Credibility analysis system for test data

Abstract

The invention discloses a credibility analysis system for test data, which provides comprehensive test result credibility prompts for test doctors through a plurality of methods such as dynamic Delta analysis, moving mean value analysis, dynamic range examination, self-adaptive correlation analysis and personalized influence factor analysis. The method assists an inspector in quickly and accurately evaluating the credibility of an inspection result, so that the risk of misdiagnosis and missed diagnosis is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of medical testing, and is particularly applicable to a credibility analysis system for testing data. Background Art

[0002] In the field of medical testing, the reliability of test results is directly related to the accuracy of clinical diagnosis. Although current testing technology is quite mature, due to the complexity of the human body's physiological state and various interference factors that may be encountered during the test, the interpretation of test results still requires a high degree of professional knowledge and experience. Traditional test result evaluation mainly relies on the professional judgment of the testing physician and is compared with the reference range set by the laboratory.

[0003] In recent years, with the development of information technology, some medical institutions have begun to try to use statistical software for preliminary data analysis to assist in the interpretation of results. However, these methods often focus only on a single indicator or simple statistics, lacking comprehensive consideration of the complex physiological mechanisms and external factors behind the test results. They either rely too much on the threshold judgment of a certain test item and ignore the mutual influence of multiple indicators; or the reference range is fixed and does not take into account the dynamic changes in data distribution, individual differences of patients and disease progression; or ignore the potential impact of external factors such as physiological factors, pathological conditions, interference factors, and drug use on the test results. Summary of the invention

[0004] The purpose of the present invention is to provide a credibility analysis system for inspection data, which is used to solve the technical problems of the current analysis of inspection data with a single evaluation index, the use of static range limitations, and the lack of consideration of external factors.

[0005] To achieve the above object, the present invention adopts the following technical solutions: The reliability analysis system of the inspection data of the present invention comprises a data processing module and a data analysis module; The data processing module obtains raw test data and clinical test information from the LIS system and performs data cleaning according to standardized rules; The data analysis module performs dynamic Delta analysis, moving mean analysis, dynamic range limit analysis, adaptive correlation analysis, and personalized influencing factor analysis on the designated project test data and clinical test information, summarizes the analysis results, and displays them visually.

[0006] Furthermore, the data cleaning cleans the original test data and clinical test information according to the configured cleaning method logic.

[0007] Furthermore, the original test data and clinical test information cover at least 2 months.

[0008] Furthermore, the dynamic Delta analysis determines the fluctuation range according to the length of the time interval between two inspections, compares the difference between the current inspection result and the previous inspection result, identifies abnormal fluctuations, and promptly warns of inspection errors.

[0009] Furthermore, the moving average analysis determines the moving average of a series of continuous test results based on a moving window, which is used to identify the trend of the test results within a specified time and eliminate the impact of short-term fluctuations.

[0010] Furthermore, the dynamic range limit analysis is used to dynamically obtain the limit range of designated item test data and clinical test information, and determine the distribution of current test results.

[0011] Furthermore, the adaptive correlation analysis evaluates the correlation between different test items by calculating an adaptive comprehensive correlation coefficient.

[0012] Furthermore, the personalized influencing factor analysis finds interference factors that match the original test data and clinical test information according to matching rules, based on influencing factor knowledge and rule engines.

[0013] Furthermore, the influencing factor knowledge and rule engine includes a knowledge base formed by collating professional books and medical guide knowledge.

[0014] The advantage of the present invention is that it provides comprehensive test result credibility prompts to test physicians through multiple methods such as dynamic Delta analysis, moving mean analysis, dynamic range check, adaptive correlation analysis and personalized influencing factor analysis, assisting test physicians in quickly and accurately evaluating the credibility of test results, thereby reducing the risk of misdiagnosis and missed diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic diagram of the principle framework of the system of the present invention.

[0016] Figure 2 It is a schematic diagram of the use process of the system of the present invention. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only 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 ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] like Figure 1 As shown, the credibility analysis system of the inspection data of the present invention includes a data processing module and a data analysis module.

[0019] The data processing module obtains the original test data and clinical test information from the LIS system, and performs data cleaning according to standardized rules; generally, the test data credibility analysis system of the present invention can obtain the original test data and clinical test information from the LIS in real time, and needs to connect to the historical test data information of two months of LIS to ensure the comprehensiveness of the test information and sufficient data volume.

[0020] After collecting the corresponding data, the test data is cleaned according to the pre-configured standardized rules, and the test data result information and clinical information are processed uniformly to provide effective data support for subsequent analysis. The cleaning rules can be configured as a whole solution such as method processing, digital calculation, object conversion and character processing, or as a workflow method for step-by-step processing.

[0021] The data analysis module performs dynamic Delta analysis, moving mean analysis, dynamic range limit analysis, adaptive correlation analysis, and personalized influencing factor analysis on the original test data and clinical test information of the specified items, summarizes the analysis results, and displays them visually. The credibility analysis system of the test data of the present invention summarizes the results of multi-factor analysis and displays them in visual charts to assist the test physician in conducting credibility assessment on the test results of the specified items.

[0022] The dynamic Delta analysis determines the fluctuation range according to the length of the time interval between two tests, compares the difference between the current test result and the previous test result, identifies abnormal fluctuations, and promptly warns of test errors. The difference from the traditional Delta analysis is that the dynamic Delta analysis in the present invention adds the time factor, and dynamically analyzes the fluctuation range according to the length of the time interval between two tests; the longer the interval time, the larger the fluctuation range allowed, and the smaller the interval time, the smaller the fluctuation range. The formula for the fluctuation range is as follows: Among them, previous result is the previous test result, present result is the current test result, and time interval is the time interval.

[0023] The moving mean analysis determines the moving average of a series of continuous test results based on a moving window, which is used to identify the trend of the test results within a specified time, eliminate the impact of short-term fluctuations, and provide a more stable basis for result evaluation. The size of the moving window can be adjusted according to actual business needs; at the same time, the distribution of the current test results is reflected based on the mean and standard deviation.

[0024] The dynamic range limit analysis is used to dynamically calculate the limit range of the test data and clinical test information of the specified project based on the historical test data in the hospital, and use this range to determine the distribution of the current test results. At the same time, the distribution of current data and historical data is visualized through charts; parameters can be adjusted dynamically or the department range can be specified to improve the accuracy of the evaluation. The range limit value is dynamically calculated, and the limit range standard can be adjusted according to the business scenario. Different data ranges are specified, and the range limit values ​​are different, which is more in line with the evaluation of the specified data set.

[0025] The adaptive correlation analysis evaluates the correlation between different test items by calculating the adaptive comprehensive correlation coefficient, identifies potential physiological mechanisms or pathological patterns, and enhances the understanding of the physiological status behind the results.

[0026] The present invention proposes an adaptive comprehensive correlation coefficient ACCC, the formula is as follows: in, and , are all decimals between 0 and 1, and their sum is 1.

[0027] w p is the weight corresponding to the Pearson correlation coefficient (Pearson correlation coefficient), and the calculation formula is: ; W s is the weight corresponding to the Spearman correlation coefficient, and the calculation formula is: ; W K is the weight corresponding to the Kendall correlation coefficient, and the calculation formula is: .W K The calculation ensures that all weights add up to 1 and do not exceed the absolute value of the Kendall correlation coefficient.

[0028] f N is the normality weight function, ,in is the Kolmogorov-Smirnov test statistic, λ and μ Parameters that need to be adjusted according to the characteristics of the data set. The normality weight function is a Sigmoid function, ensuring Between 0 and 1.

[0029] fM is a monotonic weight function, defined using the absolute value of the Spearman correlation coefficient. .

[0030] Pearson correlation coefficient, Spearman correlation coefficient, Kendall correlation coefficient are all coefficients that reflect the direction and degree of the trend of change between two variables, and their value range is -1 to +1. 0 means that the two variables are uncorrelated, positive values ​​indicate positive correlation, and negative values ​​indicate negative correlation. p (x,y),r s (x,y),t k (x, y) are the Pearson correlation coefficient, Spearman correlation coefficient, and Kendall correlation coefficient calculated based on the distribution of the test data. The present invention is based on these three correlation coefficients and combines statistical tests (such as Kolmogorov-Smirnov test) to evaluate whether the test result data is approximately normally distributed and the monotonicity between the test data, and then dynamically allocates the weights of each correlation coefficient.

[0031] The adaptive composite relevance coefficient provides a method to integrate different relevance measures into a unified framework, allowing the relevance measure to be dynamically adjusted according to different data characteristics without the need to manually maintain the relationship between items.

[0032] The personalized influencing factor analysis, according to the matching rules, influencing factor knowledge and rule engine, finds interference factors that match the original test data and clinical test information. That is, in combination with the patient's personal clinical information and test results, based on professional influencing factor knowledge and rule engine, the possible impact of external factors such as physiological factors, pathological conditions, interference factors, and drug use on the test results are dynamically matched through rules, and the corresponding detailed information can be viewed for each matched factor. Here, the influencing factor knowledge and rule engine include a knowledge graph formed by collating professional books and medical guide knowledge. When information such as clinical information and test results meets the expression conditions in the rules, the system will automatically retrieve the corresponding influencing factor knowledge and display it visually.

[0033] like Figure 2 As shown, the use process of the credibility analysis system of the inspection data of the present invention is as follows: First, the original test data and clinical information are obtained from LIS in real time.

[0034] Secondly, the test data is cleaned according to the pre-configured standardized rules, and the test data result information and clinical information are processed in a unified manner to provide effective data support for subsequent analysis.

[0035] Then, based on the test information and clinical information of the specified items, dynamic Delta analysis, moving mean analysis, dynamic range limit analysis, adaptive correlation analysis, and personalized influencing factor analysis are carried out in parallel; for personalized influencing factor analysis, personalized analysis is performed based on the influencing factor knowledge and rule engine maintained in advance.

[0036] Dynamic Delta Analysis: analyzes the fluctuation of the current result and the most recent result. Moving average analysis: analyzes the trend of historical data and the rationality of the current value. Dynamic range limit analysis: analyzes the outliers of historical data range and the current value range. Adaptive correlation analysis: analyzes the correlation of two items and the distribution of overall correlation. Personalized influencing factor analysis: organizes patent knowledge and rules to form influencing factor knowledge and rule engine. Analyzes the influencing factors of the current result by combining influencing factor knowledge and rule engine with the patient's personal situation.

[0037] Finally, the system will summarize the results of the multifactor analysis and present them in visual charts to assist laboratory physicians in evaluating the credibility of the test results of designated items.

Claims

1. A credibility analysis system for inspection data, characterized in that: Including data processing module and data analysis module; The data processing module obtains raw test data and clinical test information from the LIS system and performs data cleaning according to standardized rules; The data analysis module performs dynamic Delta analysis, moving mean analysis, dynamic range limit analysis, adaptive correlation analysis, and personalized influencing factor analysis on the original test data and clinical test information of the specified project, summarizes the analysis results, and displays them visually.

2. The system for analyzing the credibility of verification data according to claim 1, characterized in that: The data cleaning cleans the original test data and clinical test information according to the configured cleaning method logic.

3. The system for analyzing the credibility of verification data according to claim 1, characterized in that: The original test data and clinical test information shall cover at least 2 months.

4. The system for analyzing the credibility of verification data according to claim 1, characterized in that: The dynamic Delta analysis determines the fluctuation range according to the length of the time interval between two inspections, compares the difference between the current inspection result and the previous inspection result, identifies abnormal fluctuations, and promptly warns of inspection errors.

5. The system for analyzing the credibility of verification data according to claim 1, characterized in that: The moving average analysis determines the moving average of a series of continuous test results based on a moving window, which is used to identify the trend of the test results within a specified time and eliminate the impact of short-term fluctuations.

6. The system for analyzing the credibility of verification data according to claim 1, characterized in that: The dynamic range limit analysis is used to dynamically obtain the limit range of the original test data and clinical test information of the specified project and determine the distribution of the current test results.

7. The credibility analysis system of verification data according to claim 1, characterized in that: The adaptive correlation analysis evaluates the correlation between different test items by calculating the adaptive comprehensive correlation coefficient.

8. The system for analyzing the credibility of verification data according to claim 1, characterized in that: The personalized influencing factor analysis finds interference factors that match the original test data and clinical test information according to matching rules, influencing factor knowledge and rule engines.

9. The system for analyzing the credibility of verification data according to claim 1, characterized in that: The influencing factor knowledge and rule engine includes a knowledge base formed by collating professional books and medical guide knowledge.

Images