 13 constitute a biological signal generation section for generating a heartbeat signal, which is a biological signal. In each of the channels CH1 to CH4, the piezoelectric element 11, the preamplifier 12 and the low-pass filter 14 constitute a biological signal generation section for generating a respiratory signal, which is a biological signal.

The SSF processing unit 15 performs processing (period detection unit) for detecting the heartbeat period based on the heartbeat signal provided from the high-pass filter 13 (heartbeat signal extraction unit), and expresses the degree of reliability of the detected period. and a process of generating reliability (reliability determination unit). Then, the SFF processing unit 15 outputs information on the cardiac cycle Hpn (n=1, 2, 3, 4) and information on the reliability Hrn (n=1, 2, 3, 4) obtained as a result of the processing. . The PBD processing unit 16 performs processing (period detection unit) for detecting the period of respiratory motion based on the respiratory signal provided from the low-pass filter 14 (respiratory signal extraction unit), and the degree of reliability of the detected period. and a process (reliability determination unit) for generating a reliability value to represent. Then, the PBD processing unit 16 outputs information on the respiratory cycle Bpn (n=1, 2, 3, 4) and information on the reliability Brn (n=1, 2, 3, 4) obtained as a result of the processing. .

The biological signal processing device further has a biological movement cycle determination unit 20 . For example, the body movement period determination unit 20 can be configured as shown in FIG. In FIG. 4 , the living body dynamic cycle determination unit 20 has a heartbeat cycle selection unit 21 , a heartbeat cycle reliability comparison unit 22 , a respiratory cycle selection unit 23 and a respiratory cycle reliability comparison unit 24 . The heartbeat cycle reliability comparison unit 22 compares the heartbeat cycle reliability Hr1 to Hr4 output from the four processing channels CH1 to CH4 (SSF processing unit 15) described above, and determines the heartbeat corresponding to the highest reliability Hri. A selection instruction for the period Hpi is output. The heartbeat cycle selection unit 21 selects the heartbeat indicated by the selection instruction from the heartbeat cycle reliability comparison unit 22 among the heartbeat cycles Hp1 to Hp4 output from the four processing channels CH1 to CH4 (SSF processing unit 15) described above. The period Hpi is output as the determined heartbeat period Hp. The respiratory cycle reliability comparison unit 24 compares the reliability levels Br1 to Br4 of the respiratory cycle output from the four processing channels CH1 to CH4 (PBD processing unit 16) described above, and determines the breathing corresponding to the highest reliability Bri. A selection instruction for the period Bpi is output. The breathing cycle selection unit 23 selects the breathing specified by the selection instruction from the breathing cycle reliability comparison unit 24 among the breathing cycles Bp1 to Bp4 output from the four processing channels CH1 to CH4 (PBD processing unit 16) described above. The period Bpi is output as the determined respiratory period Bp.

In this way, the biological movement cycle determining unit 20 determines (selects) the most probable heartbeat cycle Hpi based on the reliability levels Hr1 to Hr4 from the four heartbeat cycles Hp1 to Hp4 output from the four processing channels CH1 to CH4. ), and outputs the heartbeat period Hpi as the determined heartbeat period Hp. Further, the biological movement cycle determination unit 20 determines (selects) the most probable respiratory cycle Bpi based on the reliability levels Br1 to Br4 from the four respiratory cycles Bp1 to Bp4 output from the four processing channels CH1 to CH4. , the respiratory cycle Bpi is output as the determined respiratory cycle Bp.

Specific processing of the SSF processing unit 15 will be described with reference to FIG.

A signal portion for a predetermined time (for example, 6 seconds) is cut out from the heartbeat signal f(x) provided from the high-pass filter 13, and the signal portion is held in a memory as a cut-out waveform portion g(x) ( signal extractor). The clipped waveform portion g(x) is moved (shifted) by a certain time width t, and each time the clipped waveform portion g(x) and the waveform portion f(x+t) of the heartbeat signal f(x) are shifted. They are compared and a degree of similarity representing the degree of matching between them is calculated (similarity calculation unit). This similarity d is, for example,
d=∫|f(x+t)−g(x)|dx (1)
is calculated as an accumulated value (digital countable value) of the difference between the clipped waveform portion g(x) and the waveform portion f(x+t) of the heartbeat signal (absolute value method). When t=0, the similarity d is such that the difference between the heartbeat signal f(x) and the clipped waveform portion g(x) is zero when they overlap each other. This degree of similarity d=0 represents a state in which the clipped waveform portion g(x) and the waveform portion f(x+0) of the heartbeat signal are most similar.

Further, the SSF processing unit 15 calculates the cutout waveform portion g(x) represented by the degree of similarity d (see formula (1) above) obtained while sequentially shifting the cutout waveform portion g(x) and the waveform portion of the heartbeat signal. The period of the heartbeat signal f(x) is determined based on the timing at which the degree of matching with f(x+t) reaches a peak (heartbeat signal determining unit). For example, the clipped waveform portion g(x) is sequentially shifted from the timing (similarity d=0: the degree of matching is peak) at which the waveform portion g(x) for a predetermined time is clipped from the heartbeat signal f(x). Then, the period of the heartbeat signal f(x) is determined based on the time until the timing at which the degree of matching represented by the degree of similarity d peaks. This means that when the heartbeat signal f(x) is, for example, an ideal waveform (eg, a sine waveform), the waveform shifted by one cycle (integer multiple cycle) is the original waveform (eg, a sine waveform). It is based on the fact that it overlaps with

The SSF processing unit 15 outputs the cycle thus determined as the heartbeat cycle Hpi. Then, the SSF processing unit 15 determines the reliability Hri representing the degree of reliability of the heartbeat period Hpi based on the degree of similarity d (the degree of coincidence is the peak) used when determining the heartbeat period Hpi. The reliability Hri is output together with the heartbeat period Hpi. The higher the degree of matching represented by the degree of similarity d, the higher the degree of reliability Hri determined.

For example, if a waveform portion obtained by sampling a heartbeat signal at 800 points at a cycle of 125 Hz is taken as a cutout waveform portion, the cutout waveform portion is a waveform portion of the heartbeat signal for about 6 seconds. In this case, there may be approximately 8 pulses (beats) in the segmented waveform portion. When the heartbeat signal is cut out every 6 seconds, if a mistake occurs in processing in a certain cut-out block, the measurement values for about 6 seconds are missing, which hinders continuous accurate cycle detection.

Therefore, in the SFF processing unit 15, for example, as shown in FIG. It can be shifted by a time length Δx shorter than the time length of the waveform portion g(x). In this way, even if a mistake occurs in the processing in the clipping block, the missing measurement value is only the short time length Δx, and continuous and more accurate period detection is possible.

According to the biological signal processing device as described above, a plurality of piezoelectric elements 11a to 11d (sensor pads 10a to 11d) are generated based on detection signals from the plurality of piezoelectric elements 11a to 11d (sensor pads 10a to 11d) installed corresponding to different parts of the person H. The most probable heartbeat period (determined heartbeat period Hp) is determined from the heartbeat signal Hpn based on the reliability Hrn obtained for each of the plurality of heartbeat signals Hpn. heartbeat signal period can be obtained with high reliability.

In the SSF processing unit 15 described above, the degree of similarity d representing the degree of matching between the clipped waveform portion g(x) and the waveform portion f(x+t) of the heartbeat signal is calculated according to Equation (1) (absolute value method), but is not limited to this. for example,
d=∫(f(x+t)*g(x))dx (2)
Even if the similarity d is calculated according to (correlation method),
d=∫(f(x+t)−g(x)) ² dx (3)
The similarity d may be calculated according to (squaring method).
In addition, similarity ( reliability), or the similarity d (reliability) can be calculated according to another method.

Next, specific processing of the PBD processing unit 16 will be described with reference to FIG.

In the PBD processing unit 16, under the control of a control unit (not shown), the signal value of the respiratory signal X(t) supplied from the low-pass filter 14 is sampled and stored in a peak hold circuit (for example, memory or register). (not shown)). While the signal value of the respiratory signal X(t) is increasing, the signal value held in the peak hold circuit is updated sequentially. Then, as shown in FIG. 7, at time t1, when the peak hold circuit holds the signal value P1 that is the peak of the respiratory signal X(t) (peak detector), the sampled signal value thereafter becomes the signal value Since it is smaller than the value P1, the state in which the signal value P1 is held in the peak hold circuit is maintained. When a predetermined time Δt elapses without the signal value P1 held in the peak hold circuit being updated, this signal value P1 is used as the peak value P1 of the respiratory signal X(t) under the control of the control section. Confirmed.

After that, in the PBD processing unit 16, switching from the peak hold circuit to the bottom hold circuit (for example, composed of a memory or a register; not shown) is performed, and the respiratory signal X(t) supplied from the low-pass filter 14 is Signal values are sampled and held sequentially in bottom hold circuits. While the signal value of the respiratory signal X(t) is decreasing, the signal value held in the bottom hold circuit is updated sequentially. Then, as shown in FIG. 7, when the bottom hold circuit holds the signal value B1 (bottom detector) at time t2, the signal value sampled thereafter is greater than the signal value B1. The state in which the signal value B1 is held in the hold circuit is maintained. When a predetermined time Δt elapses without the signal value B1 held in the bottom hold circuit being updated, this signal value B1 is used as the bottom value B1 of the respiratory signal X(t) under the control of the control section. Confirmed.

After that, in the PBD processing circuit 16, the bottom hold circuit is switched to the peak hold circuit, and the processing related to the peak hold described above is performed. Then, as shown in FIG. 7, at time t3, the signal value P2 of the respiratory signal X(t) is held as the peak value in the peak hold circuit. In a situation where the respiratory signal X(t) decreases and the signal value reaches the bottom value B2 and further increases before the predetermined time Δt elapses after the signal value P2 is held, the signal value P2 is held by the peak hold circuit. After a lapse of a predetermined time Δt from the time t3 held at , the signal value of the respiratory signal X(t) becomes larger than the signal value P2 held as the peak value in the peak hold circuit. In this case, the signal value P2 held in the peak hold circuit is not determined as the peak value. Then, without switching from the peak hold circuit to the bottom hold circuit, the signal values held in the peak hold circuit are successively updated. In this situation, the control unit does not determine the signal value P2, which is the peak of the respiratory signal X(t), as the peak value, and the signal value B2, which is the bottom of the respiratory signal X(t) at time t4, not determined as

In this way, in the process of updating the signal value held in the peak hold circuit, when the peak hold circuit holds the signal value P3 that becomes the peak of the respiratory signal X(t) at time t5, the peak The state in which the signal value P3 is held in the hold circuit is maintained. When a predetermined time Δt elapses without the signal value P3 held in the peak hold circuit being updated, this signal value P3 is used as the peak value P3 of the respiratory signal X(t) under the control of the control section. Confirmed.

In the PBD processing section 16, under the control of the control section, as described above, each process of peak value detection, peak value determination, bottom value detection, and bottom value determination is repeatedly executed. Then, the period of the respiratory signal X(t) is calculated based on the determined time interval between the two peak values (the determined time interval between the two peak values itself is the period of the respiratory signal X(t)). (breathing cycle determination unit). It is also possible to calculate the period of the respiratory signal X(t) based on the determined time interval between the peak value and the bottom value. It is also possible to calculate the period of the signal X(t).

Furthermore, the PBD processing unit 16 calculates the period of the respiratory signal X(t) as described above (period detection unit), and determines the degree of reliability representing the degree of reliability of the calculated period (reliability determination unit). part). For example, the reliability is determined based on the difference between the determined peak value and the determined bottom value included in the period. It can be said that the greater the difference between the determined peak value and the determined bottom value, the higher the reliability. In this way, even if the signal value of a false peak such as the peak value P2 (bottom value B2) in FIG. The reliability of the period determined based on this may be small.

Note that when the false peak becomes a large signal value, the accuracy of the cycle is lowered. To prevent this, the characteristics of the low-pass filter 14, eg the cutoff frequency of the respiratory signal, can be set as low as possible.

The PBD processing unit 16 uses the cycle determined as described above as the respiratory cycle Bpi, and the reliability calculated based on the determined peak value and the determined bottom value used in determining the cycle as described above. is output as the respiratory reliability Bri.

Further, according to the biomedical signal processing apparatus as described above, a plurality of piezoelectric elements 11a to 11d (sensor pads 10a to 11d) installed corresponding to different parts of the person H generate detection signals based on the Since the most probable respiratory cycle (determined respiratory cycle Hp) is determined from the plurality of respiratory signals Bpn based on the reliability Brn obtained for each of the plurality of respiratory signals Bpn, it does not take much processing time. , the period of the periodic respiratory signal can be obtained with high reliability.

In the biological signal processing apparatus described above, the preamplifier 12, the high-pass filter 13, and the low-pass filter 14 can be configured by hardware using various electronic components. Each of the SSF processing unit 15 and the PBD processing unit 16 includes hardware such as an A/D converter, memory, and processing unit (including CPU), and realizes various processes by software executed by the processing unit. .

Although the above-described biosignal processing apparatus extracts both the heartbeat signal and the respiration signal as biosignals, it is not limited to this, and either one (the heartbeat signal or the respiration signal) is extracted and the heartbeat signal is extracted. It may detect the period of either the signal or the respiratory signal.

Further, in the above-described biomedical signal processing device, four biosignals (heartbeat signal and respiratory signal, respectively) are generated from four different systems (CH1 to CH4). Biosignals are generated, and the most probable period can be determined as the biodynamic period from more cycles obtained from more biosignals. For example, a plurality of high-pass filters with different characteristics may be used, and a plurality of low-pass filters with different characteristics may be used to extract more heartbeat signals and respiratory signals. Further, the SSF processing unit 15 may extract signals of a plurality of sizes from the heartbeat signal and process each of the plurality of extracted signal waveforms (detection of period, calculation of reliability).

Next, a biological signal processing device according to a second embodiment of the invention will be described. This biological signal processing apparatus uses a single piezoelectric element (detector) and has a plurality of processing systems that process the piezoelectric element through different processes, unlike the first embodiment described above. It is different from the biological signal processing device concerned.

A biological signal processing apparatus according to a second embodiment of the present invention is configured as shown in FIG.

A single sensor pad 10 is provided on the bed 100 so as to correspond to, for example, a region around the chest of the person H lying on his back. A piezoelectric element 11 is included in the sensor pad 10 , and the piezoelectric element 11 outputs a detection signal whose amplitude changes according to vibration transmitted from the person H on the bed 100 to the sensor pad 10 . In FIG. 8, this biological signal processing apparatus includes a first amplifier 17 (HGA: signal converter) having an amplification characteristic (conversion characteristic) of a first amplification factor and a second amplification factor lower than the first amplification factor. It has a second amplifier 18 (LGA: signal converter) having amplification characteristics (conversion characteristics). In addition, this biological signal processing apparatus has two processing systems that generate biological signals (heartbeat signal, respiratory signal) through different processes: a processing system including the first amplifier 17 and a processing system including the second amplifier 18. . Each of the processing system (one processing system) including the first amplifier 17 and the processing system (the other processing system) including the second amplifier 18 is for each channel CH1 to CH1 in the biological signal processing apparatus according to the first embodiment. Like CH4 (see FIG. 3), it has a high-pass filter 13, a low-pass filter 14, an SSF processing section 15 and a PBD processing section 16. FIG. Further, this biomedical signal processing device has a biomedical movement period determining unit 20, like the channels CH1 to CH4 in the biomedical signal processing device according to the first embodiment (see FIG. 3).

In one processing system, the detection signal amplified (including impedance conversion) through the first amplifier 17 is passed through the high-pass filter 13 (biological signal extraction unit) to produce a signal that changes according to the heartbeat of the person H. A heartbeat signal containing components is extracted from the detected signal. Further, the detection signal amplified through the first amplifier 17 is passed through the low-pass filter 14 (biological signal extraction unit), whereby a respiratory signal containing a signal component that changes according to the respiratory motion of the person H is obtained from the detection signal. extracted. In the other processing system, similarly, the heartbeat signal and the respiration signal are extracted from the detection signal amplified (including impedance conversion) through the second amplifier 18 .

In each processing system, similar to channels CH1 to CH4 in the biological signal processing apparatus according to the first embodiment, the SSF processing unit 15 determines the heartbeat period based on the heartbeat signal provided from the high-pass filter 15. are detected (see FIGS. 5 and 6), and a process of generating reliability indicating the degree of reliability of the detected period is performed. Then, the SFF processing unit 15 outputs information on the cardiac cycle Hpn (n=1, 2) and information on the reliability Hrn (n=1, 2) obtained as a result of the processing. As in the first embodiment, the PBD processing unit 16 also performs processing (see FIG. 7) for detecting the period of respiratory motion based on the respiratory signal provided from the low-pass filter 16 and the detected period. and a process of generating a reliability value representing the degree of reliability of the Then, the PBD processing unit 16 outputs information on the respiratory cycle Bpn (n=1, 2) and information on the reliability Brn (n=1, 2) obtained as a result of the processing.

The biological movement cycle determination unit 20 is configured in the same manner as in the first embodiment (see FIG. 4), and from the two heartbeat cycles Hp1 and Hp2 output from the two processing systems (SSF processing unit 15), The most probable heartbeat period Hpi is determined (selected) based on the reliability levels Hr1 and Hr2, and the heartbeat period Hpi is output as the determined heartbeat period Hp. Further, the biological movement cycle determination unit 20 determines the most probable respiratory cycle Bpi based on the reliability levels Br1 and Br2 from the two respiratory cycles Bp1 and Bp2 output from the two processing systems (PBD processing unit 16) ( selected) and outputs the respiratory cycle Bpi as the determined respiratory cycle Bp.

In the biological signal processing apparatus according to the second embodiment of the present invention, as described above, the detection signal from the single piezoelectric element 11 has two different amplification characteristics (conversion characteristics), the first amplifier 17 and the second amplifier. 2 amplifier 18, and from each of the two amplified signals, a signal containing a signal component that changes according to heartbeat is extracted as a heartbeat signal (biological signal), which also changes according to respiratory motion. A signal containing a signal component is extracted as a respiratory signal (biological signal). The cycle of each of the plurality of biological signals (heartbeat signal, respiratory signal) generated through different processes by being amplified with different amplification characteristics is detected, and the cycle of the detected cycle is detected. A degree of trust is determined. Then, the period with the highest degree of reliability is determined as the period (determined heartbeat period Hp, determined respiratory period Bp) of the biological signal (heartbeat signal, respiratory signal).

According to such a biological signal processing unit, from a plurality of cycles obtained from a plurality of biological signals (heartbeat signal, respiratory signal) generated through different processes by being amplified with different amplification characteristics, the plurality of cycles Based on the degree of reliability obtained for each of the above, the most probable period is determined as the biological movement period (determined heartbeat period Hp, determined respiratory period Bp). It becomes possible to obtain the period of a certain biological motion (heartbeat, respiratory motion) with high reliability.

The amplification characteristics (amplification factor) of each of the first amplifier 17 and the second amplifier 18 can be appropriately set according to the attributes (age, sex, height, weight, etc.) of the target person (H). Further, it is also possible to provide more processing systems to determine the most probable period from more biological signal (heartbeat signal, respiratory signal) periods obtained from a single detector (piezoelectric element 11).

Further, the conversion characteristics for converting the detection signal from the detector applied in each of the plurality of processing systems are not limited to the amplification characteristics (amplification factor) as described above. It may be other transformation characteristics, such as a characteristic or a transformation characteristic that may be synthesized from multiple transformation characteristics.

Although the embodiments of the present invention have been described above, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments described above can be implemented in various other forms, and various omissions, replacements, and modifications can be made without departing from the scope of the invention. These embodiments are included in the scope and gist of the invention, and are included in the invention described in the claims.

INDUSTRIAL APPLICABILITY The biological signal processing apparatus according to the present invention has the effect of being able to obtain the period of periodic biological motion with high reliability without taking much processing time. It is useful as a biological signal processing device for processing a biological signal that changes according to the movement of the living body including the periodic movement of the body.

10a, 10b, 10c, 10d sensor pad 11a, 11b, 11c, 11d piezoelectric element 12 preamplifier 13 high-pass filter (heartbeat signal extractor)
14 low-pass filter (respiratory signal extractor)
15 SFF processing unit 16 PBD processing unit 20 Life movement cycle determination unit 21 Heartbeat cycle selection unit 22 Heartbeat cycle reliability comparison unit 23 Respiration cycle selection unit 24 Respiration cycle reliability comparison unit

In the biomedical signal processing apparatus according to the present invention, the reliability determination unit determines the respiration cycle determination unit based on the difference between the peak value and the bottom value determined by the control unit. Arrangements can be made to determine a degree of confidence in the period of the respiratory signal.

Claims (10)

a plurality of biosignal generators that generate biosignals that change according to the motion of the living body including periodic motion (heartbeat, respiratory motion) through different processes;
a cycle detection unit that detects a cycle of a signal component in the biosignal generated by each of the plurality of biosignal generation units that changes according to the periodic motion of the living body;
a reliability determination unit that determines the degree of reliability of the period detected by the period detection unit;
A most probable period is determined as a biological movement period from among the plurality of periods each obtained by the period detection unit, based on the degree of reliability obtained by the reliability determination unit for each of the plurality of periods. a biological motion cycle determination unit; and a biological signal processing device. each of the plurality of biosignal generators is a detector that outputs a detection signal according to the movement of the living body obtained through one of the plurality of different parts of the living body;
2. The biological signal processing apparatus according to claim 1, further comprising a biological signal extraction unit for extracting, as the biological signal, a signal containing a signal component that changes according to the periodic motion of the living body from the detection signal output from the detector. . the plurality of biosignal generators have a common detector that outputs a detection signal according to the movement of the living body obtained through a predetermined part of the living body;
Each of the plurality of biosignal generators,
a signal converter that converts a detection signal output from the detector according to one of a plurality of conversion characteristics assigned to the plurality of biosignal generators;
2. The biological signal processing apparatus according to claim 1, further comprising a biological signal extracting section for extracting, as the biological signal, a signal containing a signal component that changes according to the periodic motion of the living body from the signal obtained by the signal converting section. . 4. The biological signal processing apparatus according to claim 2, wherein said biological signal extracting section includes a heartbeat signal extracting section for extracting, as said biological signal, a heartbeat signal containing a signal component that changes according to the heartbeat of said living body. 4. The biomedical signal processing apparatus according to claim 2, wherein said biomedical signal extractor includes a respiratory signal extractor that extracts, as said biomedical signal, a respiratory signal including a signal component that changes according to respiratory motion of said living body. The period detection unit is
a signal extraction unit that extracts a signal portion for a predetermined time from the heartbeat signal obtained by the heartbeat signal extraction unit and holds it as an extracted waveform portion;
A similarity that expresses the degree of coincidence between the clipped waveform portion obtained by the signal clipping unit by moving the clipped waveform portion obtained by the signal clipping unit by a certain time width, and comparing the clipped waveform portion and the waveform portion of the heartbeat signal for each movement. a similarity calculator that calculates the degree of
A heartbeat period for determining a period of the heartbeat signal based on a timing at which a degree of matching between the clipped waveform portion represented by the degree of similarity obtained by the similarity degree calculation unit and the waveform portion of the heartbeat signal reaches a peak. 5. The biological signal processing device according to claim 4, further comprising a determination unit. The reliability determination unit determines the reliability of the period of the heartbeat signal determined by the heartbeat period determination unit based on the degree of similarity used when the heartbeat period determination unit determines the period of the heartbeat signal. 7. The biological signal processing device according to claim 6, which determines the degree. 8. The biomedical signal processing apparatus according to claim 6, wherein said signal clipping section shifts the clipping timing of said clipped waveform portion by a time shorter than the time width of said clipped waveform portion. The period detection unit is
a peak detection unit for detecting a signal value as a peak value of the respiratory signal obtained by the respiratory signal extraction unit;
a bottom detection unit that detects a signal value as a bottom value of the respiratory signal;
When the signal value detected as the peak value by the peak detection unit is maintained as the peak value for a predetermined time, the signal value is determined as the peak value, and the signal detected as the bottom value by the bottom detection unit a control unit that determines the signal value as the bottom value when the value is maintained as the bottom value for a predetermined time;
6. The biological signal processing apparatus according to claim 5, further comprising a respiratory cycle determination unit configured to determine the cycle of the respiratory signal based on detection timing of at least one of the peak value and the bottom value determined by the control unit. . The reliability determination unit determines the degree of reliability of the period of the respiratory signal determined by the respiratory signal determination unit based on the difference between the peak value and the bottom value respectively determined by the control unit. 10. The biological signal processing device according to claim 9, wherein

Images