Medical detection quality management system and method based on machine learning

Abstract

The invention discloses a medical detection quality management system and method based on machine learning, and relates to the technical field of machine learning. The method comprises the following steps: determining an analysis period of medical detection analysis based on medical detection requirements, and screening representative features in key clinical and laboratory data influencing medical detection quality based on the analysis period; quantitatively calculating various medical detection quality control indexes based on the representative features in combination with a long-term optimization scheduling model; performing standardization processing on the quality control index to generate a standardized quality control index; converting the standardized quality control index into annual multi-dimensional time sequence data per time period; and performing clustering analysis on the multi-dimensional time series data in each time period in the year to complete quality management staging of medical detection. According to the invention, comprehensive management of medical detection quality is realized, accuracy and management efficiency of detection quality evaluation are improved, and improvement of medical level and patient satisfaction is facilitated.

Description

Technical Field

[0001] The present invention belongs to the technical field of machine learning, and in particular relates to a medical testing quality management system and method based on machine learning. Background Art

[0002] In the field of medical testing, ensuring the accuracy and reliability of test results is crucial. However, with the continuous advancement of medical technology and the growing demand for testing, traditional medical testing quality management methods have been unable to meet current needs. Traditional quality management methods often rely on manual monitoring and regular evaluation, which is not only inefficient, but also difficult to fully and accurately reflect the true situation of medical testing quality.

[0003] First, medical testing involves a variety of disease types, patient population characteristics and testing frequencies, all of which will affect the quality of testing. Traditional quality management methods often ignore the differences and dynamics of these factors, resulting in a lack of pertinence and flexibility in quality management measures. Secondly, there is a large amount of clinical and laboratory data in the medical testing process, which contains rich information and is of great significance for evaluating the quality of testing. However, traditional quality management methods often lack effective data analysis and processing methods, and cannot make full use of these data to improve the quality of testing. In addition, the performance and environmental factors of medical testing equipment will also affect the quality of testing. Traditional quality management methods often lack real-time monitoring and evaluation of equipment performance and environmental factors, making it difficult to detect and solve potential quality problems in a timely manner.

[0004] In order to solve the above problems, people began to explore medical testing quality management methods based on machine learning. As an advanced data analysis and processing technology, machine learning can automatically extract useful information from large amounts of data and use it for prediction and decision-making. However, medical testing quality management methods based on machine learning also face some challenges. For example, how to determine reasonable analysis cycles and key indicators, how to quantitatively calculate quality control indicators, how to standardize quality control indicators, and how to convert these indicators into multidimensional time series data for cluster analysis are all issues that need to be solved.

[0005] Therefore, the present invention proposes a medical testing quality management method based on machine learning, which aims to achieve real-time monitoring and dynamic adjustment of the medical testing process and improve the accuracy and reliability of the testing quality by introducing a machine learning algorithm. Summary of the invention

[0006] In order to overcome the shortcomings and deficiencies of the above-mentioned prior art, the first purpose of the present invention is to provide a medical testing quality management system based on machine learning; the second purpose of the present invention is to provide a medical testing quality management method based on machine learning.

[0007] The first object of the present invention adopts the following technical solution:

[0008] Medical testing quality management system based on machine learning, including demand analysis and cycle determination module, quality control index calculation module, standardization processing module, multi-dimensional time series data conversion module and cluster analysis detection staging module;

[0009] Demand Analysis and Cycle Determination Module: Based on the specific needs of medical testing and the operational characteristics of hospitals or clinics, determine the analysis cycle of medical testing and screen out representative features of key clinical and laboratory data that affect the quality of testing; output a clear analysis cycle and representative feature screening list to provide a basis for subsequent steps;

[0010] Quality control index calculation module: Use machine learning algorithms for long-term optimization scheduling, and calculate the quality control indicators of various medical tests based on representative features, including the error rate of each period of the test, the impact of environmental conditions, etc.; output a series of quantitative quality control indicators for evaluating the quality of medical tests;

[0011] Standardization processing module: standardizes the quality control indicators to eliminate the dimension differences and numerical range differences between different indicators, so that different indicators can be compared and comprehensively analyzed; outputs standardized quality control indicator data to provide a basis for subsequent multi-dimensional time series data conversion and cluster analysis;

[0012] Multidimensional time series data conversion module: converts the standardized quality control indicators into multidimensional time series data for each period of the year to facilitate time series analysis and cluster analysis; outputs multidimensional time series data for each period of the year to provide data support for cluster analysis and detection staging;

[0013] Cluster analysis and testing staging module: cluster analysis is performed on multidimensional time series data for each period of the year, and medical tests are divided into different quality management stages based on the similarities and differences of the data, so as to carry out targeted quality management for different stages; the quality management staging results of medical tests are output to provide scientific quality management basis for medical institutions.

[0014] The second object of the present invention adopts the following technical solution:

[0015] The medical testing quality management method based on machine learning is used to implement a medical testing quality management system based on machine learning. The method flow is as follows:

[0016] Step 1: Determine the analysis cycle and key indicators: Based on the medical testing needs including disease type, patient population characteristics, and testing frequency, determine the analysis cycle of medical testing analysis, and based on the analysis cycle, screen representative features in key clinical and laboratory data that affect the quality of medical testing;

[0017] Step 2: Quantitatively calculate quality control indicators: Based on representative features and combined with a long-term optimization scheduling model, quantitatively calculate the quality control indicators of various medical tests; quality control indicators are multi-year sequence data of multiple influencing factors; long-term optimization scheduling includes a single-period test benefit loss objective function, sample size balance constraints, upper and lower limits of test constraints, and upper and lower limits of equipment capacity constraints;

[0018] Step 3: Standardization: Standardize the quality control indicators to generate standardized quality control indicators;

[0019] Step 4: Convert to multidimensional time series data: Convert the standardized quality control indicators into multidimensional time series data for each period of the year;

[0020] Step 5: Cluster analysis to complete the detection stage: Perform cluster analysis on the multi-dimensional time series data for each period within the year to complete the quality management stage of medical testing.

[0021] Preferably, step one specifically includes:

[0022] S11. Analyze the operational characteristics of the hospital or clinic, including diagnostic tasks, patient types, testing equipment performance, and changes in testing volume during peak and trough periods;

[0023] S12. Determine the analysis cycle: Determine the analysis period according to the detection requirements, including daily, weekly or monthly; use a shorter analysis cycle during peak periods and a longer analysis cycle during low periods;

[0024] S13. Screen representative characteristics: During the analysis cycle, screen key data that affect the quality of medical testing, including patient history, symptoms, laboratory test results, treatment response, environmental factors, and equipment operating parameters.

[0025] Preferably, step 2 specifically includes:

[0026] S21. Use machine learning algorithms for long-term optimization scheduling: Based on representative features, long-term optimization scheduling is performed to obtain the optimal operation process of each analysis cycle of medical testing; the objective function is to minimize the loss of single-period testing benefits; the constraints include sample size balance constraints, upper and lower limits of testing constraints, and upper and lower limits of equipment capacity constraints;

[0027] S22. Calculate quality control indicators: calculate the error rate of each testing period, the number of samples to be tested in each testing period, and the impact of environmental conditions on testing;

[0028] S23. Environmental condition impact index: Take the weighted average of the environmental data of each monitoring point, calculate the standardized environmental coefficient, and obtain the impact index of environmental conditions on detection.

[0029] As a preferred method, the calculation formula of the error rate of each period of the detection stage is:

[0030] E yt =max(C yt )-Q yt ;

[0031] Where y is the year; t is the analysis period; C yt is the maximum allowable detection accuracy for a certain analysis period t in a certain year y; Q yt is the actual detection accuracy of a certain analysis period t in a certain year y.

[0032] As a preferred method, the calculation formula of the environmental condition impact index is:

[0033]

[0034] Among them, ω i is the weight of each monitoring point; P i It is the environmental data of each monitoring point, including temperature and humidity.

[0035] Preferably, step three specifically includes:

[0036] S31. Determine positive effect indicators and negative effect indicators: positive effect indicators include accuracy, sensitivity, and specificity; negative effect indicators include misdiagnosis rate, missed diagnosis rate, false positive rate, and false negative rate;

[0037] S32, standardization processing: Use the standardized formula to process the positive effect index and the negative effect index to generate the standardized quality control index. Specifically, the standardized formula is as follows:

[0038]

[0039] Among them, Z jyt is the standardized quality control index; X jyt is a positive effect indicator or a negative effect indicator; j is the type number of the quality control indicator; μ j is the mean value of the quality control index; σ j is the variance of the quality control index; y is the year; t is the analysis period.

[0040] Preferably, step 4 specifically includes:

[0041] Convert to multidimensional time series data: Convert the standardized quality control indicators into multidimensional time series data for each period of the year. The specific formula is:

[0042] M T(Y,J) = reshape(Z Y(T,J) ,T,Y×J);

[0043] Among them, M T(Y,J) To convert the standardized quality control indicators with dimensions of Y×T rows and J columns into multi-dimensional time series data with dimensions of T rows and Y×J columns for each period of the year; Y is the number of years of historical data; T is the total number of analysis cycles divided within the year;

[0044] The output data is a multidimensional time series data set, in which each time point, that is, each analysis period, has a corresponding set of indicator values.

[0045] Preferably, step five specifically includes:

[0046] S51. Determine the optimal number of stages: for multi-dimensional time series data of each period within a year, use a cluster number optimization algorithm to determine the optimal number of stages for medical detection; the cluster number optimization algorithm includes a gap statistics cluster number optimization algorithm and an average profile cluster number optimization algorithm;

[0047] S52. Complete the detection staging: according to the optimal number of medical detection stages, use clustering methods to complete the detection staging; clustering methods include K-means clustering method and hierarchical clustering method.

[0048] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0049] 1. The present invention can accurately determine the analysis cycle and screen out key data features that affect the quality of medical testing by carefully analyzing the operational characteristics of hospitals or clinics, including diagnostic tasks, patient types, testing equipment performance, and fluctuations in testing demand. At the same time, through the application of long-term optimization scheduling models and machine learning algorithms, various quality control indicators such as detection error rate and environmental condition impact indicators can be quantitatively calculated, thereby achieving real-time monitoring and dynamic adjustment of the medical testing process to ensure the accuracy and stability of the test results.

[0050] 2. The present invention automatically adjusts the analysis cycle and detection strategy by detecting changes in demand to ensure that high-quality detection results can be obtained in each analysis cycle. At the same time, sample size balance constraints, detection upper and lower limit constraints, and equipment capacity upper and lower limit constraints are also considered to ensure the rational allocation and efficient use of resources. This intelligent resource allocation method can not only improve detection efficiency, but also reduce unnecessary resource consumption.

[0051] 3. The present invention can transform complex quality control indicators into a form that is easy to understand and analyze through standardized processing and conversion of multidimensional time series data. At the same time, through cluster analysis, it can automatically identify and divide different quality management stages, thereby realizing comprehensive monitoring and refined management of the medical testing process. This scientific and intelligent management method can not only improve the testing level and quality management capabilities of medical institutions, but also provide patients with safer, more reliable and efficient medical services. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0053] Figure 1 The module diagram of the medical testing quality management system based on machine learning of the present invention is shown;

[0054] Figure 2 A flow chart of a medical testing quality management method based on machine learning of the present invention is shown;

[0055] Figure 3 A flow chart of determining the analysis cycle and key indicators of the present invention is shown. DETAILED DESCRIPTION

[0056] 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 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.

[0057] In addition, the described features, structures or characteristics may be combined in one or more example embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the example embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced while omitting one or more of the specific details, or other methods, components, steps, etc. may be adopted. In other cases, well-known structures, methods, implementations or operations are not shown or described in detail to avoid obscuring various aspects of the present disclosure.

[0058] Embodiment 1:

[0059] See also Figure 1As shown, the medical testing quality management system based on machine learning in this embodiment includes a demand analysis and cycle determination module, a quality control indicator calculation module, a standardization processing module, a multidimensional time series data conversion module and a cluster analysis detection staging module.

[0060] Demand Analysis and Cycle Determination Module: Based on the specific needs of medical testing and the operational characteristics of hospitals or clinics, determine the analysis cycle of medical testing and screen out representative features of key clinical and laboratory data that affect the quality of testing. Output a clear analysis cycle and representative feature screening list to provide a basis for subsequent steps.

[0061] Quality control indicator calculation module: Use machine learning algorithms for long-term optimization scheduling, and calculate the quality control indicators of various medical tests based on representative features, including the error rate of each test stage and time period, indicators affecting environmental conditions, etc.; output a series of quantitative quality control indicators to evaluate the quality of medical tests.

[0062] Standardization processing module: standardize the quality control indicators to eliminate the dimensional differences and numerical range differences between different indicators, so that different indicators can be compared and comprehensively analyzed; output standardized quality control indicator data to provide a basis for subsequent multidimensional time series data conversion and cluster analysis.

[0063] Multidimensional time series data conversion module: converts the standardized quality control indicators into multidimensional time series data for each period of the year to facilitate time series analysis and cluster analysis; outputs multidimensional time series data for each period of the year to provide data support for cluster analysis and detection staging.

[0064] Cluster analysis and testing staging module: cluster analysis is performed on multidimensional time series data for each period of the year, and medical tests are divided into different quality management stages based on the similarities and differences of the data, so as to carry out targeted quality management for different stages; the quality management staging results of medical tests are output to provide scientific quality management basis for medical institutions.

[0065] The beneficial effects of this embodiment are: comprehensive management of medical testing quality is achieved through modular design, the assessment accuracy and management efficiency of testing quality are improved, a scientific and systematic quality management basis is provided for medical institutions, and it helps to improve medical standards and patient satisfaction.

[0066] Embodiment 2:

[0067] See also Figure 2 As shown, the medical testing quality management method based on machine learning in this embodiment has the following process:

[0068] Step 1. Determine the analysis cycle and key indicators: Based on the medical testing needs (such as disease type, patient population characteristics, testing frequency, etc.), determine the analysis cycle of medical testing analysis, and based on the analysis cycle, screen representative features in key clinical and laboratory data that affect the quality of medical testing.

[0069] See also Figure 3 As shown, the steps to determine the analysis cycle and key indicators are as follows:

[0070] S11. Investigate and analyze the operational characteristics of the hospital or clinic.

[0071] Diagnostic tasks: Clarify the main types of diagnosis (such as routine physical examinations, emergency treatment, chronic disease management, etc.).

[0072] Patient Type: Identify different types of patient populations (e.g., children, adults, elderly, patients with specific diseases, etc.).

[0073] Testing equipment performance: Evaluate the technical parameters and performance indicators of existing testing equipment.

[0074] Changes in testing volume during peak and trough periods: Analyze fluctuations in testing demand during different time periods (such as flu season, holidays, etc.).

[0075] S12. Determine the analysis period.

[0076] Determine the time period for analysis based on your testing needs, such as daily, weekly, or monthly.

[0077] During peak periods (such as flu season), use shorter analysis periods (such as daily or hourly) to allow for more frequent monitoring and adjustments.

[0078] During low periods, use a longer analysis cycle (such as weekly or monthly) to reduce unnecessary resource consumption.

[0079] S13. Screen representative features.

[0080] During the analysis cycle, key data that affect the quality of medical testing are screened, including but not limited to:

[0081] The patient’s medical history: previous illnesses, allergies, family medical history, etc.

[0082] Symptoms: Current symptoms.

[0083] Laboratory test results: blood, urine, imaging, etc. test results.

[0084] Treatment response: The response to treatment measures taken.

[0085] Environmental factors: temperature, humidity, air quality, etc.

[0086] Equipment operating parameters: detect the working status and performance parameters of the equipment.

[0087] This embodiment takes blood analysis as an example and selects the following features:

[0088] The patient’s medical history: For example, whether the patient has had blood-related diseases such as anemia or leukemia in the past, which may affect the interpretation of the blood analysis results.

[0089] Symptoms: For example, whether the patient has symptoms such as fatigue, dizziness, pale skin, etc. These symptoms may be related to the blood analysis results.

[0090] Laboratory test results: Specific blood analysis indicators such as red blood cell count, white blood cell count, hemoglobin concentration, etc.

[0091] Treatment response: Such as how a patient responds to a blood transfusion or medication, which can reflect the effectiveness of the treatment and changes in the patient's condition.

[0092] Environmental factors: such as laboratory temperature and humidity, which may affect the performance of the testing equipment and the accuracy of the test results.

[0093] Equipment operating parameters: such as the calibration status and usage time of the blood analyzer, which can reflect the operating status and performance of the equipment.

[0094] Step 2: Quantitatively calculate quality control indicators: Based on the representative characteristics and combined with the long-term optimization scheduling model, quantitatively calculate the quality control indicators of various medical tests; the quality control indicators are multi-year series data of multiple influencing factors; the long-term optimization scheduling includes a single-period detection benefit loss objective function, sample size balance constraints, detection upper and lower limit constraints, and equipment capacity upper and lower limit constraints.

[0095] Among them, the multi-year series data of multiple influencing factors include the error rate of each detection stage and time period, the number of samples to be tested in each detection stage and time period, and the impact of environmental conditions on detection.

[0096] Specifically include:

[0097] S21. Use machine learning algorithms for long-term optimization scheduling.

[0098] Dynamic programming algorithm or reinforcement learning algorithm: Based on the representative features, long-term optimization scheduling is performed to obtain the optimal operation process of each analysis cycle of medical testing.

[0099] Objective function: Minimize the loss of detection benefits in a single period.

[0100] Constraints:

[0101] Sample size balance constraint: Ensure that the sample size in each analysis cycle meets expectations.

[0102] Detection upper and lower limit constraints: set a reasonable range for each detection.

[0103] Equipment capacity upper and lower limit constraints: ensure that the equipment operates within a safe range.

[0104] S22. Calculate quality control indicators.

[0105] The error rate of each period of the detection phase is obtained by subtracting the expected accuracy rate of each analysis cycle from the actual accuracy rate to obtain the error rate E of each period of the detection phase. yt :

[0106] E yt =max(C yt )-Q yt ;

[0107] Where y is the year; t is the analysis period; C yt is the maximum allowable detection accuracy for a certain analysis period t in a certain year y; Q yt is the actual detection accuracy of a certain analysis period t in a certain year y.

[0108] S23, Environmental Condition Impact Index: Take the weighted average of the environmental data (such as temperature and humidity) of each monitoring point, calculate the standardized environmental coefficient, and obtain the environmental condition impact index I yt :

[0109]

[0110] Among them, ω i is the weight of each monitoring point; P i It is the environmental data of each monitoring point (such as temperature and humidity).

[0111] Step 3: Standardization processing.

[0112] S31. Determine positive effect indicators and negative effect indicators.

[0113] Positive effect indicators: The higher the value of these indicators, the better the quality of the medical test. For example, indicators such as accuracy, sensitivity, and specificity are all positive effect indicators. When the value of these indicators increases, it means that the accuracy of the medical test is improving.

[0114] Negative effect indicators: The higher the value of these indicators, the worse the quality of the medical test. For example, indicators such as misdiagnosis rate, missed diagnosis rate, false positive rate and false negative rate are all negative effect indicators. When the value of these indicators increases, it means that the accuracy of the medical test is decreasing.

[0115] S32: Standardization processing.

[0116] The positive effect index and the negative effect index are processed using a standardized formula to generate a standardized quality control index. Specifically, the standardized formula is as follows:

[0117]

[0118] Among them, Z jyt is the standardized quality control index; X jyt is a positive effect indicator or a negative effect indicator; j is the type number of the quality control indicator; μ j is the mean value of the quality control index; σ j is the variance of the quality control index; y is the year; t is the analysis period.

[0119] Step 4: Convert to multidimensional time series data.

[0120] The standardized quality control indicators are converted into multi-dimensional time series data for each period of the year.

[0121] M T(Y,J) = reshape(Z Y(T,J) ,T,Y×J);

[0122] Among them, M T(Y,J) To convert the standardized quality control indicators with dimensions of Y×T rows and J columns into multidimensional time series data of each period within the year with dimensions of T rows and Y×J columns; Y is the number of years of historical data; T is the total number of analysis cycles divided within the year.

[0123] The output data is a multidimensional time series data set, in which each time point (i.e. each analysis period) has a corresponding set of indicator values.

[0124] Step 5: Cluster analysis completes the detection stage.

[0125] Cluster analysis is performed on the multi-dimensional time series data of each period in the year to complete the quality management stage of medical testing, including:

[0126] S51. Determine the optimal number of installments.

[0127] For the multidimensional time series data of each period within the year, a cluster quantity optimization algorithm is used to determine the optimal number of stages for medical detection; the cluster quantity optimization algorithm includes a gap statistics cluster quantity optimization algorithm and an average profile cluster quantity optimization algorithm.

[0128] S52, completing the detection staging: according to the optimal number of medical detection stages, a clustering method is used to complete the detection staging; the clustering method includes a K-means clustering method and a hierarchical clustering method.

[0129] The beneficial effects of this embodiment are: improving the accuracy and efficiency of medical test quality management, optimizing scheduling through machine learning algorithms, and realizing quantitative calculation of multi-year sequence data of multiple influencing factors; standardization processing and cluster analysis further improve the comparability and analysis depth of data, providing a scientific basis for medical test quality management.

[0130] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

[0131] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A medical testing quality management system based on machine learning, characterized in that: The system includes a demand analysis and cycle determination module, a quality control index calculation module, a standardization processing module, a multi-dimensional time series data conversion module and a cluster analysis detection staging module; Demand analysis and cycle determination module: Based on the specific needs of medical testing and the operational characteristics of the hospital or clinic, determine the analysis cycle of medical testing and screen out representative features of key clinical and laboratory data that affect the quality of testing; Output a clear analysis cycle and representative feature screening list to provide a basis for subsequent steps; Quality control index calculation module: Use machine learning algorithms for long-term optimization scheduling, and calculate the quality control indicators of various medical tests based on representative features, including the error rate of each period of the test, the impact of environmental conditions, etc.; output a series of quantitative quality control indicators for evaluating the quality of medical tests; Standardization processing module: standardizes the quality control indicators to eliminate the dimension differences and numerical range differences between different indicators, so that different indicators can be compared and comprehensively analyzed; outputs standardized quality control indicator data to provide a basis for subsequent multi-dimensional time series data conversion and cluster analysis; Multidimensional time series data conversion module: converts the standardized quality control indicators into multidimensional time series data for each period of the year to facilitate time series analysis and cluster analysis; outputs multidimensional time series data for each period of the year to provide data support for cluster analysis and detection staging; Cluster analysis and testing staging module: cluster analysis is performed on multidimensional time series data for each period of the year, and medical tests are divided into different quality management stages based on the similarities and differences of the data, so as to carry out targeted quality management for different stages; the quality management staging results of medical tests are output to provide scientific quality management basis for medical institutions.

2. A medical testing quality management method based on machine learning, used to implement the medical testing quality management system based on machine learning as claimed in claim 1, characterized in that: The method flow is as follows: Step 1: Determine the analysis cycle and key indicators: Based on the medical testing needs including disease type, patient population characteristics, and testing frequency, determine the analysis cycle of medical testing analysis, and based on the analysis cycle, screen representative features in key clinical and laboratory data that affect the quality of medical testing; Step 2: Quantitatively calculate quality control indicators: Based on the representative characteristics and combined with the long-term optimization scheduling model, quantitatively calculate the quality control indicators of various medical tests; the quality control indicators are multi-year series data of multiple influencing factors; the long-term optimization scheduling includes a single-period detection benefit loss objective function, sample size balance constraints, detection upper and lower limit constraints, and equipment capacity upper and lower limit constraints; Step 3: Standardization: Standardizing the quality control indicators to generate standardized quality control indicators; Step 4: Converting into multidimensional time series data: converting the standardized quality control indicators into multidimensional time series data for each period of the year; Step 5: Cluster analysis to complete the detection phase: cluster analysis is performed on the multi-dimensional time series data for each period within the year to complete the quality management phase of medical detection.

3. The medical testing quality management method based on machine learning according to claim 2 is characterized in that: The step 1 specifically includes: S11. Analyze the operational characteristics of the hospital or clinic, including diagnostic tasks, patient types, testing equipment performance, and changes in testing volume during peak and trough periods; S12. Determine the analysis cycle: Determine the analysis period according to the detection requirements, including daily, weekly or monthly; use a shorter analysis cycle during peak periods and a longer analysis cycle during low periods; S13. Screening of representative characteristics: During the analysis cycle, screen key data that affect the quality of medical testing, including the patient's medical history, symptoms, laboratory test results, treatment response, environmental factors, and equipment operating parameters.

4. The medical testing quality management method based on machine learning according to claim 2 is characterized in that: The step 2 specifically includes: S21. Use machine learning algorithms for long-term optimization scheduling: Based on the representative features, long-term optimization scheduling is performed to obtain the optimal operation process of each analysis cycle of medical testing; the objective function is to minimize the loss of single-period testing benefits; the constraints include sample size balance constraints, upper and lower limit constraints for testing, and upper and lower limit constraints for equipment capacity; S22. Calculate quality control indicators: calculate the error rate of each testing period, the number of samples to be tested in each testing period, and the impact of environmental conditions on testing; S23. Environmental condition impact index: Take the weighted average of the environmental data of each monitoring point, calculate the standardized environmental coefficient, and obtain the impact index of environmental conditions on the detection.

5. The medical testing quality management method based on machine learning according to claim 2 is characterized in that: The calculation formula of the error rate of the detection stage by time period is: E yt =max(C yt )-Q yt ; Where y is the year; t is the analysis period; C yt is the maximum allowable detection accuracy for a certain analysis period t in a certain year y; Q yt is the actual detection accuracy of a certain analysis period t in a certain year y.

6. The medical testing quality management method based on machine learning according to claim 4 is characterized in that: The calculation formula of the environmental condition impact index is: Among them, ω i is the weight of each monitoring point; P i It is the environmental data of each monitoring point, including temperature and humidity.

7. The medical testing quality management method based on machine learning according to claim 2 is characterized in that: The step three specifically includes: S31. Determine positive effect indicators and negative effect indicators: positive effect indicators include accuracy, sensitivity, and specificity; negative effect indicators include misdiagnosis rate, missed diagnosis rate, false positive rate, and false negative rate; S32, standardization processing: using a standardized formula to process the positive effect index and the negative effect index to generate a standardized quality control index. Specifically, the standardized formula is as follows: Among them, Z jyt is the standardized quality control index; X jyt is a positive effect indicator or a negative effect indicator; j is the type number of the quality control indicator; μ j is the mean value of the quality control index; σ j is the variance of the quality control index; y is the year; t is the analysis period.

8. The medical testing quality management method based on machine learning according to claim 2, characterized in that: The step 4 specifically includes: Convert to multidimensional time series data: Convert the standardized quality control indicators into multidimensional time series data for each period of the year. The specific formula is: M T(Y,J) =reshape(Z Y(T,J) ,T,Y×J); Among them, M T(Y,J) To convert the standardized quality control indicators with dimensions of Y×T rows and J columns into multi-dimensional time series data with dimensions of T rows and Y×J columns for each period of the year; Y is the number of years of historical data; T is the total number of analysis cycles divided within the year; The output data is a multidimensional time series data set, in which each time point, that is, each analysis period, has a corresponding set of indicator values.

9. The medical testing quality management method based on machine learning according to claim 2, characterized in that: The step five specifically includes: S51, determining the optimal number of stages: for the multi-dimensional time series data of each period in the year, using a cluster number optimization algorithm to determine the optimal number of stages for medical detection; the cluster number optimization algorithm includes a gap statistics cluster number optimization algorithm and an average profile cluster number optimization algorithm; S52, completing the detection staging: according to the optimal number of medical detection stages, a clustering method is used to complete the detection staging; the clustering method includes a K-means clustering method and a hierarchical clustering method.

Images