JP2005198781A - Respiration monitoring apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a respiration monitoring apparatus which can monitor (observe) a state of respiration of a sleeping person with high precision. <P>SOLUTION: Peak frequencies within 256 cycles of signals acquired from individual sensors are specified for the individual sensors (S410). The specified individual peak frequencies are classified in prescribed individual frequency bands and a number of the sensors to output the peak frequencies is counted for each the frequency band (S420). The individual sensors constituting the frequency band to which the largest number of the sensors belong are selected as a sensors group SGm (S430). These individual sensors are sorted into phase groups having a phase width of π/5 on the basis of output signals from the individual sensors constituting the sensors group SGm (S440). The output signal from the sensor which belongs to the phase group having the largest number of the sensors and an inverted signal of the output signal from the sensor which belongs to a group with a phase deviation of π from the above phase group are added and averaged to calculate a respiration signal (S450-S470). <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

 The present invention relates to a respiratory monitoring device that can monitor (monitor) a respiratory state of a sleeping person.

  As a conventional monitor device for monitoring a respiratory state during sleep, there is a respiratory disease monitor device disclosed in Patent Document 1. This monitor device is installed under a bedding with a predetermined distribution, and includes a plurality of load sensors that output a load signal corresponding to an applied load, and a frequency band corresponding to a sleeper's respiration rate from the load signal of the load sensor. The signal is calculated as a respiratory signal.

When the monitor device calculates a respiratory signal, first, the load signal of the load sensor is subjected to frequency analysis to calculate a power spectrum in a frequency band corresponding to the respiratory rate. Next, a load sensor having a large power spectrum intensity is used as a reference sensor, and a load sensor that outputs at least one load signal substantially in phase with or substantially opposite in phase to the reference sensor is selected. Then, the load signal output by the selected load sensor is added as it is to the load signal output by the reference sensor, and the load signal output by the selected load sensor is the load whose phase is inverted. The respiration signal is calculated by adding the signals.
JP 2001-37742 A

  However, in the above-described conventional monitor device, when a load due to a factor other than breathing, for example, a sleeper's tremor, is applied to the load sensor, a load applied by a large motion compared to breathing is applied. There is a problem that the sensor is selected as the reference sensor and the accuracy of the respiration signal is lowered.

  The present invention has been made in view of such a problem, and an object of the present invention is to provide a respiratory monitor device capable of monitoring (monitoring) a sleeper's respiratory state with higher accuracy.

  The respiratory monitor device according to claim 1, wherein the respiratory monitor device according to claim 1 is arranged in a predetermined distribution under a sleeping person and outputs a signal corresponding to a load or vibration from the sleeping person. And a respiratory signal calculation means. The respiration signal calculating means converts a signal output from the sensor into a frequency domain, selects a sensor based on the characteristics of the converted signal, and calculates a respiration signal based on the signal output from the selected sensor. As for the sensor arrangement, the “predetermined distribution” means, for example, an evenly spaced distribution, an unevenly spaced distribution corresponding to the appearance of a sleeper's breathing motion, or the like.

  The inventors of the present application focused on the fact that the body movement of a sleeping person accompanying breathing can be confirmed at many places on the body. In a respiratory monitor device having a plurality of sensors, attention was paid to the fact that there are various characteristics in the output signal of a sensor that captures the body motion of a sleeping person caused by respiration. For example, there are periodicity and correlation with the front and rear waveforms. Therefore, in order to make it easy to capture this feature, it was considered to convert the output signal to the frequency domain and select a sensor based on the feature of the converted signal.

  If the sensor is selected and the respiration signal is calculated in this way, most of the sensors that output signals including factors other than respiration can be excluded to generate the respiration signal. As a result, the accuracy of the respiration signal can be improved as compared with the conventional respiration monitor device.

  By the way, the specific characteristics when selecting the sensor may be such that the sensor is selected based on the characteristics described in claim 2, for example. That is, the peak frequency of the spectrum of the signal output from the sensor is specified for each sensor, the frequency band in which the specified peak frequency exists most is determined, and the signal having the peak frequency included in the determined frequency band is determined. A sensor to be output may be selected. The term “peak frequency” as used herein means, for example, the frequency having the largest amplitude value when an amplitude spectrum is used as a spectrum, and the frequency having the largest power value when a power spectrum is used as a spectrum. means. Further, the frequency having the second and third largest amplitude values and power values may be regarded as the “peak frequency” even if they are not the largest amplitude value and the largest power value. The frequency band to be determined is preferably a frequency band having a width of about 0.03 Hz.

  The inventors of the present application focused on the fact that the sleeper's body movement accompanying breathing can be confirmed in many parts of the body as described above. In a respiratory monitor device having a plurality of sensors, the number of sensors that output a peak frequency corresponding to a factor other than breathing (for example, a fine movement of a sleeping person) is the number of sensors that output a peak frequency corresponding to respiration. We focused on the fact that it is less than For example, when a sleeping person moves his right arm, the sensor installed under the right arm outputs a peak frequency corresponding to the movement, but other sensors installed under the sleeping person's shoulder, back, waist, etc. The sensor continues to output the peak frequency corresponding to the sleeper's breathing.

  Therefore, as in the respiratory monitoring device according to claim 2, the frequency band in which the peak frequency exists most is determined, the sensor that outputs the signal having the peak frequency included in the determined frequency band is selected, and If the respiration signal is calculated based on the signal output from the selected sensor, most of the sensors that output signals including factors other than respiration can be excluded to generate the respiration signal. As a result, the accuracy of the respiration signal can be improved as compared with the conventional respiration monitor device.

  By the way, when determining the frequency band, the breathing signal calculation means may determine the frequency band only under the condition that the peak frequency exists most without any other conditions. In this case, for example, when a sleeping person performs an action over a wide range of the body such as turning over, the frequency band corresponding to the turning over action may be selected. Therefore, as described in claim 3, when determining the frequency band, the respiratory signal calculation means determines the frequency band using only the peak frequency included in the predetermined respiratory frequency band. It is good to have. The “predetermined respiratory frequency band” here means specifically about 0.2 Hz to 0.5 Hz.

  In this case, even when many sensors, such as when a sleeper turns over, capture the movement, the output of such a sensor is determined when determining the frequency band. Since it is determined by being excluded, the accuracy of the calculated respiratory signal is improved.

  In addition, when calculating the respiratory signal, the respiratory signal calculating means may calculate the respiratory signal by averaging all signals output from the selected sensor. In such a case, there is a risk that even valid information included in the output signal of the sensor may be distorted. This is because the movement of the body due to breathing is different for the half-cycle and phase in the part close to the shoulder and the part near the waist even if the back is taken as an example. This is because the signal will be canceled. Also, in other body parts, various phase shifts are observed in signals obtained from motion based on respiration.

  Therefore, as described in claim 4, when calculating the respiration signal, the respiration signal calculating means divides the signal output from the selected sensor based on the phase based on a predetermined phase width. The group having the largest number of signals and the group to which the phase whose phase is shifted by a half cycle from the central phase of the group belongs are selected, and one of the signals belonging to these two groups belongs to one group. The respiratory signal may be calculated by inverting the phase of each signal and adding it to each signal belonging to the other group. Here, “the signal is classified into a group divided by a phase width of a predetermined period based on the phase” means, for example, a group in which one period 2π is divided into ten equal parts (that is, π / 5). Means that the signal is distributed to the group divided according to the phase width) based on the phase. When the group is set in this way, the center phase of the group having a phase width of 0 to π / 5 is π / 10.

  When calculating the respiratory signal, the phase of the signal belonging to one of the two groups is inverted (the amplitude value is inverted) and added to the signal belonging to the other group. A respiratory signal may be calculated. Also, the respiratory signal is calculated by shifting the phase of the signal belonging to one group by π in the time direction in which there is no difference in the center phase between the two groups, and adding it to the signal belonging to the other group. Also good. Note that after the addition, the respiratory signal may be calculated by dividing by the number of added signals.

  In this way, if a part of the signal obtained from each part of the body that is out of phase is selected and the phase signal is substantially corrected to generate a respiratory signal, a more accurate respiratory signal can be obtained. Can be generated.

  By the way, even when various measures as described above are performed, there may be a case where information on the sleeper's body movement such as movement of a limb having a fixed period is included in the respiratory signal. For this reason, as in the respiratory monitoring device according to claim 5, the apparatus is further provided with a determination unit, which extracts a respiration signal for each wavelength from the respiration signal calculated by the respiration signal calculation unit. When the amplitude of the extracted respiratory signal for one wavelength is compared with the amplitude of the other one wavelength of respiratory signal, it is determined whether or not it is greater than a predetermined comparison reference value. Good. The “predetermined comparison reference value” mentioned here may be a specific value such as “3 volts” or a relative value such as “twice”.

  In this way, the determination means makes a determination, notifies the determination result, or transmits it to another device, thereby monitoring the respiratory signal that the respiratory signal contains information on body movements other than breathing. It is possible to inform the person (doctor, nurse, engineer, etc.). As a result, the person who monitors the respiratory signal can correctly understand the respiratory signal and correctly determine a disease such as apnea syndrome.

  Moreover, you may comprise a respiration monitor apparatus as described in Claim 6. In other words, in the respiration monitor device that is arranged in a predetermined distribution under the sleeping person and outputs a signal corresponding to the load or vibration from the sleeping person and outputs respiration information based on the signal, further determination It comprises so that a means may be provided. Then, the determination means selects a sensor that outputs a signal equal to or greater than a predetermined threshold among the signals output from the sensor, and whether or not the number of the selected sensors has changed by a predetermined number or more as time passes. Determine.

  According to such a determination means, it is possible to detect a slight movement of the body, such as when the sleeping person moves the body with respect to the bed. And it becomes possible to notify the person (doctor, nurse, technician, etc.) who monitors respiration information by notifying the judged result or transmitting to other devices. As a result, if the determination result of the determination means is “changed”, the person who monitors the respiratory signal excludes the respiratory information at the time of the determination from the information when diagnosing a disease such as apnea syndrome. It is possible to take measures such as

  Moreover, you may comprise a respiration monitor apparatus as described in Claim 7. In other words, in the respiration monitor device that is arranged in a predetermined distribution under the sleeping person and outputs a signal corresponding to the load or vibration from the sleeping person and outputs respiration information based on the signal, further determination It comprises so that a means may be provided. And the determination means specifies the peak frequency of the spectrum of the signal which a sensor outputs for every sensor, and determines that the sensor in which the specified peak frequency exists in a respiration frequency band is a sensor right under a sleeper.

  According to such a determination means, it is possible to exclude the sensor on which an object is placed and use only the signal of the sensor directly under the sleeping person. Various states of a sleeping person can be recognized with high accuracy.

  By the way, a sensor that outputs a signal corresponding to a load or vibration from a sleeping person, which is generally used in a respiratory monitor device, must detect a slight change in load or vibration with high accuracy. However, depending on the type of sensor, there may be large variations in resistance value output with respect to pressure due to differences in individual characteristics. For example, when considering a membrane sensor generally used in a respiration monitor device, this membrane sensor generally has a narrow pressure detection range (1 KPa to 3 KPa). However, considering use in a respiration monitor device, a minute pressure change of about 0.05 KPa to 0.6 KPa must be detected within a range of 1 KPa to 10 KPa.

  In order to achieve this, as described in claim 8, a plurality of sensors arranged in a predetermined distribution under the sleeping person and outputting signals corresponding to the load or vibration from the sleeping person. It is preferable that the respiration monitor device that has the above-mentioned signal and outputs respiration information further includes a resistance voltage dividing circuit having a sensitivity resistance for each sensor.

  Alternatively, as described in claim 9, the sensor has a plurality of sensors that are arranged in a predetermined distribution under the sleeping person and output a signal corresponding to the load or vibration from the sleeping person. The respiration monitor device that outputs may further include a resistance voltage dividing circuit having a sensitivity resistance, and switching means for switching connection between the resistance voltage dividing circuit and each sensor.

  With this configuration, the pressure-detection voltage characteristic can be made closer to linear with respect to the pressure-resistance characteristic of a sensor that is not linear in nature, and a minute pressure change can be detected with higher accuracy.

  When the switching means as described in claim 9 is provided, the resistance value of the sensitivity resistor is further changed according to the characteristics of the sensor connected to the resistance voltage dividing circuit as described in claim 10. It is preferable to provide sensitivity resistance changing means.

  In this way, since the measurement accuracy can be adjusted according to the characteristics of each sensor, respiratory information can be output with high accuracy.

  Embodiments to which the present invention is applied will be described below with reference to the drawings. The embodiments of the present invention are not limited to the following examples, and can take various forms as long as they belong to the technical scope of the present invention.

FIG. 1 is an explanatory diagram showing an installation state when using the respiratory monitor apparatus A according to the embodiment, and FIG. 2 is a schematic configuration diagram of the respiratory monitor apparatus A.
As shown in FIG. 1, the bed 1 includes a placement portion 11 for placing a bedding 10 such as a mattress and a back plate portion 12 erected from an end portion of the placement portion 11. The respiratory monitoring device A is used by being inserted into the lower part of the bedding 10 installed on the placement unit 11 of the bed 1. The respiratory monitor device A is installed on the back plate 12 side from the center of the placement unit 11 so as to correspond to the position of the abdomen from the chest of the sleeping person when the sleeping person lies on the bed 1. Is done.

  As shown in FIG. 2, the respiration monitor device A includes a seat unit 2 and a control unit 3. The sheet unit 2 is configured by sandwiching a plurality of (three in this example) sensor sheets 22 and a sensor selection unit 23 provided for each sensor sheet 22 between two sheet-like protective members 21. Has been.

  The sensor sheet 22 uses a pressure sensitive element for the electrode of the membrane switch, and a plurality of sensors 221 whose electric resistance changes (decreases) according to the applied load are arranged at approximately equal intervals (in this example, a total of 162 sheets of three sheets). ) It is arranged on a sheet-like member. In FIG. 2, illustration of a wiring pattern that electrically connects each sensor 221 and the sensor selector 23 is omitted.

The sensor selection unit 23 can electrically switch the connection with each sensor 221, acquires a voltage signal from the sensor 221, and transmits the voltage signal to the control unit 3.
That is, when a voltage is applied to a circuit including each sensor 221, the voltage drop due to the sensor 221 increases or decreases due to the change in the electrical resistance of the sensor 221 according to the applied load. Based on the change, the controller 3 can detect the applied load independently for each sensor 221.

  The control unit 3 will be described with reference to the block diagram of FIG. The control unit 3 includes an A / D converter 31 that converts an analog signal into a digital signal, a microcomputer 32 that executes various processes, a memory 33 that stores various data, and a display unit 34 that includes an LED, a liquid crystal device, and the like. The Among these, the microcomputer 32 issues a command to the sensor selection unit 23, and takes in a load signal from the sensor sheet 22 (more precisely, the sensor 221) via the sensor selection unit 23 and the A / D converter 31. A process (respiration signal output process) of calculating a respiration signal based on the load signal and outputting it to the display unit 34 is executed. This breathing signal output process will be described in detail later.

  Here, the sensor selection unit 23 will be described in detail with reference to the circuit diagram of FIG. As shown in FIG. 4, the sensor selection unit 23 includes a digital potentio 231 and a changeover switch 232. The digital potentio 231 can change the value of the sensitivity resistance Rv in accordance with the command from the microcomputer 32 described above. The command data is registered in advance in the microcomputer 32 for each sensor 221. The changeover switch 231 can switch the electrical connection between the sensor 221 and the A / D converter 31. The digital potentio 231 corresponds to the resistance voltage dividing circuit and the moving resistance changing means described in the claims. The changeover switch 232 corresponds to the changeover means described in the claims.

Here, the method of determining the value of the sensitivity resistance Rv will be described in detail for the case of the sensor 221 having the resistance Rs.
First, the detection performance necessary for the sensor 221 will be considered. When the inventors of the present application examined the pressure applied to each point of the sleeping person's body (mainly the back) and the distribution of the pressure change based on the breathing at that time, the distribution shown in the distribution diagram of FIG. 5 was obtained. It was. According to this distribution diagram, it is estimated that at least 1 KPa to 2.5 KPa needs to detect at least 0.05 KPa of pressure change (inversely about 0.05 KPa) in order to prevent the detection of pressure change from being missed to some extent. If the pressure change cannot be detected, a lot of missing data will occur.) In addition, at 2.5 KPa to 5 KPa, it is estimated that it is necessary to detect at least a pressure change of about 0.1 KPa. In addition, at 5 KPa to 10 KPa, it is estimated that it is necessary to detect at least a pressure change of about 0.2 KPa.

  By the way, a sensor using a pressure-sensitive element as an electrode of a membrane switch generally has a relationship similar to the relational expression between the pressure P and the resistance value Rs shown in the following mathematical formula 1.

  Further, the voltage division ratio of the pressure-sensitive resistor Rv is expressed by the equation shown in Equation 2.

  Therefore, the difference in resistance ratio when a pressure change of, for example, 0.05 KPa is received from a certain reference pressure is expressed by the equation shown in Equation 3.

  Further, from the distribution chart shown in FIG. 5, the value distribution is about 60% in the pressure range of 1 KPa to 2.5 KPa, 30% in the pressure range of 2.5 KPa to 5 KPa, and about 10% in the pressure range of 5 KPa to 10 KPa. Therefore, 6 times, 3 times, and 1 time are considered as weights for the respective pressure ranges.

  Then, in consideration of this weighting, in order to obtain Rv in which the difference in the voltage division ratio shown in Equation 3 is the largest, a value Rcal obtained by integrating the difference in the voltage division ratio is set as shown in Equation 4, and the value of Rcal is Find the largest Rv.

If the value of the sensitivity resistance Rv is determined so as to be the value of Rv thus obtained, the sensor 221 can capture the pressure change in a balanced manner over the required pressure range.
Next, from another viewpoint, the effect of the presence of the sensitivity resistance Rv will be described. A sensor using a pressure-sensitive element as an electrode of a membrane switch generally has a relation similar to the relational expression between the pressure P and the resistance value Rs shown in Expression 1. When Expression 1 is represented in a graph, a graph as shown in FIG. As shown in FIG. 6A, the P-Rs characteristic is not linear with respect to the pressure P. Therefore, when the sensor resistance Rs is obtained without using the sensitivity resistance Rv, the waveform when the respiration signal is calculated becomes irregular. become.

  Therefore, considering that the sensitivity resistor Rv is used as in the circuit diagram shown in FIG. 4, the AD value obtained by the A / D converter 31 is expressed by an equation shown in Equation 5.

  Then, when the relationship between the AD value shown in Formula 5 and the sensor resistance Rs is graphed, a graph as shown in FIG. 6B is obtained. In order to replace the horizontal axis with the pressure P, when the pressure P shown in Formula 1 is substituted into Formula 5, the formula shown in Formula 6 is obtained.

  If the relationship between the AD value and the pressure P shown in Equation 6 is graphed, a graph as shown in FIG. As can be seen from this graph, the P-AD value characteristic is more linear than the P-Rs characteristic shown in FIG. 6A, and therefore the waveform when the respiratory signal is calculated is less likely to be distorted.

  Next, the breathing signal output process executed by the microcomputer 32 will be described with reference to the flowchart shown in FIG. The respiration signal output process starts when a user operates an operation unit (not shown) of the respiration monitor apparatus A.

  When the execution is started, the microcomputer 23 sequentially takes in load signals for one cycle from all the sensors 221 via the sensor selection unit 23 and the A / D converter 31 (S110). The sampling frequency is preferably about 10 Hz, for example.

  Subsequently, a signal having a signal intensity equal to or higher than a predetermined threshold is selected from the acquired load signals (S115). Specifically, for example, the load signal voltage drop value is selected to be a predetermined value or less. In other words, it is intended to exclude only a load signal that represents the load of the bedding and to focus only on a load signal that is assumed to represent the sleeper's load.

  Next, the bias component is removed from the load signal (S120). Specifically, first, filtering processing by a digital filter that cuts a frequency band of 3 Hz or more is performed on the load signal, and the respiratory frequency (about 0.2 Hz to 0.5 Hz) that is a frequency corresponding to the respiratory rate is exceeded. A signal from which a high noise component is removed is generated (first signal). Next, a filtering process using a digital filter that cuts a frequency band of 0.3 Hz or higher is performed on the first signal to generate a signal from which a frequency component corresponding to the respiratory frequency is removed (second signal). Then, a signal from which the bias component has been removed is obtained by subtracting the second signal from the first signal.

  Next, the process branches depending on whether or not a load signal for 256 cycles has already been processed (S125). If the load signal for 256 cycles has already been processed, the process proceeds to S130, and if the load signal for 256 cycles has not been processed yet, the process returns to S110.

In S130, a power spectrum is obtained by using an FFT (Fast Fourier Transform) method for each signal output from each sensor 221 and processed.
Subsequently, a human or object determination process is executed (S135). This human or object determination process will be described with reference to the flowchart shown in FIG. When the execution of the human or object determination process is started, the frequency of the peak value (peak frequency) in the power spectrum group obtained in S130 of FIG. 7 is the respiratory frequency band (about 0.2 Hz to 0.5 Hz). It is determined whether or not a power spectrum exists. That is, it is determined whether or not the sensor 221 that outputs a load signal having a peak frequency in the respiratory frequency band exists in 256 cycles (S210). By making such a determination, it can be determined whether a sleeping person (person) or an object is placed on each sensor 221.

  As a result of this determination, if there is a sensor 221 that outputs a load signal having a peak frequency in the respiratory frequency band, a human determination flag is set (S220), and this processing (human or object determination processing) is terminated. Return to the breathing signal output process. On the other hand, if there is no sensor 221 that outputs a load signal whose peak frequency is in the respiration frequency band, the person determination flag is cleared (S230), this process (person or object determination process) is terminated, and a respiration signal is output. Return to processing.

  Returning to FIG. 7, when the person or object determination process (S135) ends, the process proceeds to S140, and the process branches depending on whether or not there is a sensor 221 on which a person is placed. Specifically, the process branches depending on whether the person determination flag is set or cleared. If the person determination flag is set (there is a sensor 221 on which a person is placed), the process proceeds to S145. If the person determination flag is cleared (there is no sensor 221 on which a person is placed), the process proceeds to S110. Return.

  In S145, fine movement determination processing is executed. The fine movement determination process will be described with reference to the flowchart shown in FIG. When the execution of the fine movement determination process is started, the set M0 of the sensors 221 on which a person is placed is stored in the first cycle of 256 cycles (S310). The term “sensor 221 on which a person is placed” as used herein refers to a sensor whose load signal is greater than or equal to a predetermined threshold value and is assumed to represent a sleeper's load.

  Next, in each cycle, the set of sensors 221 on which a person is placed is compared with the stored sensor set M0, and the number of different sensors 221 is counted (S320). When all the 256 cycles have been compared, the process branches depending on whether the number of different sensors 221 is 10% or more of the number of sensors 221 constituting the sensor set M0 (S330).

  When the number of different sensors 221 is 10% or more of the number of sensors 221 constituting the sensor set M0, the process proceeds to S340. When the number of sensors 221 constituting the sensor set M0 is not 10% or more, the process proceeds to S350. move on.

  In S340, the fine movement flag is set, the process (fine movement determination process) is terminated, and the process returns to the breathing signal output process. On the other hand, in S350, the fine movement flag is cleared, this process (fine movement determination process) is terminated, and the process returns to the breathing signal output process.

  Returning to FIG. 7, when the fine movement determination process (S145) is completed, the process proceeds to S150, and a respiratory signal calculation process is executed. This respiration signal calculation process will be described with reference to the flowchart shown in FIG.

  When the execution of the respiration signal calculation process is started, first, the peak frequency in 256 cycles of the signal acquired from each sensor 221 is specified for each sensor 221 (S410). Subsequently, each identified peak frequency is classified for each predetermined frequency band, and the number of sensors 221 that output the peak frequency is counted for each frequency band (S420). Here, the “predetermined frequency band” means, for example, each frequency band obtained by dividing between 0.2 Hz and 0.5 Hz in increments of 0.03 Hz.

Subsequently, each sensor 221 constituting the frequency band to which the largest number of sensors belong is selected as the sensor group SGm among the frequency bands (S430).
Subsequently, based on the output signals of the sensors 221 constituting the sensor group SGm, these sensors 221 are divided into phase groups PG1 (0 to π / 5) and PG2 (π / 5) having a phase width of π / 5. ˜2π / 5),..., PG10 (9π / 5 to 2π) (S440).

Subsequently, the sensors Sm1, Sm2, Sm3,... Belonging to the group PGmax having the largest number of sensors 221 among PG1 to PG10 are selected (S450).
Subsequently, the sensors Sn1, Sn2, Sn3,... Belonging to the group shifted by π from the phase of PGmax are selected (S460).

  Finally, a respiration signal is calculated according to the following Equation 7 (S470), this process (respiration signal calculation process) is terminated, and the process returns to the respiration signal output process. In Equation 7, addition and subtraction are performed using symbols Smx and Snx indicating the sensor for convenience, but in actual calculation, the output value of the corresponding sensor is added or subtracted with the time axis aligned.

  By calculating the respiratory signal in this way, it is possible to eliminate the influence of noise caused by body movements other than breathing as much as possible, and obtain a respiratory signal that accurately corresponds to the sleeping state of the sleeping person.

  Returning to FIG. 7, when the breathing signal calculation process (S150) is completed, the process proceeds to S155, and the limb instantaneous movement determination process is executed. This limb blink determination process will be described with reference to the flowchart shown in FIG.

  When the execution of the instantaneous limb movement determination process is started, first, the amplitude of each waveform in the respiration signal of all 256 cycles is compared with the amplitude of one waveform before and after that (S510). As a result, the process branches depending on whether or not there is a cycle whose amplitude is twice or more larger than the preceding and following cycles (S520). If there is a cycle whose amplitude is twice or more larger than that of the preceding and following cycles, the process proceeds to S530, and if there is no cycle whose amplitude is twice or more larger than that of the preceding and following cycles, the process proceeds to S540.

In step S530, the extremity blink flag is set, the present process (extremity blink determination process) is terminated, and the process returns to the breathing signal output process.
In S540, the extremity blink flag of the limb is cleared, this process (extremity blink determination process) is terminated, and the process returns to the breathing signal output process.

  Returning to FIG. 7, when the instantaneous movement determination process (S155) of the extremities is completed, the process proceeds to S160 and a respiration signal is output. This is performed by causing the display unit 34 to display the respiratory signal as a waveform. In addition, the state of the fine movement flag and the limb blink flag is also displayed on the display unit 34.

  In the subsequent S165, the process branches depending on whether or not the processing of all the cycles of the load signal output from the sensor 221 has been completed. If the processing of all cycles has been completed, this processing (breathing signal output processing) is terminated, and if the processing of all cycles has not been completed, the processing returns to S110.

  According to the respiration monitor apparatus A that performs the configuration and operation as described above, only a sensor that is likely to output a signal based on respiration is selected (S450 and S460 in FIG. 10) to generate a respiration signal. Therefore (S470 in FIG. 10), the accuracy of the respiration signal is improved as compared with the conventional respiration monitor device. In addition, since the respiratory monitor apparatus A performs fine movement determination processing (FIG. 9) and instantaneous limb movement determination processing (FIG. 11) and displays the execution results on the display unit 34, the respiratory signal It is possible to inform a person (doctor, nurse, engineer, etc.) who monitors the respiratory signal that there is a possibility that the body movement information other than breathing may be included in the information. As a result, the person who monitors the respiratory signal can correctly understand the respiratory signal and correctly determine a disease such as apnea syndrome.

Other embodiments will be described below.
In the respiratory monitoring device A of the above embodiment, in the limb blinking determination process (FIG. 11), the amplitude of each cycle is compared with the amplitude of the cycle before and after that, and the amplitude is twice or more as compared with the cycle before and after the cycle. Although it has been determined whether or not there is a large cycle, it may be determined and notified whether or not a cycle that satisfies such a condition periodically exists. If it becomes like this, it is also possible to utilize the respiration monitor apparatus A when determining that there exists a possibility of periodic limb movement abnormality (PLM).

It is explanatory drawing which shows the installation state at the time of using a respiration monitor apparatus. It is a schematic block diagram of a respiration monitor apparatus. It is a block diagram for demonstrating the control part of a respiration monitor apparatus. It is a circuit diagram for demonstrating a sensor selection part. It is a distribution map which shows the pressure concerning each point of a sleeper's body, and the distribution of the pressure change based on the breathing in that case. It is a graph for demonstrating numerical formula. It is a flowchart for demonstrating a respiration signal output process. It is a flowchart for demonstrating the determination process of a person or an object. It is a flowchart for demonstrating a fine movement determination process. It is a flowchart for demonstrating a respiration signal calculation process. It is a flowchart for demonstrating the instantaneous movement determination process of limbs.

Explanation of symbols

  A ... Respiration monitor device, 1 ... Bed, 2 ... Seat part, 3 ... Control part, 10 ... Bedding, 11 ... Placement part, 12 ... Back plate part, 21 ... Protection member, 22 ... Sensor seat, 23 ... Sensor selection , 31 ... A / D converter, 32 ... microcomputer, 33 ... memory, 34 ... display unit, 221 ... sensor, 231 ... digital potentio.

Claims (10)

A plurality of sensors arranged under a sleeping person in a predetermined distribution and outputting a signal corresponding to the load or vibration from the sleeping person;
Respiration signal calculating means for converting the signal output by the sensor into a frequency domain, selecting a sensor based on the characteristics of the converted signal, and calculating a respiration signal based on the signal output by the selected sensor; ,
A respiratory monitor device comprising: The respiratory monitoring device according to claim 1, wherein
The breathing signal calculation means specifies the peak frequency of the spectrum of the signal output from the sensor for each sensor, determines the frequency band in which the specified peak frequency is the most, and includes the determined frequency band A respiration monitor device that selects the sensor that outputs the signal having the peak frequency and calculates a respiration signal based on the signal output by the selected sensor. In the respiratory monitoring device according to claim 2,
The respiration signal calculating means, when determining the frequency band, determines the frequency band using only the peak frequency included in a predetermined respiration frequency band. In the respiratory monitoring device according to claim 2 or 3,
The respiratory signal calculation means classifies the signals output from the selected sensor into groups divided by a predetermined phase width based on the phase when calculating the respiratory signal, Select the group with the largest number and the group to which the phase whose phase is shifted by a half cycle from the center phase of the group, and invert the phase of the signal belonging to one of the signals belonging to these two groups Then, the respiration signal is calculated by adding the signal belonging to the other group and calculating the respiration signal. In the respiratory monitoring device according to any one of claims 1 to 4,
Furthermore, a determination means is provided,
The determination means extracts a respiration signal for each wavelength from the respiration signal calculated by the respiration signal calculation means, and the amplitude of the extracted respiration signal for one wavelength is the respiration signal for another wavelength. A respiration monitor device for determining whether or not the amplitude is greater than a predetermined comparison reference value compared with an amplitude In a respiratory monitor device that is arranged in a predetermined distribution under a sleeping person, has a plurality of sensors that output signals corresponding to the load or vibration from the sleeping person, and outputs respiratory information based on the signals,
Furthermore, a determination means is provided,
The determination means selects a sensor that outputs a signal equal to or greater than a predetermined threshold from the signals output by the sensor, and whether or not the number of the selected sensors has changed by a predetermined number or more over time. A respiratory monitor device characterized by determining whether or not. In a respiratory monitor device that is arranged in a predetermined distribution under a sleeping person, has a plurality of sensors that output signals corresponding to the load or vibration from the sleeping person, and outputs respiratory information based on the signals,
Furthermore, a determination means is provided,
The determination means specifies the peak frequency of the spectrum of the signal output from the sensor for each sensor, and determines that the sensor having the specified peak frequency in the respiratory frequency band is a sensor directly under the sleeping person. Respiratory monitor device characterized by In a respiratory monitor device that is arranged in a predetermined distribution under a sleeping person, has a plurality of sensors that output signals corresponding to the load or vibration from the sleeping person, and outputs respiratory information based on the signals,
Furthermore,
A respiratory monitor device comprising a resistance voltage dividing circuit having a sensitivity resistance for each sensor. In a respiratory monitor device that is arranged in a predetermined distribution under a sleeping person, has a plurality of sensors that output signals corresponding to the load or vibration from the sleeping person, and outputs respiratory information based on the signals,
Furthermore,
A resistance voltage divider having a sensitivity resistance;
Switching means for switching the connection between the resistance voltage dividing circuit and each of the sensors;
A respiratory monitor device comprising: The respiratory monitoring device according to claim 9,
Furthermore,
A respiratory monitor device comprising sensitivity resistance changing means for changing a resistance value of the sensitivity resistor in accordance with characteristics of the sensor connected to the resistance voltage dividing circuit.