A clinical trial data analysis method and system

Abstract

The present invention relates to the field of data processing technology, and in particular to a clinical trial data analysis method and system. The method includes the following steps: presetting the observation range of blood oxygen saturation data points in a human body's blood oxygen saturation time series sequence and the respiratory frequency data points in a respiratory frequency time series sequence; obtaining a mutation index by removing the deviation of the coefficient of variation before and after the blood oxygen saturation data point from the observation range of a blood oxygen saturation data point, and obtaining the mutation factor of the blood oxygen saturation data point according to the mutation index and the data change of all adjacent blood oxygen saturation data points in the observation range; calculating the false factor of the blood oxygen saturation data point; obtaining a corrected target threshold by multiplying the false factor of the blood oxygen saturation data point by a preset threshold; using the target threshold in the Chebyshev inequality to obtain the analysis result of the blood oxygen saturation data point, which effectively improves the accuracy of the analysis result.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a clinical trial data analysis method and system. Background Art

[0002] Medical R&D can promote technological innovation in the medical industry. Through continuous research and experimentation, new treatment principles can be discovered, new technical means can be developed, and new solutions can be provided for the development of the medical industry, thereby improving the quality and efficiency of medical services. In the medical R&D stage, by analyzing clinical data, researchers can evaluate the R&D results.

[0003] In clinical trials, human blood oxygen saturation data is one of the important indicators, which is used to characterize the effect of drugs or treatment methods on the oxygen content in the patient's blood. Therefore, the research and development effect can be evaluated by monitoring blood oxygen saturation data. There are many monitoring and analysis methods for blood oxygen saturation data in the prior art. For example, the patent application document with publication number CN117598700A discloses an intelligent blood oxygen saturation detection system and method. The application collects the blood oxygen saturation data of the target object, captures the long-term trend and change pattern of blood oxygen saturation, obtains the fluctuation embedded in the blood oxygen saturation time series semantic features, and then uses the local time series fluctuation semantic features embedded in the local time series change features of blood oxygen saturation for classification processing, so as to determine whether the target object's blood oxygen saturation is abnormal.

[0004] The above-mentioned existing technologies can improve the efficiency of blood oxygen saturation data monitoring and analysis to a certain extent. However, the collected blood oxygen saturation data may have different changing trends due to different individual behaviors. If neural network training is performed on all categories of target objects, it will result in a large amount of data processing.

[0005] Based on this, how to accurately obtain the monitoring results of blood oxygen saturation data is a problem that needs to be solved urgently by technical personnel in this field. Summary of the invention

[0006] In order to solve the technical problem of how to accurately obtain monitoring results of blood oxygen saturation data, the present invention provides a clinical trial data analysis method and system.

[0007] In a first aspect, the present invention provides a clinical trial data analysis method, which adopts the following technical solution:

[0008] A clinical trial data analysis method comprising the steps of:

[0009] Preset the observation ranges of the blood oxygen saturation data points in the human body's blood oxygen saturation time series and the respiratory rate data points in the respiratory rate time series; record the normalized value of the difference in the coefficient of variation before and after excluding the blood oxygen saturation data point in the observation range of a blood oxygen saturation data point as a mutation index, and obtain the mutation factor of the blood oxygen saturation data point by multiplying the mutation index by the mean value of the data change between all adjacent blood oxygen saturation data points in the observation range; calculate the false factor of the blood oxygen saturation data point:

[0010] ;

[0011] is the false factor of the i-th blood oxygen saturation data point, is the mutation factor of the i-th blood oxygen saturation data point, is the data value ranking of the i-th blood oxygen saturation data point in its observation range, is the length of the observation range, is the data value ranking of the i-th respiratory frequency data point in its observation range, is a hyperparameter, is the variance of the Z-score scores of all respiratory rate data points in the observation range of the i-th respiratory rate data point, is the preset adjustment factor, is a linear normalization function; the corrected target threshold is obtained by multiplying the false factor of the blood oxygen saturation data point by the preset threshold; the target threshold is used in the Chebyshev inequality to obtain the analysis result of the blood oxygen saturation data point.

[0012] The present invention analyzes abnormal data in the blood oxygen saturation time series through Chebyshev inequality, and can filter out abnormal data from blood oxygen saturation time series in various distribution forms. In this process, the present invention takes into account that there may be false pseudo data in the collected blood oxygen saturation data. Based on this, the present invention analyzes the prominent performance of the blood oxygen saturation data points in the local range to obtain the possibility that it is pseudo data, so as to realize the correction of the preset threshold and improve the accuracy of the analysis result; on this basis, the present invention also takes into account that the abnormal data generated by the human body may be similar to the pseudo data. In order to further distinguish, the present invention also obtains the respiratory rate data that is correlated with the blood oxygen saturation data to correct the prominent performance, so that the target threshold corresponding to each blood oxygen saturation data point can be accurately obtained, effectively improving the accuracy of the analysis result.

[0013] According to a clinical trial data analysis method provided by the present invention, the observation range of the blood oxygen saturation data points in the preset human body's blood oxygen saturation time series sequence and the respiratory rate data points in the respiratory rate time series sequence also includes: collecting the human body's blood oxygen saturation data and respiratory rate data, and preprocessing the blood oxygen saturation data and respiratory rate data to obtain the blood oxygen saturation time series sequence and the respiratory rate time series sequence.

[0014] The present invention takes into account the possibility of missing data values ​​in the collected original data, and therefore improves the overall quality of the data through preprocessing to prepare for subsequent data processing.

[0015] According to a clinical trial data analysis method provided by the present invention, the method for obtaining the observation range includes: presetting the length of the observation range , taking the blood oxygen saturation data point or respiratory rate data point as the center, obtain blood oxygen saturation data points or respiratory rate data points to construct an observation range of the blood oxygen saturation data points or respiratory rate data points.

[0016] By acquiring data points on both sides of the blood oxygen saturation data point or the respiratory rate data point, the present invention can accurately describe the characteristic performance of the blood oxygen saturation data point or the respiratory rate data point relative to the data on both sides, so that its mutation factor can be accurately obtained based on this later.

[0017] According to a clinical trial data analysis method provided by the present invention, the method for obtaining the coefficient of variation includes: obtaining the percentage of the ratio of the standard deviation to the mean value before and after excluding the blood oxygen saturation data point in the observation range of the blood oxygen saturation data point, respectively, and recording it as the coefficient of variation before elimination and the coefficient of variation after elimination.

[0018] The present invention can obtain the contribution of the blood oxygen saturation data point to the coefficient of variation by obtaining the difference in the coefficient of variation before and after the current blood oxygen saturation data point is eliminated in the observation range. If the difference in the coefficient of variation in the observation range before and after elimination is large, it means that the existence of the blood oxygen saturation data point makes the coefficient of variation in the observation range large, and the blood oxygen saturation data point is likely to be false data.

[0019] According to a clinical trial data analysis method provided by the present invention, a method for obtaining a data value ranking of a blood oxygen saturation data point or a respiratory rate data point in its observation range includes: arranging the data values ​​in the observation range in ascending order to obtain a data value ranking corresponding to the blood oxygen saturation data point or the respiratory rate data point.

[0020] According to a clinical trial data analysis method provided by the present invention, the target threshold is used in the Chebyshev inequality to obtain the analysis result of the blood oxygen saturation data point, including: obtaining a historical data set of blood oxygen saturation data points, and determining the mean and standard deviation of the historical data set; if the absolute value of the difference between the data value of the blood oxygen saturation data point and the mean is less than the product of the target threshold of the blood oxygen saturation data point and the standard deviation, then the blood oxygen saturation data point is normal data; otherwise, the blood oxygen saturation data point is abnormal data.

[0021] According to a clinical trial data analysis method provided by the present invention, the analysis results of the blood oxygen saturation data points are obtained, and then the normal data and the abnormal data are stored in a normal database and an abnormal database respectively, and an analysis model is constructed based on the data values ​​corresponding to the blood oxygen saturation data points and the analysis results.

[0022] By constructing a normal database, an abnormal database and an analysis model, the present invention can obtain relevant data from the database in a timely and accurate manner when the analysis model detects abnormal data, so that medical staff can make a diagnosis in a timely manner.

[0023] In a second aspect, the present invention provides a clinical trial data analysis system, which adopts the following technical solution:

[0024] A clinical trial data analysis system comprises: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the clinical trial data analysis method described above is implemented.

[0025] By adopting the above technical solution, the above clinical trial data analysis method is generated into a computer program and stored in a memory so as to be loaded and executed by a processor, thereby making a terminal device based on the memory and the processor for easy use.

[0026] The present invention has the following technical effects:

[0027] Based on the above technical solution, when determining the analysis results of clinical data, the present invention analyzes the abnormal data in the blood oxygen saturation time series through the Chebyshev inequality, and can screen out abnormal data from the blood oxygen saturation time series of various distribution forms. In this process, the present invention takes into account the possibility of false pseudo data in the collected blood oxygen saturation data. Based on this, the present invention analyzes the prominent performance of the blood oxygen saturation data points in the local range, and corrects the preset threshold through the possibility of pseudo data, thereby improving the accuracy of the analysis results; on this basis, the present invention also obtains the respiratory rate data that is correlated with the blood oxygen saturation data to correct the prominent performance, reducing the misjudgment of abnormal data generated by the human body, so that the target threshold corresponding to each blood oxygen saturation data point can be accurately obtained, effectively improving the accuracy of the analysis results. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts.

[0029] Figure 1 A schematic diagram of a flow chart of a clinical trial data analysis method provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0031] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.

[0032] It should be noted that medical R&D can promote technological innovation in the medical industry. Through continuous research and experimentation, new treatment principles can be discovered, new technical means can be developed, and new solutions can be provided for the development of the medical industry, thereby improving the quality and efficiency of medical services. In the medical R&D stage, by analyzing clinical data, researchers can evaluate the R&D effect.

[0033] Chebyshev's inequality is a threshold judgment method in probability theory. This method believes that within a certain range, the value of a random variable will be close to the average value. If it deviates from the average value, it is likely to be abnormal data. Chebyshev's inequality is applicable to data of any distribution shape, which means that no matter what the distribution shape of blood oxygen saturation data is, this inequality can be used for abnormality detection.

[0034] However, the blood oxygen saturation data collected clinically may have different changing trends due to different individual behaviors. For example, the blood oxygen saturation data of a person just after exercise or in a tense state may have different changing trends from that of a person just after sleep. If the blood oxygen saturation data is directly judged as abnormal based on a fixed threshold, some data may be misjudged.

[0035] Based on this, the present invention obtains the change characteristics in the blood oxygen saturation data, adjusts the threshold in the Chebyshev inequality, and obtains the analysis result of the blood oxygen saturation data, so as to accurately screen out abnormal data of various distribution forms.

[0036] For details, please refer to Figure 1 As shown, Figure 1 A flowchart of a clinical trial data analysis method provided in an embodiment of the present invention is provided. The method specifically includes the following steps S1 to S4.

[0037] S1: preset the observation range of the blood oxygen saturation data points in the human body's blood oxygen saturation time series.

[0038] For example, in an embodiment of the present invention, an observation range of blood oxygen saturation data points in a blood oxygen saturation time series sequence of a human body is preset, and the process also includes: collecting blood oxygen saturation data of the human body, preprocessing the blood oxygen saturation data, and obtaining a blood oxygen saturation time series sequence.

[0039] The preprocessing may be data format conversion, missing data interpolation, etc., which may be specifically configured according to actual needs, and the embodiment of the present invention does not impose too many limitations thereon.

[0040] For example, the collection duration can be set to 2 hours, and the collection frequency can be once every 5 seconds; the collection duration and the collection frequency can be set according to actual needs, and the embodiment of the present invention does not impose too many restrictions on this.

[0041] Specifically, when collecting the blood oxygen saturation data of the human body, the blood oxygen saturation data of the human body can be collected through a pulse oximeter.

[0042] It should be noted that when collecting human blood oxygen saturation data, the electromagnetic field generated by other electronic devices around may overlap with the signal of the blood oxygen saturation data, resulting in the presence of false pseudo data in the collected blood oxygen saturation data, which is less accurate. When using Chebyshev's inequality for abnormality judgment, such pseudo data may be mistakenly detected as abnormal data, resulting in poor results in subsequent drug efficacy evaluation based on this.

[0043] It should be further explained that the combined effect of the heart's autonomy and the autonomic nervous system causes the blood oxygen saturation data to change gradually rather than instantly. However, the false pseudo data is more prominent and irregular compared to the surrounding data.

[0044] Based on this, the embodiment of the present invention obtains the mutation of the blood oxygen saturation data point by analyzing the numerical performance and numerical change trend of the blood oxygen saturation data point in the local range of the blood oxygen saturation time series sequence.

[0045] For example, in an embodiment of the present invention, when obtaining an observation range for observing numerical performance and numerical change trend, the following two possible implementation methods are included:

[0046] In one possible implementation, the length of the observation range can be obtained , taking the current blood oxygen saturation data point as the center, obtain blood oxygen saturation data points to construct the observation range of the current blood oxygen saturation data point.

[0047] Among them, the length of the observation range of the blood oxygen saturation data point is the number of blood oxygen saturation data points contained in the observation range; the length of the observation range can be preset to 51, and can be set according to actual needs. The embodiment of the present invention does not impose too many restrictions on this.

[0048] In this way, the embodiment of the present invention can accurately describe the performance of the blood oxygen saturation data point relative to the data on both sides by considering the data points on both sides of the blood oxygen saturation data point, so that its mutation factor can be accurately obtained based on this later.

[0049] In another possible implementation, the length of the observation range can be obtained , taking the current blood oxygen saturation data point as the end point, obtain the historical data of the blood oxygen saturation data point blood oxygen saturation data points to construct the observation range of the current blood oxygen saturation data point.

[0050] In this way, the embodiment of the present invention obtains historical data as the local observation range of the blood oxygen saturation data point and analyzes the rules and trends in the historical data to accurately describe the deviation of the blood oxygen saturation data point from the historical data and reduce the data waiting time.

[0051] After obtaining the observation range of the blood oxygen saturation data point based on any of the above possible implementation methods, the mutation of the blood oxygen saturation data point can be obtained based on the numerical performance and numerical change trend of the blood oxygen saturation data point in its observation range, that is, continue to execute the following steps.

[0052] S2: Obtain the mutation factor of the blood oxygen saturation data point by eliminating the data changes before and after the blood oxygen saturation data point in the observation range of the blood oxygen saturation data point, as well as the data changes between all adjacent blood oxygen saturation data points in the observation range.

[0053] It should be noted that the larger the gap between the blood oxygen saturation data point and the overall data in the observation range and the more drastic the data change trend in the observation range, the higher the possibility that the blood oxygen saturation data point is false pseudo data, and the more obvious the corresponding mutation situation.

[0054] By way of example, in an embodiment of the present invention, when determining the mutation factor of a blood oxygen saturation data point, the normalized value of the difference in the coefficient of variation before and after excluding the blood oxygen saturation data point in the observation range of the blood oxygen saturation data point can be recorded as the mutation index, and the mutation factor of the blood oxygen saturation data point is obtained by multiplying the mutation index by the mean of the data change amount between all adjacent blood oxygen saturation data points in the observation range.

[0055] For example, when obtaining the data variation between adjacent blood oxygen saturation data points, the absolute value of the data value difference between adjacent blood oxygen saturation data points can be used as the data variation between the adjacent blood oxygen saturation data points.

[0056] Take an example to illustrate the mean value of the data change between all adjacent blood oxygen saturation data points in the observation range: if there are 3 blood oxygen saturation data points in the observation range, there are 2 groups of adjacent blood oxygen saturation data points, among which the 1st and 2nd blood oxygen saturation data points are adjacent blood oxygen saturation data points, and the 2nd and 3rd blood oxygen saturation data points are adjacent blood oxygen saturation data points, and the mean value of the data change of the 2 groups of adjacent blood oxygen saturation data points is obtained.

[0057] By way of example, in an embodiment of the present invention, a method for obtaining a coefficient of variation includes: respectively obtaining the percentage of the ratio of the standard deviation to the mean value before and after excluding the blood oxygen saturation data point in the observation range of the blood oxygen saturation data point, recorded as the coefficient of variation before excluding and the coefficient of variation after excluding.

[0058] Specifically, the percentage of the ratio of the standard deviation to the mean value in the observation range of the blood oxygen saturation data point is obtained, which is recorded as the coefficient of variation before elimination; the percentage of the ratio of the standard deviation to the mean value after eliminating the blood oxygen saturation data point in the observation range of the blood oxygen saturation data point is obtained, which is recorded as the coefficient of variation after elimination.

[0059] It is understandable that the specific steps of obtaining the coefficient of variation of data are prior art and will not be described in detail in the embodiment of the present invention.

[0060] For example, the mutation factor of the blood oxygen saturation data point is determined, and the specific relationship can be seen in the following formula:

[0061] ;

[0062] In the formula, is the mutation factor of the i-th blood oxygen saturation data point, is the coefficient of variation of the observation range of the i-th blood oxygen saturation data point, is the coefficient of variation after removing the i-th blood oxygen saturation data point from the observation range of the i-th blood oxygen saturation data point, is the mean value of the data change between all adjacent blood oxygen saturation data points in the observation range of the i-th blood oxygen saturation data point, is a linear normalization function.

[0063] In the above formula, It represents the normalized value of the difference in coefficient of variation before and after the i-th blood oxygen saturation data point is removed from the observation range, that is, the mutation index. The larger the value is, the larger the coefficient of variation in the observation range is due to the existence of the i-th blood oxygen saturation data point, the more prominent the i-th blood oxygen saturation data point is, the greater the possibility that the corresponding i-th blood oxygen saturation data point is false data, and the larger its mutation factor is.

[0064] The mean value of the data change between all adjacent blood oxygen saturation data points in the observation range of the i-th blood oxygen saturation data point The larger it is, the more dramatic the change in the value of the i-th blood oxygen saturation data point in the observation range is, the greater the credibility of the outstanding performance of the i-th blood oxygen saturation data point is, and the larger its mutation factor is.

[0065] After obtaining the mutation of each blood oxygen saturation data point within its observation range based on the above steps, continue to perform the following steps.

[0066] S3: Obtain the observation range of the respiratory rate data points in the respiratory rate time series of the human body, and calculate the false factor of the blood oxygen saturation data points.

[0067] It should be further explained that the mutation factor of the blood oxygen saturation data point obtained based on the above steps can reflect the data mutation performance in the local range. However, in actual situations, the blood oxygen saturation data of the human body will produce abnormal data due to its own physiological reasons, etc. The data value of the abnormal data may be close to the data value of the pseudo data, affecting the accuracy of the data analysis.

[0068] It should be further explained that blood oxygen saturation reflects the ability of human tissues to obtain oxygen. When the human body detects a decrease in blood oxygen saturation, it activates the body's compensatory response through multiple mechanisms. In order to increase oxygen intake and reduce carbon dioxide accumulation, the respiratory rate will increase significantly, thereby ensuring that various organs of the body can obtain sufficient oxygen to maintain normal human activities.

[0069] Based on this, the embodiment of the present invention verifies the accuracy of the mutation factor of the blood oxygen saturation data point by obtaining the respiratory rate data that is correlated with the blood oxygen saturation data. If the blood oxygen saturation data decreases, the respiratory rate data also increases, indicating that the stronger the negative correlation between the two, the stronger the authenticity of the blood oxygen saturation data.

[0070] For example, in an embodiment of the present invention, obtaining the observation range of the respiratory frequency data points in the respiratory frequency time series sequence of the human body also includes: collecting the respiratory frequency data of the human body, preprocessing the respiratory frequency data, and obtaining the respiratory frequency time series sequence.

[0071] The acquisition frequency and duration may be the same as those of the blood oxygen saturation data, and may be set according to actual needs. The embodiment of the present invention does not impose too many restrictions on this.

[0072] Specifically, when collecting the respiratory rate data of the human body, the respiratory rate data of the human body can be collected through a respiratory monitor.

[0073] It can be understood that, at the same acquisition moment, the blood oxygen saturation data points in the blood oxygen saturation time series sequence and the respiratory rate data points in the respiratory rate time series sequence correspond one to one.

[0074] To facilitate synchronous analysis, the observation range of the respiratory rate data points may be obtained based on a similar method. For specific steps, reference may be made to the method for obtaining the observation range of the blood oxygen saturation data points described above, which will not be described in detail in the embodiments of the present invention.

[0075] For example, when verifying the accuracy of the mutation factor of the blood oxygen saturation data point through the correlation between the respiratory rate data and the blood oxygen saturation data, the data value rankings of the blood oxygen saturation data points and the respiratory rate data points at the same collection time in their observation ranges can be obtained; then the variance of the respiratory rate data point in the observation range is obtained, and the false factor of the blood oxygen saturation data point is obtained by combining the data value ranking and the variance.

[0076] Among them, the method for obtaining the data value ranking of the blood oxygen saturation data point or the respiratory rate data point in its observation range includes: arranging the data values ​​in the observation range in ascending order to obtain the data value ranking corresponding to the blood oxygen saturation data point or the respiratory rate data point.

[0077] For example, in the embodiment of the present invention, the false factor of the blood oxygen saturation data point is determined, and the specific relationship can be referred to as follows:

[0078] ;

[0079] In the formula, is the false factor of the i-th blood oxygen saturation data point, is the mutation factor of the i-th blood oxygen saturation data point, is the data value ranking of the i-th blood oxygen saturation data point in its observation range, is the length of the observation range, is the data value ranking of the i-th respiratory frequency data point in its observation range, is a hyperparameter, is the variance of the Z-score scores of all respiratory rate data points in the observation range of the i-th respiratory rate data point, is the preset adjustment factor, is a linear normalization function.

[0080] The hyperparameter can be 0.1. The purpose of the hyperparameter is to prevent the numerator of the above formula from being 0. The value of the hyperparameter can be set according to actual needs. The preset adjustment factor can be set to 1. The purpose of setting the preset adjustment factor is to adjust the value of the false factor to the interval , which is convenient for subsequent data processing, the value of the preset adjustment factor can be set according to actual needs.

[0081] In the above formula, the larger the false factor of the blood oxygen saturation data point is, the greater the possibility that the blood oxygen saturation data point is false data.

[0082] It indicates the negative correlation between the blood oxygen saturation data point and the respiratory rate data point at the same collection time. The larger the value, the worse the negative correlation between the blood oxygen saturation data point and the respiratory rate data point. The blood oxygen saturation data point is less likely to be abnormal data generated by the human body, and the more likely it is to be false data, and the corresponding false factor is also larger.

[0083] The smaller the variance of the Z-score scores of all respiratory frequency data points in the observation range of the i-th respiratory frequency data point, the more stable the data changes in the observation range of the i-th respiratory frequency data point, that is, the possibility that the human body is active or in a state of tension in the acquisition time period corresponding to the observation range is small, the more likely the outstanding performance of the i-th blood oxygen saturation data point is caused by pseudo data, the less likely it is caused by physiological changes of the human body itself, and the larger the corresponding mutation factor. Among them, the variance of the Z-score scores of all respiratory frequency data points in the observation range of the respiratory frequency data point can be obtained by existing technology, and the embodiments of the present invention are not repeated here.

[0084] After obtaining the false factors of each blood oxygen saturation data point based on the above steps, the analysis results can be obtained through the false factors of the blood oxygen saturation data point, that is, continue to perform the following steps.

[0085] S4: The corrected target threshold is obtained by multiplying the false factor of the blood oxygen saturation data point by the preset threshold; the target threshold is used in the Chebyshev inequality to obtain the analysis result of the blood oxygen saturation data point.

[0086] Among them, the preset threshold can be set to 2; the preset threshold can be set according to actual needs, and the embodiment of the present invention does not impose too many restrictions on this.

[0087] It should be noted that, based on the above steps, the false factor of the blood oxygen saturation data point can be obtained. The larger the false factor, the greater the possibility that the blood oxygen saturation data point is false pseudo data. When performing abnormal monitoring based on Chebyshev's inequality, the threshold needs to be increased to avoid the blood oxygen saturation data point being classified as abnormal data.

[0088] After the target threshold of each blood oxygen saturation data point is obtained based on the above steps, the analysis result of each blood oxygen saturation data point can be obtained based on the target threshold.

[0089] By way of example, in an embodiment of the present invention, a target threshold is used in the Chebyshev inequality to obtain an analysis result of a blood oxygen saturation data point, including: obtaining a historical data set of blood oxygen saturation data points, and determining the mean and standard deviation of the historical data set; if the absolute value of the difference between the data value of the blood oxygen saturation data point and the mean is less than the product of the target threshold of the blood oxygen saturation data point and the standard deviation, then the blood oxygen saturation data point is normal data; otherwise, the blood oxygen saturation data point is abnormal data.

[0090] It can be understood that each blood oxygen saturation data point in the historical data set of blood oxygen saturation data points is the observed value of the blood oxygen saturation data point, and the mean is the expected value of the blood oxygen saturation data point. By analyzing the deviation of the blood oxygen saturation data point from the mean in its historical data set, abnormal data can be accurately identified.

[0091] It should be further explained that in order to improve the efficiency of subsequent management based on data analysis results, a database and analysis model can also be built.

[0092] For example, in an embodiment of the present invention, the analysis results of the blood oxygen saturation data points are obtained, and then the method also includes: storing normal data and abnormal data in a normal database and an abnormal database respectively, and constructing an analysis model based on the data values ​​and analysis results corresponding to the blood oxygen saturation data points.

[0093] Specifically, when constructing an analysis model, the data values ​​of the human body's blood oxygen saturation data points and the corresponding analysis results can be used as the main features, and the corresponding basic information of the human body can be used as auxiliary features, such as age, gender, medical history, status when data was collected, etc.; the main features and auxiliary features are combined into a feature vector, and the autoregressive integrated moving average model is trained to obtain an analysis model that can be used to predict and analyze human blood oxygen saturation data.

[0094] Among them, the specific steps of training the autoregressive integrated moving average model through the feature vector to obtain the analysis model that can be used to predict and analyze human blood oxygen saturation data can be achieved through the existing technology, and the embodiments of the present invention are not described in detail here.

[0095] In this way, the embodiment of the present invention can optimize the storage structure of the database, reduce unnecessary redundancy, and improve the speed and efficiency of data retrieval by storing normal data and abnormal data separately. When the analysis model detects that the blood oxygen saturation data is abnormal, the system can quickly extract relevant information from the abnormal database to provide doctors with timely medical decision support.

[0096] It can be seen that in the embodiment of the present invention, when determining the analysis result of the clinical trial data, the observation ranges of the blood oxygen saturation data points in the blood oxygen saturation time series sequence of the human body and the respiratory rate data points in the respiratory rate time series sequence can be preset; the normalized value of the difference in the coefficient of variation before and after the blood oxygen saturation data point is removed from the observation range of a blood oxygen saturation data point is recorded as a mutation index, and the mutation factor of the blood oxygen saturation data point is obtained by multiplying the mutation index with the mean value of the data change amount between all adjacent blood oxygen saturation data points in the observation range; the false factor of the blood oxygen saturation data point is calculated; the corrected target threshold is obtained by multiplying the false factor of the blood oxygen saturation data point and the preset threshold; the target threshold is used in the Chebyshev inequality to obtain the analysis result of the blood oxygen saturation data point.

[0097] In this way, the embodiment of the present invention analyzes the abnormal data in the blood oxygen saturation time series through the Chebyshev inequality, and can filter out abnormal data from the blood oxygen saturation time series of various distribution forms. In this process, the embodiment of the present invention takes into account the signal overlap of the collected blood oxygen saturation data, resulting in the possibility of false pseudo data in the blood oxygen saturation data. Based on this, the embodiment of the present invention analyzes the prominent performance of the blood oxygen saturation data points in the local range, and corrects the preset threshold value through the possibility of it being pseudo data, thereby improving the accuracy of the analysis results; on this basis, the embodiment of the present invention also takes into account that the abnormal data generated by the human body may be similar to pseudo data. In order to further distinguish, the embodiment of the present invention also obtains the respiratory rate data that is correlated with the blood oxygen saturation data to correct the prominent performance, and can accurately obtain the target threshold corresponding to each blood oxygen saturation data point, effectively improving the accuracy of the analysis results.

[0098] An embodiment of the present invention further discloses a clinical trial data analysis system, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a clinical trial data analysis method provided by the present invention is implemented.

[0099] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0100] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM, a dynamic random access memory DRAM, a static random access memory SRAM, an enhanced dynamic random access memory EDRAM, a high bandwidth memory HBM, a hybrid memory cube HMC, etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device.

[0101] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

[0102] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A clinical trial data analysis method, characterized in that: include: Presetting the observation range of the blood oxygen saturation data points in the human body's blood oxygen saturation time series and the respiratory rate data points in the respiratory rate time series; The normalized value of the difference in coefficient of variation before and after excluding a blood oxygen saturation data point from the observation range is recorded as the mutation index. The mutation factor of the blood oxygen saturation data point is obtained by multiplying the mutation index by the mean of the data changes between all adjacent blood oxygen saturation data points in the observation range. Calculate the spurious factor for the oxygen saturation data point: ; is the false factor of the i-th blood oxygen saturation data point, is the mutation factor of the i-th blood oxygen saturation data point, is the data value ranking of the i-th blood oxygen saturation data point in its observation range, is the length of the observation range, is the data value ranking of the i-th respiratory rate data point in its observation range, is a hyperparameter, is the variance of the Z-score scores of all respiratory rate data points in the observation range of the i-th respiratory rate data point, is the preset adjustment factor, is a linear normalization function; The corrected target threshold is obtained by multiplying the false factor of the blood oxygen saturation data point by the preset threshold; the target threshold is used in the Chebyshev inequality to obtain the analysis result of the blood oxygen saturation data point.

2. A clinical trial data analysis method according to claim 1, characterized in that: The observation range of the blood oxygen saturation data points in the preset human body's blood oxygen saturation time series and the respiratory rate data points in the respiratory rate time series also includes: The blood oxygen saturation data and respiratory rate data of the human body are collected, and the blood oxygen saturation data and respiratory rate data are preprocessed to obtain a blood oxygen saturation time series sequence and a respiratory rate time series sequence.

3. A clinical trial data analysis method according to claim 1, characterized in that: The method for obtaining the observation range includes: Preset observation range length , taking the blood oxygen saturation data point or respiratory rate data point as the center, obtain blood oxygen saturation data points or respiratory rate data points to construct an observation range of the blood oxygen saturation data points or respiratory rate data points.

4. A clinical trial data analysis method according to claim 1, characterized in that: The method for obtaining the coefficient of variation includes: The percentages of the ratios of the standard deviation to the mean value before and after the blood oxygen saturation data point is removed from the observation range of the blood oxygen saturation data point are respectively obtained, and recorded as the coefficient of variation before and after removal.

5. A clinical trial data analysis method according to claim 1, characterized in that: A method for obtaining a data value ranking of a blood oxygen saturation data point or a respiratory rate data point within its observation range includes: Arrange the data values ​​in the observation range in ascending order to obtain the data value ranking corresponding to the blood oxygen saturation data point or the respiratory rate data point.

6. A clinical trial data analysis method according to claim 1, characterized in that: The target threshold is used in the Chebyshev inequality to obtain analysis results of the blood oxygen saturation data points, including: Obtain a historical data set of blood oxygen saturation data points and determine the mean and standard deviation of the historical data set; If the absolute value of the difference between the data value of the blood oxygen saturation data point and the mean is less than the product of the target threshold value and the standard deviation of the blood oxygen saturation data point, then the blood oxygen saturation data point is normal data; otherwise, the blood oxygen saturation data point is abnormal data.

7. A clinical trial data analysis method according to claim 6, characterized in that: The analysis result of the obtained blood oxygen saturation data point further includes: Normal data and abnormal data are stored in a normal database and an abnormal database respectively, and an analysis model is constructed based on the data values ​​and analysis results corresponding to the blood oxygen saturation data points.

8. A clinical trial data analysis system, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a clinical trial data analysis method according to any one of claims 1 to 7 is implemented.

Images