KR101866517B1 - Apparatus and Method for diagnosing sleep apnea using UWB radar - Google Patents

Abstract

The present invention relates to an apparatus and method for detecting apnea using an ultra-wideband radar, and more particularly, to an apparatus and method for detecting an apnea by removing impulse noise and a peak noise inherent in a respiration signal, And an apparatus and method for apnea detection using radar.
According to another aspect of the present invention, there is provided an ultrawideband radar module for generating and transmitting ultrawideband pulse signals at regular intervals, receiving ultra-wideband pulse signals reflected from a human body to generate and output raw data, A human body sensing unit for grasping a human body position in the raw data, accumulating raw data of the detected human body position according to time, and outputting the accumulated raw data; And a controller for acquiring a signal of a human body chest position from the raw data accumulated in the human body sensing unit as a raw signal, selecting a maximum value as a source signal based on the distance, removing the impulse noise from the selected original signal, An apnea detection module for detecting a peak value and detecting an apnea interval, and an apnea detection device and method using the ultra wideband radar.

Description

TECHNICAL FIELD [0001] The present invention relates to an apparatus and method for detecting apnea using UWB radar,

The present invention relates to an apparatus and method for detecting apnea using an ultra-wideband radar, and more particularly, to a method and apparatus for detecting an apnea by eliminating impulse noise and a peak noise inherent in a respiration signal, And an apparatus and method for apnea detection using radar.

Apnea refers to stopping external breathing. During apnea, there is no muscle movement to breathe air and the size of the lungs does not change.

The cause of this apnea is as follows.

1. When a person intentionally holds breath

2. Induction by drug

3. Blocked airway

4. Results from neurological disease

In particular, if apnea occurs during sleep, it becomes a serious disease that can lead to death.

Sleep Apnea is a sleeping disorder that causes shallow breathing or breathing stops during sleep. Apnea stops breathing and hypopnea is called shallow breathing.

There are four kinds of sleep apnea.

1. Obstructive Sleep Apnea (OSA)

2. OH (Obstructive Hypopnea)

3. CSA (Central Sleep Apnea)

4. Multiple apnea

Among these, obstructive sleep apnea is the most common form of sleep apnea, which has various causes such as tonsillar hypertrophy, adenoid hypertrophy, enlargement of the study group, and size and location of the tongue.

These obstructive sleep apnea are accompanied by severe snoring. This can be explained by the fact that during the sleeping, the position of the study dog and tongue is sagging downward, and the tension of the muscles attached to the tongue and mandible is decreased.

Obstructive sleep apnea accounts for 84% of the total. CSA occurs in the central nervous system without respiratory signals and accounts for 15%.

A conventional sleep apnea management device for preventing sleep apnea is provided as Patent No. 10-1456461.

However, in the related art, since the user must take a sleeping state while wearing his / her neck at all times when sleeping, it is not only troublesome to use but also takes a sleeping state while the user's body is restrained. There is a problem that it is inhibited.

Many apneic sensing devices have been developed to improve this.

Such an improved apnea sensing device is disclosed in Korean Patent No. 10-1669453.

The improved prior art relates to a sleep apnea management apparatus, and more particularly, to a sleep apnea management apparatus that includes a sensor member for sensing a blood oxygen saturation of a user when sleeping. A first operation for determining a sleep apnea state by the oxygen saturation detected by the sensor member and a first operation for securing airway by raising a user's neck portion in a sleep apnea state and a second operation for supplying a vibration to correct a user's sleeping position, absence; And a smartphone having a sleeping app that receives and stores information on the operating state of the sensor member and the pillow member and provides the stored information to the user.

However, such an improved conventional technique has a problem in that the accuracy is lowered by an indirect method based on the user's blood coral saturation.

Another prior art that improves this is the sleep monitoring system and method for monitoring the sleep apnea and sleep stages of Laid-open No. 10-2014-0003867.

The above-described other prior art technique monitors the sleep apnea of a user by generating respiration data of a user using a bio-radar, monitors a user's sleep phase from motion data generated by a motion sensor, It is possible to monitor the sleeping phase together with the sleep apnea of the user so that the sleeping state of the user can be monitored without interfering with the sleep due to the inconvenience or discomfort of the user.

However, such a conventional technology has a problem that it is expensive to manufacture due to a motion detection sensor, raises the price of the product, and the precision of the motion detection sensor is low.

Korean Publication No. 10-2006-0087852 Korean Patent No. 10-1456461 Korean Patent No. 10-1669453 Publication No. 10-2014-0003867

SUMMARY OF THE INVENTION In order to solve the above problems, the present invention has been made to solve the above problem, and it is an object of the present invention to provide an apnea detection device using an ultra wideband radar which can detect an impulse noise and a peak noise inherent in a respiration signal, And a method.

According to an aspect of the present invention, there is provided an ultrawideband radar module for generating and transmitting ultrawideband pulse signals at regular intervals, receiving ultra-wideband pulse signals reflected from a human body to generate raw data, and outputting the generated raw data; A human body sensing unit for grasping a human body position in the raw data, accumulating raw data of the detected human body position according to time, and outputting the accumulated raw data; And a controller for acquiring a signal of a human body chest position from the raw data accumulated in the human body sensing unit as a raw signal, selecting a maximum value as a source signal based on the distance, removing the impulse noise from the selected original signal, And an apnea measurement module for detecting a peak value and detecting an apnea interval.

In addition, one aspect of the present invention further includes a preprocessor for preprocessing the raw data and outputting the preprocessed data to the human body sensor.

In addition, the human body detecting unit of the present invention detects the human body using the distance and the signal size in the raw data, grasps the detected human body position, and accumulates and outputs the raw data in the vicinity of the human body .

According to another aspect of the present invention, there is provided an apnea measurement module including: a raw signal acquisition unit for acquiring a signal of a human breast position from original data accumulated in the human body sensing unit as a raw signal and outputting the raw signal; A noise removing unit for removing impulse noise from the original signal acquired by the original signal acquiring unit; A respiration signal calculation unit for acquiring a respiration signal from the original signal from which the impulse noise is removed in the noise elimination; A peak detector for dividing a peak signal of a predetermined section in the respiration signal calculated by the respiration signal calculator into a high peak or a low peak to calculate a breathing number; And an apnea detection unit for calculating an amplitude using the peak signal detected by the peak detection unit and detecting an apnea interval.

The noise eliminating unit of the present invention applies the median filter to the original signal acquired by the original signal acquiring unit to remove the impulse noise and applies the impulse noise removing logic to samples larger than the average value in the filter window .

The noise canceling unit of the present invention may be configured such that when a start value of a sample of a filter window is x0 and a final value of a sample of a filter window is xn, a replacement value xi 'of a sample xi corresponding to impulse noise is xi '= x0 + (xn-x0) * (i / window size), where i represents the rank of the sample.

In addition, the peak detecting unit of the present invention may distinguish whether a high peak or a low peak is caused by a slope change of a predetermined section in an input waveform of a respiration signal passed through the respiration signal calculating unit Calculate the number of breaths.

Further, the apnea detection unit of the present invention compares the amplitude of the waveform having the high peak value with the average of the previous amplitude, and if it exceeds the first predetermined value of the average of the previous amplitude, it is determined as a valid high peak value, If it is below the second predetermined value of the average of the previous amplitudes as compared with the average, it is judged that it has entered the apnea period.

In another aspect of the present invention, the apnea detection unit determines that the apnea interval is out of order when the apnea interval is greater than or equal to a second predetermined value of the average of the peak values before the apnea interval progresses.

In addition, the apnea detector of the present invention compares the amplitude of a waveform having a high peak value with an average of previous amplitudes, and judges that the peak value is not valid unless the first predetermined value of the average of the previous amplitudes is exceeded.

According to another aspect of the present invention, there is provided a method of controlling an ultrawideband radar module, the method comprising: (A) generating and transmitting an ultrawideband pulse signal at a constant period, receiving ultrasound pulse signals reflected from the human body to generate and output raw data; (B) the human body detecting unit grasps the human body position in the raw data, and accumulates and outputs the raw data of the detected human body position with time; (C) acquiring a human chest position signal as a raw signal from raw data accumulated in the human body sensing unit by the apnea measurement module, and selecting a maximum value as a raw signal based on the distance; And (D) the apnea measurement module detects an apnea interval by calculating a respiration signal after removing the impulse noise from the original signal selected by the apnea measurement module, and detecting a peak value.

In another aspect of the present invention, the step (D) includes the steps of: (D-1) removing the impulse noise from the original signal by the apnea measurement module; (D-2) the apnea measurement module acquiring a respiration signal from the original signal from which the impulse noise is removed; (D-3) dividing whether the apnea measurement module is a high peak or a low peak with respect to a peak signal of a predetermined interval in the respiration signal to calculate a breathing number; And (D-4) the apnea measurement module calculates an amplitude using a peak signal to detect an apnea interval.

In another aspect of the present invention, in the step (D-2), the apnea measurement module removes impulse noise by applying a median filter to the original signal, and performs impulse noise removal logic on a sample larger than an average value in the filter window To be applied.

In another aspect of the present invention, in the step (D-2), the apnea measurement module calculates the apex of the impulse noise when the start value of the sample of the filter window is x0 and the final value of the sample of the filter window is xn Xi '= x0 + (xn-x0) * (i / window size), where i represents the rank of the sample.

In another aspect of the present invention, in the step (D-3), the apnea measurement module measures a slope change of a predetermined section in an input waveform of the respiration signal and discriminates whether the peak is a high peak or a low peak Calculate the number of breaths.

Further, in the step (D-4) of the other aspect of the present invention, when the apnea measurement module compares the amplitude of the waveform having the high peak value with the average of the previous amplitude, If it is determined that the peak value is equal to or less than the second predetermined value of the average of the previous amplitudes, it is determined that the user has entered the apnea period.

In another aspect of the present invention, in the step (D-4) after (D-5), the apnea measurement module measures the second predetermined value or more of the average of the high peak values before the apnea interval progresses It is determined that it is out of the apnea interval.

In another aspect of the present invention, in the step (D-4), the apnea measurement module compares the amplitude of a waveform having a high peak value with an average of previous amplitudes, and if the first predetermined value of the average of the previous amplitudes is not exceeded It is determined as a high peak value.

The present invention can eliminate the impulse noise generated by peripheral wireless devices such as a Wi-Fi device, thereby enabling accurate determination of apnea.

In addition, the present invention eliminates the peak noise inherent in the respiration signal and enables accurate apnea determination.

1 is a block diagram of an apnea sensor using an ultra-wideband radar according to a preferred embodiment of the present invention.
FIG. 2 is a diagram illustrating an ultrawideband pulse signal generated by the UWB radar module of FIG. 1. Referring to FIG.
3 is a diagram illustrating an example of raw data output from the UWB radar module of FIG.
FIG. 4 is an exemplary diagram illustrating a noise-canceled signal through the signal preprocessing in FIG.
FIG. 5 is a view showing a state in which the human body detection unit of FIG. 1 accumulates and stores raw data near a human body according to time progression.
6 is a partial magnified view of the stored signal of FIG.
7 is a diagram showing the original signal acquired by the original signal acquisition unit of FIG.
FIG. 8 is a diagram illustrating a process of removing impulse noise in the noise elimination of FIG. 1; FIG.
FIG. 9 is a diagram showing a signal in which impulse noise is removed from the original signal of FIG. 7; FIG.
10 is an exemplary view of a respiration signal calculated by the respiration signal calculation unit of FIG.
11 is an exemplary diagram for explaining a process of calculating a peak by the peak calculating section of FIG.
FIG. 12 is an exemplary diagram for explaining a process of calculating an apnea interval of the apnea calculation unit of FIG. 1;
13 is a flowchart of an apnea detection method using an UWB radar according to a preferred embodiment of the present invention.
14 is a graph showing a change in a signal according to the movement of an object when measuring an object signal using an UWB radar.
Fig. 15 shows a signal for a space having no motion.
FIG. 16 is a diagram showing a change with time of the signal of FIG. 15; FIG.

BRIEF DESCRIPTION OF THE DRAWINGS The above and other objects, features and advantages of the present invention will become more apparent from the following detailed description of the present invention when taken in conjunction with the accompanying drawings.

First, the terminology used in the present application is used only to describe a specific embodiment, and is not intended to limit the present invention, and the singular expressions may include plural expressions unless the context clearly indicates otherwise. Also, in this application, the terms "comprise", "having", and the like are intended to specify that there are stated features, integers, steps, operations, elements, parts or combinations thereof, But do not preclude the presence or addition of features, numbers, steps, operations, components, parts, or combinations thereof.

In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear.

1 is a block diagram of an apnea sensor using an ultra-wideband radar according to a preferred embodiment of the present invention.

Referring to FIG. 1, an apparatus for detecting apnea using an UWB radar according to an embodiment of the present invention includes an UWB radar module 100, a preprocessor 200, a human body sensor 300, and an apnea measurement module 400. .

The apnea measurement module 400 includes an original signal acquisition unit 410, a noise removal unit 420, a respiration signal calculation unit 430, a peak detection unit 440, and an apnea detection unit 450.

In this configuration, the UWB radar module 100 includes a pulse generator, a transmitting antenna, a receiving antenna, a time delay, a sampler, a preamplifier, and a microcontroller.

The UWB radar module 100 receives a signal from the microcontroller, generates a UWB pulse signal from the pulse generator, and transmits the UWB pulse signal through a transmission antenna.

An example of the UWB pulse signal generated by the UWB radar module 100 is shown in FIG. 2. The UWB pulse signal is generated at 90 to 150 Hz.

The UWB radar module 100 receives the signal reflected from the human body through the receiving antenna, amplifies the signal received from the preamplifier, and samples the received signal after the delay time in the delay unit by the sampler And generates and outputs raw data.

3, the raw data output from the UWB radar module 100 is shown in FIG. 3, where the X axis represents time (unit: ps) and the Y axis represents the size of the signal (unit is V, the unit of voltage).

Here, since the time delay is set so that the UWB radar module 100 processes the received signal after the UWB pulse signal travels 50 cm, the time starting point (0 in FIG. 3) can be seen as 50 cm in the distance, The number of samples is 256 times per 1m, so the number of 512 samples is 2m.

Next, the preprocessor 200 preprocesses the raw data output from the UWB radar module 100 to remove noise.

Since the signal output from the UWB radar module 100 includes power source noise and thermal noise, the preprocessor 200 removes noise using a band pass filter having a band of 5 to 10 GHz, The signal is shown in FIG.

Then, the human body sensing unit 300 senses the human body based on the distance and the signal size in a state where the noise is removed from the raw data subjected to the preprocessing, and grasps the position of the human body.

That is, as described above, when the number of 256 samples is 1m and the number of 512 samples is 2m, the human body detector 300 detects a signal having a certain size or more between 200 samples and 400 samples, It is judged that the human body exists and it is judged that the human body is positioned at the position where the largest signal is detected.

When the human body is sensed and the position of the human body is recognized, the human body sensing unit 300 accumulates and stores the raw data in the vicinity of the human body as time progresses. At this time, a signal to be stored is shown in Fig. 5, and Fig. 6 is a partially enlarged view.

Next, the apnea measurement module 400 selects the maximum value as the original signal based on the distance using the accumulated raw data (original signal) at the human body position, and calculates the respiration signal after removing the impulse noise from the selected original signal And the peak value is detected to detect the apnea interval.

In the apnea measurement module 400, the original signal acquiring unit 410 acquires a signal having the largest size as a source signal, based on the distance from the accumulated raw data near the human body (the expression is based on the number of samples) .

The original signal acquired by the original signal acquisition unit 410 is a respiration signal and a noise combined signal. The waveform of the original signal accumulated for N hours is shown in FIG.

In the waveform of FIG. 7, it can be seen that the noise is mixed in addition to the waveform that appears as respiration. These noises are caused by the surrounding environment (wifi, board power noise, etc.).

7, it is necessary to reduce such impulse noise caused by the surrounding environment in order to calculate the respiration position and the breathing number.

It is difficult to obtain normal results when the respiration position and respiration rate are searched through frequency analysis without removing the impulse noise.

The noise removing unit 420 removes the impulse noise using a median filter.

At this time, the noise removing unit 420 can determine the impulse noise through the same conditions as in FIG. 8 (a).

The noise remover 420 applies impulse noise removal logic only when the standard deviation value in the filter window exceeds a standard deviation threshold to reduce the risk of misinterpreting the normal signal as impulse noise.

That is, in FIG. 8 (a), when the size of the filter window is n + 1, the average value in the filter window is indicated by a dotted line.

The noise remover 420 applies the impulse noise removal logic only to xi and x (i + 1) which are larger than the average value in the filter window.

In this case, for example, when the start value of the sample of the filter window is x0 and the final value of the sample of the filter window is xn, the noise removal section 420 substitutes the replacement value xi 'of the sample xi corresponding to the impulse noise by xi '= x0 + (xn-x0) * (i / window size). Here, i represents the rank of the sample corresponding to the impulse noise.

As described above, the noise removing unit 420 may replace the average value of the normal signal samples around the sample corresponding to the impulse noise, the sample start value x0 of the filter window, and the sample final value xn of the filter window, .

The waveform in which the impulse noise is removed by the noise remover 420 is shown in FIG.

Next, the breathing signal calculator 430 obtains the respiration signal shown in FIG. 10 using a band-pass filter having a band of 0.1 to 1.5 Hz or a moving average window (MAW).

In this way, the respiration signal calculator 430 uses a band-pass filter or the like to measure the respiratory rate in the waveform from which the impulse noise has been removed, thereby removing signals other than the respiration rate corresponding to the respiration rate of the human.

11, the peak detector 440 detects a change in the slope of a predetermined section of the input waveform of the breathing signal passing through the breathing signal calculator 430 and outputs a high peak or a low peak peak) to determine the number of breaths.

The peak detector 440 determines that the peak is a high peak of the waveform if the slope of the waveform in the peak detection window changes from + to -, and if the slope of the waveform changes from - to + peak).

The number of breaths can be obtained using the number of high peaks and time Δt. For example, in Fig. 11, when? T1 is 5000 ms, the respiratory rate per minute is (60 * 1000) / (5000/2) = 24 BPM (beat per minute).

On the other hand, when the respiratory rate is obtained through real-time waveform analysis, the following consideration is necessary.

1. Minimize the influence of noise

2. Breathing in no-breathing state

3. When there is movement

In order to minimize the influence of noise, it is necessary to distinguish whether there is an actual breath or not by setting an amplitude threshold between a high peak and a low peak, and to reduce the influence of small noise by adjusting the window size for tilt calculation.

For this, the peak detector 440 uses a window size of 9 when the UWB radar signal is acquired at a frame rate of 10 fps. If this value is small, it is affected by noise. If this value is large, it is impossible to measure respiration rate for fast breathing.

On the other hand, the apnea detector 450 compares the amplitude of the waveform having the high peak value with the average of the previous amplitude, as shown in FIG. 12, to calculate a first constant value of the average of the previous amplitudes (for example, , It is judged as a valid high peak value. If it is judged to be a valid peak value, it is compared with the average of the previous amplitudes, and if it is less than the second predetermined value of the average of the previous amplitudes (for example, A predetermined value, preferably 50%), it is determined that the user has entered the apnea period.

Then, the apnea detector 450 determines that the apnea interval is out of range when the apnea interval is longer than the second predetermined value of the average of the peak values before the apnea interval progresses.

The first constant value should be determined after the noise measurement in the empty space in the UWB radar equipment. In FIG. 12, a high peak value that does not exceed the first predetermined value can be identified as a peak due to noise, not a peak of actual respiration.

The apnea detector 450 can determine whether the apnea is caused by respiration through the first constant value even during the apnea period.

In addition, the apnea detection unit 450 measures to cope with the motion or severe noise.

1. The peak position of the waveform is memorized and it is judged that it is not a respiratory waveform when overlapping high peak-high peak, low peak-low peak occurs.

2. Calculate the number of breaths using the high-peak interval and the low-peak interval. If the error is more than 10%, it is judged that it is not the breathing waveform.

3. If the number of breaths calculated through real-time waveform analysis does not fall within the normal breathing range, it is judged that it is not a breathing waveform.

13 is a flowchart of an apnea detection method using an UWB radar according to a preferred embodiment of the present invention.

Referring to FIG. 13, a method for detecting apnea using an UWB radar according to an exemplary embodiment of the present invention includes receiving a signal from a microcontroller of an UWB radar module, generating a UWB pulse signal from a pulse generator, And receives the ultra-wideband radar signal reflected by the human body and returned to the receiving antenna (S100).

The ultra-wideband pulse signal generated by the ultra-wideband radar module is transmitted at 90 to 150 Hz.

The UWB radar module receives the signal reflected from the human body through the receiving antenna, amplifies the signal received by the preamplifier, samples the received signal after the delay time in the delayer, and samples the raw data And outputs it.

Next, the preprocessor preprocesses the raw data output from the UWB radar module to remove noise (S110).

The signal output from the UWB radar module includes power source noise, thermal noise, and the like, and the preprocessor removes noise using a band pass filter having a band of 5 to 10 GHz.

Then, the human body sensing unit senses the human body based on the distance and the signal size in a state where the noise is removed from the raw data subjected to the preprocessing, and determines the position of the human body (S120).

When the human body is sensed and the position of the human body is detected, the human body sensing unit accumulates raw data near the human body in accordance with the progress of time.

Next, the original signal acquiring unit of the apnea measurement module acquires the signal of the largest magnitude as the original signal (S130), based on the distance from the accumulated raw data near the human body (which is equivalent to the expression of the number of samples). As described above, the original signal acquired by the original signal acquisition unit is a signal in which a respiration signal, a heartbeat signal, and noise are combined.

As described above, the original signal acquired by the original signal acquisition unit is a signal obtained by combining a breathing signal and noise, and these noises are generated by the influence of the surrounding environment (wifi, board power noise, etc.).

To calculate the respiration location and respiratory rate, it is necessary to reduce such impulse noise caused by the surrounding environment.

It is difficult to obtain normal results when the respiration position and respiration rate are searched through frequency analysis without removing the impulse noise.

The noise removing unit removes the impulse noise using a median filter (S140).

At this time, the noise removing unit applies the impulse noise removal logic only when the standard deviation value in the filter window exceeds the standard deviation threshold (threshold) in order to reduce the risk of mistaking the normal signal as impulse noise.

The noise removal unit applies impulse noise removal logic only for xi and x (i + 1), which is larger than the average value in the filter window.

In this case, for example, when the start value of the sample of the filter window is x0 and the final value of the sample of the filter window is xn, the noise elimination unit substitutes xi 'of the sample xi corresponding to the impulse noise by xi' = x0 + xn-x0) * (i / window size). Here, i represents the rank of the sample corresponding to the impulse noise.

In this way, the noise removing unit can remove the noise value by replacing the mean value of the normal signal samples around the sample corresponding to the impulse noise with the new value using the sample start value x0 of the filter window and the sample final value xn of the filter window.

Next, the breathing signal calculator obtains a breathing signal using a band-pass filter or a moving average window (MAW) having a band of 0.1 to 1.5 Hz (S150).

In this way, the respiration signal calculator uses a band-pass filter or the like to measure the respiration rate in the waveform from which the impulse noise has been removed, thereby removing signals other than the frequency range corresponding to the human respiration rate.

Meanwhile, the peak detector calculates the number of respirations by distinguishing the high peak or the low peak according to the slope change of the predetermined section in the input waveform of the respiration signal passing through the respiration signal calculator (S160) .

The peak detection section determines that the peak is a high peak of the waveform if the gradient of the waveform in the peak detection window changes from + to -, and if the gradient of the waveform changes from - to + .

The number of breaths can be obtained using the number of high peaks and time Δt. For example, in Fig. 11, when? T1 is 5000 ms, the respiratory rate per minute is (60 * 1000) / (5000/2) = 24 BPM (beat per minute).

On the other hand, when the respiratory rate is obtained through real-time waveform analysis, the following consideration is necessary.

1. Minimize the influence of noise

2. Breathing in no-breathing state

3. When there is movement

In order to minimize the influence of noise, it is necessary to distinguish whether there is an actual breath or not by setting an amplitude threshold between a high peak and a low peak, and to reduce the influence of small noise by adjusting the window size for tilt calculation.

For this purpose, the peak detector uses a window size of 9 when acquiring an ultra wideband radar signal at a frame rate of 10 fps. If this value is small, it is affected by noise. If this value is large, it is impossible to measure respiration rate for fast breathing.

On the other hand, the apnea detection unit compares the amplitude of the waveform having the high peak value with the average of the previous amplitude (S170), and determines the first predetermined value of the amplitude of the previous amplitude (for example, any value in the range of 10% to 40% %), It is determined that the peak value is a valid peak value (S180). If it is determined that the peak value is greater than the second predetermined value of the average of the previous amplitudes (for example, from 40% Value, preferably 50%), it is determined that the user has entered the apnea interval (S200).

In contrast, the apnea detection unit compares the amplitude of the waveform having the high peak value with the average of the previous amplitude (S170), and determines the first predetermined value of the amplitude of the previous amplitude (for example, any value in the range of 10% to 40% %), It is judged as an invalid high peak value and removed from the list of high peak values or the like (S182).

On the other hand, the apnea detection unit compares the amplitude of the waveform having the high peak value with the average of the previous amplitude (S170), and determines the first predetermined value of the amplitude of the previous amplitude (for example, any value in the range of 10% to 40% (S180). If it is greater than or equal to the second predetermined value of the average of the previous amplitudes (for example, within a range from 40% to 60% inclusive) Value, preferably 50%), it is determined that it is still in the breath interval (S202).

Next, the apnea detection unit repeats the determination of S180. If it is determined that the apnea interval exceeds the second predetermined value of the average of the high peak values before the apnea interval progresses, the apnea detection unit determines that the apnea interval has passed and enters the breathing interval (S202).

The first constant value should be determined after the noise measurement in the empty space in the UWB radar equipment.

The apnea detection unit may determine whether a peak due to breathing is present through the first constant value even in the apnea period.

In addition, the apnea detection unit measures the following measures in case of motion or severe noise.

1. The peak position of the waveform is memorized and it is judged that it is not a respiratory waveform when overlapping high peak-high peak, low peak-low peak occurs.

2. Calculate the number of breaths using the high-peak interval and the low-peak interval. If the error is more than 10%, it is judged that it is not the breathing waveform.

3. If the number of breaths calculated through real-time waveform analysis does not fall within the normal breathing range, it is judged that it is not a breathing waveform.

On the other hand, according to the conventional technology using radar, if the difference between the maximum value and the second maximum value is below a predetermined value in the frequency analysis of breath analysis, it is detected as apnea.

Such a conventional technique does not perform a clock analysis and thus has a low accuracy and various other difficulties.

For example, the frequency analysis can not collect the data of a certain time (5 seconds or 10 seconds) to confirm the apnea time.

As shown in FIG. 12, for the respiration signal such as the rim, the accurate apnea time point is a time point when a very small signal size comes out.

However, in the conventional frequency analysis, the respiratory signal was continuously detected until the signal became smaller, and the accurate apnea time point could not be known.

However, according to the present invention, by using time analysis, a point at which the signal starts to be made smaller can be accurately detected so that the apnea time can be known.

In addition, according to the related art, even if a living body signal is acquired in a state where the measurer is not moving, it can be moved to a small size of 1 cm or less. In this case, the signal moves as a whole as shown in FIG.

On the other hand, according to the related art, a small signal changes due to thermal noise or noise from other equipment even in a space free from motion in an ultra-wideband radar. Such noise causes a problem of detecting a motion or changing the size of a motion when there is no motion when acquiring an actual biological signal. Also, the noise may increase depending on the radar installation environment. Fig. 15 shows a signal for a space having no motion. The change with time of this signal is shown in Fig.

Also, when there is a radio wave using a similar frequency band such as Wi-Fi (see Fig. 7), a large noise is generated.

The present invention solves such a problem of the related art by using a media filter that removes impulse noise and removes an ineffective peak signal by using an amplitude threshold value so that an accurate apnea interval can be detected.

The foregoing description is merely illustrative of the technical idea of the present invention, and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention. Therefore, the embodiments of the present invention are not intended to limit the scope of the present invention but to limit the scope of the present invention. The scope of protection of the present invention should be construed according to the following claims, and all technical ideas within the scope of equivalents thereof should be construed as being included in the scope of the present invention.

100: ultra-wideband radar module 200: preprocessing section
300: Human body sensing part 400: Apnea measurement module
410: original signal acquisition unit 420: noise rejection unit
430: respiration signal calculation unit 440: peak detection unit
450: Apnea detection unit

Claims (18)

An ultrawide broadband radar module for generating and transmitting an ultrawideband pulse signal at a predetermined cycle, receiving ultra-wideband pulse signals reflected from a human body to generate raw data, and outputting the raw data;
A human body sensing unit for grasping a human body position in the raw data, accumulating raw data of the detected human body position according to time, and outputting the accumulated raw data; And
A human body chest position signal is obtained from the raw data accumulated in the human body sensing unit as a raw signal, a maximum value is selected as a raw signal based on the distance, an impulse noise is removed from the selected original signal, An apnea measurement module for detecting a peak value and detecting an apnea interval,
The apnea measurement module
An original signal acquisition unit for acquiring a signal of a human body chest position as a raw signal from the raw data accumulated in the human body sensing unit and outputting the raw signal;
A noise removing unit for removing impulse noise from the original signal acquired by the original signal acquiring unit;
A respiration signal calculation unit for acquiring a respiration signal from the original signal from which the impulse noise is removed in the noise elimination;
A peak detector for dividing a peak signal of a predetermined section in the respiration signal calculated by the respiration signal calculator into a high peak or a low peak to calculate a breathing number; And
And an apnea detection unit for detecting an apnea interval by calculating an amplitude using the peak signal detected by the peak detector,
The peak detector calculates a respiratory rate by distinguishing a high peak or a low peak according to a slope change of a predetermined section in an input waveform of a respiration signal passing through the respiration signal calculator,
The apnea detection unit compares the amplitude of the waveform having the high peak value with the average of the previous amplitudes to determine a valid peak value when the amplitude exceeds a first predetermined value of the average of the previous amplitudes, If the second predetermined value is less than the second predetermined value, it is determined that the robot enters the apnea interval. The method according to claim 1,
And a preprocessor for preprocessing the raw data and outputting the preprocessed data to the human body sensing unit. The method according to claim 1,
Wherein the human body detection unit detects the human body using the distance and the signal size from the raw data and grasps the detected human body position and accumulates the raw data near the human body according to the time progression and outputs the accumulated data. Apnea detection device. delete The method according to claim 1,
The noise removing unit
Wherein the impulse noise eliminating unit applies a median filter to the original signal acquired by the original signal acquiring unit and applies impulse noise elimination logic to samples larger than an average value in the filter window. The method of claim 5,
The noise canceller calculates xi '= x0 + (xn-x0) (xi') as a substitute value of the sample xi corresponding to the impulse noise when the starting value of the sample of the filter window is x0 and the final value of the sample of the filter window is xn. ) * (i / window size), where i is the ultra-wideband radar indicating the position of the sample. delete delete The method according to claim 1,
The apnea detection unit
Wherein the apnea detection unit determines that the apnea interval is out of range when the apnea interval progresses to a second predetermined value or more of an average of high peak values before the apnea interval progresses. The method according to claim 1,
Wherein the apnea detection unit compares an amplitude of a waveform having a high peak value with an average of previous amplitudes to determine an ineffective peak value if the amplitude does not exceed a first predetermined value of the average of the previous amplitudes. (A) generating and transmitting an ultrawideband pulse signal at an interval of a predetermined period, receiving ultrasound pulse signals reflected from a human body to generate and output raw data, and
(B) the human body detecting unit grasps a human body position in the raw data, and accumulates and outputs raw data of the detected human body position according to time;
(C) acquiring a human chest position signal as a raw signal from raw data accumulated in the human body sensing unit by the apnea measurement module, and selecting a maximum value as a raw signal based on the distance; And
(D) calculating a breathing signal after the impulse noise is removed from the original signal selected by the apnea measurement module, detecting a peak value, and detecting an apnea interval,
The step (D)
(D-1) removing the impulse noise from the original signal by the apnea measurement module;
(D-2) the apnea measurement module acquiring a respiration signal from the original signal from which the impulse noise is removed;
(D-3) dividing whether the apnea measurement module is a high peak or a low peak with respect to a peak signal of a predetermined interval in the respiration signal to calculate a breathing number; And
(D-4) The apnea measurement module calculates an amplitude using a peak signal to detect an apnea interval,
In step (D-3), the apnea measurement module calculates a respiratory rate by distinguishing a high peak or a low peak according to a slope change of a predetermined interval in an input waveform of a respiration signal,
In the step (D-4), the apnea measurement module compares the amplitude of the waveform having the high peak value with the average of the previous amplitude, and if it exceeds the first predetermined value of the average of the previous amplitude, Of the average of the previous amplitudes is less than a second predetermined value, it is determined that the robot enters the apnea period. delete 12. The method of claim 11,
In the step (D-2), the apnea measurement module applies a median filter to the original signal to remove impulse noise, and applies an impulse noise removal logic to samples larger than an average value in a filter window. Detection method. The method of claim 13,
In step (D-2), the apnea measurement module calculates a replacement value xi 'of the sample xi corresponding to the impulse noise when the starting value of the sample of the filter window is x0 and the final value of the sample of the filter window is xn, Is calculated as xi '= x0 + (xn-x0) * (i / window size), where i is the rank of the sample. delete delete 12. The method of claim 11,
After the step (D-4)
(D-5) determining that the apnea interval has exceeded a second predetermined value of the average of the high peak values before the apnea interval progresses, . 12. The method of claim 11,
In step (D-4), the apnea measurement module compares the amplitude of the waveform having the high peak value with the average of the previous amplitude, and determines that it is not an effective high peak value unless the first predetermined value of the average of the previous amplitude is exceeded A method for detecting apnea using ultra wideband radar.

Images