Dynamic algorithm for blood culture positive judgment

Abstract

The invention discloses a dynamic algorithm for blood culture positive judgment, which comprises the following steps: S1, data acquisition: a blood volume monitoring module is arranged on a blood culture bottle, and the module adopts at least one of a pressure sensor, a weight sensor or a visual system to monitor the liquid level height and is used for accurately measuring the blood collection volume; after a collected blood sample is injected into the blood culture bottle, the blood volume monitoring module obtains blood collection volume data in real time. According to the invention, the blood volume monitoring module is added to realize accurate measurement of the blood collection volume, a scientific relation model is established by using big data analysis, and a dynamic positive judgment model based on sample information is constructed. According to the algorithm, a dynamic positive judgment threshold value can be calculated according to real-time data, the false positive rate caused by too much blood sampling amount or too much leukocytes of a patient is effectively reduced, the accuracy and timeliness of blood culture detection are remarkably improved, a more reliable basis is provided for clinical diagnosis, doctors are assisted to formulate a reasonable treatment scheme in time, and the treatment effect is improved. The treatment effect and the survival rate of patients are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of blood culture, in particular to a dynamic algorithm for judging blood culture positive. BACKGROUND

[0002] Sepsis has become a major health care problem due to its high incidence and high mortality in hospitals, and bloodstream infection (BSI) is a major cause of sepsis. In hospital clinics and medical research trials, blood culture is an important microbiological examination method for detecting the presence of bacteria in blood samples. It plays a crucial role in quickly detecting whether bacteria grow in the blood of patients with clinically severe sepsis and bacteremia to confirm the diagnosis, and is one of the most important and common bacterial examination items.

[0003] Existing automatic blood culture systems can detect the presence of microorganisms in blood within 1-48 hours and exclude the presence of microorganisms within 5 days. However, in medical examination practice, the positive judgment algorithm of existing blood culture instruments mostly adopts a fixed threshold method, which has obvious defects.

[0004] On the one hand, the amount of blood collected is an important factor affecting the test results. When the amount of blood collected is too much, the carbon dioxide contained in the blood will cause changes in the culture bottle sensor, which may interfere with the sensor and cause false positive results.

[0005] On the other hand, the patient's own white blood cell count can also affect the test. When the patient has too many white blood cells, the metabolic activity of white blood cells may produce signals similar to microbial growth, thereby interfering with the culture bottle detection and also causing false positives.

[0006] Currently, there is no good solution to this problem. The existing processing method is only limited to standardizing the blood collection process, or analyzing the causes when false positive results occur, and does not realize the dynamic correlation of patient's basic blood volume information and white blood cell count with culture bottle detection information. This makes the existing technology unable to comprehensively and accurately judge the positive results of blood culture, resulting in a high false positive rate of blood culture, which greatly disturbs clinical diagnosis and affects the timeliness and accuracy of disease diagnosis, which is not conducive to effective treatment of patients. SUMMARY

[0007] The purpose of the present application is to provide a dynamic algorithm for blood culture positive judgment, which realizes accurate measurement of blood sampling volume through the addition of a blood volume monitoring module, establishes a scientific relationship model through big data analysis, automatically obtains LIS system data to realize the dynamic correlation of patient's basic blood volume information, white blood cell value and culture bottle detection information, and further constructs a dynamic positive judgment model based on sample information. The algorithm can calculate the dynamic positive judgment threshold according to real-time data, effectively reduce the false positive rate caused by excessive blood sampling volume or excessive white blood cells of patients, significantly improve the accuracy and timeliness of blood culture detection, provide more reliable basis for clinical diagnosis, help doctors to develop reasonable treatment plans in time, and improve the treatment effect and survival rate of patients.

[0008] In order to achieve the above purpose, the main technical solutions adopted by the present application include:

[0009] A dynamic algorithm for blood culture positive judgment, comprising the following steps:

[0010] S1, data acquisition:

[0011] A blood volume monitoring module is arranged on the blood culture bottle, which uses at least one of a pressure sensor, a weight sensor or a visual system to monitor the liquid level height for accurate measurement of blood sampling volume; after collecting the blood sample and injecting it into the blood culture bottle, the blood volume monitoring module obtains the blood sampling volume data in real time;

[0012] When the blood culture bottle is connected to the machine, the white blood cell count detection result of the patient is automatically obtained through the interface with the hospital laboratory information system (LIS), which includes the number of white blood cells and other related information;

[0013] S2, model establishment:

[0014] Using big data analysis technology, the historical blood culture data is processed to establish a relationship model between blood sampling volume and culture bottle receptor signal; a large amount of detection data of blood culture bottles under different blood sampling volumes are collected, including signal strength, change trend, etc., and through statistical methods or machine learning algorithms, the influence law of blood sampling volume on receptor signal is analyzed to determine the normal signal range and abnormal signal characteristics corresponding to different blood sampling volumes;

[0015] The white blood cell detection result of the patient is correlated and analyzed with the detection information of the blood culture bottle to establish a relationship model between the number of white blood cells and the interference degree of the receptor signal; the change of the detection signal of the blood culture bottle under different white blood cell numbers is analyzed to determine the influence coefficient of the number of white blood cells on the detection result;

[0016] S3, dynamic judgment:

[0017] The real-time collected blood sampling data and obtained white blood cell detection results are input into the above established relationship model, and a dynamic positive judgment model is constructed in combination with the real-time detected receptor signals of the culture bottle;

[0018] According to the dynamic positive judgment model, the dynamic positive judgment threshold of each culture bottle is calculated, which is adjusted according to the blood sampling amount and the number of white blood cells of the current sample;

[0019] The real-time detected receptor signals of the culture bottle are compared with the calculated dynamic threshold value, and if the signals exceed the dynamic threshold value, the blood culture is judged to be positive; otherwise, it is judged to be negative.

[0020] The above dynamic algorithm for blood culture positive judgment, wherein the blood volume monitoring module in S1 adopts a pressure sensor, a weight sensor or a visual system to monitor the liquid level height, which can be selected and combined according to the specific structure and requirements of the blood culture bottle.

[0021] The above dynamic algorithm for blood culture positive judgment, wherein the blood volume monitoring module in S1 has the characteristics of real-time and precision when obtaining blood sampling data, and can transmit the data to the subsequent processing system in time.

[0022] The above dynamic algorithm for blood culture positive judgment, wherein the interface has stable and efficient data transmission capability when automatically obtaining the white blood cell count detection results of the patient's blood routine through the interface with the hospital laboratory information system (LIS) in S1, ensuring the accuracy and timeliness of the obtained results.

[0023] The above dynamic algorithm for blood culture positive judgment, wherein the data cleaning and preprocessing method used when processing historical blood culture data by big data analysis technology in S2 can effectively remove noise data and abnormal data, ensuring the accuracy of the established relationship model.

[0024] The above dynamic algorithm for blood culture positive judgment, wherein the algorithm selected when analyzing the influence law of blood sampling amount on receptor signals by statistical methods or machine learning algorithms in S2 has high goodness of fit and generalization ability, and can accurately determine the normal signal range and abnormal signal characteristics corresponding to different blood sampling amounts.

[0025] The above dynamic algorithm for blood culture positive judgment, wherein the data analysis method used when analyzing the change of blood culture bottle detection signals under different white blood cell counts in S2 can accurately quantify the influence coefficient of white blood cell count on detection results.

[0026] The dynamic algorithm for judging blood culture positivity, wherein, in S3, when constructing the dynamic positive judgment model, the blood sampling amount data, white blood cell detection results and real-time sensor signals detected by the culture bottle are comprehensively considered, a reasonable model construction algorithm is adopted, and the accuracy and reliability of the model are ensured.

[0027] The dynamic algorithm for judging blood culture positivity, wherein, in S3, when calculating the dynamic positive judgment threshold value of each culture bottle, the algorithm for dynamically adjusting according to the blood sampling amount and the number of white blood cells of the current sample is scientific and reasonable, and can effectively reduce the occurrence of false positives and false negatives.

[0028] The dynamic algorithm for judging blood culture positivity, wherein, in S3, when comparing the real-time sensor signals detected by the culture bottle with the calculated dynamic threshold value, the comparison method adopted has the characteristics of being fast and accurate, and can timely give the judgment result of blood culture positivity.

[0029] Compared with the prior art, the advantages and positive effects of the present application are that:

[0030] 1、The present application realizes precise measurement of blood sampling amount by adding a blood amount monitoring module, establishes a scientific relationship model by using big data analysis, automatically obtains LIS system data to realize dynamic correlation of patient's basic blood amount information, white blood cell value and culture bottle detection information, and further constructs a dynamic positive judgment model based on sample information. The algorithm can calculate a dynamic positive judgment threshold value according to real-time data, effectively reduces the false positive rate caused by excessive blood sampling amount or excessive white blood cells of the patient, significantly improves the accuracy and timeliness of blood culture detection, provides a more reliable basis for clinical diagnosis, helps doctors to timely develop a reasonable treatment plan, and improves the treatment effect and survival rate of the patient.

[0031] 2、Precise measurement of blood sampling amount: by adding a blood amount monitoring module, using a pressure sensor, a weight sensor or a visual system to monitor the liquid level, precise measurement of blood sampling amount is realized. This provides reliable basic data for subsequent data analysis based on blood sampling amount, and avoids detection errors caused by inaccurate blood sampling amount measurement.

[0032] 3、Establishing a scientific relationship model: by means of big data analysis technology, historical blood culture data of the hospital are collected, including blood sampling amount, white blood cell number, culture bottle sensor signal, final culture result, etc., and machine learning algorithms such as random forest, support vector machine, etc. are used to establish the relationship model of blood sampling amount and sensor signal and the relationship model of white blood cell number and signal interference degree. These models are continuously trained and optimized, have high accuracy and generalization ability, and can accurately reflect the influence law of different factors on blood culture detection results.

[0033] 4. Realize data dynamic correlation: When the blood culture bottle is connected to the computer, a special communication module is used to connect with the LIS system of the hospital, and the standard communication protocol (such as HL7 protocol) is adopted to automatically obtain the white blood cell detection results of the corresponding patient. The patient's basic blood volume information, white blood cell value and culture bottle detection information are dynamically correlated to construct a dynamic positive judgment model based on sample information. This dynamic correlation method fully considers the influence of individual differences of patients on the detection results, so that the positive judgment is more in line with the actual situation.

[0034] 5. Reduce false positive rate: Compared with the prior art, the present application can calculate the dynamic positive judgment threshold value of each culture bottle according to the real-time acquisition of blood volume data, the acquisition of white blood cell detection results and the real-time detection of the sensor signal of the culture bottle. The threshold value is adjusted according to the blood volume and white blood cell count of the current sample, which effectively reduces the false positive situation caused by excessive blood volume or excessive white blood cells of the patient, and significantly reduces the false positive rate of blood culture.

[0035] 6. Improve detection accuracy: Through the above series of innovative methods and technical means, the present application can more accurately judge the positive result of blood culture, improve the accuracy of blood culture detection, and provide a more reliable basis for clinical diagnosis, which helps doctors to diagnose the patient's condition in time and accurately, and to develop a reasonable treatment plan, so as to improve the treatment effect and survival rate of patients.

[0036] 7. Improve timeliness: The dynamic algorithm of the present application can acquire relevant data in real time and perform dynamic analysis and judgment during the blood culture detection process, without the need for additional complex operations and long waiting time, and can give accurate positive judgment results in a short time, improving the timeliness of the blood culture detection project and meeting the needs of clinical rapid diagnosis. BRIEF DESCRIPTION OF DRAWINGS

[0037] The drawings described herein are used to provide further understanding of the present application, and form a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application, and do not constitute an improper limitation on the present application. In the drawings:

[0038] Figure 1 The flow chart of the dynamic algorithm for blood culture positive judgment of the present application. DETAILED DESCRIPTION

[0039] The embodiments of the present application will be described in detail below with reference to the drawings and examples, so that the realization process of how to apply technical means to solve technical problems and achieve technical effects of the present application can be fully understood and implemented.

[0040] Example 1

[0041] Please refer to Figure 1As shown, the embodiment of the application provides a dynamic algorithm for judging blood culture positivity, which includes three steps of data collection, model establishment and dynamic judgment.

[0042] The data collection step includes:

[0043] A blood volume monitoring module is arranged on the blood culture bottle, which uses at least one of a pressure sensor, a weight sensor or a visual system to monitor the liquid level, for accurately measuring the blood sampling volume; after the blood sample is collected and injected into the blood culture bottle, the blood volume monitoring module acquires real-time blood sampling volume data.

[0044] When the blood culture bottle is connected to the machine, the white blood cell count detection result of the patient is automatically obtained through the interface with the hospital laboratory information system (LIS), and the result includes the information related to the number of white blood cells.

[0045] The model establishment step includes establishing a relationship model between the blood sampling volume and the signal of the culture bottle receptor.

[0046] The method for establishing the relationship model between the blood sampling volume and the signal of the culture bottle receptor is:

[0047] A large amount of detection data of the blood culture bottle under different blood sampling volumes is collected, including the signal strength and change trend of the receptor; and the influence law of the blood sampling volume on the signal of the receptor is analyzed by statistical methods or machine learning algorithms to determine the normal signal range and abnormal signal characteristics corresponding to different blood sampling volumes.

[0048] The model establishment step further includes establishing a relationship model between the number of white blood cells and the interference degree of the signal of the receptor.

[0049] The method for establishing the relationship model between the number of white blood cells and the interference degree of the signal of the receptor is:

[0050] The change of the detection signal of the blood culture bottle under different white blood cell counts is analyzed, and the influence coefficient of the white blood cell count on the detection result is determined.

[0051] The dynamic judgment step includes:

[0052] The real-time collected blood sampling volume data and the obtained white blood cell detection result are input into the above-established relationship model, and the dynamic positive judgment model is constructed in combination with the real-time detected signal of the receptor of the culture bottle.

[0053] The dynamic judgment step further includes:

[0054] According to the dynamic positive judgment model, the dynamic positive judgment threshold of each culture bottle is calculated, and the threshold is adjusted according to the blood sampling volume and the number of white blood cells of the current sample.

[0055] The threshold adjustment mode is:

[0056] When the blood collection amount is large, the signal intensity threshold for judging positive is appropriately increased to reduce false positives caused by excessive blood collection; when the white blood cell count is high, the detection signal is corrected to deduct the interference signal that may be generated by white blood cell metabolic activity, and then compared with the adjusted threshold value; if the signal exceeds the dynamic threshold value, the blood culture is judged to be positive; otherwise, it is judged to be negative.

[0057] By adopting the above technical solutions, the embodiment verifies the effectiveness of a dynamic algorithm for blood culture positive judgment through detailed data collection, model establishment and dynamic judgment process. The algorithm realizes accurate blood collection amount measurement by adding a blood amount monitoring module, constructs a dynamic positive judgment model in combination with big data analysis and patient white blood cell detection results, can calculate a dynamic positive judgment threshold value according to real-time data, effectively reduces the false positive rate of blood culture, improves the accuracy of blood culture detection, and provides a more reliable basis for clinical diagnosis.

[0058] Embodiment 2

[0059] Please refer to Figure 1 The dynamic algorithm for blood culture positive judgment provided by the embodiment of the application includes:

[0060] (I) Data collection:

[0061] 1. Blood amount monitoring module setting and data acquisition:

[0062] A blood amount monitoring module is arranged on the blood culture bottle, and a weight sensor is used in the embodiment. When the blood sample is collected and injected into the blood culture bottle, the weight in the bottle changes, and the blood collection amount is calculated by measuring the weight change of the blood culture bottle, with the blood density being 1.055 g / mL by default. For example, data on the influence of different blood collection amounts (0, 1, 3, 5, 8, 10, 15, 18, 20, 25, 28, 30, 35 ml) on detection data change are collected, and the change of detection values under different blood collection amounts is recorded, as shown in the following table:

[0063]

[0064] 2. White blood cell detection result acquisition:

[0065] When the blood culture bottle is uploaded, the patient's routine white blood cell count test result is automatically obtained through the interface with the hospital laboratory information system (LIS). A special communication module is set on the blood culture instrument to connect with the hospital LIS system, and a standard communication protocol (such as HL7 protocol) is used to realize automatic data acquisition. When the blood culture bottle is uploaded, the communication module sends a request to the LIS system to obtain the white blood cell test result (including the number of white blood cells and other related information) of the corresponding patient, and transmits the result to the data processing unit;

[0066] (ii) Model establishment

[0067] Blood volume and culture bottle sensor signal relationship model establishment

[0068] Using big data analysis technology, the collected historical blood culture data (including blood culture bottle detection data under different blood volumes, such as sensor signal strength, change trend, etc.) are processed. Linear, binomial, trinomial, tetranomial, pentanomial, etc. Fitting method is used to establish the curve of the influence of different blood volumes on the detection data, so as to establish the relationship of calculating the influence of detection data according to the blood volume. This embodiment adopts binomial fitting method with intercept 0, and establishes the relationship between blood volume and detection value change as: y = 0.0489x + 87.114x, R = 0.9986. Through the model, the normal signal range and abnormal signal characteristics corresponding to different blood volumes can be determined. 2 2

[0069] White blood cell count and sensor signal interference degree relationship model establishment

[0070] The patient's white blood cell test result is associated with the blood culture bottle detection information for analysis, and a large amount of data on the change of blood culture bottle detection signal under different white blood cell counts is collected. Through statistical methods or machine learning algorithms (such as random forest, support vector machine, etc.), the influence law of different white blood cell counts on the detection result is analyzed, the interference coefficient of white blood cell count on the detection signal is determined, and the relationship model of white blood cell count and sensor signal interference degree is established.

[0071] (iii) Dynamic judgment

[0072] 1. Dynamic positive judgment model construction:

[0073] ​​The data processing unit receives the blood volume data collected by the blood volume monitoring module, the white blood cell detection result obtained by the LIS system, and the sensor signal detected by the culture bottle in real time. First, according to the blood volume data and the established relationship model of blood volume and sensor signal, the normal signal reference range and the adjustment coefficient under the blood volume are determined; then, according to the white blood cell detection result and the relationship model of white blood cell quantity and signal interference degree, the interference value of white blood cells to the detection signal is calculated; then, the real-time detected sensor signal is corrected, and the interference value is subtracted, and the judgment threshold is adjusted according to the adjustment coefficient; finally, a dynamic positive judgment model is constructed.

[0074] 2. Dynamic positive judgment threshold calculation and result judgment:

[0075] According to the dynamic positive judgment model, the dynamic positive judgment threshold of each culture bottle is calculated. Assuming that the blood volume of a patient is x milliliters, the initial detected sensor signal intensity is S0, and the real-time detected sensor signal intensity is S. Then the dynamic threshold T = S0 + 0.0489x 2 +87.114x. Compare the real-time detected sensor signal of the culture bottle with the calculated dynamic threshold value, if the signal exceeds the dynamic threshold value, it is judged as blood culture positive; otherwise, it is judged as negative;

[0076] (4) Verification comparison

[0077] The following samples are selected for dynamic threshold verification and comparison with the existing navy blue threshold algorithm (static threshold, the present embodiment assumes 3000). The real-time signal value, dynamic threshold, dynamic threshold result, static threshold and static threshold result under different blood volumes are recorded as shown in the following table:

[0078]

[0079]

[0080]

[0081] From the test data in the above table, compared with the existing static threshold, the dynamic threshold has no false positive data, and the static threshold has 16 false positive samples in the high blood volume samples. This shows that the dynamic threshold algorithm can effectively avoid false positives caused by excessive blood volume, and improves the accuracy of blood culture.

[0082] The working principle of the present application is:

[0083] 1. Data acquisition:

[0084] Blood volume monitoring module setting and data acquisition: Set up a blood volume monitoring module on the blood culture bottle. In Example 1, at least one of the following methods is used to monitor the liquid level: pressure sensor, weight sensor, or visual system. In Example 2, a weight sensor is used. The blood culture bottle is placed on a weighing platform, and the blood volume is calculated by measuring the weight change of the blood culture bottle (assuming a blood density of 1.055 g / mL). The changes in detection values at different blood volumes are recorded. This method can accurately measure the blood volume, providing accurate data for subsequent model establishment.

[0085] White blood cell detection result acquisition: When the blood culture bottle is placed in the machine, the patient's routine white blood cell count detection result is automatically obtained through the interface with the hospital laboratory information system (LIS). A special communication module is set up on the blood culture instrument, and a standard communication protocol (such as HL7 protocol) is used to connect with the hospital LIS system to realize automatic data acquisition. This ensures that the patient's white blood cell count and other related information can be obtained in time for subsequent analysis of the interference of white blood cell count on the detection signal.

[0086] 2. Model establishment:

[0087] Blood volume and culture bottle sensor signal relationship model establishment: Using big data analysis technology, the collected historical blood culture data (including detection data of blood culture bottles under different blood volumes, such as sensor signal strength, change trend, etc.) are processed. In Example 2, a binomial fitting method with an intercept of 0 is used to establish the relationship between blood volume and detection value change: y = 0.0489x + 87.114x, R = 0.9986. Through this model, the normal signal range and abnormal signal characteristics corresponding to different blood volumes can be determined, and the quantitative relationship between blood volume and detection signal is clear. 2 2

[0088] White blood cell count and sensor signal interference degree relationship model establishment: Correlation analysis is performed between the patient's white blood cell detection results and the blood culture bottle detection information, and a large amount of data on the change of blood culture bottle detection signal under different white blood cell counts is collected. Through statistical methods or machine learning algorithms (such as random forest, support vector machine, etc.), the influence of different white blood cell counts on the detection results is analyzed, the interference coefficient of white blood cell count on the detection signal is determined, and the relationship model between white blood cell count and sensor signal interference degree is established. This helps to understand the interference degree of white blood cell metabolic activity on the detection signal, so as to make corrections in subsequent judgments.

[0089] 3. Dynamic judgment:

[0090] ​​Dynamic positive judgment model construction: The data processing unit receives the blood volume data collected by the blood volume monitoring module, the white blood cell detection results obtained by the LIS system, and the sensor signals detected by the culture bottle in real time. First, according to the blood volume data and the established relationship model between blood volume and sensor signals, the baseline range of normal signals and the adjustment coefficient under the blood volume are determined; then, according to the white blood cell detection results and the relationship model between the number of white blood cells and the degree of signal interference, the interference value of white blood cells on the detection signal is calculated; then, the real-time detected sensor signal is corrected, and the interference value is subtracted, and the judgment threshold is adjusted according to the adjustment coefficient; finally, the dynamic positive judgment model is constructed. This process considers the influence of blood volume and white blood cell number on the detection signal, so that the judgment model is more in line with the actual situation.

[0091] Dynamic positive judgment threshold calculation and result judgment: According to the dynamic positive judgment model, the dynamic positive judgment threshold of each culture bottle is calculated. For example, assuming that the blood volume of a patient is x milliliters, the initial detected sensor signal strength is S0, and the real-time detected sensor signal strength is S. Then the dynamic threshold T = S0 + 0.0489x 2 +87.114x. Compare the real-time detected sensor signal of the culture bottle with the calculated dynamic threshold value, if the signal exceeds the dynamic threshold value, it is judged as blood culture positive; otherwise, it is judged as negative.

[0092] Verification comparison

[0093] The dynamic threshold is verified by selecting samples and compared with the existing static threshold algorithm. From the test data, compared with the existing static threshold, the dynamic threshold has no false positive data, and the static threshold has false positive in samples with high blood volume. This shows that the dynamic threshold algorithm considers factors such as blood volume and white blood cell number, effectively avoids false positives due to excessive blood volume, improves the accuracy of blood culture, and provides a more reliable basis for clinical diagnosis.

[0094] The above description shows and describes several preferred embodiments of the present application, but as mentioned above, it should be understood that the present application is not limited to the forms disclosed herein, and should not be considered as excluding other embodiments, but can be used in various other combinations, modifications and environments, and can be modified within the scope of the inventive concept described herein by the above teaching or related art or knowledge. Any modification and change made by those skilled in the art without departing from the spirit and scope of the present application shall be within the protection scope of the appended claims of the present application.

Claims

1. A dynamic algorithm for blood culture positivity determination, characterized in that, Comprising the following steps: S1, data collection: A blood volume monitoring module is arranged on the blood culture bottle, which uses at least one of pressure sensor, weight sensor or visual system to monitor liquid level for precise measurement of blood collection volume; after the blood sample is collected and injected into the blood culture bottle, the blood volume monitoring module acquires blood collection volume data in real time; When the blood culture bottle is put into the machine, the patient's routine white blood cell count test results are automatically obtained through the interface with the hospital laboratory information system (LIS), which includes relevant information such as the number of white blood cells; S2, model establishment: Using big data analysis technology, historical blood culture data is processed to establish a relationship model between blood collection volume and culture bottle sensor signal; a large amount of detection data of blood culture bottles under different blood collection volumes are collected, including sensor signal strength, change trend, etc., the influence of blood collection volume on sensor signal is analyzed by statistical method or machine learning algorithm, and the normal signal range and abnormal signal characteristics corresponding to different blood collection volumes are determined; The patient's white blood cell test results are associated with the detection information of the blood culture bottle to establish a relationship model between white blood cell count and sensor signal interference degree; the change of blood culture bottle detection signal under different white blood cell counts is analyzed to determine the influence coefficient of white blood cell count on detection results; S3, dynamic judgment: The real-time collected blood collection volume data and obtained white blood cell test results are input into the above established relationship model, and the real-time detected sensor signal of the culture bottle is combined to construct a dynamic positive judgment model; According to the dynamic positive judgment model, the dynamic positive judgment threshold of each culture bottle is calculated, which is adjusted according to the current sample blood collection volume and white blood cell count; The real-time detected sensor signal of the culture bottle is compared with the calculated dynamic threshold value, if the signal exceeds the dynamic threshold value, it is judged as blood culture positive; otherwise, it is judged as negative.

2. The dynamic algorithm for blood culture positivity determination according to claim 1, wherein: The blood volume monitoring module in S1 uses pressure sensor, weight sensor or visual system to monitor liquid level, which can be selected and combined according to the specific structure and needs of the blood culture bottle.

3. The dynamic algorithm for blood culture positivity determination according to claim 2, wherein: The blood volume monitoring module in S1 has the characteristics of real-time and precision when acquiring blood collection volume data, and can transmit the data to the subsequent processing system in time.

4. The dynamic algorithm for blood culture positivity determination according to claim 3, wherein: When the interface automatically obtains the patient's routine white blood cell count test results through the interface with the hospital laboratory information system (LIS) in S1, the interface has stable and efficient data transmission capability to ensure the accuracy and timeliness of the obtained results.

5. The dynamic algorithm for blood culture positivity determination according to claim 4, wherein: When the historical blood culture data is processed by big data analysis technology in S2, the data cleaning and preprocessing method can effectively remove noise data and abnormal data to ensure the accuracy of the established relationship model.

6. The dynamic algorithm for blood culture positivity determination according to claim 5, wherein: When the influence of blood collection volume on sensor signal is analyzed by statistical method or machine learning algorithm in S2, the selected algorithm has high goodness of fit and generalization ability, which can accurately determine the normal signal range and abnormal signal characteristics corresponding to different blood collection volumes.

7. The dynamic algorithm for blood culture positivity determination according to claim 6, wherein: When the change of blood culture bottle detection signal under different white blood cell counts is analyzed in S2, the data analysis method can accurately quantify the influence coefficient of white blood cell count on detection results.

8. The dynamic algorithm for blood culture positivity determination according to claim 7, wherein: In the S3, when the dynamic positive judgment model is constructed, the blood sampling data, the white blood cell detection result and the real-time sensor signal detected by the culture bottle are comprehensively considered, a reasonable model construction algorithm is adopted, and the accuracy and reliability of the model are ensured.

9. The dynamic algorithm for blood culture positivity determination according to claim 8, wherein: In the S3, when the dynamic positive judgment threshold of each culture bottle is calculated, the algorithm for dynamically adjusting according to the blood sampling amount and the white blood cell quantity of the current sample is scientific and reasonable, and can effectively reduce the occurrence of false positives and false negatives.

10. The dynamic algorithm for blood culture positivity determination according to claim 9, wherein: In the S3, when the real-time sensor signal detected by the culture bottle is compared with the calculated dynamic threshold, the comparison method adopted has the characteristics of rapidness and accuracy, and can give the blood culture positive judgment result in time.

Images