Breathing segment positioning method and device under voice instruction guidance, equipment and medium

Abstract

The application relates to the field of respiratory monitoring, in particular to a respiratory segment positioning method and device guided by voice instructions, equipment and a medium. The application adopts radar non-contact detection technology, generates a test signal by capturing chest and abdominal breathing operation, does not need a contact type blowing operation, discards the dependence on a contact type collection device, solves the problem of high operation threshold, secondly, the test action is guided by voice instructions, replaces professional medical staff guidance, reduces the operation professional requirement through standardized instructions, then, the radar is installed flexibly and is suitable for various postures, breaks through the limitation of a professional medical scene, finally, the respiratory segment is automatically extracted and the respiratory index of the measured object is calculated, without complex manual intervention, the process is simplified, and the equipment and use cost are reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of respiratory monitoring, in particular to a method and device for positioning a respiratory segment under the guidance of a voice instruction, an equipment and a medium. BACKGROUND

[0002] In the field of diagnosis and evaluation of chronic obstructive pulmonary disease, pulmonary function test (PFT) is a clinically recognized gold standard. Its core is to collect the expiratory flow rate and volume data of the subject's mouth blowing through a lung function instrument, generate a respiratory mechanics curve, and then calculate the forced vital capacity (FVC), forced expiratory volume in one second (FEV1), and one-second rate (FEV1 / FVC) and other key indicators. Among them, when the one-second rate is less than 70% in clinical practice, it usually indicates the presence of persistent airflow limitation, which can be used as an important diagnostic basis for COPD. However, the existing detection method based on the lung function instrument has obvious limitations in actual application:

[0003] On the one hand, the detection process requires the subject to hold the mouthpiece and perform a forced breathing action, which is difficult for the elderly and other groups with weak physical functions, leading to detection failure or data distortion.

[0004] On the other hand, the test scene is highly restricted and can only be carried out in hospitals or large medical institutions, which cannot meet the rapid detection needs in community screening, home daily monitoring and other scenarios. At the same time, professional medical personnel are required to guide the operation during the detection process, making it difficult to achieve contactless, low-cost, and convenient wide application. SUMMARY

[0005] Therefore, the embodiments of the present application provide a method and device for positioning a respiratory segment under the guidance of a voice instruction, an equipment and a medium to solve the problem that the current lung function detection equipment uses a contact type blowing collection method and relies on the detection scene and professional operation, which cannot realize contactless, low-cost, and convenient lung function detection.

[0006] In a first aspect, the embodiments of the present application provide a method for positioning a respiratory segment under the guidance of a voice instruction, which comprises:

[0007] Detecting the test action performed by the measured object according to the voice instruction using a radar to obtain a test signal;

[0008] Extracting the respiratory speed waveform and the respiratory signal of the chest and abdomen of the measured object during the test process based on the test signal;

[0009] Analyzing the respiratory signal to obtain a segment index representing the chest and abdomen of the measured object from inhaling to exhaling after the lungs are filled, extracting a target respiratory segment from the respiratory speed waveform according to the segment index, and calculating the respiratory index corresponding to the measured object based on the target respiratory segment.

[0010] Further, the analysis of the respiratory signal obtains an index of a segment of the chest and abdomen after the subject inhales to the lung fullness and exhales, comprising:

[0011] The respiratory signal is averaged according to the distance gate dimension to obtain a displacement signal, and the displacement signal is slidingly averaged according to the time dimension to obtain a smoothed respiratory signal;

[0012] Detecting a plurality of different types of key points in the smoothed respiratory signal;

[0013] Pairing based on a plurality of different types of key points to obtain a candidate key point combination;

[0014] According to a predetermined condition, a target signal segment representing the chest and abdomen after the subject inhales to the lung fullness and exhales is screened from the respiratory signal corresponding to the candidate key point combination, and the segment index corresponding to the target signal segment is extracted.

[0015] Further, the pairing based on a plurality of different types of key points to obtain a candidate key point combination, comprising:

[0016] Classifying a plurality of different types of key points to obtain a key point set corresponding to each type;

[0017] For each type of key point set, three adjacent key points are sequentially selected for combination and pairing to obtain a basic candidate unit;

[0018] Based on the pairing results of the adjacent three points of all types of keys, a candidate key point combination containing a complete respiratory cycle is constructed.

[0019] Further, the screening of the segment representing the chest and abdomen after the subject inhales to the lung fullness and exhales from the respiratory signal corresponding to the candidate key point combination according to the predetermined condition, comprising:

[0020] Obtaining the signal characteristics of the respiratory signal segment corresponding to each candidate key point combination;

[0021] Comparing the signal characteristics with the preset signal characteristics corresponding to the inhale-to-lung fullness and exhale action;

[0022] If the signal characteristics match the preset signal characteristics, the respiratory signal segment corresponding to the corresponding candidate key point combination is taken as the target signal segment.

[0023] Further, the extraction of the target respiratory segment from the respiratory speed waveform according to the segment index, comprising:

[0024] Determining the time range corresponding to the segment index in the respiratory signal;

[0025] extracting a corresponding target respiratory segment from the respiratory speed waveform according to the segment index.

[0026] Further, after extracting the target respiratory segment from the respiratory speed waveform according to the segment index, the method further comprises:

[0027] extracting an expiration segment in the target respiratory segment, and calculating a first integral value of a first respiratory speed signal in the expiration segment and a baseline;

[0028] calculating a second respiratory speed signal within one second after the start of the expiration segment, and calculating a second integral value of the second respiratory speed signal and the baseline;

[0029] calculating a one-second rate based on the first integral value and the second integral value.

[0030] Further, the method further comprises:

[0031] calculating an energy proportion of each regional respiratory speed waveform based on the chest and abdominal region signals corresponding to different distance gates in the test signal;

[0032] judging the current body position of the measured object based on the energy proportions of the regional respiratory speed waveforms;

[0033] extracting a stable segment signal before inspiration from the target respiratory segment according to the current body position, and calculating a mean value of the stable segment as an initial baseline under the current body position;

[0034] correcting the initial baseline by using a baseline offset empirical value corresponding to the current body position to obtain the baseline.

[0035] In a second aspect, an embodiment of the present application provides a respiratory segment positioning device under the guidance of a voice instruction, and the device comprises:

[0036] a test module configured to obtain a test signal by detecting a test action performed by a measured object according to a voice instruction by using a radar;

[0037] an extraction module configured to extract a respiratory speed waveform and a respiratory signal of a chest and abdominal region of the measured object in a test process based on the test signal;

[0038] an analysis module configured to analyze the respiratory signal to obtain a segment index representing the chest and abdominal region of the measured object from inspiration to expiration after lung filling, extract a target respiratory segment from the respiratory speed waveform according to the segment index, and calculate a respiratory index corresponding to the measured object based on the target respiratory segment.

[0039] In a third aspect, an embodiment of the present application provides a computer device, comprising a memory and a processor, which are communicatively connected with each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the method in the first aspect or any of the corresponding embodiments.

[0040] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, which stores computer instructions, and the computer instructions are used to make a computer perform the method in the first aspect or any of the corresponding embodiments.

[0041] The present application adopts radar non-contact detection technology, generates a test signal by capturing chest and abdominal respiration, does not need contact blowing operation, and discards the dependence on contact collection equipment, thereby solving the problem of high operation threshold. Secondly, the test action is guided by voice instructions, and the professional medical staff guidance is replaced, and the operation professional requirement is reduced through standardized instructions. Then, the radar is installed flexibly and is adapted to multiple postures, thereby breaking through the limitation of professional medical scene. Finally, the respiration segment is automatically extracted, and the respiration index of the measured object is calculated, without complex manual intervention, thereby simplifying the process and reducing the cost of equipment and use. BRIEF DESCRIPTION OF DRAWINGS

[0042] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the drawings needed in the specific embodiments or prior art description will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0043] Figure 1 is a flowchart of a respiration segment positioning method under voice instruction according to some embodiments of the present application;

[0044] Figure 2 is a schematic diagram of radar non-contact respiration test according to some embodiments of the present application;

[0045] Figure 3 is a schematic diagram of respiration speed waveform and respiration Doppler signal obtained after pre-processing of test signal according to some embodiments of the present application;

[0046] Figure 4 is a schematic diagram of target respiration segment of effective deep inhalation and fast exhalation according to some embodiments of the present application;

[0047] Figure 5 is a flowchart of another respiration segment positioning method according to some embodiments of the present application;

[0048] Figure 6is a schematic diagram of calculating a lung function index based on an exhalation segment according to some embodiments of the present application;

[0049] Figure 7 is a structural block diagram of a method and device for positioning a respiratory segment according to an embodiment of the present application;

[0050] Figure 8 is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present application. DETAILED DESCRIPTION

[0051] To make the objects, technical solutions and advantages of embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the present application.

[0052] According to embodiments of the present application, a method and device for positioning a respiratory segment under the guidance of a voice instruction, equipment and media are provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0053] In the present embodiment, a method for positioning a respiratory segment under the guidance of a voice instruction is provided, Figure 1 is a flowchart of a method for positioning a respiratory segment under the guidance of a voice instruction according to an embodiment of the present application, as shown in Figure 1 The flowchart includes the following steps:

[0054] In step S101, a radar is used to detect a test action performed by a measured object according to a voice instruction to obtain a test signal.

[0055] In the embodiment of the present application, in the test preparation stage, the radar installation position and the posture of the test object are determined according to the physical condition and comfort requirement of the test object, and a test scene that makes the test object relaxed in body and mind is preferentially selected. If the test object is inconvenient to move, a lying posture can be adopted, the radar is installed directly above the chest of the testee, for example, at a distance of 50 cm from the bed surface, is fixed through a support to keep the detection direction directly opposite the chest and abdominal region, and ensures that the radar beam can completely cover the chest and abdominal movement region. If the test object is in good condition, a sitting or standing posture can also be selected, and the radar is installed at a suitable height directly in front of the chest to avoid clothes wrinkles and surrounding obstacles from causing signal interference. After the test is started, the voice system will guide throughout the process according to the preset process. First, a relaxation prompt such as “please adjust the sitting posture, relax your whole body, and maintain natural breathing for 30 seconds” is played to make the breathing state of the test object stable. At this time, the radar starts to warm up and enters the signal collection standby mode.

[0056] When the test object's breathing is stable, the voice instruction will clearly issue the test action requirement, for example, “please inhale deeply to the lung and completely fill it, and then restore natural breathing”, and at the same time, the radar starts to collect signals synchronously, and real-time captures the slight displacement changes of the chest and abdominal parts caused by breathing. The whole test contains multiple groups of repeated deep inhalation and exhalation actions. Usually, 3 groups of core test actions are set. The voice system will insert a recovery prompt between each group of actions, such as “please breathe naturally for 5 times, and prepare for the next group of actions”, to ensure that the test object will not have rapid breathing due to continuous testing. During the collection process, the radar converts the echo signal into a test signal, transmits the data to the computer through a data transmission module, the implementation mode of the transmission module includes but is not limited to Bluetooth communication, WiFi connection, data line direct connection, etc., and anti-interference coding technology is adopted in the transmission process to avoid signal loss or distortion, and finally a complete test signal set containing the natural breathing segment, the deep inhalation and exhalation segment, and the action interval segment is formed, which provides the original basis for subsequent data processing.

[0057] For example, as shown in FIG. 1, Figure 2 As shown in FIG. 1, a scene of non-contact respiratory test using radar is shown: the testee lies on the bed, the radar is installed 50 cm away from the bed surface and directly above the chest of the testee, and monitors the chest movement of the testee through the emitted signal; at the same time, the voice device issues a standardized voice instruction “inhale…exhale…” to guide the testee to test; the radar echo signal is transmitted to the computer in real time through data transmission (such as Bluetooth), and the computer automatically extracts the deep inhalation and exhalation segment and calculates the lung function index.

[0058] In step S102, the breathing speed waveform and the breathing signal of the chest and abdominal part of the test object in the test process are extracted based on the test signal.

[0059] In the embodiments of the present application, the test signal received at the beginning is mixed with environmental noise (such as air flow, equipment running noise), human body motion interference (such as limb micro-motion, heartbeat vibration) and electromagnetic interference, so the first step is to filter out the noise.

[0060] After the signal is purified, the processed signal is subjected to frequency spectrum decomposition through time-frequency transformation, the time domain signal is converted into a time-frequency two-dimensional graph, the frequency component corresponding to the respiratory motion is identified from the graph, and the normal respiratory frequency is usually 0.1-0.3 Hz. The frequency of the deep inhalation and rapid exhalation action will appear a temporary increase, and accordingly the respiratory related signal interval can be preliminarily locked.

[0061] Then the displacement information corresponding to the respiratory motion is extracted through a phase demodulation algorithm, the displacement change is converted into the motion speed data of the chest and abdomen site in combination with the Doppler effect principle of the radar, and after data smoothing and standardization processing, a continuous respiratory speed waveform is generated. The peak value and the valley value of the waveform correspond to the maximum speed point and the stationary point of the respiratory motion respectively.

[0062] At the same time, the amplitude feature of the signal is extracted, the signal amplitude change caused by the respiratory motion is quantitatively processed, and after removing the abnormal fluctuation points, the respiratory signal is formed. The period and amplitude of the signal are directly related to the depth and frequency of the respiration. In the extraction process, the time stamp corresponding to the signal needs to be labeled synchronously, so as to ensure that the respiratory speed waveform and the respiratory signal are completely aligned in the time dimension.

[0063] Exemplarily, Figure 3 The respiratory speed waveform and the respiratory Doppler signal obtained after the pre-processing of the test signal are shown in the upper graph (respiratory speed waveform) and the lower graph (respiratory Doppler heat map) of Figure 3 The upper graph (respiratory speed waveform) shows that the blue curve presents the speed change waveform of the chest and abdomen of the measured person due to the respiration. The fluctuation of the waveform reflects the speed difference of the chest and abdomen motion in the respiration process, and the characteristic points such as the peak value and the valley value can be used to identify the phase of the respiratory action, such as the deep inhalation and rapid exhalation action which will show more obvious waveform fluctuation.

[0064] Figure 3 The lower graph (respiratory Doppler heat map) shows that the horizontal axis is time (second), the vertical axis is distance gate (which can be understood as different spatial regions monitored by the radar), and the respiratory Doppler signal intensity at different distance gates and different time points is presented through the color depth (the color scale is from-0.5 to 0.5). The color change reflects the motion characteristics of the chest and abdomen at different spatial positions and time points, which is helpful for analyzing the spatial distribution and time law of the respiratory motion from multiple dimensions.

[0065] Step S103, analyze the respiratory signal to obtain a segment index representing the chest and abdomen after the subject inhales to the lung filling and exhales, and extract the target respiratory segment from the respiratory speed waveform according to the segment index, and calculate the corresponding respiratory index of the subject based on the target respiratory segment.

[0066] In the embodiments of the present application, analyzing the respiratory signal to obtain a segment index representing the chest and abdomen after the subject inhales to the lung filling and exhales includes the following steps A1-A4:

[0067] Step A1, average the respiratory signal according to the distance gate dimension to obtain a displacement signal, and average the displacement signal according to the time dimension to obtain a smoothed respiratory signal.

[0068] Specifically, the respiratory signal is generated by radar multi-distance gate collection, and each distance gate corresponds to a different spatial position of the chest and abdomen of the subject. Due to the slight difference in the amplitude of chest and abdominal movement at different positions, and the interference noise between distance gates mixed in the signal, directly using the original signal will cause feature extraction deviation. Therefore, first, the distance gate dimension average processing is performed: a preset signal fusion algorithm is called, all effective distance gate collected respiratory signal data is traversed, the signal amplitude of different distance gates at the same timestamp is subjected to arithmetic average operation, the local noise interference is offset through the fusion of multi-space dimension data, and the core features of the overall chest and abdominal movement are retained, and finally a continuous displacement signal is output. The amplitude change of the signal directly corresponds to the displacement fluctuation of the chest and abdomen due to breathing.

[0069] After distance gate averaging, the time domain noise and random fluctuations in the displacement signal need to be further eliminated, which requires sliding average processing in the time dimension. First, set the sliding window parameters, combine the physiological characteristics of the respiratory signal (normal breathing period is about 3-5 seconds, deep inhalation and fast exhalation period is about 3 seconds), set the window length to 0.2 seconds, and the window sliding step to 0.05 seconds, to ensure that the key details of the respiratory action are not lost while the noise is smoothed. During processing, take the time axis as the reference, and make the sliding window move along the displacement signal. Weighted average operation is performed on all signal points in each window, and the signal points in the center of the window have the highest weight, which gradually decreases towards both ends. By this weighting method, the phase shift in the signal smoothing process is reduced. When the window slides over the entire displacement signal, the generated smoothed respiratory signal will exhibit continuous and smooth waveform characteristics. The original burr and jump caused by respiratory airflow disturbance and radar signal slight fluctuation are effectively filtered out, and the rising edge, falling edge and peak position of the chest and abdominal movement are more clear. This provides high signal-to-noise ratio data support for accurately identifying various key points in the subsequent steps. The entire processing process is automatically completed by the signal processing module, and the processing result is stored in real time and synchronized to the subsequent analysis unit.

[0070] Step A2, detect a plurality of different types of key points in the smoothed respiratory signal.

[0071] Firstly, the key point types to be detected are determined, and in combination with the waveform characteristics and physiological significance of the respiratory signal, five types of key points including zero-crossing points, peak points, valley points, and upper 30% and lower 30% quantile points are determined. The zero-crossing point corresponds to the transition moment of the respiratory action (inspiration to expiration or expiration to inspiration), the peak point and the valley point correspond to the limit positions of lung filling and emptying respectively, and the upper and lower 30% quantile points are used to assist in defining the start and end stages of the respiratory action. Before detection, the smoothed respiratory signal needs to be preprocessed, and the signal amplitude is uniformly mapped to the 0-1 interval through mean normalization to eliminate the detection deviation caused by the signal amplitude difference of different measured objects, and a first-order difference algorithm is used to calculate the signal rate to provide rate characteristic basis for key point recognition.

[0072] According to the characteristics of different types of key points, different detection algorithms are used:

[0073] For the zero-crossing point, the critical point where the signal amplitude changes from positive to negative or from negative to positive is located by traversing the signal sequence, and the sign change of the first-order difference signal is verified. If the sign of the first-order difference signal jumps simultaneously, it is confirmed that the point is a valid zero-crossing point, avoiding false judgment caused by small fluctuations in the signal.

[0074] For the peak point and the valley point, a sliding window extreme value detection method is used. A 0.3-second window is used to traverse the signal. When the signal amplitude at the center position of the window is greater than (or less than) all other points in the window, and the first-order difference signal of the point changes from positive to negative (or from negative to positive) and the second-order difference signal is less than (or greater than) zero, it is determined as a peak point (or a valley point).

[0075] For the upper and lower 30% quantile points, the maximum and minimum values of the signal amplitude are first calculated to determine the amplitude dynamic range, and then the amplitude thresholds corresponding to 30% and 70% of the range are calculated. When traversing the signal, the points first reaching the threshold and first leaving the threshold are located, and the signal rate is used to confirm the quantile point position.

[0076] All detected key points need to record their accurate time stamps and amplitude information to form a classified key point sequence. At the same time, abnormal points are removed through neighborhood verification. If the time interval between a key point and its adjacent key point of the same type is less than 0.1 second, it is determined as a false key point and deleted. The final output key point sequence will fully reflect the waveform characteristics of the respiratory signal.

[0077] Step A3, pairing based on multiple different types of key points to obtain a candidate key point combination.

[0078] Specifically, pairing is performed based on multiple different types of key points to obtain candidate key point combinations, including: classifying the multiple different types of key points to obtain a key point set corresponding to each type; for the key point set corresponding to each type, sequentially selecting three adjacent key points for combination and pairing to obtain a basic candidate unit; and based on the pairing results of the three adjacent key points of all types of key points, a candidate key point combination containing a complete respiratory cycle is constructed.

[0079] The multiple different types of detected key points (such as zero-crossing points, peak points, and 30% upper quantile points) are classified and arranged according to types, and key point sets corresponding to respective types are divided according to the characteristic attributes of the key points (such as zero-crossing points representing respiratory conversion time and peak points representing lung filling limit). The key points in each set retain their original time stamp, amplitude, and signal position information, ensuring the integrity of the characteristics of the classified key points of each type. Then, for the key point set corresponding to each type, the key points in the set are sorted according to the time stamp in chronological order. Then, three adjacent key points are sequentially selected as a group for combination and pairing to form a basic candidate unit. For example, in the peak point set, if the sorted order is P1, P2, P3, P4, …, then the basic candidate units (P1, P2, P3) and (P2, P3, P4) are constructed, and each unit corresponds to the characteristic changes of the key points of the type in a continuous time interval. Finally, the pairing results of the three adjacent key points of all types of key points are summarized, and the physiological laws of the respiratory cycle (such as a complete respiratory cycle containing key stages such as the start of inspiration, inspiration process, end of inspiration, expiration process, and end of expiration) are combined to associate and integrate the basic candidate units of different types. Invalid combinations that are not continuous in time and cannot form a complete respiratory stage are removed, and finally a candidate key point combination that can completely cover the entire process of “inspiration-expiration” or “expiration-inspiration” is constructed, ensuring that each combination corresponds to a signal segment with complete respiratory characteristics, providing comprehensive candidate samples for subsequent screening of deep inspiration and rapid expiration segments.

[0080] Step A4: filtering a target signal segment representing the chest and abdomen of the measured object from inspiration to lung filling and then expiration from the candidate key point combination corresponding respiratory signal according to the preset condition, and extracting the segment index corresponding to the target signal segment.

[0081] Specifically, the segment representing the chest and abdomen of the measured object from inspiration to lung filling and then expiration is filtered from the candidate key point combination corresponding respiratory signal according to the preset condition, including: obtaining the signal characteristics of the signal segment corresponding to each candidate key point combination; comparing the signal characteristics with the preset signal characteristics corresponding to the action from inspiration to lung filling and then expiration; if the signal characteristics match the preset signal characteristics, the respiratory signal segment corresponding to the corresponding candidate key point combination is taken as the target signal segment.

[0082] Firstly, for each candidate key point combination, the corresponding signal segment is accurately cut from the complete respiratory signal according to the timestamp information of each key point in the combination, and then the core signal features of the segment are extracted, including time features (inspiration phase duration, expiration phase duration and total duration of the segment), amplitude features (amplitude rise amplitude in the inspiration process, difference between peak point amplitude and valley point amplitude, and proportion of peak point relative to the overall maximum value of the signal) and rate features (average rising rate of signal amplitude in the inspiration phase, average falling rate of signal amplitude in the expiration phase, and respiratory rate ratio), and these feature parameters are arranged into a standardized feature vector.

[0083] Then, the pre-stored preset signal feature threshold corresponding to the "inspiration to lung filling after expiration" action is called, which is based on a large number of clinical sample data statistics, for example, the inspiration phase duration needs to be 0.8-1.5 seconds, the expiration phase duration needs to be 1.2-3.5 seconds, the peak and valley amplitude difference needs to be more than 1.8 times the average amplitude difference of the signal, and the expiration rate and inspiration rate ratio needs to be greater than 1.2.

[0084] Finally, the extracted feature vector of each candidate segment is compared with the preset feature threshold one by one, if all the feature parameters of a segment fall within the preset threshold range, and the key features (such as peak amplitude proportion, respiratory rate ratio) meet the core matching conditions, it is determined that the signal features of the segment match the target action features, and it is determined as the target signal segment representing "inspiration to lung filling after expiration", and the corresponding candidate key point combination information and signal start and end position of the segment are recorded.

[0085] In the embodiments of the present application, the target respiratory segment is extracted from the respiratory speed waveform according to the segment index, including the following steps B1-B2:

[0086] Step B1, determine the time range corresponding to the segment index in the respiratory signal.

[0087] Specifically, first, the basic parameters during respiratory signal acquisition are called, including signal sampling frequency (such as 100 Hz representing 100 signal points collected per second) and collection start timestamp, to establish the "segment index-time" conversion formula - the start time of the time range = the collection start time + (the start index ÷ the sampling frequency), and the end time = the collection start time + (the end index ÷ the sampling frequency). Then, the segment index corresponding to the target signal segment (such as z1 for the start index and z2 for the end index) is extracted, and the specific range of the index interval on the time axis is calculated according to the above formula, for example, if the sampling frequency is 100 Hz, the collection start time is 0 seconds, the start index z1 = 500, and the end index z2 = 1200, then the corresponding time range is 5 seconds to 12 seconds.

[0088] Meanwhile, the calculated time range needs to be verified by combining the timestamp record of the respiratory signal. The timestamp of the key points in the signal (such as the time information of the peak point and the zero-crossing point) in the index interval is compared. If the error is within ±0.01 seconds, the time range is confirmed to be valid. If the error exceeds the threshold, the sampling frequency and index value are rechecked to ensure that the correspondence between the segment index and the respiratory signal time range is accurate and error-free, providing accurate time positioning basis for subsequent extraction of respiratory speed waveform segments.

[0089] Step B2, the corresponding time range of the segment index is used to extract the corresponding target respiratory segment from the respiratory speed waveform.

[0090] Specifically, first, the respiratory speed waveform data is sorted by timestamp to construct a waveform time sequence completely synchronized with the respiratory signal time axis, ensuring one-to-one correspondence in the time dimension. Then, according to the time range obtained in step B1 (such as 5 seconds to 12 seconds), the start and end positions of the interval are located in the time sequence of the respiratory speed waveform, and the corresponding waveform data point index is marked - if the waveform sampling frequency is consistent with the respiratory signal, the segment index z1, z2 can be directly used for positioning; if the sampling frequency is different, the corresponding index of the waveform data is obtained through time proportion conversion.

[0091] Subsequently, based on the located index interval, the data in this segment is extracted from the complete respiratory speed waveform, while the amplitude, rate of change and other core feature information of the waveform are preserved. During the extraction process, a lossless data cutting method should be used to avoid waveform distortion caused by interpolation or compression. After extraction, the target respiratory segment also needs to be verified for integrity, checking whether the segment contains the complete waveform characteristics corresponding to the respiratory action (such as the rising edge, peak point in the inspiration stage and the falling edge in the expiration stage), and comparing the segment with the corresponding respiratory signal segment for time synchronization to ensure that they match completely in the action stage division. Finally, the target respiratory segment containing time markers and feature parameters is output.

[0092] An exemplary, Figure 4 The target respiratory segment representing effective deep inhalation and rapid exhalation after condition screening is shown in Table 1, which includes three deep inhalation and rapid exhalation segments: period 1, period 2, and period 3. Each period is labeled with the inspiration and expiration duration (e.g., period 1 inspiration 6.3s, expiration 1.1s). After extracting each respiratory segment through the corresponding index of the deep inhalation and rapid exhalation segment, the inspiration and expiration segments can be distinguished, and the respiratory speed waveform characteristics of each period are presented.

[0093] In the embodiments of the present application, after the target respiratory segment is extracted, the respiratory frequency (unit: times / minute) can be calculated by counting the number of target segments and the total duration, that is, the total number of segments divided by the total duration and converted to minute dimension; the respiratory depth can also be calculated by extracting the maximum and minimum values of the respiratory speed in each target segment, that is, the absolute value of the difference between the two values; at the same time, the integral area of the speed waveform in the segment can be further analyzed to obtain the estimated value of the ventilation of each breath, or the respiratory rhythm stability can be evaluated by calculating the coefficient of variation of the waveform to obtain the respiratory index of the measured object.

[0094] The present application adopts radar non-contact detection technology to capture the test signal generated by the respiratory movement of the chest and abdomen without the need for contact blowing operation, thereby eliminating the dependence on contact collection equipment and solving the problem of high operation threshold. Secondly, the test action is guided by voice instructions to replace the guidance of professional medical personnel, and the operation professional requirement is reduced through standardized instructions. Then, the radar is installed flexibly and is suitable for multiple postures, thereby breaking through the limitation of professional medical scenes. Finally, the respiratory segment is automatically extracted and the respiratory index of the measured object is calculated without the need for complex manual intervention, thereby simplifying the process and reducing the cost of equipment and use.

[0095] In the embodiments of the present application, after the target respiratory segment is extracted from the respiratory speed waveform according to the segment index, as shown in Figure 5 the method further includes:

[0096] Step S201, an exhalation segment is extracted from the target respiratory segment, and a first integral value of the first respiratory speed signal in the exhalation segment and the baseline is calculated.

[0097] First, the exhalation segment is divided according to the characteristics of the respiratory speed waveform from the determined target respiratory segment of the “deep inhale fast exhale” mode, and the exhalation segment is combined with the respiratory physiological law and the waveform characteristics. When the respiratory speed waveform starts to show a rapid downward trend from the peak point and crosses the baseline y=0 until the next valley point, the interval is the exhalation segment. The start and end positions of the interval are accurately positioned by the segment index and the extraction is completed. Then, taking the baseline y=0 as the reference datum, a numerical integral algorithm (such as the trapezoidal integral method) is called to perform integral operation on the respiratory speed signal in the exhalation segment: the time axis of the exhalation segment is divided into a plurality of small time intervals, the trapezoidal area surrounded by the respiratory speed signal and the baseline y=0 in each interval is calculated, and the preliminary integral result is obtained by accumulating all the trapezoidal areas. Since the integral result may present a negative value due to the direction of the signal, the absolute value thereof is taken, which is the first integral value representing the forced vital capacity (FVC). During the calculation process, sufficient numerical precision needs to be preserved to ensure the accuracy of the index.

[0098] Step S202, a second respiratory speed signal in the first second of the exhalation segment is calculated, and a second integral value of the second respiratory speed signal and the baseline is calculated.

[0099] In a specific implementation, the precise start time stamp of the exhalation segment (i.e. the time corresponding to the peak point of the waveform) is first determined, and then a signal interval of 1 second in length is intercepted from this starting point, obtaining a second respiratory rate signal. If the total length of the exhalation segment is less than 1 second, the entire exhalation segment signal is used as the second respiratory rate signal, and the actual length of the intercepted signal is recorded for subsequent description. Then, the baseline y = 0 is used as a reference, and the signal in the 1 second is integrated: the 1 second time axis is also subdivided into multiple small intervals, the area enclosed by the signal and the baseline in each interval is calculated and accumulated, and the absolute value of the integral result of the interval is obtained. This absolute value is the second integral value representing the forced expiratory volume in one second (FEV1). After the calculation is completed, the signal length corresponding to the integral value needs to be marked, and if it is less than 1 second, the data record needs to be clearly stated to ensure the integrity and traceability of the index.

[0100] In step S203, the one-second rate is calculated based on the first integral value and the second integral value.

[0101] Based on the first integral value (FVC) and the second integral value (FEV1), the effectiveness is checked: if FVC is 0, the integral result is invalid, and the exhalation segment needs to be re-extracted and the integral needs to be calculated; if the signal length corresponding to FEV1 is less than 1 second, the note "FEV1 is calculated based on less than 1 second signal" needs to be marked in the calculation result. After the check is passed, the one-second rate is calculated according to the definition of the one-second rate, i.e. one-second rate = second integral value (FEV1) ÷ first integral value (FVC), and two decimal places are retained during the operation to meet the accuracy requirements of clinical diagnosis. After the calculation is completed, the one-second rate is compared with the reference threshold value (usually 70%) of COPD clinical diagnosis, and the specific value and the corresponding diagnostic reference opinion are output, providing data support for subsequent medical diagnosis, and the specific values of FVC, FEV1 and one-second rate, calculation time and other information are also stored in the database.

[0102] An example is shown in the following table: Figure 6 The process of calculating lung function indicators based on the exhalation segment is shown in the following table: Figure 6 The upper graph in the following table is a respiratory rate waveform, with the baseline y = 0 as a reference, and the deep inhalation and rapid exhalation segment is marked for positioning the exhalation segment; Figure 6 The lower graph in the following table presents the integral operation visually, including the forced expiratory volume in one second (FEV1) and the forced vital capacity (FVC), and the one-second rate (FEV1 / FVC) can be calculated from these two parameters to assist in COPD diagnosis.

[0103] In the embodiment of the present application, the method further comprises: calculating the energy proportion of the respiratory speed waveform of each region based on the chest and abdominal region signals corresponding to different distance gates in the test signal; judging the current body position of the measured object based on the energy proportion of the respiratory speed waveform of each region; extracting the stable segment signal before inspiration from the target respiratory segment according to the current body position, and calculating the mean value of the stable segment as the initial baseline under the current body position; and correcting the initial baseline by using the baseline offset empirical value corresponding to the current body position to obtain the baseline.

[0104] Specifically, first, for the chest and abdominal region signals corresponding to different distance gates in the test signal, the energy value of the respiratory speed waveform of each region is calculated by an energy calculation method such as Fourier transform, and then the energy proportion is obtained by the ratio of the energy value of each region to the total energy; then, the current body position of the measured object is judged according to the distribution characteristics of the energy proportion of each region (such as the difference in energy proportion of different regions of the chest and abdomen under lying, sitting, and standing positions); then, the stable breathing stage before the inspiration action occurs is positioned from the target respiratory segment according to the determined current body position, the stable segment signal is extracted, and the mean value is calculated by using a sliding average algorithm as the initial baseline under the current body position; finally, the baseline offset empirical value corresponding to the current body position (the empirical value is obtained by statistical analysis of a large amount of test data under the same body position) is called, and the initial baseline is operated (such as addition or adjustment) with the empirical value to correct the initial baseline, and the final baseline used for lung function index calculation is obtained.

[0105] The embodiment of the present application uses the energy proportion calculated by the signals of different distance gate regions to judge the body position, which can accurately adapt to different states of the measured object such as lying, sitting, etc., and avoid baseline deviation caused by body position difference; secondly, the stable segment signal before inspiration is extracted and the mean value is used as the initial baseline, which ensures that the baseline fits the current respiratory basic state and reduces the interference of respiratory fluctuation; finally, the initial baseline is corrected by combining the baseline offset empirical value corresponding to the body position, which further offsets the system error related to the body position, makes the baseline more fit the actual physiological scene, and effectively improves the accuracy of subsequent lung function index calculation.

[0106] In the embodiment, a respiratory segment positioning device under voice instruction guidance is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments, and will not be described again. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware, or a combination of software and hardware is also possible and contemplated.

[0107] The embodiment provides a respiratory segment positioning device under voice instruction guidance, as shown in Figure 7 , comprising:

[0108] The test module 701 is configured to detect a test action performed by the object under test according to the voice instruction by using the radar, and obtain a test signal.

[0109] The extraction module 702 is configured to extract a respiratory speed waveform of the chest and abdomen of the object under test in the test process and a respiratory signal based on the test signal.

[0110] The analysis module 703 is configured to analyze the respiratory signal to obtain a segment index representing a segment from inhaling to exhaling after the chest and abdomen of the object under test are filled with air, and extract a target respiratory segment from the respiratory speed waveform according to the segment index.

[0111] In the embodiment of the present application, the analysis module 703 is configured to average the respiratory signal according to the distance gate dimension to obtain a displacement signal, and average the displacement signal according to the time dimension to obtain a smoothed respiratory signal; detect a plurality of different types of key points in the smoothed respiratory signal; pair the plurality of different types of key points to obtain a candidate key point combination; and filter a target signal segment representing a segment from inhaling to exhaling after the chest and abdomen of the object under test are filled with air from the respiratory signal corresponding to the candidate key point combination according to a preset condition, and extract a segment index corresponding to the target signal segment.

[0112] In the embodiment of the present application, the analysis module 703 is specifically configured to classify the plurality of different types of key points to obtain a key point set corresponding to each type; and for the key point set corresponding to each type, sequentially select three adjacent key points to combine and pair to obtain a basic candidate unit; and based on the pairing results of the adjacent three points corresponding to all types of key points, construct a candidate key point combination containing a complete respiratory cycle.

[0113] In the embodiment of the present application, the analysis module 703 is specifically configured to obtain a signal feature of a signal segment corresponding to each candidate key point combination; compare the signal feature with a preset signal feature corresponding to the action from inhaling to exhaling after the chest and abdomen of the object under test are filled with air; and if the signal feature matches the preset signal feature, take the signal segment corresponding to the corresponding candidate key point combination as the target signal segment.

[0114] In the embodiment of the present application, the analysis module 703 is specifically configured to determine a time range corresponding to the segment index in the respiratory signal; and extract a corresponding target respiratory segment from the respiratory speed waveform by using the time range corresponding to the segment index.

[0115] In the embodiment of the present application, the device further includes a first calculation module configured to extract an exhale segment in the target respiratory segment, calculate a first integral value of a first respiratory speed signal in the exhale segment and a baseline, calculate a second respiratory speed signal in the first second of the start of the exhale segment, calculate a second integral value of the second respiratory speed signal and the baseline, and calculate a one-second rate based on the first integral value and the second integral value.

[0116] In this embodiment of the application, the device further includes: a second calculation module, used to calculate the energy proportion of the respiratory velocity waveform of each region based on the chest and abdominal region signals corresponding to different distance gates in the test signal; determine the current body position of the test subject based on the energy proportion of the respiratory velocity waveform of each region; extract the steady segment signal before inspiration from the target respiratory segment according to the current body position, and calculate the mean of the steady segment as the initial baseline under the current body position; and correct the initial baseline using the baseline offset empirical value corresponding to the current body position to obtain the baseline.

[0117] Please see Figure 8 , Figure 8 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 8 As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system).

[0118] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0119] The memory 20 stores instructions executable by at least one processor 10 to cause the at least one processor 10 to perform the method shown in the above embodiments.

[0120] The memory 20 can include a program storage area and a data storage area. The program storage area can store an operating system and applications required by at least one function. The data storage area can store data created by the computer device according to the presentation of a small program landing page, and the like. In addition, the memory 20 can include a high-speed random access memory, and can also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some optional embodiments, the memory 20 can optionally include a memory disposed remotely relative to the processor 10, and these remote memories can be connected to the computer device through a network. Examples of the network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0121] The memory 20 can include a volatile memory, such as a random access memory, and can also include a non-volatile memory, such as a flash memory, a hard disk, or a solid-state disk. The memory 20 can also include a combination of the above-mentioned kinds of memories.

[0122] The computer device further includes a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0123] The embodiments of the present application also provide a computer readable storage medium. The above-mentioned method according to the embodiments of the present application can be implemented in hardware, firmware, or recorded in a storage medium, or stored in a remote storage medium or a non-transitory machine readable storage medium and downloaded to a local storage medium through network downloading, so that the method described herein can be processed by such software on a storage medium using a general-purpose computer, a special-purpose processor, or programmable or special-purpose hardware. The storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid-state disk, and the like. Further, the storage medium can also include a combination of the above-mentioned kinds of memories. It can be understood that the computer, the processor, the microprocessor controller, or the programmable hardware includes a storage component that can store or receive software or computer code, and when the software or computer code is accessed and executed by the computer, the processor, or the hardware, the method shown in the above embodiments is implemented.

[0124] Although the embodiments of the present application are described in conjunction with the accompanying drawings, various modifications and changes can be made by those skilled in the art without departing from the spirit and scope of the present application, and such modifications and changes fall within the scope defined by the appended claims.

Claims

1. A method for locating a respiratory segment under voice command guidance, characterized in that, The method comprises: detecting a test action performed by a measured object according to a voice instruction by using a radar to obtain a test signal; extracting a breathing speed waveform and a breathing signal of a chest and abdomen of the measured object during a test based on the test signal; analyzing the breathing signal to obtain a segment index representing a segment in which the chest and abdomen of the measured object exhale after the measured object inhales to fill the lungs, extracting a target breathing segment from the breathing speed waveform according to the segment index, and calculating a breathing index corresponding to the measured object based on the target breathing segment; wherein the analysis of the breathing signal to obtain the segment index representing the segment in which the chest and abdomen of the measured object exhale after the measured object inhales to fill the lungs comprises: averaging the breathing signal according to a distance gate dimension to obtain a displacement signal, and performing time dimension sliding average on the displacement signal to obtain a smoothed breathing signal; detecting a plurality of different types of key points in the smoothed breathing signal; pairing the plurality of different types of key points to obtain a candidate key point combination; filtering a target signal segment representing the segment in which the chest and abdomen of the measured object exhale after the measured object inhales to fill the lungs from the breathing signal corresponding to the candidate key point combination according to a preset condition, and extracting a segment index corresponding to the target signal segment.

2. The method of claim 1, wherein, The pairing of the plurality of different types of key points to obtain the candidate key point combination comprises: classifying the plurality of different types of key points to obtain a key point set corresponding to each type; for each type of key point set, sequentially selecting three adjacent key points to form a basic candidate unit; based on the pairing results of the adjacent three key points of all types of key points, a candidate key point combination containing a complete breathing cycle is constructed.

3. The method of claim 1, wherein, The filtering of the segment in which the chest and abdomen of the measured object exhale after the measured object inhales to fill the lungs from the breathing signal corresponding to the candidate key point combination according to the preset condition comprises: obtaining a signal feature of a breathing signal segment corresponding to each candidate key point combination; comparing the signal feature with a preset signal feature corresponding to the action of inhaling to fill the lungs and exhaling; if the signal feature matches the preset signal feature, the breathing signal segment corresponding to the corresponding candidate key point combination is taken as the target signal segment.

4. The method of claim 1, wherein, The extraction of the target breathing segment from the breathing speed waveform according to the segment index comprises: determining a time range corresponding to the segment index in the breathing signal; extracting a corresponding target breathing segment from the breathing speed waveform by using the time range corresponding to the segment index.

5. The method of claim 1, wherein, After the extraction of the target breathing segment from the breathing speed waveform according to the segment index, the method further comprises: extracting an exhalation segment in the target breathing segment, and calculating a first integral value of a first breathing speed signal in the exhalation segment and a baseline; calculating a second breathing speed signal within one second at the beginning of the exhalation segment, and calculating a second integral value of the second breathing speed signal and the baseline; calculating a one-second rate based on the first integral value and the second integral value.

6. The method of claim 5, wherein, The method further comprises: Based on the chest and abdomen region signals corresponding to different distance gates in the test signal, the energy proportion of each region respiratory speed waveform is calculated; Based on the energy proportion of each region respiratory speed waveform, the current body position of the measured object is determined; According to the current body position, a stable segment signal before inspiration is extracted from the target respiratory segment, and the mean value of the stable segment is calculated as the initial baseline under the current body position; The initial baseline is corrected by using the baseline offset empirical value corresponding to the current body position, to obtain the baseline.

7. An apparatus for locating breath segments under voice command guidance, the apparatus comprising: a microphone for receiving voice commands; a processor for processing the voice commands; a memory for storing a plurality of breath segments; and a display for displaying the plurality of breath segments. The device comprises: A test module for detecting the test action performed by the measured object according to the voice instruction by using radar to obtain a test signal; An extraction module for extracting the respiratory speed waveform and the respiratory signal of the chest and abdomen of the measured object during the test based on the test signal; An analysis module for analyzing the respiratory signal to obtain a segment index representing the chest and abdomen of the measured object after inspiration to lung filling and exhalation, extracting a target respiratory segment from the respiratory speed waveform according to the segment index, and calculating the respiratory index corresponding to the measured object based on the target respiratory segment; wherein the analysis of the respiratory signal to obtain the segment index representing the chest and abdomen of the measured object after inspiration to lung filling and exhalation comprises: The respiratory signal is averaged according to the distance gate dimension to obtain a displacement signal, and the displacement signal is averaged according to the time dimension to obtain a smoothed respiratory signal; Multiple different types of key points in the smoothed respiratory signal are detected; Based on multiple different types of key points, a candidate key point combination is obtained; According to a predetermined condition, a target signal segment representing the chest and abdomen of the measured object after inspiration to lung filling and exhalation is screened from the respiratory signal corresponding to the candidate key point combination, and the segment index corresponding to the target signal segment is extracted.

8. A computer device, comprising: It comprises: A memory and a processor, which are communicatively connected, the memory stores computer instructions, and the processor executes the computer instructions to perform the method of any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions, which are used to make the computer execute the method of any one of claims 1 to 6.

Images