CN118133209B - Data processing method based on breath trainer data analysis - Google Patents
Data processing method based on breath trainer data analysis Download PDFInfo
- Publication number
- CN118133209B CN118133209B CN202410547588.6A CN202410547588A CN118133209B CN 118133209 B CN118133209 B CN 118133209B CN 202410547588 A CN202410547588 A CN 202410547588A CN 118133209 B CN118133209 B CN 118133209B
- Authority
- CN
- China
- Prior art keywords
- data
- data segment
- period
- segment
- airflow
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000007405 data analysis Methods 0.000 title claims abstract description 13
- 238000003672 processing method Methods 0.000 title claims abstract description 11
- 230000002159 abnormal effect Effects 0.000 claims abstract description 74
- 230000005856 abnormality Effects 0.000 claims abstract description 66
- 238000012545 processing Methods 0.000 claims abstract description 21
- 238000001514 detection method Methods 0.000 claims abstract description 15
- 238000012937 correction Methods 0.000 claims abstract description 14
- 230000000737 periodic effect Effects 0.000 claims description 133
- 238000000034 method Methods 0.000 claims description 27
- 230000008569 process Effects 0.000 claims description 17
- 230000000241 respiratory effect Effects 0.000 claims description 16
- 239000013598 vector Substances 0.000 claims description 14
- 238000010606 normalization Methods 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 10
- 230000011218 segmentation Effects 0.000 claims description 4
- 230000009466 transformation Effects 0.000 claims description 3
- 230000029058 respiratory gaseous exchange Effects 0.000 description 23
- 230000008859 change Effects 0.000 description 12
- 230000001788 irregular Effects 0.000 description 5
- 230000004199 lung function Effects 0.000 description 5
- 238000011156 evaluation Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000003247 decreasing effect Effects 0.000 description 3
- 210000004072 lung Anatomy 0.000 description 3
- 230000000630 rising effect Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 208000023504 respiratory system disease Diseases 0.000 description 2
- 208000000884 Airway Obstruction Diseases 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 208000006673 asthma Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000035565 breathing frequency Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002040 relaxant effect Effects 0.000 description 1
- 230000004202 respiratory function Effects 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000000758 substrate Substances 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B23/00—Exercising apparatus specially adapted for particular parts of the body
- A63B23/18—Exercising apparatus specially adapted for particular parts of the body for improving respiratory function
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B2230/00—Measuring physiological parameters of the user
- A63B2230/40—Measuring physiological parameters of the user respiratory characteristics
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B2230/00—Measuring physiological parameters of the user
- A63B2230/40—Measuring physiological parameters of the user respiratory characteristics
- A63B2230/42—Measuring physiological parameters of the user respiratory characteristics rate
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- Physical Education & Sports Medicine (AREA)
- General Health & Medical Sciences (AREA)
- Pulmonology (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention relates to the technical field of data processing, in particular to a data processing method based on data analysis of a breath trainer, which comprises the following steps: acquiring airflow flow data, and carrying out sectional processing on the airflow flow data to obtain data segments of each period; obtaining characteristic parameters of each period data segment according to the data distribution fluctuation condition and the data difference condition of the left side and the right side of each airflow flow data in each period data segment; obtaining the data abnormality degree of each period data segment according to the difference between the characteristic parameters and the whole of each period data segment and the distribution amplitude of the airflow flow data in the period data segment; acquiring the data distance between each period data segment and the period data segment in the neighborhood; and acquiring an abnormal period data segment by using an abnormal detection algorithm based on the data distance, and correcting the airflow data in the abnormal period data segment by using the data distance between the abnormal period data segment and the period data segment in the neighborhood. The invention can obtain more accurate correction data.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a data processing method based on data analysis of a respiratory trainer.
Background
Respiratory rehabilitation training is an important rehabilitation means for improving respiratory function and preventing and treating respiratory diseases. The respiratory rehabilitation training is carried out through the respiratory trainer, so that a comprehensive respiratory rehabilitation solution can be provided, the patient is helped to improve the lung function, the life quality is improved, and the occurrence of respiratory diseases is reduced. Is widely applied in the medical field or in usual life. Therefore, the related data collected by the breath trainer has wide application, and it is particularly important to ensure the accuracy of the feedback data of the breath trainer.
Under normal conditions, there may be a case that the breathing trainer is not worn normally, and thus air leakage is caused, so that the collected data is abnormal, and abnormal detection and correction of feedback data of the breathing trainer are required. In the prior art, the feedback data is usually subjected to abnormality detection through an abnormality detection algorithm, but due to the fact that the feedback data with irregular wear has certain aggregation, the abnormality score for carrying out abnormality evaluation through the abnormality detection algorithm is relatively inaccurate, so that an abnormality detection result is relatively inaccurate, and further, the result of data correction is relatively inaccurate.
Disclosure of Invention
In order to solve the technical problems that an abnormal detection result is inaccurate and a data correction result is inaccurate, the invention aims to provide a data processing method based on data analysis of a breath trainer, and the adopted technical scheme is as follows:
Acquiring airflow data of the respiratory trainer at each moment in a set time period, and carrying out sectional processing on the airflow data according to the distribution characteristics of the airflow data at each moment to obtain each period data segment;
Obtaining characteristic parameters of each period data segment according to the data distribution fluctuation condition and the data difference condition of the left side and the right side of each airflow flow data in each period data segment;
Obtaining the data abnormality degree of each period data segment according to the difference between the characteristic parameters and the whole of each period data segment and the distribution amplitude of the airflow flow data in the period data segment;
Obtaining the data distance between the periodic data segments according to the difference distribution condition of the data abnormality degree of each periodic data segment and the periodic data segment in the neighborhood and the difference condition of the characteristic parameters of each periodic data segment and the periodic data segment in the neighborhood;
and acquiring an abnormal period data segment by using an abnormal detection algorithm based on the data distance, and correcting the airflow data in the abnormal period data segment by using the data distance between the abnormal period data segment and the period data segment in the neighborhood.
Preferably, the obtaining the characteristic parameter of each period data segment according to the data distribution fluctuation condition and the data difference condition of the left side and the right side of each airflow flow data in each period data segment specifically includes:
Taking any one airflow data in any one period data segment as target airflow data, and obtaining a stability coefficient of the target airflow data according to the data difference between two adjacent airflow data on the left side of the target airflow data in the period data segment and the fluctuation condition of the data difference;
Calculating the difference value between each airflow data and the next adjacent airflow data in the period data segment to obtain difference data; taking the quantity ratio of the difference data on the right side of the target airflow flow data, which is smaller than the corresponding difference data of the preset difference threshold value, as a trend coefficient of the target airflow flow data;
And taking the binary group formed by the stable coefficient and the trend coefficient as a characteristic vector of the target airflow flow data, and obtaining the characteristic parameters of the periodic data segment according to the data distribution characteristic of the characteristic vector of each airflow flow data in the periodic data segment.
Preferably, the obtaining the stability coefficient of the target airflow data according to the data difference between two adjacent airflow data on the left side of the target airflow data in the period data segment and the fluctuation condition of the data difference specifically includes:
Calculating the mean value of all the difference data on the left side of the target airflow flow data and marking the mean value as a first coefficient, and calculating the variance of all the difference data on the left side of the target airflow flow data and marking the variance as a second coefficient; fitting all airflow data on the left side of the target airflow data by adopting a horizontal function to obtain a correction decision coefficient;
And calculating the ratio of the negative correlation normalized value of the first coefficient to the second coefficient, and taking the product of the ratio and the correction decision coefficient as the stable coefficient of the target airflow flow data.
Preferably, the obtaining the characteristic parameter of the periodic data segment according to the data distribution characteristic of the characteristic vector of each airflow data in the periodic data segment specifically includes:
and taking the characteristic vector corresponding to the maximum value of the L2 norms of the characteristic vectors of all the airflow flow data in the period data segment as the characteristic parameter of the period data segment.
Preferably, the obtaining the data anomaly degree of each period data segment according to the difference between the characteristic parameter and the whole of each period data segment and the distribution amplitude of the airflow data in the period data segment specifically includes:
for any periodic data segment, calculating the product of the L2 norm of the characteristic parameter of the periodic data segment and the maximum value of the airflow flow data in the periodic data segment to obtain a third coefficient, and carrying out negative correlation normalization processing on the third coefficient to obtain a possibility index of the periodic data segment;
and obtaining the data abnormality degree of the periodic data segment according to the difference of the possibility indexes between the periodic data segment and the adjacent periodic data segment and the airflow flow data at the last moment in the periodic data segment.
Preferably, the obtaining the data anomaly degree of the periodic data segment according to the difference of the likelihood indexes between the periodic data segment and the adjacent periodic data segment and the airflow flow data at the last moment in the periodic data segment specifically includes:
Recording any one period data segment as a selected period data segment, wherein the selected period data segment and the period data segment adjacent to the selected period data segment form a data segment set;
Calculating the average value of the possibility indexes of all the periodic data segments except the data segment set to obtain a first average value, and calculating the average value of the possibility indexes of all the periodic data segments in the data segment set to obtain a second average value; taking the absolute value of the difference value of the first average value and the second average value as a first difference value; taking the difference value between the maximum value of the probability indexes of all the period data segments and the probability index of the selected period data segment as a second difference value, and taking the ratio of the first difference value to the second difference value as a first ratio;
calculating the ratio of the airflow data at the last moment in the selected period data segment to the stable coefficient of the airflow data as a second ratio; taking the product of the first ratio and the second ratio as the data abnormality degree of the selected period data segment.
Preferably, the obtaining the data distance between the periodic data segments according to the difference distribution condition of the data anomaly degree of each periodic data segment and the periodic data segment in the neighborhood and the difference condition of the characteristic parameters of each periodic data segment and the periodic data segment in the neighborhood specifically includes:
Recording any one period data segment as a first period data segment, recording any one period data segment in the neighborhood of the first period data segment as a second period data segment, and calculating the ratio of the sum value between the data anomaly degrees of the first period data segment and the second period data segment to the variance of the data anomaly degrees of all the period data segments in the neighborhood of the first period data segment to obtain a third ratio;
Taking the second norm of the difference between the characteristic parameters of the first period data segment and the second period data segment as a first characteristic coefficient, taking the product of the second norm of the characteristic parameters of the first period data segment and the second norm of the characteristic parameters of the second period data segment as a second characteristic coefficient, and calculating the ratio of the first characteristic coefficient to the second characteristic coefficient to obtain a fourth ratio;
the product of the third ratio and the fourth ratio is taken as the data distance between the first period data segment and the second period data segment.
Preferably, the correcting the airflow data in the abnormal period data segment by using the data distance between the abnormal period data segment and the period data segment in the neighborhood specifically includes:
for any abnormal period data segment, carrying out normalization processing on the average value of the data distance between the abnormal period data segment and all period data segments in the neighborhood to obtain an adjustment coefficient of the abnormal period data segment;
and calculating the product of the sum of the constant 1 and the adjustment coefficient and the airflow data in the abnormal period data segment to obtain the corrected airflow data in the abnormal period data segment.
Preferably, the acquiring the abnormal periodic data segment by using an abnormality detection algorithm based on the data distance specifically includes:
For any one period data segment, carrying out normalization processing on the average value of the data distance between the period data segment and all period data segments in the neighborhood to obtain an abnormal score of the period data segment; and taking the period data segment corresponding to the abnormality score larger than the preset abnormality threshold value as an abnormal period data segment.
Preferably, the step of processing the airflow data in segments according to the distribution characteristics of the airflow data at each moment to obtain each period data segment specifically includes:
and converting the airflow flow data into frequency domain data by adopting Fourier transformation to obtain a spectrogram, acquiring a period length based on the frequency corresponding to the maximum amplitude in the spectrogram, and carrying out segmentation processing on the airflow flow data by utilizing the period length to obtain each period data segment.
The embodiment of the invention has at least the following beneficial effects:
The invention collects airflow flow data, segments the periodic data segments obtained by segmenting the airflow flow data, and utilizes the periodic data segments to represent each complete breathing process, thereby providing a data basis for the subsequent analysis of the data change trend of each breathing process. Firstly, analyzing the data distribution fluctuation condition and the data difference condition of the left side and the right side of each airflow flow data in each period data segment, and quantifying the data distribution characteristic of each period data segment, namely, the characteristic parameter reflects the stability and the fluctuation of the data in the period data segment. And then, obtaining the data abnormality degree by analyzing the difference between the characteristic parameters and the whole of each period data segment and the distribution amplitude of the airflow data in the period data segment, wherein the abnormality degree represents the possibility and the abnormality degree of the air leakage in each period data segment. Further, the difference distribution condition of the data abnormality degree of each period data segment and the period data segment in the neighborhood and the difference condition of the characteristic parameters of each period data segment and the period data segment in the neighborhood are analyzed, the data distance between the two period data segments can be quantized, namely, the period data segment with the abnormal air leakage can be more accurately reflected based on the distance parameter between the period data segments analyzed by the abnormality degree, based on the fact, the abnormal period data segment is finally obtained by using an abnormality detection algorithm based on the data distance, and further, the data in each period data segment can be adaptively corrected, so that the corrected data is more accurate, and the data abnormality phenomenon caused by the air leakage phenomenon is eliminated.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of steps in a data processing method based on data analysis of a breath trainer according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following is a detailed description of a specific implementation, structure, characteristics and effects of a data processing method based on breath training device data analysis according to the invention with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of a data processing method based on data analysis of a breath training device provided by the invention with reference to the accompanying drawings.
The specific implementation scene aimed by the invention is as follows: according to the method, the abnormal situation of the data in each breathing process period is quantified by analyzing the change trend of the air flow in the breathing training device exhalation process and combining the data characteristic representation of the air leakage caused by the irregular wearing, the abnormal score condition is used as an adjustment coefficient, the abnormal data caused by the air leakage behavior is corrected, and the abnormal data can be effectively corrected.
Referring to fig. 1, a flowchart of a data processing method based on breath trainer data analysis according to an embodiment of the invention is shown, the method includes the following steps:
Step one, acquiring airflow data of a respiratory training device at each moment in a set time period, and carrying out sectional processing on the airflow data according to the distribution characteristics of the airflow data at each moment to obtain each period data segment.
Firstly, through the air flow data of the breathing trainer, the breathing frequency, the breathing depth and other parameters of a user can be monitored, and the user is helped to know the breathing mode of the user, so that training targets such as relaxing mind and body, adjusting emotion and the like are realized. And simultaneously, the lung function detection parameters such as the lung capacity, the maximum inspiration air flow and the like of the user can be reflected, and important information is provided for the diagnosis and monitoring of the diseases such as airway obstruction symptoms, asthma and the like of the relevant medical staff evaluation user.
In this embodiment, by collecting the airflow data in the respiratory trainer at each moment in a set period of time, in order to eliminate the influence of dimensions on subsequent data analysis, the collected data is subjected to normalization processing. The time length of the set time period is set to be 5 minutes, the time interval between two adjacent moments in the set time period is 0.2 seconds, and an implementer can set according to a specific implementation scene.
The user continuously performs different long breathing operations within a set period, the change of the air flow has a certain periodic characteristic in the whole breathing process, after the whole long breath, the air in the lung is almost completely removed, the air flow can be reduced to an extremely low level, the period from the end of expiration to the beginning of the next breath is called a breathing period, and the air flow data can be maintained to be at a relatively stable level in the period.
Based on the method, the characteristic that the airflow flow data has periodic distribution can be utilized, the acquired data is divided into data segments, namely, the airflow flow data is subjected to segmentation processing according to the distribution characteristic of the airflow flow data at each moment to obtain each period data segment. Specifically, fourier transformation is adopted to convert airflow flow data into frequency domain data to obtain a spectrogram, the period length is obtained based on the frequency corresponding to the maximum amplitude in the spectrogram, and the period length is utilized to segment the airflow flow data to obtain each period data segment.
In other embodiments, the period may be adjusted according to the actual data segment partitioning process, taking into account the time period variability of the different breathing processes. Specifically, assuming that the period length is L, from the first moment, acquiring airflow data corresponding to the moment equal to the period length forms a first data segment, that is, the airflow data from the first moment to the L moment forms the first data segment, and considering that a certain time difference exists between respiratory processes, at the same time, at the stage that the respiratory process is about to end or about to start, the airflow data distribution is stable, so that the minimum value of surrounding data of the airflow data at the last moment of the first data segment is acquired as the end position of the data segment, that is, the minimum value of the airflow data from the L-5 moment to the l+5 moment is acquired, and the first period data segment is the airflow data from the first moment to the l+2 moment at this moment.
Further, from the L+3rd moment, the L airflow data are acquired to form a second data segment, namely the primarily acquired second data segment is the airflow data from the L+3rd moment to the 2L+3rd moment, and then the minimum value in the airflow data from the 2L-2 th moment to the 2L+8th moment is acquired as the end position of the second period data segment. And the like, until all time airflow data are divided.
So far, the airflow data at each moment in the set time period is divided, and each period data segment can represent the airflow data corresponding to the process period of breathing each time.
And step two, obtaining characteristic parameters of each period data segment according to the data distribution fluctuation condition and the data difference condition of the left side and the right side of each airflow flow data in each period data segment.
Because wear the irregular condition that leads to breathing training ware to have the gas leakage, the air current to breathing process exists the influence this moment, leads to can get into partial air from the gap in the inspiration process, and the air source is not all from the equipment pipeline, leads to the monitoring result of air current flow data to be less. The expired gas may flow out from the gap during expiration, resulting in slower decrease of the airflow data during monitoring, and a trend difference from the normal situation, and the expiration behavior is actually finished, but the airflow data is still slowly decreased, and the data change feature of the next respiration process is entered without showing the data feature of relatively smooth distribution. Based on the characteristics, each piece of the periodic data can be firstly used as an analysis object respectively for each piece of the airflow data, and the fluctuation condition of the airflow data on the left side and the airflow data on the right side of the data can be analyzed to find the most representative data characteristic distribution in the periodic data piece.
In the period data segment corresponding to each breathing process, generally, a relatively stable data change is included, at this time, a user is ready to perform a deep breathing operation, then a relatively obvious data inflection point exists, the airflow flow data starts to have a relatively obvious change trend, namely a monotonically increasing trend, and the data feature of the airflow flow data at the inflection point is used as a representative data feature to represent the data feature distribution of the period data segment.
First, the stationarity distribution of the left data of each airflow data is analyzed. And taking any one airflow data in any one period data segment as target airflow data, and obtaining a stability coefficient of the target airflow data according to the data difference between two adjacent airflow data on the left side of the target airflow data in the period data segment and the fluctuation condition of the data difference.
Specifically, calculating the difference value between each airflow data and the next adjacent airflow data in the period data segment to obtain difference data; calculating the mean value of all the difference data on the left side of the target airflow flow data and marking the mean value as a first coefficient, and calculating the variance of all the difference data on the left side of the target airflow flow data and marking the variance as a second coefficient; fitting all airflow data on the left side of the target airflow data by adopting a horizontal function to obtain a correction decision coefficient; and calculating the ratio of the negative correlation normalized value of the first coefficient to the second coefficient, and taking the product of the ratio and the correction decision coefficient as the stable coefficient of the target airflow flow data.
In this embodiment, taking any one period data segment as an example, taking the t-th airflow data in the i-th period data segment as the target airflow data, the stability coefficient of the target airflow data can be expressed as: Wherein, A stationary factor representing the tth airflow rate data in the ith period data segment,Represents the tth airflow data in the ith period data segment,Representing the (t + 1) th airflow rate data in the ith period data segment,Representing difference data corresponding to the tth airflow rate data in the ith period data segment,For the first coefficient, the average value of all difference data on the left side of the t-th airflow data in the ith period data segment is represented,For the second coefficient, the variance of all difference data on the left side of the tth airflow rate data in the ith period data segment is represented,The correction decision coefficient obtained by fitting all airflow data on the left side of the tth airflow data in the ith period data segment is represented, and exp () represents an exponential function based on a natural constant e.
Difference dataThe data difference condition between two adjacent moments in the period data segment is reflected, and when the target airflow flow data is a data inflection point in the period data segment, the data distribution on the left side of the target airflow flow data is more stable, namely the data fluctuation degree is smaller, and the airflow flow data amplitude is smaller.
Thus, the first and second substrates are bonded together,The smaller the value of (c) is, the smaller the overall difference distribution of the data on the left side of the target airflow flow data is,The smaller the value of the corresponding stationary coefficient is, the smaller the fluctuation condition of the difference data on the left side of the target airflow flow data is, and the larger the value of the corresponding stationary coefficient is.
It is to be noted that it is assumed that the ith period data segment is from the ith periodFrom moment to momentThe left data of the t-th airflow data in the ith period data segment can be the thFrom moment to momentAirflow data between moments.
Fitting correction decision coefficients using horizontal functionsThe larger the value, the greater the likelihood that the data appears as a straight line in the horizontal direction as a whole, so that the larger the value of the corresponding stationary coefficient is, thereby avoiding the influence of the larger data difference but smaller final average value. It should be noted that, the fitting model when fitting the correction determining coefficient is a constant or a function independent of an independent variable, for example, the value of y=c, and the value of c may be obtained empirically to obtain the average value of the airflow data when the airflow data is not in the deep breathing operation.
The stability coefficient reflects the characteristic degree of data stability of each airflow flow data at the left side of the corresponding period data section, and the larger the value of the stability coefficient is, the more the corresponding airflow flow data accords with the characteristic distribution of inflection point data in the period data section.
Then, the fluctuation condition of the data on the right side of each airflow rate data is analyzed. And taking the quantity ratio of the difference data on the right side of the target airflow data, which is smaller than the preset difference threshold value, corresponding to the difference data as a trend coefficient of the target airflow data.
In the present embodiment, the difference threshold has a value of 0, i.e. when the difference dataWhen the airflow data is reduced with the increase of time, the data change shows a decreasing trend. And when the data are differentWhen the flow rate data is increased with time, the data change shows an upward trend. And further, the ratio of the data quantity showing the rising trend to all the airflow data on the right side is obtained, the ratio of the difference data showing the rising trend on the right side of the target airflow data can be obtained, and the larger the ratio is, the higher the rising degree is, the more the corresponding change trend on the right side of each airflow data accords with the characteristic distribution of the inflection point data in the period data segment.
Based on the characteristic data distribution conditions of the left side and the right side of each airflow data, the information parameters of each airflow data are characterized, namely, the binary group formed by the stable coefficient and the trend coefficient is used as the characteristic vector of the target airflow data, and the characteristic parameters of the periodic data segment can be obtained according to the data distribution characteristics of the characteristic vector of each airflow data in the periodic data segment.
Specifically, the feature vector corresponding to the maximum value of the L2 norms of the feature vectors of all the airflow data in the period data segment is used as the feature parameter of the period data segment. And taking the characteristic representation corresponding to the airflow flow data with the maximum characteristic representation degree in the period data segment as the characteristic parameter of the period data segment. The characteristic parameters then characterize the normal condition of the data distribution in the periodic data segment.
And thirdly, obtaining the data abnormality degree of each period data segment according to the difference between the characteristic parameters and the whole of each period data segment and the distribution amplitude of the airflow flow data in the period data segment.
The characteristic parameter is used as the characteristic representation of the periodic data segment, and certain difference exists between the characteristic representation of the periodic data segment where the normal change airflow flow data is located and the characteristic representation of the periodic data segment where the abnormal change airflow flow data is located. For example, the value of the stability coefficient in the characteristic parameter of the periodic data segment with the air leakage condition is smaller, i.e. the data stability in the early stage of breathing is poorer. Meanwhile, due to the air leakage condition, airflow data in the period data section are smaller, so that the peak value of the expiratory flow in the breathing process is reduced.
Based on the characteristic, the initial abnormality possibility of the periodic data segment is quantified firstly through the characteristic parameter of the periodic data segment and the peak airflow flow data in the periodic data segment. Specifically, for any periodic data segment, calculating the product of the L2 norm of the characteristic parameter of the periodic data segment and the maximum value of the airflow flow data in the periodic data segment to obtain a third coefficient, and carrying out negative correlation normalization processing on the third coefficient to obtain a possibility index of the periodic data segment.Wherein,A likelihood indicator representing the ith period data segment,Characteristic parameters representing the ith period data segment,An L2 norm representing the characteristic parameter of the i-th period data segment,Representing the maximum value of the airflow data in the ith period data segment, norm () represents the linear normalization function.
And as the third coefficient, reflecting the product between the data representation of the characteristic parameter and the peak value data in the period data segment, when the L2 norm of the characteristic parameter is smaller, the stability of the data in the period data segment is poorer, the fluctuation is probably poorer, and meanwhile, the smaller the value of the peak value is, the greater the possibility of air leakage abnormality exists in the corresponding period data segment is, namely the greater the value of the possibility index is. The likelihood indicator for the periodic data segment characterizes a likelihood that there is a data anomaly within the periodic data segment.
And then, combining the abnormal possibility of the initial quantized periodic data segment, and further analyzing the abnormal situation of the periodic data segment, namely obtaining the data abnormality degree of the periodic data segment according to the difference of the possibility indexes between the periodic data segment and the adjacent periodic data segment and the airflow flow data at the last moment in the periodic data segment.
Specifically, any one period data segment is marked as a selected period data segment, the selected period data segment and a period data segment adjacent to the selected period data segment form a data segment set, in this embodiment, the ith period data segment is used as a selected period data segment, and the ith period data segment adjacent to the ith period data segment is the ith-1 th period data segment and the (i+1) th period data segment, namely, the data segment set comprises three period data segments.
Calculating the average value of the possibility indexes of all the periodic data segments except the data segment set to obtain a first average value, and calculating the average value of the possibility indexes of all the periodic data segments in the data segment set to obtain a second average value; taking the absolute value of the difference value of the first average value and the second average value as a first difference value; taking the difference between the maximum value of the probability indexes of all the period data segments and the probability index of the selected period data segment as a second difference value, and taking the ratio of the first difference value to the second difference value as a first ratio.
Calculating the ratio of the airflow data at the last moment in the selected period data segment to the stable coefficient of the airflow data as a second ratio; taking the product of the first ratio and the second ratio as the data abnormality degree of the selected period data segment.
In this embodiment, taking any one period data segment as an example for explanation, the calculation formula of the data anomaly degree of the ith period data segment, i.e. the selected period data segment, can be expressed as: Wherein, Indicating the degree of data abnormality of the i-th period data segment,The first mean value, the mean value of the probability indicators representing all periodic data segments in the set of data segments,And a second mean value representing a mean value of the likelihood indicator for all periodic data segments except the set of data segments,A likelihood indicator representing the ith period data segment,Representing the maximum value of the likelihood indicator for all periodic data segments,Airflow data representing the last instant in the ith period data segment,A stationary coefficient representing airflow data at the last instant in the ith period data segment.
And as the first difference value, the difference between the abnormal probability condition in the neighborhood of the data segment of the selected period and the abnormal probability of the whole data is reflected, and the larger the value of the first difference value is, the larger the abnormal probability of the selected period compared with the whole data distribution is, and the larger the value of the corresponding data difference degree is.
The second difference value reflects the difference between the probability index of the selected period data segment and the maximum value, and the larger the difference is, the smaller the probability of data abnormality of the selected period data segment is, the smaller the corresponding degree of data abnormality is, and the smaller the data abnormality of the selected period data segment is.
In the case of air leakage, the airflow data show a slow decreasing trend in the end of the breath and have a poor stability coefficient in the beginning of the breath, so thatThe larger the value of the corresponding data abnormality degree is, which indicates that the abnormality degree exists in the data section of the selected period is larger.
The data abnormality degree of the periodic data segment comprehensively reflects the abnormality degree of the airflow flow data in the periodic data segment, and further reflects the possibility of data abnormality.
And step four, obtaining the data distance between the periodic data segments according to the difference distribution condition of the data abnormality degree of each periodic data segment and the periodic data segment in the neighborhood and the difference condition of the characteristic parameters of each periodic data segment and the periodic data segment in the neighborhood.
In practice, the KNN algorithm is often used for detecting data anomalies, and the central idea of the algorithm is that by calculating the average distance between each sample point and the K nearest samples, the calculated average distance is compared with a threshold value, and if the calculated average distance is greater than the threshold value, the sample point is considered as an anomaly point. In consideration of directly utilizing airflow data to perform abnormality evaluation, whether the data is abnormal or real due to air leakage cannot be distinguished, so that the abnormality score of the data is relatively inaccurate, and therefore the embodiment performs distance calculation based on the data abnormality index capable of representing one respiratory cycle.
Firstly, the K-neighbor periodic data segment of each periodic data segment needs to be determined, in this embodiment, the euclidean distance between the degree of data abnormality of each two periodic data segments is used as a distance evaluation index, and then the nearest K periodic data segments of each periodic data segment are acquired to form a neighborhood of each periodic data segment.
And then, analyzing the difference distribution condition of the data abnormality degree of each periodic data segment and the periodic data segment in the neighborhood and the difference condition of the characteristic parameters of each periodic data segment and the periodic data segment in the neighborhood, and quantifying the data distance between the periodic data segments, namely optimizing the distance parameters between every two periodic data segments.
Specifically, any one period data segment is recorded as a first period data segment, any one period data segment in the neighborhood of the first period data segment is recorded as a second period data segment, and a third ratio is obtained by calculating the ratio of the sum value between the data abnormality degrees of the first period data segment and the second period data segment to the variance of the data abnormality degrees of all the period data segments in the neighborhood of the first period data segment.
Taking the second norm of the difference between the characteristic parameters of the first period data segment and the second period data segment as a first characteristic coefficient, taking the product of the second norm of the characteristic parameters of the first period data segment and the second norm of the characteristic parameters of the second period data segment as a second characteristic coefficient, and calculating the ratio of the first characteristic coefficient to the second characteristic coefficient to obtain a fourth ratio; the product of the third ratio and the fourth ratio is taken as the data distance between the first period data segment and the second period data segment.
In this embodiment, taking the nth period data segment as the first period data segment and taking the mth period data segment in the neighborhood of the first period data segment as the second period data segment, a calculation formula of the data distance between the first period data segment and the second period data segment may be expressed as: Wherein, Representing the data distance between the nth period data segment and the mth period data segment within the neighborhood of the first period data segment,Indicating the degree of data abnormality of the nth period data segment,Representing the degree of data abnormality of the mth period data segment in the neighborhood of the nth period data segment,A variance representing the degree of data anomalies for all periodic data segments within the neighborhood of the nth periodic data segment,Characteristic parameters representing an nth period data segment, characteristic parameters representing an mth period data segment within a neighborhood of the nth period data segment,Representing the L2 norm.
For the third ratio, the smaller the variance is, the more concentrated the abnormal data in the neighborhood is, but the smaller the variance is for the data representing the normal breathing cycle, so the difference in the numerical value of the abnormal degree in the k neighborhood is utilized to determine to realize the distinction between the abnormal data and the normal data, namely, the larger the value of the abnormal degree of the abnormal data is, namely, the larger the value of the sum of the abnormal degrees of the data of the two cycle data segments is, the larger the corresponding third ratio is, and the larger the corresponding distance between the two is.
Reflecting the difference in the characteristic representation between two periodic data segments,And the fourth ratio reflects the characteristic difference distance between the two periodic data segments, and the larger the value is, the larger the data characteristic distribution difference between the two periodic data segments is, and the larger the corresponding data distance is.
The data distance between the first period data segment and the second period data segment reflects the difference distance between the two through the abnormal data distribution condition and the characteristic distribution difference distance.
And fifthly, acquiring an abnormal period data segment by using an abnormal detection algorithm based on the data distance, and correcting airflow flow data in the abnormal period data segment by using the data distance between the abnormal period data segment and the period data segment in the neighborhood.
In this embodiment, the KNN algorithm is used to perform anomaly detection on the periodic data segments, that is, calculate the average data distance between each periodic data segment and the K-neighbor periodic data segment of the periodic data segment, and then compare the calculated average data distance with the threshold value to screen out the periodic data segment with an anomaly condition.
Specifically, for any one period data segment, carrying out normalization processing on the average value of the data distances between the period data segment and all period data segments in the neighborhood to obtain an abnormal score of the period data segment; the anomaly score for a periodic data segment reflects the differential distance distribution between the periodic data segment and the data segment within the neighborhood. When the value of the abnormality score is larger, the larger the data distribution difference between the periodic data segment and the periodic data segment in the neighborhood is, the larger the possibility that the periodic data segment is abnormal is further, and the greater the abnormality degree is. When the value of the abnormality score is smaller, the data distribution difference between the period data segment and the period data segment in the neighborhood is smaller, and the possibility that the period data segment is abnormal is further smaller, and the degree of abnormality is smaller.
Based on the above, the period data segment corresponding to the abnormality score larger than the preset abnormality threshold is used as the abnormal period data segment. In this embodiment, the value of the anomaly threshold value is 0.8. Since the value of the anomaly score of the period data segment is a normalized value, the value range of the anomaly threshold value is (0, 1).
When the anomaly score of the periodic data segment is greater than the anomaly threshold, the difference between the data distribution of the periodic data segment and the data distribution of the periodic data segment in the neighborhood is larger, and the greater the anomaly degree of the periodic data segment compared with the periodic data segment in the neighborhood is, the greater the possibility of air leakage anomaly is. When the abnormality score of the periodic data segment is smaller than or equal to the abnormality threshold, the data distribution characteristic of the periodic data segment is compounded with the normal data change characteristic, and the probability of the occurrence of air leakage abnormality of the periodic data segment is further reduced.
Further, the data weight of each period data segment can be determined through the abnormality score of each period data segment, so that correction of the lung function monitoring parameter can be realized. And correcting the airflow flow data in the abnormal period data segment by utilizing the data distance between the abnormal period data segment and the period data segment in the neighborhood.
Specifically, for any abnormal period data segment, carrying out normalization processing on the average value of the data distances between the abnormal period data segment and all period data segments in the neighborhood to obtain an adjustment coefficient of the abnormal period data segment; and calculating the product of the sum of the constant 1 and the adjustment coefficient and the airflow data in the abnormal period data segment to obtain the corrected airflow data in the abnormal period data segment.
In this embodiment, in order to monitor the lung capacity in the lung function, the maximum respiratory flow in the respiratory process, that is, the maximum value of the airflow data in each period data segment, is required to be obtained, so the maximum value of the airflow data is corrected by using the abnormality score of the corresponding period data segment, which can be expressed as: Wherein, Representing the maximum value of the airflow data of the modified r-th abnormal period data segment,A maximum value of the airflow data representing the r-th abnormal period data segment,Representing the adjustment coefficient of the data segment of the r-th anomaly period.
It should be noted that, the abnormal period data segment is that the wearing is irregular and the air leakage phenomenon is caused, so that the measured value of the maximum value of the airflow flow data in the abnormal period data segment is smaller than the actual value, and the airflow flow data is enlarged by using the adjustment coefficient, so that the corrected airflow flow data eliminates the data abnormality caused by the problem of the irregular wearing, and the final data is more accurate.
Further, the relevant medical personnel can conduct further data investigation on the lung function performance of the user based on more accurate airflow flow data. The user can be timely reminded to go to the hospital for relevant examination.
The above embodiments are only for illustrating the technical solution of the present application, and not for limiting the same; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the application and are intended to be included within the scope of the application.
Claims (3)
1. A data processing method based on breath trainer data analysis, the method comprising the steps of:
Acquiring airflow data of the respiratory trainer at each moment in a set time period, and carrying out sectional processing on the airflow data according to the distribution characteristics of the airflow data at each moment to obtain each period data segment;
Obtaining characteristic parameters of each period data segment according to the data distribution fluctuation condition and the data difference condition of the left side and the right side of each airflow flow data in each period data segment;
Obtaining the data abnormality degree of each period data segment according to the difference between the characteristic parameters and the whole of each period data segment and the distribution amplitude of the airflow flow data in the period data segment;
Obtaining the data distance between the periodic data segments according to the difference distribution condition of the data abnormality degree of each periodic data segment and the periodic data segment in the neighborhood and the difference condition of the characteristic parameters of each periodic data segment and the periodic data segment in the neighborhood;
Acquiring an abnormal period data segment by using an abnormal detection algorithm based on the data distance, and correcting airflow flow data in the abnormal period data segment by using the data distance between the abnormal period data segment and the period data segment in the neighborhood;
the method for obtaining the characteristic parameters of each period data segment according to the data distribution fluctuation condition and the data difference condition of the left side and the right side of each airflow flow data in each period data segment specifically comprises the following steps:
Taking any one airflow data in any one period data segment as target airflow data, and obtaining a stability coefficient of the target airflow data according to the data difference between two adjacent airflow data on the left side of the target airflow data in the period data segment and the fluctuation condition of the data difference;
Calculating the difference value between each airflow data and the next adjacent airflow data in the period data segment to obtain difference data; taking the quantity ratio of the difference data on the right side of the target airflow flow data, which is smaller than the corresponding difference data of the preset difference threshold value, as a trend coefficient of the target airflow flow data;
Taking the binary group formed by the stable coefficient and the trend coefficient as a characteristic vector of the target airflow flow data, and obtaining characteristic parameters of the periodic data segment according to the data distribution characteristics of the characteristic vector of each airflow flow data in the periodic data segment;
obtaining a stability coefficient of the target airflow data according to the data difference between two adjacent airflow data on the left side of the target airflow data in the period data section and the fluctuation condition of the data difference, wherein the stability coefficient specifically comprises the following steps:
Calculating the mean value of all the difference data on the left side of the target airflow flow data and marking the mean value as a first coefficient, and calculating the variance of all the difference data on the left side of the target airflow flow data and marking the variance as a second coefficient; fitting all airflow data on the left side of the target airflow data by adopting a horizontal function to obtain a correction decision coefficient;
Calculating the ratio of the negative correlation normalized value of the first coefficient to the second coefficient, and taking the product of the ratio and the correction decision coefficient as a stable coefficient of the target airflow flow data;
the method for obtaining the characteristic parameters of the periodic data segment according to the data distribution characteristics of the characteristic vector of each airflow flow data in the periodic data segment specifically comprises the following steps:
taking the characteristic vector corresponding to the maximum value of the L2 norms of the characteristic vectors of all the airflow flow data in the period data segment as the characteristic parameter of the period data segment;
The method for obtaining the data anomaly degree of each period data segment according to the difference between the characteristic parameters and the whole of each period data segment and the distribution amplitude of the airflow flow data in the period data segment comprises the following steps:
for any periodic data segment, calculating the product of the L2 norm of the characteristic parameter of the periodic data segment and the maximum value of the airflow flow data in the periodic data segment to obtain a third coefficient, and carrying out negative correlation normalization processing on the third coefficient to obtain a possibility index of the periodic data segment;
obtaining the data abnormality degree of the periodic data segments according to the difference of the possibility indexes between the periodic data segments and the adjacent periodic data segments and the airflow flow data at the last moment in the periodic data segments;
the method for obtaining the data abnormality degree of the periodic data segment according to the difference of the possibility indexes between the periodic data segment and the adjacent periodic data segment and the airflow flow data at the last moment in the periodic data segment specifically comprises the following steps:
Recording any one period data segment as a selected period data segment, wherein the selected period data segment and the period data segment adjacent to the selected period data segment form a data segment set;
Calculating the average value of the possibility indexes of all the periodic data segments except the data segment set to obtain a first average value, and calculating the average value of the possibility indexes of all the periodic data segments in the data segment set to obtain a second average value; taking the absolute value of the difference value of the first average value and the second average value as a first difference value; taking the difference value between the maximum value of the probability indexes of all the period data segments and the probability index of the selected period data segment as a second difference value, and taking the ratio of the first difference value to the second difference value as a first ratio;
Calculating the ratio of the airflow data at the last moment in the selected period data segment to the stable coefficient of the airflow data as a second ratio; taking the product of the first ratio and the second ratio as the data abnormality degree of the selected period data segment;
The method for obtaining the data distance between the periodic data segments according to the difference distribution condition of the data abnormality degree of each periodic data segment and the periodic data segment in the neighborhood and the difference condition of the characteristic parameters of each periodic data segment and the periodic data segment in the neighborhood specifically comprises the following steps:
Recording any one period data segment as a first period data segment, recording any one period data segment in the neighborhood of the first period data segment as a second period data segment, and calculating the ratio of the sum value between the data anomaly degrees of the first period data segment and the second period data segment to the variance of the data anomaly degrees of all the period data segments in the neighborhood of the first period data segment to obtain a third ratio;
Taking the second norm of the difference between the characteristic parameters of the first period data segment and the second period data segment as a first characteristic coefficient, taking the product of the second norm of the characteristic parameters of the first period data segment and the second norm of the characteristic parameters of the second period data segment as a second characteristic coefficient, and calculating the ratio of the first characteristic coefficient to the second characteristic coefficient to obtain a fourth ratio;
Taking the product of the third ratio and the fourth ratio as the data distance between the first period data segment and the second period data segment;
the correcting the airflow data in the abnormal period data segment by utilizing the data distance between the abnormal period data segment and the period data segment in the neighborhood specifically comprises the following steps:
for any abnormal period data segment, carrying out normalization processing on the average value of the data distance between the abnormal period data segment and all period data segments in the neighborhood to obtain an adjustment coefficient of the abnormal period data segment;
and calculating the product of the sum of the constant 1 and the adjustment coefficient and the airflow data in the abnormal period data segment to obtain the corrected airflow data in the abnormal period data segment.
2. The method for processing data based on data analysis of a respiratory training device according to claim 1, wherein the acquiring the abnormal periodic data segment by using an abnormality detection algorithm based on the data distance specifically comprises:
For any one period data segment, carrying out normalization processing on the average value of the data distance between the period data segment and all period data segments in the neighborhood to obtain an abnormal score of the period data segment; and taking the period data segment corresponding to the abnormality score larger than the preset abnormality threshold value as an abnormal period data segment.
3. The data processing method based on data analysis of a respiratory training device according to claim 1, wherein the step of performing a segmentation process on the airflow data according to the distribution characteristics of the airflow data at each moment to obtain each period data segment specifically comprises:
and converting the airflow flow data into frequency domain data by adopting Fourier transformation to obtain a spectrogram, acquiring a period length based on the frequency corresponding to the maximum amplitude in the spectrogram, and carrying out segmentation processing on the airflow flow data by utilizing the period length to obtain each period data segment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410547588.6A CN118133209B (en) | 2024-05-06 | 2024-05-06 | Data processing method based on breath trainer data analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410547588.6A CN118133209B (en) | 2024-05-06 | 2024-05-06 | Data processing method based on breath trainer data analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118133209A CN118133209A (en) | 2024-06-04 |
| CN118133209B true CN118133209B (en) | 2024-07-05 |
Family
ID=91239318
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410547588.6A Active CN118133209B (en) | 2024-05-06 | 2024-05-06 | Data processing method based on breath trainer data analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118133209B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118416354B (en) * | 2024-07-05 | 2024-09-13 | 长沙心诺医疗器械有限公司 | Semiconductor laser oxygen therapy instrument and respiration state detection method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116956188A (en) * | 2023-07-06 | 2023-10-27 | 中国电信股份有限公司技术创新中心 | Data anomaly detection method and device, electronic equipment and storage medium |
| CN117153416A (en) * | 2023-10-26 | 2023-12-01 | 天津医科大学第二医院 | Diabetes screening data processing and management system |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5301310B2 (en) * | 2009-02-17 | 2013-09-25 | 株式会社日立製作所 | Anomaly detection method and anomaly detection system |
| CN117349778B (en) * | 2023-12-04 | 2024-02-20 | 湖南蓝绿光电科技有限公司 | Online real-time monitoring system of consumer based on thing networking |
-
2024
- 2024-05-06 CN CN202410547588.6A patent/CN118133209B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116956188A (en) * | 2023-07-06 | 2023-10-27 | 中国电信股份有限公司技术创新中心 | Data anomaly detection method and device, electronic equipment and storage medium |
| CN117153416A (en) * | 2023-10-26 | 2023-12-01 | 天津医科大学第二医院 | Diabetes screening data processing and management system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118133209A (en) | 2024-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN118133209B (en) | Data processing method based on breath trainer data analysis | |
| Abeyratne et al. | Pitch jump probability measures for the analysis of snoring sounds in apnea | |
| US11304624B2 (en) | Method and apparatus for performing dynamic respiratory classification and analysis for detecting wheeze particles and sources | |
| US20190088367A1 (en) | Method and apparatus for training and evaluating artificial neural networks used to determine lung pathology | |
| US9357953B2 (en) | System and method for diagnosing sleep apnea | |
| US20120037159A1 (en) | Detection of asynchrony | |
| US20060011200A1 (en) | Auto CPAP system profile information | |
| WO2012060107A1 (en) | Biometric sound testing device and biometric sound testing method | |
| US11529072B2 (en) | Method and apparatus for performing dynamic respiratory classification and tracking of wheeze and crackle | |
| US20160228037A1 (en) | Homecare asthma management | |
| CN102215750B (en) | System for detecting respiratory muscle activity of a patient receiving assisted breathing | |
| CN117497181B (en) | Chronic disease information management system based on artificial intelligence | |
| US10426426B2 (en) | Methods and apparatus for performing dynamic respiratory classification and tracking | |
| CN111281349A (en) | Sleep disordered breathing monitoring method and system | |
| Ghaemmaghami et al. | Normal probability testing of snore signals for diagnosis of obstructive sleep apnea | |
| CN114711749A (en) | Lung function state classification method based on quantitative report template | |
| TW201544074A (en) | Breathing waveform recognition method and system thereof | |
| CN117174294A (en) | Method and system for constructing slow-resistance lung evaluation model | |
| Ozdemir et al. | A time-series approach to predict obstructive sleep apnea (OSA) Episodes | |
| CN112971762B (en) | Respiratory signal quality evaluation method | |
| CN114767089A (en) | Voice guidance detection method and system for pulmonary function examination | |
| Staykov et al. | Nasal pressure derived airflow limitation and ventilation measurements are resilient to reduced signal quality | |
| Pal et al. | Automatic breathing phase identification based on the second derivative of the recorded lung sounds | |
| US20230380719A1 (en) | Method and apparatus for simultaneous collection, processing and display of audio and flow events during breathing | |
| US20230380792A1 (en) | Method and apparatus for determining lung pathologies and severity from a respiratory recording and breath flow analysis using a convolution neural network (cnn) |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |