KR20180105425A - Biological signal-based pain depth classification apparatus and thereof method - Google Patents

Abstract

The present invention relates to a biological signal based pain depth classifying device and a method thereof capable of classifying pain depth by using an electrocardiogram (ECG) and photoplethysmograph (PPG) measuring device. The present invention extracts energy and a zero-crossing rate which are main features of an ECG signal or a PPG signal depending on reduction of blood flow and a frequent pulse when a sympathetic nervous system is excited by pain impulses from the ECG signal or the PPG signal, normalizes the same, and shows a pain depth index by using a value calculated by applying weight values to the normalized energy and the zero-crossing rate through a sigmoid function, thereby simply classifying the pain depth with high accuracy. The PPG signal or the ECG signal detected from a PPG sensor unit or an ECG sensor unit is converted into a digital signal through an A/D converting unit, and is transmitted to a calculation processing unit. The calculation processing unit detects the pain depth index from the received PPG signal or the ECG signal. The calculation processing unit detects the energy and the zero-crossing rate from the PPG signal or the ECG signal to normalize the same, calculates a sigmoid function of the energy and a sigmoid function of the zero-crossing function, adds a value of multiplying a predetermined energy weighted value by the sigmoid function of the energy and a value of multiplying a predetermined zero-crossing rate weighted value by the sigmoid function of the zero-crossing rate, and outputs the same as the pain depth index.

Description

[0001] The present invention relates to a biological signal-based pain depth classification apparatus and method,

The present invention relates to a bio-signal-based pain intensity classifying apparatus and method for classifying pain intensity using an electrocardiogram (ECG), a photoplethysmograph (PPG) measuring apparatus, and more particularly, When the sympathetic nervous system is excited by stimulation, the energy and zero crossing rates, which are the main characteristics of the ECG signal or the PPG signal, are extracted and normalized according to the reduced blood flow and tachycardia, and the normalized energy and zero crossing rate The present invention relates to a bio-signal-based pain intensity classifying apparatus and method capable of performing simple and highly accurate classification of pain intensity by showing a pain intensity index through a sigmoid function.

Generally, depending on the cause of the pain, the pain can be divided into nociceptive pain, neurogenic pain, non-tempered pain or psychogenic pain. It is divided into somatic pain, visceral pain, and central pain according to the site of occurrence, and the somatic tub is divided into the superficial deep vein tubules.

It also classifies pain as either a fast pain or a slow pain.

The depth of pain is a kind of pain that occurs at a certain depth of a predetermined part of the body, that is, the pain is classified according to the site of occurrence. It is generally difficult to classify pain because it depends on the patient's subjectivity. At present, there is no way to classify the depth of pain as an overview, numerical, or exact. Therefore, there is a need for an objective, numerical, simple, and highly accurate bio-signal-based pain intensity classification apparatus and method for determining the depth of pain.

Prior art is disclosed in Korean Patent Laid-Open No. 10-2006-0017510, which discloses a method for screening, diagnosing, and classifying the risk of a clinical onset such as acute coronary syndrome in patients having symptoms and symptoms of a suspected heart disease will be. The present invention detects ECG and, in addition, extracts a substance stream (sample) of a patient's body fluid or the like, obtains sample data using in-vivo diagnostic analysis, and uses them to diagnose and classify risk. Particularly, in the present invention, an ECG having no ST segment movement or T-wave change in an ECG parameter is regarded as " negative ", and a " positive " ECG is an ECG having a ST segment inhibition or increase of 2 mV or more, (Eg left ventricular block, constant rhythm, extensive pathological Q-wave, and / or continuous ST intercept increase after previous AMI measurement) are considered "negative".

Korean Patent Laid-Open No. 10-2006-0017510 analyzes the risk due to pain using an electrocardiogram, but the depth of pain can not be known only by analysis of ST segment and T-wave in ECG parameters.

In another prior art, Korean Patent No. 10-1000761 relates to a medical device for measuring the pain and consciousness level of a surgical patient, and more particularly to a medical device for accurately measuring the degree of surgery intoxication / consciousness of a surgical patient with a minimum error The present invention relates to an apparatus and method for measuring a pain / consciousness level of a surgical patient so as to prevent the pain of a surgical patient. The device for measuring a pain / consciousness level of a surgical patient according to the present invention includes a stimulus signal generator for applying a stimulus signal to a first part of a surgical patient and a second part for detecting the electric potential of the stimulation signal at a second part of the patient A bio-signal detector for amplifying the detected potential and outputting a somatosensory evoked potential (SSEP) signal; a bio-signal conversion processor for converting the SSEP signal to generate digital potential data; index) to generate and display pain / awareness level information of the surgical patient.

In the case of Korean Patent No. 10-1000761, a stimulus is applied to a predetermined region to detect the degree of pain, and the induced potential generated in the other region according to the stimulus is detected to determine the pain along the region. In this case, however, it is necessary to use a stimulation signal generating unit or the like which directly applies a stimulus. Further, since the biological signal detecting unit is located at a predetermined position, there is a danger that only a person having a high level of expertise can use it. In addition, it is considerably troublesome to attach the stimulating electrode and the signal detecting electrode to the skin of the patient.

It is an object of the present invention to provide an apparatus and a method for extracting energy or zero crossing from ECG or PPG signals, which is a main characteristic of an ECG signal or PPG signal according to reduction of blood flow and tachycardia when a sympathetic nervous system is excited by a pain stimulus, rate, and the normalized energy and the zero crossing rate are weighted and the calculated value is represented by the sigmoid function. Thus, it is possible to classify the depth of pain based on the bio-signal, which is simple and highly accurate, Classification apparatus and method.

In order to solve the above-described problems, the present invention provides a blood pressure monitor which is configured to convert a pulse wave signal or electrocardiogram signal detected from a pulse wave (PPG) sensor or an ECG sensor part into a digital signal through an A / And the arithmetic processing unit detects a pain intensity index from the received pulse wave signal or electrocardiogram signal, wherein the arithmetic processing unit calculates energy energy and zero crossing rate from the pulse wave signal or the electrocardiogram signal The zero-crossing rate is detected and normalized, and the pain intensity index is detected using a sigmoid function of normalized energy and zero crossing rate.

 The arithmetic processing section detects and normalizes the energy and zero crossing rate from the pulse wave signal or the electrocardiogram signal, obtains a sigmoid function of the sigmoid function and the zero crossing rate of energy, multiplies the sigmoid function of energy by a predetermined energy weight And the zero-crossing rate weight multiplied by the sigmoid function of the zero crossing rate, and outputs it as the pain intensity index.

The arithmetic processing unit performs filtering to remove noise from the pulse wave signal or electrocardiogram signal received from the A / D converting unit, and outputs a pulse wave signal or an electrocardiogram signal for a predetermined energy detection reference time period in the filtered pulse wave signal or electrocardiogram signal To obtain the energy of the pulse wave signal per reference time or the electrocardiogram signal per reference time.

The arithmetic processing unit performs filtering to remove noise from the pulse wave signal or electrocardiogram signal received from the A / D converting unit, and calculates a maximum value, a minimum value, and a maximum value in a filtered pulse wave signal or electrocardiogram A point time and a minimum point time as a minimum value are detected and the number of times the sign of the pulse wave signal or the electrocardiogram signal changes during the zero crossing reference time period is obtained as a zero crossing rate.

The zero-crossing reference time interval is from the previous time point to the maximum point time point by 15 seconds in the future from the maximum point time point.

The operation processing unit normalizes the energy divided by the standard Gaussian distribution of energy, normalizes the zero crossing rate by dividing the standard normal distribution of the zero crossing rate.

In order to obtain a standard normal distribution of energy, the arithmetic processing unit performs filtering to remove noise from the pulse wave signal or electrocardiogram signal received from the A / D converter during the standard normal distribution detection time interval at the beginning of signal detection, The pulse wave signal or the electrocardiogram signal is obtained by summing the squares of the pulse wave signal or the electrocardiogram signal during the energy detection reference time period to obtain energies of the pulse wave signal per reference time or the electrocardiogram signal per reference time, In energies, the standard deviation and averages are obtained to obtain the standard normal distribution of energy.

In order to obtain a standard normal distribution of the zero crossing rate, the calculation processing section performs filtering to remove noise from the pulse wave signal or electrocardiogram signal received from the A / D conversion section during the standard normal distribution detection time interval at the beginning of signal detection, The zero crossing rate, which is the number of times the sign of the pulse wave signal or electrocardiogram signal changes during the zero crossing reference time interval, is obtained from the received pulse wave signal or electrocardiogram signal, and the standard deviation and the average are obtained to obtain the standard normal distribution of the zero crossing rate.

The arithmetic processing unit performs filtering to remove noise from the pulse wave signal or electrocardiogram signal received from the A / D converter during the standard normal distribution detection time interval at the beginning of signal detection, and calculates a standard deviation And an average are obtained to obtain a standard normal distribution of the pulse wave signal or electrocardiogram signal and the standard normal distribution of the obtained pulse wave signal or electrocardiogram signal is used instead of the standard normal distribution of the standard normal distribution and the zero crossing ratio of energy.

In the present invention, a pulse wave signal or electrocardiogram signal detected from a pulse wave (PPG) sensor or an electrocardiogram (ECG) sensor unit is converted into a digital signal through an A / D conversion unit and transmitted to an arithmetic processing unit, And a processing unit for detecting a pain intensity index from the received pulse wave signal or electrocardiogram signal, wherein the operation processing unit calculates energy and zero-crossing ratio from a pulse wave signal or an electrocardiogram signal, crossing rate, and normalized, and using the sigmoid function of the normalized energy and zero crossing ratio, the pain intensity index is detected.

In the present invention, a pulse wave signal or electrocardiogram signal detected from a pulse wave (PPG) sensor or an electrocardiogram (ECG) sensor unit is converted into a digital signal through an A / D conversion unit and transmitted to an arithmetic processing unit, The processor is configured to perform filtering using a reshape filter to remove noise from the received pulse wave signal or electrocardiogram signal and to detect a pain intensity index from the filtered pulse wave signal or electrocardiogram signal, In the apparatus driving method, the operation processing unit calculates the energy of the pulse wave signal per reference time or the electrocardiogram signal per reference time by summing squares of the filtered pulse wave signal or the electrocardiogram signal during the energy detection reference time period in the filtered pulse wave signal or electrocardiogram signal , An energy calculating step; The operation processing unit detects a minimum point time which is a time when the maximum value, the minimum value, and the maximum point time, which is the time when the maximum value is the minimum value, in the filtered pulse wave signal or the electrocardiogram signal, A zero crossing rate calculating step of obtaining, as a zero crossing rate, the number of times the sign of the pulse wave signal or the electrocardiogram signal changes during the zero crossing reference time period; The energy obtained in the energy computing step is normalized by dividing the energy obtained in the energy computing step by the standard normal distribution of energy and the normalized normalizing of the zero crossing rate with respect to the zero crossing rate obtained in the zero crossing rate computing step, Normalizing step of dividing and normalizing the zero crossing rate obtained in step

 The operation processing unit obtains a sigmoid function of the energy sigmoid function and the zero crossing ratio by applying the normalized energy and the normalized zero crossing ratio output from the normalization step to the sigmoid function, And a sigmoid function calculating step of summing a value obtained by multiplying the weight function by a predetermined energy weight and a value obtained by multiplying a sigmoid function of the zero crossing ratio by a predetermined zero crossing rate weight, .

The reshape filter is a filter that allows a frequency of 0.5 Hz to 8 Hz to pass.

In the sigmoid function arithmetic step, the pain intensity index,

w ₁ / (1 + exp (-αE)) + w ₂ / (1 + exp (-αZ)

(Where E is the normalized energy, Z is the normalized zero crossing rate, a is the slope parameter, w ₁ is the energy weight, and w ₂ is the zero crossing rate weight).

The slope parameter α is 1.5 to 1.8, and the sum of the energy weight w1 and the zero crossing weight w2 is 1.

The pain intensity index has a value between 0 and 1.

The present invention also provides a computer-readable recording medium having recorded thereon a program for implementing a method of operating a bio-signal based pain intensity classifier according to the present invention.

According to the bio-signal-based pain intensity classification apparatus and method of the present invention, when the sympathetic nervous system is excited by the pain stimulus from the ECG or PPG signal, energy, zero (0) It is possible to classify simple and highly accurate pain depth classification by extracting and normalizing the zero crossing rate, and expressing the pain intensity index through the sigmoid function which is calculated by weighting the normalized energy and zero crossing rate .

Further, the present invention can be used even for a beginner, and there is no danger such as stimulation.

1 is a block diagram for explaining a schematic configuration of a bio-signal-based pain intensity classifying apparatus according to the present invention.
FIG. 2 is a flowchart for explaining a pain intensity classification process performed by the arithmetic processing unit of FIG. 1;

Hereinafter, a light signal method of a bio-signal based pain intensity classifier according to the present invention will be described with reference to the accompanying drawings.

FIG. 1 is a block diagram showing a schematic structure of a bio-signal-based pain intensity classifying apparatus according to the present invention, including a block road bio-signal detector 100, a biological signal preprocessor 150, an A / D converter 180, 200, a display unit 210, and a memory unit 220.

The bio-signal detection unit 100 includes a pulse wave sensor (PPG) sensor unit 110 or an electrocardiogram (ECG) sensor unit and detects a pulse wave (pulse wave pulse) or electrocardiogram.

The pulse wave (PPG) sensor unit 110 includes a light emitting unit (not shown) and a light receiving unit (not shown) and emits light to a blood vessel (blood flow) ) Or receives light reflected from a blood vessel (blood flow), converts it into an electric signal, and outputs the electric signal. The output signal represents an optical pulse wave of a blood vessel.

The light emitting unit (not shown) may be composed of red light (wavelength: 650 to 750 nm) and light emitting diode (LED) having infrared light (wavelength: 850 to 1000 nm) and emits red light or infrared light to the blood vessel.

The light receiving unit (not shown) comprises a photodiode or a light receiving sensor (light sensor), and receives light reflected or transmitted through the blood vessel, that is, red light or infrared light, and outputs it as an electric signal.

The pulse wave (PPG) sensor unit 110 may be mounted on any part of the blood vessel of the human body (finger, toe, ear ball, upper body, lower body, chest, hands, feet, etc.). And is preferably mounted on a finger.

The electrocardiogram (ECG) sensor unit 120 includes an electrocardiogram electrode (not shown) and a reference electrode (or ground electrode) (not shown), and detects an electrocardiogram signal. In some cases, the ECG electrode may be two. The electrocardiogram (ECG) sensor unit 120 may be mounted on the wrist, ankle, chest, upper body, lower body, and the like.

The biological signal preprocessing unit 150 performs preprocessing for removing noise from the electrocardiogram signal or the pulse wave signal and amplifying the noise.

The pulse wave (PPG) signal preprocessing unit 160 amplifies the pulse wave signal detected by the pulse wave sensor (PPG) sensor unit 110 and performs preprocessing to remove noise.

The ECG signal preprocessing unit 170 amplifies the pulse wave signal detected by the electrocardiogram (ECG) sensor unit 120 and performs preprocessing to remove noise.

The A / D conversion unit 180 converts the pulse wave signal from the pulse wave (PPG) signal preprocessing unit 160 into a digital signal and converts the electrocardiogram signal from the ECG signal preprocessing unit 170 into a digital signal .

The arithmetic processing unit 200 performs filtering to remove noise from an electrocardiogram (ECG) signal or a pulse wave signal, and generates energy, zero crossing rate, or the like from a filtered ECG signal or a pulse wave signal. ) Is extracted and normalized, and a value obtained by weighting the normalized energy and the zero crossing ratio is output through a sigmoid function to express the pain intensity index.

In other words, the arithmetic processing unit 200 receives the pulse wave (PPG) signal or the electrocardiogram (ECG) signal from the A / D converter 180 and outputs the received pulse wave (PPG) Filtering is performed through a filter (HPF) and a digital low-pass filter (LPF), and energy and a zero crossing rate are obtained from a filtered pulse wave (PPG) signal or an electrocardiogram (ECG) signal.

In order to obtain the energy, the arithmetic processing unit 200 adds the squares of a pulse wave (PPG) signal or an electrocardiogram (ECG) signal for a predetermined energy detection reference time period in a filtered pulse wave (PPG) signal or an electrocardiogram (PPG) signal per reference time or the ECG signal per reference time (S120).

In order to obtain the zero crossing rate, the arithmetic processing unit 200 calculates a maximum value, a minimum value, a maximum point time which is a time when a maximum value is obtained, and a minimum time which is a time when the pulse rate is the minimum value, in a filtered pulse wave (PPG) signal or an electrocardiogram (ECG) And the number of times that the number of times of passing the zero (0), that is, the number of times the sign of the pulse wave (PPG) signal or the electrocardiogram (ECG) signal is changed, is set as the zero crossing reference time interval based on the minimum point or the maximum point. It is obtained as a crossing rate. Here, the zero-crossing reference time interval can be set to a time from the point in time which is a predetermined time (for example, 15 seconds) of the maximum point time to the point of time of the maximum point time.

In order to normalize the energy, the operation processing unit 200 divides the energy into a standard Gaussian distribution of energy, normalizes the energy, and further, in order to normalize the zero crossing rate, The distribution is normalized by dividing the zero crossing rate. Here, the standard normal distribution of the energy normalization and the zero crossing ratio is the value detected and stored at the beginning of signal detection.

The arithmetic processing unit 200 obtains the sigmoid function value using the normalized energy and the zero crossing rate using the slope parameter, the energy weight and the zero crossing rate weight, and outputs the obtained sigmoid function value as the pain intensity value to the display unit 210 and the memory unit 220. [

Here, the operation processing unit 200 may be a computer.

FIG. 2 is a flowchart for explaining a pain intensity classification process performed by the arithmetic processing unit of FIG. 1;

In the signal reception step, the operation processing unit 200 receives a pulse wave (PPG) signal or an electrocardiogram (ECG) signal from the A / D conversion unit 180 (S110).

In the filtering step, a pulse wave (PPG) signal or an electrocardiogram (ECG) signal received in the signal receiving step is filtered through a digital high pass filter (HPF) and a digital low pass filter (LPF) And performs reshape filtering. At this time, the cut-off frequency of the digital high-pass filter HPF and the cut-off frequency of the digital high-pass filter HPF are set to a predetermined frequency, and the cutoff frequency of the digital high-pass filter HPF is the cut- Lt; / RTI > The reshape filter consists of a digital high-pass filter (HPF) and a digital low-pass filter (LPF). For example, the cut-off frequency of the digital high-pass filter (HPF) may be 0.5 Hz and the digital low-pass filter (LPF) may be a filter that passes 8 Hz, i.e., 0.5 Hz to 8 Hz.

(PPG) signal or an electrocardiogram (ECG) signal during the energy detection reference time period in a pulse wave (PPG) signal or an electrocardiogram (ECG) signal in a filtering step (S120) (PPG) signal or an ECG signal per reference time (S120).

Here, the energy detection reference time interval is a time interval set by the user at the initial stage of use or set at the time of factory shipment, and is a time interval for obtaining energy in the energy calculating step. The time interval is a time interval Is called a sliding window. That is, the energy detection reference time interval refers to a sliding window having a predetermined time interval based on the time of reception of a pulse wave (PPG) signal or an electrocardiogram (ECG) signal in the signal receiving step, Obtain energy from the movement window. For example, the energy detection reference time can be set to an interval of 10 seconds. That is, in this case, the energy from the time point of the current signal to the point of time of the current signal can be obtained in the energy calculation step, which is 10 seconds past from the present point of time.

In the zero crossing rate calculation step, a maximum point, a minimum value, a maximum point time which is a time when a maximum value is obtained, and a minimum point time which is a time when a minimum value is used in a pulse wave (PPG) signal or an electrocardiogram (ECG) And determines the number of times that the zero crossing, that is, the number of times the sign of the pulse wave (PPG) signal or the electrocardiogram (ECG) signal changes, during the zero crossing reference time interval based on the minimum point or the maximum point.

Here, the waveform of a pulse wave (PPG) signal or an electrocardiogram (ECG) signal has a hill and a valley, and a peak point is referred to as a maximum point in a portion representing a hill and a minimum point in a portion representing a valley is referred to as a minimum point , The amplitude value of the maximum point is the maximum value, the amplitude value of the minimum point is the minimum value, the time when the maximum value is the maximum point time, and the time when the minimum value is the minimum point time.

Here, the zero crossing reference time interval is a time interval for obtaining the zero crossing rate in the zero crossing rate calculation step, and this time interval is referred to as a sliding window since the time interval changes every time the zero crossing rate is obtained. That is, the zero-crossing reference time interval may be set to a predetermined time based on the minimum point time, or may be changed from the minimum point time to the maximum point time, Of the maximum point time, or until the point of time that is the predetermined point in time after the maximum point time from the point of time which is the predetermined time in the future of the maximum point time.

Preferably, the zero-crossing reference time interval may be from a time point between a predetermined time (for example, 15 seconds) and a maximum point time point in the future from the maximum point time.

In the present invention, the energy detection reference time and the zero crossing reference time can be extended up to 5 minutes through the sliding window in consideration of the heartbeat interval.

In the normalization step, the energy obtained in the energy calculation step is normalized by dividing the energy obtained in the energy calculation step with the standard normal distribution of energy, and the standard normal distribution of the zero crossing ratio , The zero crossing rate obtained in the zero crossing rate calculation step is divided and normalized (S160).

The standard normal distribution of the standard normal distribution and the zero crossing rate of energy is obtained during the standard normal distribution detection time interval at the initial stage of starting signal detection using the bio-signal based pain depth classifier 10.

In general, the standard normal distribution can be expressed as (X-μ) / σ. Where X is the energy signal or the zero crossing rate signal, σ is the standard deviation of the energy or zero crossing rate, and μ is the average of the energy or zero crossing rate.

Here, the standard normal distribution detection time interval is used to obtain a standard normal distribution of the standard normal distribution and the zero crossing rate of energy for a predetermined time from the initial point of time when the signal is detected using the bio-signal based pain classification device 10 (PPG) signal or an electrocardiogram (ECG) signal, for example, 10 seconds to 2 minutes after the reception of the first pulse wave (PPG) signal or the electrocardiogram (ECG) signal. In some cases, the standard normal distribution detection time interval can be adjusted in consideration of the energy detection reference time, the zero crossing reference time, and the heartbeat interval.

(PPG) signal or an electrocardiogram (ECG) signal for the standard normal distribution detection time interval at the beginning of signal detection for the standard normal distribution of the standard normal distribution of the energy and the zero crossing rate, D converter 180 and performs reshape filtering of the received pulse wave (PPG) signal or electrocardiogram (ECG) signal through a digital high-pass filter (HPF) and a digital low-pass filter (LPF). (PPG) signal or an electrocardiogram (ECG) signal per reference time in a filtered pulse wave (PPG) signal or an electrocardiogram (ECG) signal, (ECG) signal, and obtains a standard normal distribution and an average of the energies of the pulse wave per unit time (PPG) signal or the ECG signal per reference time to obtain a standard normal distribution of energy. In the filtered pulse wave (PPG) signal or the electrocardiogram (ECG) signal, the number of times that the pulse (PPG) signal or the electrocardiogram (ECG) signal is changed over the zero crossing reference time interval, The zero crossing rate is obtained, and the standard deviation and the average are obtained to obtain the standard normal distribution of the zero crossing rate.

The standard normal distribution of the standard normal distribution and the zero crossing ratio of the energy may be a pulse wave (PPG) signal or an electrocardiogram (ECG) signal for the normal normal distribution detection time interval at the beginning of signal detection, (PPG) signal or electrocardiogram (ECG) signal through a digital high-pass filter (HPF) and a digital low-pass filter (LPF), performs reshape filtering on the received pulse wave (PPG) (PPG) signal or electrocardiogram (ECG) signal is obtained from a filtered pulse wave (PPG) signal or an electrocardiogram (ECG) signal to obtain a standard normal distribution and an average of zero energy It can be used instead of the normal distribution.

Sigmoid function mapping step, the sigmoid function value is obtained by using normalized energy (E) and zero crossing rate in the normalization step, slope parameter, energy weighting and zero crossing weighting weight (S170).

In general, a sigmoid function is a function with a sigmoid curve (S), and is also a special form of a logistic function. It is a function that mainly shows the learning curve, and is a function that approaches a constant finite value at a value close to zero. In general, the sigmoid function f (x) for x is represented by f (x) = 1 / (1 + e ^{- ?} X) = 1 / (1 + exp to be.

That is, in the normalization step, the normalized energy E and the normalized zero crossing rate Z are multiplied by the slope parameter α and the energy weight w ₁ and the zero crossing weight w ₂ are used to calculate the energy And the sigmoid function w ₁ · f (E) + w ₂ · f (Z) for the addition. In other words, a normalized energy E having a sigmoid function of the normalized energy E with an energy weight w ₁ and a slope parameter a, and a normalized zero with a zero crossing rate weight w ₂ and a slope parameter a, in the form Sig plus feeders dE function of crossing rate (Z), that is, can be obtained as _{w 1 / (1 + exp (} -αE)) + w 2 / (1 + exp (-αZ)).

Here, the energy weight, the zero crossing rate weight, and the slope parameter may be values set at the time of factory shipment or at the beginning of use. The slope parameter alpha may be between 1.5 and 1.8, and the sum of the energy weight w1 and the zero crossing rate weight w2 should be one.

The sigmoid function value has a value between 0 and 1, with 0 being the maximum pain, and 1 being without pain. In addition, inverse calculations can be performed to show that the pain is zero when the pain is absent, and the pain is maximized when the pain is absent.

In the pain intensity classification step, the sigmoid function value obtained in the sigmoid function mapping step is output as the degree of pain (S180). The w ₁ · f (E) + w ₂ · f (Z) equations in the sigmoid function mapping step can be called the pain depth formula.

The present invention has a smoothing effect because it is strong against noise, simple in calculation, and squared at the time of energy detection.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, Modification is possible. Accordingly, the spirit of the present invention should be understood only in accordance with the following claims, and all equivalents or equivalent variations thereof are included in the scope of the present invention.

100: biological signal detection unit 110: pulse wave (PPG) sensor unit
120: Electrocardiogram (ECG) sensor unit 150: Biological signal preprocessing unit
160: Pulse wave (PPG) signal preprocessor 170: Electrocardiogram (ECG) signal preprocessor
180: A / D conversion unit 200:
210: display unit 220: memory unit

Claims (21)

A pulse wave signal or electrocardiogram signal detected from a pulse wave (PPG) sensor or an electrocardiogram (ECG) sensor part is converted into a digital signal through an A / D conversion part and transmitted to an arithmetic processing part. Or a cardiac depth index from an electrocardiogram signal, the apparatus comprising:
The arithmetic processing unit detects energy and zero crossing rate from a pulse wave signal or an electrocardiogram signal and normalizes it and detects a pain intensity index using a sigmoid function of normalized energy and zero crossing ratio Wherein the bio-signal-based pain intensity classification apparatus comprises: The method according to claim 1,
The arithmetic processing section detects and normalizes the energy and zero crossing rate from the pulse wave signal or the electrocardiogram signal, obtains a sigmoid function of the sigmoid function and the zero crossing rate of energy, multiplies the sigmoid function of energy by a predetermined energy weight And a sum of values obtained by multiplying a sigmoid function of a zero crossing rate by a preset zero crossing rate weight, and outputs the sum as a pain intensity index. 3. The method of claim 2,
The arithmetic processing unit performs filtering to remove noise from the pulse wave signal or electrocardiogram signal received from the A / D converting unit, and outputs a pulse wave signal or an electrocardiogram signal for a predetermined energy detection reference time period in the filtered pulse wave signal or electrocardiogram signal And the energy of the electrocardiogram signal per reference time or the pulse wave signal per reference time is obtained by summing the squares of the electrocardiogram signal. The method of claim 3,
The arithmetic processing unit performs filtering to remove noise from the pulse wave signal or electrocardiogram signal received from the A / D converting unit, and calculates a maximum value, a minimum value, and a maximum value in a filtered pulse wave signal or electrocardiogram And a minimum point time which is a time when the pulse time or the minimum value is detected, and the number of times the sign of the pulse wave signal or the electrocardiogram signal changes during the zero crossing reference time period is obtained as the zero crossing rate. 5. The method of claim 4,
Wherein the zero-crossing reference time interval is from the time point forward by 15 seconds in the maximum point time to the point of maximum point time. 6. The method of claim 5,
Wherein the calculation processing unit classifies the energy into a standard Gaussian distribution of energy and normalizes the divided energy. The method according to claim 6,
Wherein the calculation processing section normalizes the zero crossing rate by dividing the zero crossing rate into the standard normal distribution of the zero crossing rate. 8. The method of claim 7,
In order to obtain a standard normal distribution of energy, the arithmetic processing unit performs filtering to remove noise from the pulse wave signal or electrocardiogram signal received from the A / D converter during the standard normal distribution detection time interval at the beginning of signal detection, The pulse wave signal or the electrocardiogram signal is obtained by summing the squares of the pulse wave signal or the electrocardiogram signal during the energy detection reference time period to obtain energies of the pulse wave signal per reference time or the electrocardiogram signal per reference time, Characterized in that, in the energies, a standard deviation and an average are obtained to obtain a standard normal distribution of energy. 9. The method of claim 8,
In order to obtain a standard normal distribution of the zero crossing rate, the calculation processing section performs filtering to remove noise from the pulse wave signal or electrocardiogram signal received from the A / D conversion section during the standard normal distribution detection time interval at the beginning of signal detection, Zero crossing rate, which is the number of times the sign of the pulse wave signal or the electrocardiogram signal changes during the zero crossing reference time interval in the received pulse wave signal or electrocardiogram signal, and obtaining a standard deviation and an average to obtain a standard normal distribution of the zero crossing rate. Signal - based pain depth classification device. 8. The method of claim 7,
The arithmetic processing unit performs filtering to remove noise from the pulse wave signal or electrocardiogram signal received from the A / D converter during the standard normal distribution detection time interval at the beginning of signal detection, and calculates a standard deviation And the standard normal distribution of the obtained pulse wave signal or electrocardiogram signal is used instead of the standard normal distribution of the standard normal distribution and the zero crossing rate of the energy. A bio - signal based pain depth classification system. A pulse wave signal or electrocardiogram signal detected from a pulse wave (PPG) sensor or an electrocardiogram (ECG) sensor part is converted into a digital signal through an A / D conversion part and transmitted to an arithmetic processing part. Or a cardiac depth index from an electrocardiogram signal, the method comprising:
The arithmetic processing unit detects energy and zero crossing rate from a pulse wave signal or an electrocardiogram signal and normalizes it and detects a pain intensity index using a sigmoid function of normalized energy and zero crossing ratio Wherein the bio-signal-based pain intensity classification apparatus comprises: 12. The method of claim 11,
The arithmetic processing section detects and normalizes the energy and zero crossing rate from the pulse wave signal or the electrocardiogram signal, obtains a sigmoid function of the sigmoid function and the zero crossing rate of energy, multiplies the sigmoid function of energy by a predetermined energy weight And a value obtained by multiplying a sigmoid function of a zero crossing rate by a predetermined zero crossing rate weight and outputting the sum as a pain intensity index. 13. The method of claim 12,
The arithmetic processing section performs filtering to remove noise from the pulse wave signal or electrocardiogram signal received from the A / D conversion section,
The pulse wave signal or the electrocardiogram signal is summed with the squares of the pulse wave signal or the electrocardiogram signal for a predetermined energy detection reference time period to obtain the energy of the pulse wave signal per reference time or the electrocardiogram signal per reference time,
A pulse wave signal or an electrocardiogram signal during a zero-crossing reference time period, and detects a pulse wave signal or an electrocardiogram signal in the filtered pulse wave signal or the electrocardiogram signal as a minimum point time that is a time when a maximum value, a minimum value, As the zero crossing rate, the number of times that the sign of the heartbeat signal is changed. 14. The method of claim 13,
Wherein the zero-crossing reference time interval is from the time point forward by 15 seconds from the maximum point time to the time point of the maximum point time. 14. The method of claim 13,
The operation processing unit divides and normalizes the energy into a standard Gaussian distribution of energy,
Wherein the zero crossing rate is normalized by dividing the zero crossing rate by the standard normal distribution of the zero crossing rate. A pulse wave signal or electrocardiogram signal detected from a pulse wave (PPG) sensor or an electrocardiogram (ECG) sensor part is converted into a digital signal through an A / D conversion part and transmitted to an arithmetic processing part. Or the electrocardiogram signal by using a reshape filter to remove noise, and detecting a pain intensity index from the filtered pulse wave signal or electrocardiogram signal, the method comprising: ,
Calculating an energy of a pulse-wave signal per reference time or an electrocardiogram signal per reference time by summing the squares of the filtered pulse-wave signal or the electrocardiogram signal during the energy detection reference time period in the filtered pulse-wave signal or electrocardiogram signal;
The operation processing unit detects a minimum point time which is a time when the maximum value, the minimum value, and the maximum point time, which is the time when the maximum value is the minimum value, in the filtered pulse wave signal or the electrocardiogram signal, A zero crossing rate calculating step of obtaining, as zero crossing rate, the number of times the sign of the pulse wave signal or the electrocardiogram signal changes during the zero crossing reference time period;
The energy obtained in the energy computing step is normalized by dividing the energy obtained in the energy computing step by the standard normal distribution of energy and the normalized normalizing of the zero crossing rate with respect to the zero crossing rate obtained in the zero crossing rate computing step, Normalizing step of dividing and normalizing the zero crossing rate obtained in step
The operation processing unit obtains a sigmoid function of the energy sigmoid function and the zero crossing ratio by applying the normalized energy and the normalized zero crossing ratio output from the normalization step to the sigmoid function, A sigmoid function arithmetic step of summing a value obtained by multiplying a de function by a predetermined energy weight and a value obtained by multiplying a sigmoid function of a zero crossing ratio by a predetermined zero crossing ratio weight;
Wherein the bio-signal-based pain intensity classification apparatus comprises: 17. The method of claim 16,
Wherein the reshape filter is a filter that allows a frequency of 0.5 Hz to 8 Hz to pass therethrough. 17. The method of claim 16,
In the sigmoid function arithmetic step, the pain intensity index,
w ₁ / (1 + exp (-αE)) + w ₂ / (1 + exp (-αZ)
(Where, E is the normalized energy, Z is a normalized zero crossing rate, α is the slope parameter, w ₁ is a weight energy, zero crossing rate w ₂ are weight Im)
Based on the intensity of the light emitted from the light source. 19. The method of claim 18,
Wherein the slope parameter alpha is 1.5 to 1.8 and the sum of the energy weight w1 and the zero crossing weight weight w2 is 1. 19. The method of claim 18,
Wherein the pain intensity index has a value between 0 and 1. A computer-readable recording medium on which a computer-implemented method for driving a bio-signal-based pain intensity classification apparatus according to any one of claims 11 to 20 is recorded.

Images