WO2008029587A1

WO2008029587A1 - Medical measurement device for outputting information useful for diagnosis

Info

Publication number: WO2008029587A1
Application number: PCT/JP2007/065496
Authority: WO
Inventors: Kazunobu Itonaga; Shozo Takamatsu; Hironori Sato
Original assignee: Omron Healthcare Co., Ltd.
Priority date: 2006-09-07
Filing date: 2007-08-08
Publication date: 2008-03-13
Also published as: TW200816953A; JP2008061824A

Abstract

A pulse wave detection device as a medical measurement device calculates a plurality of types of feature amounts such as the AI value, ET, the pulse wave cycle, the MSP conspicuity, and the base line fluctuation ratio for each of delimited pulse waveforms (S1, 3) from the measured continuous pulse wave (S5). Upon completion of the measurement, a representative waveform is searched according to the plurality of types of feature amounts for each of the calculated pulse waveform (S7).

Description

Specification

Medical measuring instrument that outputs information useful for diagnosis

Technical field

The present invention relates to a medical measuring instrument, a biological signal waveform extraction method, and a medium on which a biological signal waveform extraction program is recorded, and in particular, a medical measuring instrument that continuously detects a waveform obtained from a biological signal, The present invention relates to a biological signal waveform extraction method and a medium on which a biological signal waveform extraction program is recorded.

Background art

[0002] Some pulse wave detection devices as medical measuring instruments continuously detect a pulse wave that is a waveform obtained from a biological signal. In such a pulse wave detection device, in order to make an accurate diagnosis in a short time using the pulse wave waveform output from the pulse wave detection device, a representative waveform (hereinafter referred to as a “waveform”) is selected from a large number of detected pulse wave waveforms. , Representative waveform) must be extracted and displayed.

[0003] The arteriosclerosis inspection apparatus disclosed in Japanese Patent Application Laid-Open No. 2004-136107 (hereinafter referred to as Patent Document 1) extracts a representative waveform without displaying a waveform mixed with noise due to arrhythmia or body movement. In order to display, a method of selecting and displaying a waveform in which the sharpness of the peak of the ejection wave is close to the average from the sequentially detected waveforms is adopted.

Patent Document 1: JP 2004-136107 A

Disclosure of the invention

Problems to be solved by the invention

[0004] However, in the method disclosed in Patent Document 1 above, the feature quantity of the waveform is limited to the sharpness of the peak of the ejection wave, and the feature quantity other than the sharpness is lost. The extracted representative waveform may be insufficient depending on the purpose of diagnosis.

[0005] As another method, there may be a method of calculating and displaying an average waveform from a detected waveform using Fourier transform or the like, but the waveform displayed by this method is actually a living body. The waveform is not detected. Therefore, there is a problem that the original feature amount may be lost. [0006] The present invention has been made in view of such a problem, and extracts a waveform corresponding to a diagnostic purpose from a plurality of detected waveforms obtained from a biological signal and outputs the same. An object is to provide a measuring instrument, a biological signal waveform extraction method, and a medium on which a biological signal waveform extraction program is recorded.

Means for solving the problem

[0007] In order to achieve the above object, according to one aspect of the present invention, a medical measuring instrument includes: a detection unit that detects a plurality of waveforms obtained from a biological signal; and a plurality of waveforms from each of the plurality of waveforms. A calculation unit that calculates one feature value and a second feature value, a plurality of waveform forces, a relationship of the first feature value to the first index obtained from the plurality of calculated first feature values, and a plurality A search unit for searching for a representative waveform from the plurality of waveforms based on the relationship of the second feature quantity to the second index obtained from the plurality of second feature quantities calculated from the waveform of An output processing unit that performs processing for outputting a waveform.

[0008] Preferably, the waveform obtained from the biological signal is a pulse waveform, and the first feature value and the second feature value include an AI (Augmentation Index) value, a pulse wave cycle, a baseline fluctuation rate, and a sharpness. , And at least one of ET (Ejection Time).

[0009] Preferably, the search unit sets a weighting coefficient for each of the first feature value and the second feature value in one waveform, and the first feature for the first index And a determination unit that determines a representative waveform in consideration of a weighting factor for each of the relationship between the quantities and the relationship between the second feature quantity and the second index.

[0010] Preferably, the search unit uses the average value of the first feature value for the plurality of waveforms as a first index, and the second value for the plurality of waveforms as a second index. An average calculation unit for calculating an average value of feature quantities, a first difference that is a difference between the first index and the first feature quantity, and a second index and a second feature for each of the plurality of waveforms. A difference calculation unit that calculates a second difference that is a difference from the quantity, a normalization processing unit that normalizes the calculated first difference and second difference, and the normalized first difference and second difference And a determination unit for determining a representative waveform based on the determination waveform.

[0011] More preferably, the search unit includes a coefficient setting unit that sets a weighting factor for each of the first feature amount and the second feature amount in one waveform, and the normalized first difference and first feature amount. A weight processing unit that multiplies each of the two differences by a set weighting coefficient, and the determination unit determines the representative waveform based on the weighted first difference and second difference.

[0012] More preferably, the search unit further includes an addition processing unit that adds the weighted first difference and the second difference for each of the plurality of waveforms, and the determination unit has a minimum added difference. Is determined as a representative waveform.

[0013] Preferably, the search processing in the search unit is performed based on the relationship between each of the first feature value and the second feature value and the first-stage index for the plurality of waveforms. Based on the first-stage search process for selecting the target waveform from the above, and the relationship between the first and second feature quantities of the target waveform and the second-stage index, And a second-stage search process for searching for a representative waveform from the waveforms.

[0014] Preferably, the output processing unit performs a process for displaying the value used for the process of detecting the representative waveform in the search unit or the information specifying the representative waveform together with the representative waveform.

[0015] Preferably, each of the first index and the second index includes an average value, a median value, a mode value, and a maximum value of a plurality of corresponding feature amounts for the plurality of waveforms. The difference between the minimum value and any threshold! / Straight! /.

[0016] Preferably, an index setting unit that sets the first index and the second index is further provided.

According to another aspect of the present invention, a biological signal waveform extraction method is a method of extracting a representative waveform from waveforms obtained from a biological signal, the step of acquiring a plurality of continuous waveforms, A step of dividing a unit waveform from a plurality of continuous waveforms, a step of calculating a first feature amount and a second feature amount from each of the plurality of unit waveforms, and a plurality of components calculated from the plurality of waveforms The relationship of the first feature quantity to the first index obtained from the first feature quantity, and the second feature for the second index obtained from a plurality of second feature quantities calculated from a plurality of waveforms Based on the quantity relationship, a step of extracting a representative waveform from the plurality of unit waveforms and a step of outputting the representative waveform are provided.

[0017] According to still another aspect of the present invention, a medium on which a biological signal waveform extraction program is recorded executes a process of extracting a representative waveform from waveforms obtained from a biological signal. Each of the plurality of unit waveforms, the step of obtaining a plurality of consecutive waveforms, the step of dividing a unit waveform from the plurality of consecutive waveforms, and the second feature waveform. A feature amount of the first feature amount, a plurality of waveform forces, a relationship of the first feature amount to the first index obtained from the plurality of first feature amounts calculated, and a plurality of waveforms calculated from the plurality of waveforms. The step of extracting the representative waveform from the plurality of unit waveforms and the step of outputting the representative waveform are executed based on the relationship of the second feature amount to the second index obtained from the second feature amount of It is a computer-readable recording medium on which the program to be recorded is recorded.

The invention's effect

[0018] In the medical measuring instrument, which is the power of the present invention, the representative waveform is extracted from the waveform obtained from the biological signal force measured using a plurality of feature quantities, so that a stable waveform can be extracted. .

Furthermore, by considering the importance of each feature quantity, a waveform that matches the purpose of diagnosis can be extracted, and a useful waveform can be provided by diagnosis.

Brief Description of Drawings

FIG. 1 is a diagram showing a specific example of a device configuration of a pulse wave detection device.

FIG. 2 is a block diagram showing a specific example of a functional configuration of the pulse wave detection device.

FIG. 3 is a diagram showing a specific example of a pulse wave waveform.

FIG. 4 is a diagram showing a specific example of a pulse wave waveform.

FIG. 5 is a diagram showing a specific example of a pulse wave waveform.

FIG. 6 is a diagram showing a specific example of a pulse wave waveform.

FIG. 7 is a block diagram showing one specific example of a functional configuration included in the representative waveform search processing unit 109.

FIG. 8 is a flowchart showing processing in the pulse wave detection device.

FIG. 9 is a flowchart showing an example of a pulse wave segmentation process executed in step S3.

FIG. 10 is a flowchart showing an example of representative waveform search processing executed in step S7.

FIG. 11 is a diagram for explaining display contents on the display screen. FIG. 12 is a flowchart showing an example of a representative waveform search process executed in step S 7 in the pulse wave detection device according to a modification.

Explanation of symbols

[0021] 1 Sensor unit, 3 Display unit, 5 Air tube, 7 Fixed base unit, 11 CPU, 12

ROM, 13 RAM, 14 Control circuit, 15 Pressure pump, 16 Negative pressure pump, 17 Switching valve, 18 Pressing cuff, 19 Semiconductor pressure sensor, 20 Multiplexer, 21 Amplifier, 22 Characteristic variable filter, 23 A / D converter, 24 operation unit, 25 display unit, 101 pulse wave measurement value acquisition unit, 103 pulse wave segmentation processing unit, 105 feature amount calculation unit, 107 feature amount storage unit, 109 representative waveform search processing unit, 111 representative waveform display control unit, 1050 AI calculation unit, 1051 ET calculation unit, 1052 pulse wave period calculation unit, 1053 MSP calculation unit, 1054 baseline fluctuation rate calculation unit, 1091 feature value reading unit, 1092 average calculation unit, 1093 difference calculation unit, 1094 normalization processing unit, 1095 coefficient determination unit, 1096 weighting processing unit, 1097 addition processing unit, 1098 representative waveform determination unit.

BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the following description, the same parts and components are denoted by the same reference numerals. Their names and functions are the same.

In the present embodiment, the medical measuring instrument according to the present invention is employed in a pulse wave detection device that detects a pulse wave that is a waveform obtained from a biological signal. However, the medical measuring instrument according to the present invention is not limited to the pulse wave detection device, and may be adopted in any other device as long as it is a device that detects a waveform obtained from a biological signal.

Referring to FIG. 1, the pulse wave detection device according to the present embodiment generally includes, as one specific example, sensor unit 1, display unit 3, and fixed base unit 7. Composed.

[0025] The display unit 3 is provided so as to be operable from the outside, and is operated to input various information related to pulse wave analysis, and outputs various information such as pulse wave analysis results to the outside. It includes a display unit 25 consisting of a light emitting diode (LED) and a liquid crystal display (LCD).

[0026] The fixed base unit 7 stores data and programs for controlling the pulse wave detection device. ROM (Read Only Memory) 12 and RAM (Random Access Memory) 13, CPU (Central Processing Unit) 11 that performs various processes including calculations to centrally control the pulse detector, pressurization pump 15, negative pressure pump 16, switching valve 17, control circuit 14 for receiving signal from CPU11 and sending to pressure pump 15, negative pressure pump 16, switching valve 17, variable characteristics that can be changed to at least two values Includes filter 22 and A / D converter 23.

[0027] The CPU 11 accesses the ROM 12, reads the program, develops it on the RAM 13, executes it, and controls the entire pulse wave detection device. The CPU 11 receives an operation signal from the user from the operation unit 24, and performs control processing for the entire pulse wave detection device based on the operation signal. That is, the CPU 11 sends a control signal to the control circuit 14, the multiplexer 20, and the characteristic variable filter 22 based on the operation signal input from the operation unit 24. Further, the CPU 11 performs control for displaying the pulse wave analysis result on the display unit 25.

[0028] The pressurizing pump 15 is a pump for pressure-causing an internal pressure (hereinafter referred to as "cuff pressure") of a press cuff (air bag) 18 to be described later. The negative pressure pump 16 is a pump for reducing the cuff pressure. The switching valve 17 selectively connects one of the pressurizing pump 15 and the negative pressure pump 16 to the air pipe 5. The control circuit 14 controls these in accordance with a control signal from the CPU 11.

[0029] The sensor unit 1 includes a semiconductor pressure sensor 19 including a plurality of sensor elements, a multiplexer 20 that selectively derives pressure signals output from the plurality of sensor elements, and a pressure signal output from the multiplexer 20. And a pressure cuff 18 including an air bag that is pressurized to press the semiconductor pressure sensor 19 onto the wrist.

[0030] The semiconductor pressure sensor 19 is configured to include a plurality of sensor elements arranged at predetermined intervals in one direction on a semiconductor chip made of single crystal silicon or the like. The semiconductor pressure sensor 19 is pressed against the measurement site such as the wrist of the subject under measurement by the pressure of the pressing cuff 18. In this state, the semiconductor pressure sensor 19 detects the pulse wave of the subject via the radial artery. The semiconductor pressure sensor 19 detects the pulse wave to detect The pressure signal output from the sub-element is input to the multiplexer 20 for each sensor element channel. For example, 40 sensor elements are arranged.

[0031] The multiplexer 20 selectively outputs a pressure signal output from each sensor element. The pressure signal sent from the multiplexer 20 is amplified by the amplifier 21 and selectively supplied to the A / D converter 23 via the variable characteristic filter 22.

In the present embodiment, until an optimal sensor element for pulse wave detection is selected, multiplexer 20 outputs a plurality of pressure signals output from each sensor element in accordance with a control signal from CPU 11. Switch sequentially and output. After the optimum sensor element for pulse wave detection is selected, it is fixed to the corresponding channel according to the control signal from CPU11. Accordingly, at this time, the multiplexer 20 selects and outputs the pressure signal output from the selected sensor element.

[0033] The characteristic variable filter 22 is a low-pass filter for blocking signal components of a predetermined value or more, and can be changed to at least two values.

The A / D converter 23 converts the pressure signal, which is an analog signal derived from the semiconductor pressure sensor 19, into digital information, and provides the digital information to the CPU 11. The A / D converter 23 simultaneously acquires the pressure signals output from the sensor elements included in the semiconductor pressure sensor 19 via the multiplexer 20 until the channel of the multiplexer 20 is fixed by the CPU 11. After the CPU 11 fixes the channel of the multiplexer 20, the A / D converter 23 acquires the pressure signal output from the corresponding sensor element. The period in which the pressure signal is sampled (hereinafter referred to as “sampling period”) is, for example, 2 ms.

The variable characteristic filter 22 changes the value of the cutoff frequency until the channel of the multiplexer 20 is fixed and after it is fixed. Until the channel of the multiplexer 20 is fixed, sampling is performed by switching a plurality of pressure signals. Therefore, the variable characteristic filter 22 selects a cutoff frequency value higher than the sampling frequency (for example, 20 kHz) at this time. As a result, the force S can be prevented from causing rounding after A / D conversion, and the optimum sensor element can be selected appropriately. After the channel is fixed, the characteristic variable filter 22 follows the control signal from the CPU 11. Thus, a value is selected that results in a cutoff frequency that is less than half of the sampling frequency (eg, 500 Hz) for a single pressure signal. As a result, aliasing noise can be reduced, and pulse wave analysis can be performed with high accuracy. Note that aliasing noise is a frequency component of 1/2 or more of the sampling frequency that appears in an area of 1/2 or less of the sampling frequency due to the aliasing phenomenon when an analog signal is converted to a digital signal by the sampling theorem. Refers to noise with

In the present embodiment, since the CPU 11, ROM 12 and RAM 13 are provided in the fixed base unit 7! /, The display unit 3 can be downsized.

[0037] It should be noted that the display unit 3 may be built in the force fixing base unit 7 provided separately from the fixing base unit 7 and the display unit 3. Conversely, the display unit 3 may be provided with the CPU 11, ROM 12, and RAMI 3. Also, it can be connected to a PC (Personal Computer) to perform various controls.

[0038] In the pulse wave detection device according to the present embodiment, a representative pulse wave is searched from a plurality of detected pulse waves, extracted, and displayed. FIG. 2 is a block diagram showing a specific example of a functional configuration for searching, extracting, and displaying a representative pulse wave from a plurality of detected pulse waves in the pulse wave detection device. Each function shown in FIG. 2 is performed by CPU 11 accessing ROM 12, reading a program, developing it on RAM 13, and executing it. At least a part of the functions formed in CPU 11 is shown in FIG. It may be a function that is exhibited by the device.

Referring to FIG. 2, the above functions of the pulse wave detection device include a pulse wave measurement value acquisition unit 101 for acquiring a pressure signal (sensor signal) from a sensor element included in the semiconductor pressure sensor 19, 1 is divided by a pulse wave delimiter processing unit 103 and a pulse wave delimiter processing unit 103 that process the pressure signal acquired by the wave measurement value acquisition unit 101 to perform a process of dividing a continuous pulse wave into one beat at a time. Feature quantity calculation unit 105 that calculates multiple types of feature quantities for each unit waveform that represents the pulse wave of acupuncture, feature quantity storage unit 10 7 that stores multiple types of calculated feature quantities one beat at a time, calculation for each unit waveform A representative waveform search processing unit 109 for performing processing for searching and extracting a pulse waveform as a representative waveform (hereinafter referred to as a representative waveform) from a continuous pulse wave using the plurality of types of feature values obtained, To display the extracted representative waveform on the display unit 25. It includes a representative waveform display control unit 111 that performs processing for generating and displaying signals.

Yes

[0040] The types of feature amounts calculated by the feature amount calculation unit 105 include, for example, AI value, ET value, pulse period, rising sharpness MSP, and baseline fluctuation rate. In the present embodiment, the feature quantity calculation unit 105 includes an AI calculation unit 1050, an ET calculation unit 1051, a pulse wave period calculation unit 1052, an MSP calculation unit 1053, and a baseline fluctuation rate calculation unit 1054. It is assumed that the type of feature amount is calculated.

[0041] Here, AI is a known index indicating the ratio of the reflected wave to the pulse pressure, and is an index for mainly evaluating the hardening of the arterial vascular wall of the central blood vessel. The AI value is the level a that is the amplitude difference between the lowest pulse wave position (that is, the start position of the pulse wave waveform) and the first peak, and the level that is the amplitude difference between the lowest pulse wave position and the second peak. For example, AI (%) = b / a X 100. An example in which the pulse waveform shown in FIG. 3 is detected will be described. In FIG. 3, the vertical axis represents the cuff pressure, and the horizontal axis represents the passage of time. The same applies to the diagrams showing the following waveforms. Level a shows the pressure value due to the ejection wave of blood due to the heartbeat, and level b shows the pressure value due to the reflected wave with respect to the ejection wave due to the heartbeat. The intensity and appearance time phase of this reflected wave change corresponding to the hardening of the blood vessel, and as the blood vessel hardens, the intensity of the reflected wave increases and the appearance time phase becomes earlier (shifts to the left). In other words, the higher the AI value, the more the blood vessel is cured. As a method for determining levels a and b, there is a method of performing an operation such as differentiation on the pulse waveform. That is, referring to FIG. 3, a differential curve obtained by fourth-order differentiation is superimposed on the detected waveform, and level a and level b can be determined using the amplitude corresponding to the extreme point position.

[0042] ET refers to the time from the opening point of the aortic valve to the closing point, and is an index related to cardiac contraction force, stroke volume, outflow tract resistance, and peripheral resistance. The ET value is obtained from the maximum point of the second derivative waveform of the waveform representing one beat of the pulse wave to the second maximum point. For example, when the pulse waveform shown in FIG. 4 is detected, the length c of the second derivative waveform up to the maximum point PA force and the second maximum point PB represents the ET value. The maximum point PA of the second derivative waveform is the rising point of the pulse wave, and the second maximum point PB represents the point at which the aortic valve begins to close. In other words, the smaller the ET value, the lower the cardiac function [0043] The pulse wave period indicates a period of a pulse wave waveform in one section, and is used as an index for detecting the deformation of a time component of the waveform shape due to the influence of body movement, noise, or the like, or arrhythmia. One example of a method for determining the period of the pulse wave waveform is as follows. Referring to Fig. 5, when the local maximum value of the original waveform is detected, the local minimum point (PB point) corresponding to the previous rising zero crossing point to the previous local minimum rising point (PB point) (PB point) Referring to the original waveform up to (PA point), if it is confirmed that the maximum point (PP point) exists between them and the PA point is the minimum value between the PA point and the PP point, PA The point is determined as the “rising point”, in other words, the “pulse wave start point” of one beat. Then, it is determined as the period of the pulse wave waveform of the length D force from the PA point to the PB point.

[0044] Further, the rising sharpness MSP is a parameter that significantly represents the pressing force of the pressing cuff 18 and evaluates the distortion of the waveform. When the rising point of the output change of the semiconductor pressure sensor 19 is sharp, it indicates that the pressing force of the pressing cuff 18 is appropriate, and when the rising point approaches a flat change, the pressing force of the pressing cuff 18 is inappropriate. It is shown that the subcutaneous artery at the measurement site is pressurized more than necessary. Rising sharpness MSP is a numerical value of the sharpness of the rising point of the output change of the semiconductor pressure sensor 19. For example, referring to FIG. 6, let Ta be the time interval between two points that are 10% higher than the maximum amplitude from the amplitude at a predetermined breakpoint (TDIA) set around the rising position of the pulse waveform. If the time from the end point between two points to the peak of the pulse waveform is defined as Tb, the rising sharpness MSP can be obtained by the ratio of Ta and Tb, for example, MSP = Ta / Tb.

[0045] The baseline fluctuation rate represents the fluctuation of the baseline of the adjacent pulse wave waveform, and is an index for detecting the deformation of the amplitude component of the waveform shape due to the influence of body movement, noise, and the like. The baseline fluctuation rate is obtained by the ratio of the maximum value of the pulse wave and the difference between the “pulse wave start point” and the “pulse wave start point” of the next pulse wave. For example, if the pulse wave waveform shown in Fig. 6 is detected, the pressure difference between the pulse wave start point S1 and the pulse wave maximum point S2 is A, and the pulse wave start point S1 and the next pulse wave start point S3 Assuming that the pressure difference with B is B, the baseline fluctuation rate (%) = 8 / eight 100. The “pulse wave start point” can be determined by a method similar to the method described in the description of the pulse period.

[0046] The feature amount storage unit 107 stores the five types of feature amounts calculated for each pulse wave. . Note that the types of feature amounts calculated by the feature amount calculation unit 105 are not limited to the above five types, and may be at least two of them. Further, other feature amounts may be calculated. Other feature quantities include, for example, the time interval between the maximum traveling wave point and the maximum reflected wave point, which is an index for evaluating the temporal component of the reflected wave and the degree of arterial vessel wall hardening, and the rising point of the traveling wave A well-known index such as TR (Traveling time to Reflected wave) representing the time interval from the rising point of the reflected wave can be mentioned. In addition, any feature quantity useful for diagnosis using the pulse wave waveform can be adopted.

Referring to FIG. 7, representative waveform search processing section 109 has feature quantity reading section 1091 for reading out the feature quantity stored from feature quantity storage section 107 for each pulse wave waveform, and for each type of feature quantity. An average calculation unit 1092 that calculates an average value and sets it as an index, a difference calculation unit 1093 that calculates a difference of the average value for each feature quantity as a relationship with the above index, and the calculated difference value as the type of feature quantity A normalization processing unit 1094 for normalization according to the above, a coefficient determination unit 1095 for determining a weighting factor for determining a weighting factor for the type of feature amount to be considered when extracting a representative waveform, and the normalized difference value A weighting processing unit 1096 for weighting each pulse waveform using the determined weighting factor, an addition processing unit 1097 for calculating the sum of the weighted difference values for each pulse waveform, and a difference for each pulse wave waveform Determine the representative waveform from the above sum of values. A representative waveform determining unit 1098 is included.

[0048] The weighting here refers to the importance of the feature quantity used when extracting the representative waveform using a plurality of types of feature quantities as described above. For the feature quantity corresponding to the extraction, the value is used as it is, and the weighting coefficient is set to 0.5. By using 50% of the value for the feature quantity corresponding to the representative waveform extraction, the weighting coefficient is used. By setting to 0, it is possible to suppress the use of the corresponding feature value when extracting the representative waveform. For this reason, by setting a combination of weighting factors here, it is possible to extract a representative waveform using only one feature quantity (for example, AI value) out of a plurality of calculated feature quantities. It is also possible to extract a representative waveform by using all of the plurality of calculated feature amounts equally. In addition, according to a certain purpose, it is possible to extract a representative waveform with particular emphasis on a specific feature amount among a plurality of calculated feature amounts and reducing the importance of other feature amounts. Such a weight coefficient is determined by the coefficient determination unit 1095. Is done. The following methods can be used as a method for determining the weighting factor by the coefficient determination unit 1095.

[0049] As a premise of the first method of determining the weighting coefficient in the coefficient determination unit 1095, the operation unit 24 is provided with a specific button. Alternatively, the setting screen is displayed on the display unit 25. Users such as doctors can select the weights (numbers, importance (small, medium, large) selection, etc.) corresponding to each feature according to the purpose of diagnosis by operating the specific buttons or following the setting screen. To input. The coefficient determination unit 1095 receives operation signals from these operations and determines a weight coefficient for each feature quantity.

[0050] As a second method, the input information capability of the user such as a doctor such as the medical history of the subject and the purpose of diagnosis, which is input prior to the main measurement and stored in a predetermined area such as RAMI 3, is used. A method of automatically setting the coefficient. Another example is a method of automatically setting a weighting factor of a feature amount suitable for a symptom having a high incidence by age and gender from input information on the subject such as age and gender. When automatically setting the weighting factor, the relationship between the input information and the weighting factor for each feature amount is stored in advance in a storage area such as ROM 12, and the coefficient determination unit 1095 stores the above-described storage as necessary. This is determined by referring to the area and reading the corresponding weighting factor.

FIG. 8 is a flowchart showing processing for displaying a representative waveform after measuring a pulse wave in the pulse wave detection device according to the present exemplary embodiment. The processing shown in the flowchart of FIG. 8 is realized by the CPU 11 accessing the ROM 12, reading the program, developing it on the RAM 13, executing it, and controlling each function shown in FIGS.

Referring to FIG. 8, first, pulse wave measurement value acquisition unit 101 acquires a pressure signal from semiconductor pressure sensor 19 to measure a pulse wave (step S 1). The pulse wave measured in step SI is divided into pulse waves by the pulse wave delimiter processing unit 103 (step S3), and the feature amount calculating unit 105 The above five types of feature values are calculated by this method (step S5). The processing in steps S1 to S5 may be performed in parallel with pulse wave measurement. In that case, the feature amount calculation in step S5 is performed for each pulse wave waveform until the pulse wave measurement is completed. When the measurement is completed, the representative waveform search processing unit 109 performs the above five types of pulse wave waveforms calculated in step S5. A representative waveform is searched based on the feature amount, and a corresponding pulse waveform is extracted (step S7). The representative waveform display control unit 111 generates and displays a display signal for displaying the pulse waveform extracted as the representative waveform in step S7 together with other calculated values on the display unit 25 as a measurement result (step S9). .

[0053] An example of the pulse wave segmentation process executed in step S3 is shown in the flowchart of FIG. In the following, this process is described as an analysis process after the channel of the multiplexer 20 is fixed.

[0054] Referring to FIG. 9, when it is first detected that a pressure signal is transmitted from the semiconductor pressure sensor 19 (S301), the mano-replexer 20 is fixed according to the control signal of the CPU 11 force and the like. Select the sensor signal from the sensor element corresponding to the channel and input it to amplifier 21.

[0055] The input pressure signal is amplified to a predetermined frequency in the amplifier 21 (S303), and analog filter processing is performed in the characteristic variable filter 22 (S305).

[0056] At this time, the variable characteristic filter 22 blocks signal components that are 1/2 or more of the sampling frequency. If the sampling frequency is 500 Hz, for example, signal components with a frequency exceeding 250 Hz are blocked. That is, here, the signal component corresponding to the pulse waveform outside the standard range is removed.

[0057] The pressure signal that has passed through the variable characteristic filter 22 is digitized by the A / D converter 23 (S307), and is subjected to digital filtering to extract a predetermined range of frequencies for the purpose of noise removal and the like. (S309). Then, the A / D converter 23 transfers the digitized pressure signal to the CPU 11.

[0058] With respect to the pressure signal received from the A / D converter 23, the pulse wave segmentation processing unit 103 performs N-order differentiation on the pulse wave waveform obtained from the pressure signal by taking the difference of each data (S311). Then, the pulse wave division processing unit 103 extracts the first pulse wave waveform by dividing the pulse wave waveform based on the differential result of step S311 (S313).

FIG. 10 is a flowchart showing an example of the representative waveform search process executed by the representative waveform search processor 109 in step S7. Referring to FIG. 7, the average value is calculated for each type of feature value in the average calculation unit 1092 for the feature value read from the feature value storage unit 107 by the feature value reading unit 1091 (Step S70 Do, then the difference calculation unit In 1093, the difference from the above average value is calculated for each feature amount (step S703), and the difference calculated in step S703 is normalized in the normalization processing unit 1094 (step S705). The normalization process is a general normalization process and is not limited to a specific process in the present invention, and the difference normalized in step S705 is weighted by the coefficient determination unit 1095 in the weighting processing unit 1096. Weighting is performed using a coefficient (step S707), specifically, the weighting factor determined for each type of feature amount by the coefficient determination unit 10 95 is used as the feature amount with respect to the difference normalized in step S705. Seeds The difference between the above five types of feature values for each waveform of 1 柏 that has been weighted is added by the addition processing unit 1097 to calculate the sum of the differences for each waveform (step S709). Then, in the representative waveform determining unit 1098, the waveform having the smallest sum, that is, the waveform closest to the average as a whole for all the above five types of feature values is determined as the representative waveform and extracted (step). S711) ₀

FIG. 11 is a diagram for explaining the display contents on the display screen displayed by the display signal from the representative waveform display control unit 111 in step S9. Referring to Fig. 11, on the display screen, the pulse waveform 200 for the beat extracted as a representative waveform and the average value of blood pressure and AI are valid as the pulse wave measurement result 300. The average value of the blood pressure and the average value of the AI values are displayed for all the waveforms! In addition, the display screen displays information 201 relating to the process of extracting the representative waveform. The display 201 indicates the force corresponding to the number of the effective waveform of the extracted representative waveform, and the representative screen. It is preferable to include marks 203, 205, 207 representing the characteristic positions and values of the feature values used when extracting the waveform on the waveform, and a display 209 of the feature values. In the example of Fig. 11, AI value, ET value, and pulse wave period are used among the above five types of features when extracting the representative waveform, and mark 203 is used to calculate the AI value. The above-mentioned level a and level 1 are represented, the mark 205 represents the above-mentioned length c corresponding to ET, and the mark 207 represents the above-described pulse wave start point which is a pulse wave period. Furthermore, those specific numbers are Displayed in 209. In this way, by displaying information related to the process of extracting the representative waveform together with the representative waveform on the display screen, how the representative waveform is extracted, and whether the representative waveform is the entire waveform. It is possible to find out which position (the number) is among them and provide useful information for diagnosis.

[0061] In the example of FIG. 11, as the result 300 of the pulse wave measurement, the average value of blood pressure and the average value of AI values obtained for all valid waveforms are displayed! You can also display the blood pressure and AI values obtained! By displaying in this way, the representative waveform and the numerical value related to the representative waveform can be obtained from the display screen.

In the above embodiment, an average value for each type of feature quantity is used as the index, and a difference from the average value is calculated for each feature quantity as a relationship with the index, and all the feature quantities are calculated. As a whole, it is closest to the average value! /, And the waveform is determined as the representative waveform! /. In this way, a stable waveform close to the average as a whole can be used as the representative waveform.

[0063] The above index is not limited to an average value, and it is possible to use other straight lines. Similar to the case where the average value is used as an index, the median value or the mode value can be used as another index for the purpose of making a stable waveform a representative waveform.

More specifically, when the median is used as an index, the representative waveform search processing unit 109 normalizes the feature amount read from the feature amount storage unit 107, and then performs the above-described multiple types of one waveform. The normalized feature values are added and arranged in order of magnitude, and the waveform with the magnitude order in the center is determined as the representative waveform.

[0065] When the mode value is used as an index, the representative waveform search processing unit 109 is a feature amount storage unit.

After normalizing the feature values read out from 107, check the distribution by adding the above-mentioned multiple types of normalized feature values to the graph and plotting them on a graph, etc., and the sum of the feature values is the largest. A waveform located in the existing range is determined as a representative waveform.

[0066] Further, as another specific example of the index, an arbitrary threshold value for an arbitrary type of feature amount among the plurality of types of feature amounts may be set as the index. For example, a threshold value with an AI value of 10% or more can be set as an index. In this case, the representative waveform search processing unit 109 determines a waveform having a read feature amount equal to or greater than the threshold, a waveform equal to or less than the threshold, or a waveform within a predetermined range from the threshold as the representative waveform. Use such indicators Thus, a characteristic waveform useful for a specific diagnosis can be extracted by selecting a characteristic waveform.

[0067] As another specific example of the index, a minimum value or a maximum value can be used. For example, the minimum baseline fluctuation rate or the maximum AI value can be used as an indicator. In this case, the representative waveform search processing unit 109 determines a waveform whose read feature amount satisfies the condition as a representative waveform. By using such an index, it is possible to select a stable waveform or to select a characteristic waveform.

[0068] Furthermore, the representative waveform search processing unit 109 may determine which value is used as an index. As a method for determining the index, a method similar to the method for determining the weighting factor described above can be used. That is, as a first method, it may be determined according to an operation signal by an operator's operation, or input information of a user such as a doctor such as a medical history of a subject or a diagnosis purpose, or input regarding a subject such as age and gender. It may be set automatically from the information.

[0069] [Modification]

In the pulse wave detection device according to the modification, the representative waveform search processing unit 109 performs the representative waveform search processing in two stages.

FIG. 12 is a flowchart showing an example of representative waveform search processing executed in representative waveform search processing unit 109 in step S 7 in the pulse wave detection device according to the modification.

[0071] Referring to FIG. 12, in the pulse wave detection device that is powerful as a modified example, as a first-stage search, the average value of each type of feature amount is used as the first-stage index as described above. Steps S701 to S709 are performed, and in step S712, the representative waveform determination unit 1098 searches for a waveform in which the sum of the difference of feature values for one beat waveform is within a predetermined range from the average value in the next second stage search. Are selected as target waveforms.

[0072] Further, in the modified example, the representative waveform determination unit 1098 performs a second-stage search on the target waveform selected in step S712, for example, using the maximum value of the AI value as the second-stage index. The waveform with the largest AI value is determined as the representative waveform from the target waveforms (step S713).

As described above, in the pulse wave detection device that is powerful as a modification, a representative waveform is searched and extracted from a plurality of pulse wave waveforms in two or more stages. In multi-stage search, By selecting, as a target waveform, a waveform that matches the condition in the relationship with the index from a plurality of pulse wave waveforms, and extracting a waveform that matches the condition in the relationship with the further index as a representative waveform. It is possible to extract a waveform that is more suited to the purpose, such as a pulse waveform with a small and prominent AI value, as a representative waveform. This makes it possible to extract representative waveforms that are more useful for diagnosis.

[0074] The biological signal waveform extraction method, which is a method for extracting the representative waveform in the above-described pulse wave detection device, is not limited to the analysis of the pulse wave, and is generated by the first waveform and expansion caused by the heart contraction. Any other biological signal waveform obtained by synthesizing the second waveform can be used in the process of extracting the representative waveform.

Furthermore, it is possible to provide a program for realizing the biological signal waveform extraction method in the pulse wave detection device according to the present embodiment. Such a program can be recorded on a computer-readable recording medium such as a flexible disk, CD-ROM (Compact Disk-Read Only Memory), ROM, RAM, or memory card attached to the computer as a program product. It can also be provided. Alternatively, the program can be provided by being recorded on a recording medium such as a hard disk built in the computer. The program can also be provided by downloading via the network.

[0076] The program according to the present invention is a program module provided as a part of a computer operation system (OS) that calls necessary modules in a predetermined arrangement at a predetermined timing to execute processing. It may be. In that case, the program itself does not include the module, and the process is executed in cooperation with the OS. A program that does not include such a module can also be included in the program that is effective in the present invention.

[0077] Further, the program according to the present invention may be provided by being incorporated in a part of another program such as a normal pulse wave measurement program. Even in this case, the program itself does not include the module included in the other program, and the process is executed in cooperation with the other program. Programs incorporated in such other programs can also be included in the programs that are effective in the present invention.

[0078] The provided program product is installed in a program storage unit such as a hard disk. To be executed. The program product includes the program itself and a recording medium on which the program is recorded.

It should be considered that the embodiments disclosed herein are illustrative and non-restrictive in every respect. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

Claims

The scope of the claims

[1] A detection unit (103) for detecting a plurality of waveforms obtained from a biological signal;

A calculation unit (105) that calculates a first feature value and a second feature value from each of the plurality of waveforms;

The relationship of the first feature quantity to the first index obtained from the plurality of first feature quantities calculated from the plurality of waveforms, and the plurality of second feature quantities calculated from the plurality of waveforms. A search unit (109) for performing a process for searching for a representative waveform from among the plurality of waveforms based on a relationship of the second feature quantity with respect to the second index obtained, and outputting the representative waveform A medical measuring instrument including an output processing unit (111) for performing processing for the purpose.

[2] The waveform obtained from the biological signal is a pulse waveform,

The first feature quantity and the second feature quantity include at least one of an AI (Augmentation Index) value, a pulse period, a baseline fluctuation rate, a sharpness, and an ET (Ejection Time) value. Medical measuring instrument according to paragraph 1 of the scope.

[3] The search unit

A coefficient setting unit (1095) for setting a weight coefficient for each of the first feature quantity and the second feature quantity in one waveform;

For each of the relationship between the first feature quantity with respect to the first index and the relationship with the second feature quantity with respect to the second index, the representative waveform is determined in consideration of the weighting factor. The medical measuring instrument according to claim 1, further comprising a determining unit (1098) for determining.

[4] The search unit includes:

Average calculation for calculating an average value of the first feature quantity for the plurality of waveforms as the first index and an average value of the second feature quantity for the plurality of waveforms as the second index. Part (1092),

For each of the plurality of waveforms, a first difference that is a difference between the first index and the first feature quantity, and a second difference that is a difference between the second index and the second feature quantity. A difference calculating unit (1093) for calculating

A normalization processing unit (1094) for normalizing the calculated first difference and the second difference When,

The medical measuring instrument according to claim 1, further comprising: a determining unit (1098) that determines the representative waveform based on the normalized first difference and the second difference.

[5] The search unit includes:

A weighting processing unit (1096) that multiplies each of the normalized first difference and the second difference by the weighting factor;

5. The medical measuring instrument according to claim 4, wherein the determining unit determines the representative waveform based on the weighted first difference and the second difference.

[6] The search unit includes:

For each of the plurality of waveforms, further includes an addition processing unit (1097) for adding the weighted first difference and the second difference,

6. The medical measuring instrument according to claim 5, wherein the determining unit determines a waveform having the smallest difference as the representative waveform.

[7] The search process in the search unit is:

A first step of selecting a target waveform from the plurality of waveforms based on a relationship between each of the first feature amount and the second feature amount of the plurality of waveforms and a first step index; Search processing (S701 to S712)

Searching the representative waveform from the target waveform based on the relationship between each of the first feature quantity and the second feature quantity of the target waveform and the second stage index Second stage search The medical measuring instrument according to claim 1, further comprising a process (S713).

[8] The output processing unit performs a process for displaying, together with the representative waveform, a value used in the process of detecting the representative waveform in the search unit or information specifying the representative waveform. The medical measuring instrument according to item 1.

[9] The first index and the second index are both an average value, a median value, a mode value, and a maximum value of the corresponding feature quantities of the plurality of waveforms. The medical measuring instrument according to claim 1, wherein the medical measuring instrument is one of a minimum value and an arbitrary threshold value.

[10] The medical measuring instrument according to claim 1, further comprising an index setting unit (109) for setting the first index and the second index.

[11] A method for extracting a representative waveform from waveforms obtained from biological signals,

Obtaining a plurality of consecutive waveforms (S1);

A step (S3) of dividing a unit waveform from the plurality of continuous waveforms;

A step (S5) for calculating a first feature value and a second feature value from each of the plurality of unit waveforms;

The relationship of the first feature quantity to the first index obtained from the plurality of first feature quantities calculated from the plurality of waveforms, and the plurality of second feature quantities calculated from the plurality of waveforms. A step (S7) of extracting a representative waveform from the plurality of unit waveforms based on the relationship of the second feature quantity to the second index obtained from

A biological signal waveform extraction method comprising: (S9) outputting the representative waveform.

[12] A medium storing a program for causing a computer to execute processing for extracting a representative waveform from waveforms obtained from a biological signal,

Obtaining a plurality of consecutive waveforms (S1);

A computer-readable medium recording a biological signal waveform extraction program for executing the step (S9) of outputting the representative waveform.