Drug sensitivity evaluation method and system for riemerella anatipestifer

Abstract

The invention relates to the technical field of drug sensitivity evaluation, in particular to a riemerella anatipestifer drug sensitivity evaluation method which comprises the following steps: S1, bacterial liquid preparation: sampling a bacterial strain, inoculating the bacterial strain into a nutrient solution, and culturing to 108CFU / mL bacterial concentration; s2, automatic drug dilution and addition: a system automatically dilutes to-be-tested drugs to different concentrations according to experimental design, and the to-be-tested drugs are added into the bacterial liquid; s3, real-time optical detection: collecting optical density data once every 30 minutes by an optical detection module, and recording growth changes of the bacterial liquid under different drug concentrations; s4, data analysis and output: the system calculates the optical density data, draws a bacterium growth curve and calculates the minimum inhibitory concentration, according to the scheme, a micro liquid dilution method and an automatic drug sensitivity evaluation system are adopted, drug sensitivity data can be obtained within 6-8 hours, the detection time is greatly shortened, and the detection efficiency is improved. Automatic data processing: through optical density detection, image processing and data analysis technologies, the inhibition effect of drugs on bacteria is monitored in real time, and data is more objective and accurate.

Description

Technical Field

[0001] The invention relates to the technical field of drug sensitivity assessment, and in particular to a drug sensitivity assessment method and system for Riemerella anatipestifer. Background Art

[0002] Riemerella anatipestifer disease is a contact infectious disease caused by Riemerella anatipestifer, which is a serious threat to the world's duck farming industry. Meat ducks generally show systemic septic and exudative pathological changes after being infected with RA. Autopsy shows fibrinous exudates in the pericardium, liver, and air sacs, a large amount of caseous exudate in the oviduct, and congestion and edema of the meninges. The autopsy features vary at different stages of infection development. Clinically, the main manifestations are lying on the back with the legs in a paddling state and slow movements. With the rapid spread of bacteria, meat ducks become emaciated and the mortality rate increases, causing huge economic losses to farmers.

[0003] Riemerella anatipestifer is a Gram-negative bacterium that is resistant to many antibiotics, so the drug susceptibility assessment of Riemerella anatipestifer is of great significance;

[0004] Conventional drug sensitivity test methods (such as paper diffusion method) are cumbersome and time-consuming (usually more than 24 hours). In emergency situations, drug sensitivity results cannot be obtained quickly. In addition, different experimental conditions (such as temperature, incubation time, culture medium, etc.) have a great impact on the results of drug sensitivity tests, resulting in poor reproducibility of the results.

[0005] Therefore, a method and system for drug susceptibility assessment of Riemerella anatipestifer were proposed. Summary of the invention

[0006] The purpose of the present invention is to provide a method and system for evaluating the drug sensitivity of Riemerella anatipestifer, which can scientifically determine the most suitable drug and dosage for treatment through high-precision drug sensitivity evaluation and data processing, improve the treatment effect and reduce the cost of medication, so as to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solution: a method for evaluating the drug sensitivity of Riemerella anatipestifer, comprising the following steps:

[0008] S1. Preparation of bacterial solution: Take the sample strain and inoculate it into the nutrient solution and culture it to a bacterial concentration of 10^8 CFU / mL;

[0009] S2. Automatic drug dilution and addition: The system automatically dilutes the drug to be tested to different concentrations according to the experimental design and adds it to the bacterial solution;

[0010] S3, real-time optical detection: the optical detection module collects optical density data every 30 minutes to record the growth changes of bacterial solution under different drug concentrations;

[0011] S4, data analysis and output: The system calculates the optical density data, draws the bacterial growth curve and calculates the minimum inhibitory concentration, and outputs the drug sensitivity report;

[0012] The drug susceptibility assessment system for Riemerella anatipestifer includes:

[0013] Main control unit: responsible for overall system control, including data collection, analysis, storage, and result output;

[0014] Drug dilution module: equipped with an automatic drug dilution device, which automatically adds drug solutions of different concentrations into the bacterial solution to form a series of drug gradients;

[0015] Optical detection module: using optical density detection instrument to detect the inhibitory effect of different concentrations of drugs on bacterial solution in real time, obtain bacterial growth curve and automatically calculate the inhibitory concentration;

[0016] Data analysis module: Utilizes statistical models and machine learning algorithms to automatically analyze drug sensitivity data, grade and evaluate drug sensitivity, and generate charts and reports;

[0017] Environmental regulation module: The system is equipped with a temperature and humidity regulator to ensure that the entire drug sensitivity assessment process is carried out under stable environmental conditions, thereby improving the reliability of experimental data.

[0018] Preferably, the data analysis module introduces deep learning and big data analysis technology to optimize the data processing flow and analysis method, wherein the introduction of the deep learning model includes data imaging analysis based on convolutional neural networks. In the process of drug sensitivity assessment, the bacterial liquid growth curve and antibacterial image are converted into image data and analyzed through convolutional neural networks to capture the microscopic characteristics of the bacterial growth state;

[0019] The use of long short-term memory networks for time series analysis was introduced to capture the long-term dependencies of bacterial growth, which is predictive in the analysis of optical density data.

[0020] Preferably, the data analysis module further includes an abnormal data detection and elimination unit, which mainly detects abnormal values ​​through an isolation forest algorithm, and further eliminates abnormal data detection results based on a local abnormal factor algorithm;

[0021] The steps of the isolation forest algorithm include:

[0022] a. Random data partitioning: Each data point is randomly partitioned multiple times, and outliers are identified by calculating the depth of the data point in the partition tree;

[0023] b. Anomaly score calculation: The anomaly score S(x,n) of each data point is:

[0024]

[0025] Where: E(h(x)) is the average path length of data point x in the partition tree, c(n) is the average path length of normal points;

[0026] c. Threshold judgment: When the anomaly score is higher than the threshold, it is marked as abnormal data;

[0027] Among them, the anomaly detection based on the local anomaly factor algorithm identifies local abnormal data by comparing the density difference between the data point and its neighboring points. The anomaly score calculation formula is:

[0028]

[0029] Where: LOF(A) is the local outlier factor of data point A, LDR(B) is the local density of data point B, and |N(A)| is the number of neighborhood points of data point A.

[0030] Preferably, the data analysis module includes a dynamic adjustment and multidimensional data clustering analysis unit, which includes dynamic data weighting and multidimensional data clustering, and clusters multiple groups of data according to different conditions through dynamic weighting and multidimensional clustering analysis;

[0031] Dynamic Weighting Model:

[0032]

[0033] Where: W i is the weighted value of the ith data point, e i (t) is the error value at time t, α is the learning rate;

[0034] Multidimensional data clustering: Use the K-means algorithm to cluster data of different time periods, drug concentrations, and strains, and readjust the drug concentration scheme based on the clustering results of each group of data;

[0035] Clustering formula:

[0036]

[0037] Where d is the distance between cluster centers, which is used to aggregate similar data.

[0038] Preferably, the data analysis module includes a historical data accumulation and model self-update unit, which specifically includes incremental learning model update and data storage and feature extraction optimization, and continuously updates the model based on the incremental learning method, so that the system automatically adapts to different experimental conditions and strain characteristics;

[0039] Incremental update formula:

[0040]

[0041] Where: θt+1 and θ t are the parameters of the new and old models, J(θ) is the loss function;

[0042] The data storage and feature extraction are optimized, historical drug sensitivity data is stored, features are regularly analyzed and extracted, and principal component analysis is used to reduce dimensionality to reduce data dimensions and improve model prediction efficiency;

[0043] Principal component analysis dimensionality reduction formula:

[0044] X′=XW

[0045] Where: X is the original data matrix, W is the projection matrix, and X′ is the data matrix after dimensionality reduction.

[0046] Preferably, the drug dilution module includes a real-time feedback and early warning mechanism unit, which is mainly used for threshold monitoring: setting the minimum effective concentration threshold of the drug concentration, and giving real-time reminders when the drug has no obvious inhibitory effect on the bacterial solution, so as to avoid the use of ineffective drugs; early warning of inconsistent results: when there is a deviation between the experimental results and the historical data, the system can issue an early warning, prompting retesting or checking the experimental conditions to ensure data accuracy.

[0047] Preferably, the threshold monitoring adopts dynamic threshold adjustment, real-time analysis of the monitored optical density, temperature, and humidity parameters, and dynamic adjustment of the minimum inhibitory concentration threshold of the key drug;

[0048] The dynamic adjustment model is:

[0049]

[0050] Where: Threshold dynamic is the dynamically adjusted threshold, Threshold baseline is the basic threshold, and ΔP is the deviation percentage caused by environmental parameters such as temperature and humidity.

[0051] Preferably, the inconsistent result warning includes a graded warning, which is divided into three levels and different feedback measures are taken according to the degree of abnormality;

[0052] Level 1 warning (slight deviation): When the data deviates from the normal threshold by 0% to 5%, the system will issue a prompt to recommend a review;

[0053] Level 2 warning (moderate deviation): When the data deviation reaches 5% to 10%, the system automatically performs parameter correction and reminds you to retest;

[0054] Level 3 warning (serious deviation): If the deviation exceeds 10% to 15%, the test will be stopped and a prompt will be given to check the environment and operating procedures.

[0055] Preferably, the drug dilution module includes real-time feedback data intelligent analysis, introduces real-time data analysis and anomaly detection algorithms, and performs real-time monitoring and anomaly analysis of key data in the experimental process;

[0056] Anomaly Detection Model:

[0057]

[0058] Where: Z is the standard score used to judge anomalies, X is the current value, μ and σ are the mean and standard deviation of historical data respectively. When Z>3, the data point is considered abnormal;

[0059] Real-time feedback strategy: Once an abnormal data point is identified, the system will immediately record it and take two steps:

[0060] Short-term feedback: Instantly adjust current spraying volume, temperature and humidity control parameters;

[0061] Long-term feedback: Abnormal data points are stored in the database for subsequent model optimization and early warning algorithm training.

[0062] Preferably, the inconsistent result warning includes adaptive adjustment of the warning frequency. On the basis of multiple tests, the system can adaptively adjust the warning frequency according to different experimental conditions, strain types and test drugs to avoid excessive feedback and waste of resources;

[0063] Frequency adaptation formula:

[0064]

[0065] Where: f alert is the current warning frequency, T mean is the feedback time interval recommended by the system, Δσ is the current data fluctuation, σ baseline The frequency is adjusted to reduce the feedback frequency under stable conditions and increase the feedback frequency during abnormal fluctuations.

[0066] Compared with the prior art, the present invention has the following beneficial effects:

[0067] 1. Rapid drug sensitivity testing: This program uses a micro-liquid dilution method and an automated drug sensitivity assessment system to obtain drug sensitivity data within 6-8 hours, greatly shortening the testing time.

[0068] 2. Automated data processing: Through optical density detection, image processing and data analysis technology, the inhibitory effect of drugs on bacteria can be monitored in real time, and the data is more objective and accurate.

[0069] 3. Systematic data accumulation and analysis: The system stores the results of each drug sensitivity test, and can perform historical data analysis on the sensitivity of different batches, different strains and different drugs, and provide medication trends and optimization suggestions for multiple antimicrobial drug combinations.

[0070] 4. Strong environmental adaptability: The system uses built-in temperature and humidity adjustment modules to ensure that drug sensitivity tests are carried out under stable environmental conditions, thereby improving the stability and repeatability of test results. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0072] Figure 1 The present invention is a flow chart of the drug sensitivity assessment method of Riemerella anatipestifer of the present invention;

[0073] Figure 2 It is a comparison table of parameters of different antibiotics of the present invention;

[0074] Figure 3 The block diagram of the drug sensitivity assessment system of Riemerella anatipestifer of the present invention;

[0075] Figure 4 This is a data analysis module diagram of the present invention. DETAILED DESCRIPTION

[0076] 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.

[0077] See also Figures 1 to 4 , the present invention provides a technical solution:

[0078] The method for evaluating the drug sensitivity of Riemerella anatipestifer comprises the following steps:

[0079] S1. Preparation of bacterial solution: Take the sample strain and inoculate it into the nutrient solution and culture it to a bacterial concentration of 10^8 CFU / mL;

[0080] S2. Automatic drug dilution and addition: The system automatically dilutes the drug to be tested to different concentrations according to the experimental design and adds it to the bacterial solution;

[0081] S3, real-time optical detection: the optical detection module collects optical density data every 30 minutes to record the growth changes of bacterial solution under different drug concentrations;

[0082] S4, data analysis and output: The system calculates the optical density data, draws the bacterial growth curve and calculates the minimum inhibitory concentration, and outputs the drug sensitivity report;

[0083] The drug susceptibility assessment system for Riemerella anatipestifer includes:

[0084] Main control unit: responsible for overall system control, including data collection, analysis, storage, and result output;

[0085] Drug dilution module: equipped with an automatic drug dilution device, which automatically adds drug solutions of different concentrations into the bacterial solution to form a series of drug gradients;

[0086] Optical detection module: using optical density detection instrument to detect the inhibitory effect of different concentrations of drugs on bacterial solution in real time, obtain bacterial growth curve and automatically calculate the inhibitory concentration;

[0087] Data analysis module: Utilizes statistical models and machine learning algorithms to automatically analyze drug sensitivity data, grade and evaluate drug sensitivity, and generate charts and reports;

[0088] Environmental regulation module: The system is equipped with a temperature and humidity regulator to ensure that the entire drug sensitivity assessment process is carried out under stable environmental conditions, thereby improving the reliability of experimental data.

[0089] Specifically, the data analysis module introduces deep learning and big data analysis technologies to optimize the data processing flow and analysis methods. The deep learning model introduced includes data imaging analysis based on convolutional neural networks. In the process of drug sensitivity assessment, the bacterial liquid growth curve and antibacterial image are converted into image data and analyzed through convolutional neural networks to capture the microscopic characteristics of bacterial growth status.

[0090] CNN model structure:

[0091] Input layer: Convert the growth curve data detected by optical density into a two-dimensional grayscale image or heat map.

[0092] Convolutional layer and pooling layer: automatically extract local features of data changes (such as the expansion of the inhibition zone, color changes, etc.) to improve data recognition and discrimination capabilities.

[0093] Fully connected layer and Softmax output layer: convert the extracted features into classification results of antibacterial effects and drug sensitivity scores.

[0094] Model training: The CNN model is trained using historical drug sensitivity data so that the model can recognize the characteristics of different strains and drug concentrations.

[0095] Through the graphical analysis of the CNN model, the system can more accurately determine changes in bacterial growth status, identify data fluctuations with higher accuracy, and reduce false alarm rates.

[0096] In order to make the drug susceptibility assessment system of Riemerella anatipestifer more intelligent, accurate and practical, deep learning and big data analysis technologies are introduced in the data processing and analysis modules to optimize the data processing flow and analysis methods, ensure the stability and real-time nature of the detection data, and thus obtain more reliable and predictive data analysis results;

[0097] Introducing the use of long short-term memory networks for time series. Long short-term memory networks for time series are suitable for processing time series data, analyzing and capturing the long-term dependence of bacterial growth, and have stronger predictive power in the analysis of optical density data;

[0098] Long short-term memory network model structure:

[0099] Input layer: Input optical density data at different time points.

[0100] Hidden layer: The memory unit is used to capture the long-term trend changes in the bacterial growth curve, such as the gradual emergence of the antibacterial effect.

[0101] Output layer: Outputs bacterial growth prediction results or drug sensitivity scores.

[0102] Long short-term memory network model:

[0103] h t =σ(W x x t +W h h t-1 +b)

[0104] Where: h t is the further hidden state at the current time, x t is the input data (optical density value), W x and W h is the weight matrix and b is the bias.

[0105] Long short-term memory networks have strong memory capabilities when processing time series data, can accurately analyze the dynamic changes of antibacterial effects, reduce the time lag of data processing, and reflect drug sensitivity in real time.

[0106] Specifically, the data analysis module also includes an abnormal data detection and elimination unit, which mainly detects outliers through the isolation forest algorithm, and further eliminates the abnormal data detection results based on the local anomaly factor algorithm; the isolation forest algorithm is suitable for processing anomaly detection of multidimensional data and can identify abnormal data points (such as extreme optical density values) during the experiment.

[0107] The steps of the isolation forest algorithm include:

[0108] a. Random data partitioning: Each data point is randomly partitioned multiple times, and outliers are identified by calculating the depth of the data point in the partition tree;

[0109] b. Anomaly score calculation: The anomaly score S(x,n) of each data point is:

[0110]

[0111] Where: E(h(x)) is the average path length of data point x in the partition tree, c(n) is the average path length of normal points;

[0112] c. Threshold judgment: When the anomaly score is higher than the threshold, it is marked as abnormal data;

[0113] The isolation forest algorithm is used to automatically remove extreme abnormal data in the experiment to ensure the stability and accuracy of data analysis.

[0114] Among them, the anomaly detection based on the local anomaly factor algorithm identifies local abnormal data by comparing the density difference between the data point and its neighboring points. The anomaly score calculation formula is:

[0115]

[0116] Where: LOF(A) is the local outlier factor of data point A, LDR(B) is the local density of data point B, and |N(A)| is the number of neighborhood points of data point A.

[0117] The local abnormal factor algorithm is used to automatically identify and eliminate slightly abnormal data, thus avoiding the impact of small deviations on the results during drug sensitivity analysis.

[0118] Specifically, the data analysis module includes a dynamic adjustment and multidimensional data clustering analysis unit, which includes dynamic data weighting and multidimensional data clustering. Through dynamic weighting and multidimensional clustering analysis, multiple groups of data are clustered according to different conditions; through dynamic weighting and multidimensional clustering analysis, multiple groups of data are clustered according to different conditions to improve the accuracy of drug sensitivity assessment.

[0119] Dynamic Weighting Model:

[0120]

[0121] Where: W i is the weighted value of the ith data point, e i (t) is the error value at time t, α is the learning rate, and multidimensional data clustering: K-means algorithm is used to cluster data of different time periods, drug concentrations, and strains, and the drug concentration scheme is readjusted according to the clustering results of each group of data;

[0122] Multidimensional data clustering: Use the K-means algorithm to cluster data of different time periods, drug concentrations, and strains, and readjust the drug concentration scheme based on the clustering results of each group of data;

[0123] Clustering formula:

[0124]

[0125] Where d is the distance between cluster centers, which is used to aggregate similar data. Dynamic weighting and cluster analysis can significantly improve the sensitivity analysis of different concentrations of drugs on different strains, forming a more comprehensive drug sensitivity data evaluation.

[0126] Specifically, the data analysis module includes a historical data accumulation and model self-update unit, which specifically includes incremental learning model update and data storage and feature extraction optimization, and continuously updates the model based on the incremental learning method, so that the system automatically adapts to different experimental conditions and strain characteristics;

[0127] Incremental update formula:

[0128]

[0129] Where: θ t+1 and θ t are the parameters of the new and old models, J(θ) is the loss function;

[0130] The data storage and feature extraction are optimized, historical drug sensitivity data is stored, features are regularly analyzed and extracted, and principal component analysis is used to reduce dimensionality to reduce data dimensions and improve model prediction efficiency;

[0131] Principal component analysis dimensionality reduction formula:

[0132] X′=XW

[0133] Among them: X is the original data matrix, W is the projection matrix, and X′ is the data matrix after dimensionality reduction. Incremental learning and feature dimensionality reduction enable the model to adapt to changes in various strain types and experimental environmental conditions, ensuring the long-term reliability of the test results.

[0134] Through the comprehensive application of deep learning, anomaly detection, multidimensional clustering and incremental learning, the data processing and analysis modules of the drug sensitivity assessment system are made more intelligent, flexible and precise. Specific advantages include:

[0135] 1. Data analysis accuracy is significantly improved: The application of CNN and LSTM in data visualization and time series analysis enables the system to identify complex data features and provide higher recognition accuracy.

[0136] 2. Automatic anomaly detection and elimination: Through the isolation forest and LOF algorithms, the system can automatically identify and eliminate outliers when the data is abnormal, reducing misjudgments caused by errors.

[0137] 3. Dynamic adaptability and multidimensional clustering analysis: The system can dynamically adjust drug dosage according to different conditions and optimize medication strategy based on multidimensional data clustering.

[0138] 4. Historical data accumulation and model optimization: The system optimizes historical data through incremental learning and PCA, so that the system can gradually improve its analysis efficiency and adaptability in long-term use.

[0139] In order to improve the real-time feedback and early warning capabilities of the drug susceptibility assessment system for Riemerella anatipestifer, more detailed data analysis, intelligent feedback mechanism, dynamic threshold adjustment and multi-level early warning are combined to ensure that the system can quickly respond to abnormal situations and provide optimization suggestions during the detection process, thereby improving the reliability and anti-interference ability of the system.

[0140] Specifically, the drug dilution module includes a real-time feedback and early warning mechanism unit, which is mainly used for threshold monitoring: setting the minimum effective concentration threshold of the drug concentration, and giving real-time reminders when the drug has no obvious inhibitory effect on the bacterial solution, to avoid the use of ineffective drugs; early warning of inconsistent results: when there is a deviation between the experimental results and the historical data, the system can issue an early warning, prompting retesting or checking the experimental conditions to ensure data accuracy.

[0141] Preferably, the threshold monitoring adopts dynamic threshold adjustment, real-time analysis of the monitored optical density, temperature, and humidity parameters, and dynamic adjustment of the minimum inhibitory concentration threshold of the key drug;

[0142] The dynamic adjustment model is:

[0143]

[0144] Where: Threshold dynamic is the dynamically adjusted threshold, Threshold baseline is the basic threshold, and ΔP is the deviation percentage caused by environmental parameters such as temperature and humidity.

[0145] Preferably, the inconsistent result warning includes a graded warning, which is divided into three levels and different feedback measures are taken according to the degree of abnormality;

[0146] Level 1 warning (slight deviation): When the data deviates from the normal threshold by 0% to 5%, the system will issue a prompt to recommend a review;

[0147] Level 2 warning (moderate deviation): When the data deviation reaches 5% to 10%, the system automatically performs parameter correction and reminds you to retest;

[0148] Level 3 warning (serious deviation): If the deviation exceeds 10% to 15%, the test will be stopped and a prompt will be given to check the environment and operating procedures.

[0149] The combination of dynamic thresholds and graded warnings can effectively avoid unnecessary false alarms and missed alarms, making the system more accurate and stable.

[0150] Specifically, the drug dilution module includes intelligent analysis of real-time feedback data, introduces real-time data analysis and anomaly detection algorithms, and performs real-time monitoring and anomaly analysis of key data during the experiment;

[0151] Anomaly Detection Model:

[0152]

[0153] Where: Z is the standard score used to judge anomalies, X is the current value, μ and σ are the mean and standard deviation of historical data respectively. When Z>3, the data point is considered abnormal;

[0154] Real-time feedback strategy: Once an abnormal data point is identified, the system will immediately record it and take two steps:

[0155] Short-term feedback: Instantly adjust current spraying volume, temperature and humidity control parameters;

[0156] Long-term feedback: store abnormal data points in the database for subsequent model optimization and early warning algorithm training;

[0157] Real-time feedback and intelligent anomaly analysis can effectively reduce errors caused by environmental fluctuations or data fluctuations, ensuring the accuracy and consistency of data.

[0158] Specifically, the inconsistent result warning includes adaptive adjustment of the warning frequency. On the basis of multiple tests, the system can adaptively adjust the warning frequency according to different experimental conditions, strain types and test drugs to avoid excessive feedback and waste of resources;

[0159] Frequency adaptation formula:

[0160]

[0161] Where: falert is the current warning frequency, T mean is the feedback time interval recommended by the system, Δσ is the current data fluctuation, σ baseline The frequency is adjusted to reduce the feedback frequency under stable conditions and increase the feedback frequency under abnormal fluctuations.

[0162] The adaptive warning frequency reduces the frequency of unnecessary feedback while ensuring the real-time and accuracy of the detection results.

[0163] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for evaluating the drug sensitivity of Riemerella anatipestifer, characterized in that: The steps include: S1. Preparation of bacterial solution: Take the sample strain and inoculate it into the nutrient solution and culture it to a bacterial concentration of 10^8 CFU / mL; S2. Automatic drug dilution and addition: The system automatically dilutes the drug to be tested to different concentrations according to the experimental design and adds it to the bacterial solution; S3, real-time optical detection: the optical detection module collects optical density data every 30 minutes to record the growth changes of bacterial solution under different drug concentrations; S4, data analysis and output: The system calculates the optical density data, draws the bacterial growth curve and calculates the minimum inhibitory concentration, and outputs the drug sensitivity report; The drug susceptibility assessment system for Riemerella anatipestifer includes: Main control unit: responsible for overall system control, including data collection, analysis, storage, and result output; Drug dilution module: equipped with an automatic drug dilution device, which automatically adds drug solutions of different concentrations into the bacterial solution to form a series of drug gradients; Optical detection module: using optical density detection instrument to detect the inhibitory effect of different concentrations of drugs on bacterial solution in real time, obtain bacterial growth curve and automatically calculate the inhibitory concentration; Data analysis module: Utilizes statistical models and machine learning algorithms to automatically analyze drug sensitivity data, grade and evaluate drug sensitivity, and generate charts and reports; Environmental regulation module: The system is equipped with a temperature and humidity regulator to ensure that the entire drug sensitivity assessment process is carried out under stable environmental conditions, thereby improving the reliability of experimental data.

2. The method for drug sensitivity assessment of Riemerella anatipestifer according to claim 1, characterized in that: The data analysis module introduces deep learning and big data analysis technologies to optimize the data processing flow and analysis methods. The deep learning model introduced includes data imaging analysis based on convolutional neural networks. In the process of drug sensitivity assessment, the bacterial liquid growth curve and antibacterial image are converted into image data and analyzed through convolutional neural networks to capture the microscopic characteristics of bacterial growth status; The use of long short-term memory networks for time series analysis was introduced to capture the long-term dependencies of bacterial growth, which is predictive in the analysis of optical density data.

3. The method for drug sensitivity assessment of Riemerella anatipestifer according to claim 2, characterized in that: The data analysis module also includes an abnormal data detection and elimination unit, which mainly detects abnormal values ​​through an isolation forest algorithm and further eliminates abnormal data detection results based on a local abnormal factor algorithm; The steps of the isolation forest algorithm include: a. Random data partitioning: Each data point is randomly partitioned multiple times, and outliers are identified by calculating the depth of the data point in the partition tree; b. Anomaly score calculation: The anomaly score S(x,n) of each data point is: Where: E(h(x)) is the average path length of data point x in the partition tree, c(n) is the average path length of normal points; c. Threshold judgment: When the anomaly score is higher than the threshold, it is marked as abnormal data; Among them, the anomaly detection based on the local anomaly factor algorithm identifies local abnormal data by comparing the density difference between the data point and its neighboring points. The anomaly score calculation formula is: Where: LOF(A) is the local outlier factor of data point A, LDR(B) is the local density of data point B, and |N(A)| is the number of neighborhood points of data point A.

4. The method for drug sensitivity assessment of Riemerella anatipestifer according to claim 3, characterized in that: The data analysis module includes a dynamic adjustment and multidimensional data clustering analysis unit, which includes dynamic data weighting and multidimensional data clustering. Through dynamic weighting and multidimensional clustering analysis, multiple groups of data are clustered according to different conditions; Dynamic Weighting Model: Where: W i is the weighted value of the ith data point, e i (t) is the error value at time t, α is the learning rate; Multidimensional data clustering: Use the K-means algorithm to cluster data of different time periods, drug concentrations, and strains, and readjust the drug concentration scheme based on the clustering results of each group of data; Clustering formula: Where d is the distance between cluster centers, which is used to aggregate similar data.

5. The method for drug sensitivity assessment of Riemerella anatipestifer according to claim 4, characterized in that: The data analysis module includes a historical data accumulation and model self-update unit, which specifically includes incremental learning model update and data storage and feature extraction optimization, and continuously updates the model based on the incremental learning method, so that the system automatically adapts to different experimental conditions and strain characteristics; Incremental update formula: Where: θ t+1 and θ t are the parameters of the new and old models, J(θ) is the loss function; The data storage and feature extraction are optimized, historical drug sensitivity data is stored, features are regularly analyzed and extracted, and principal component analysis is used to reduce dimensionality to reduce data dimensions and improve model prediction efficiency; Principal component analysis dimensionality reduction formula: X′=XW Where: X is the original data matrix, W is the projection matrix, and X′ is the data matrix after dimensionality reduction.

6. The method for drug sensitivity assessment of Riemerella anatipestifer according to claim 1, characterized in that: The drug dilution module includes a real-time feedback and early warning control unit, which is mainly used for threshold monitoring: setting the minimum effective concentration threshold of the drug concentration, and giving real-time reminders when the drug has no obvious inhibitory effect on the bacterial solution to avoid the use of ineffective drugs; early warning of inconsistent results: when there is a deviation between the experimental results and historical data, the system can issue an early warning, prompting re-testing or checking the experimental conditions to ensure data accuracy.

7. The method for drug sensitivity assessment of Riemerella anatipestifer according to claim 6, characterized in that: The threshold monitoring adopts dynamic threshold adjustment, analyzes the monitored optical density, temperature, and humidity parameters in real time, and dynamically adjusts the minimum inhibitory concentration threshold of key drugs; The dynamic adjustment model is: Where: Threshold dynamic is the dynamically adjusted threshold, Threshold baseline is the basic threshold, and ΔP is the deviation percentage caused by environmental parameters such as temperature and humidity.

8. The method for drug sensitivity assessment of Riemerella anatipestifer according to claim 7, characterized in that: The result inconsistency warning includes a graded warning, which is divided into three levels and different feedback measures are taken according to the degree of abnormality; Level 1 warning (slight deviation): When the data deviates from the normal threshold by 0% to 5%, the system will issue a prompt to recommend a review; Level 2 warning (moderate deviation): When the data deviation reaches 5% to 10%, the system automatically performs parameter correction and reminds you to retest; Level 3 warning (serious deviation): If the deviation exceeds 10% to 15%, the test will be stopped and a prompt will be given to check the environment and operating procedures.

9. The method for drug sensitivity assessment of Riemerella anatipestifer according to claim 8, characterized in that: The drug dilution module includes real-time feedback data intelligent analysis, introduces real-time data analysis and anomaly detection algorithms, and performs real-time monitoring and anomaly analysis of key data in the experimental process; Anomaly Detection Model: Where: Z is the standard score used to judge anomalies, X is the current value, μ and σ are the mean and standard deviation of historical data respectively. When Z>3, the data point is considered abnormal; Real-time feedback strategy: Once an abnormal data point is identified, the system will immediately record it and take two steps: Short-term feedback: Instantly adjust current spraying volume, temperature and humidity control parameters; Long-term feedback: Abnormal data points are stored in the database for subsequent model optimization and early warning algorithm training.

10. The method for drug sensitivity assessment of Riemerella anatipestifer according to claim 9, characterized in that: The inconsistent result warning includes adaptive adjustment of the warning frequency. Based on multiple tests, the system can adaptively adjust the warning frequency according to different experimental conditions, strain types and test drugs to avoid excessive feedback and waste of resources; Frequency adaptation formula: Where: f alert is the current warning frequency, T mean is the feedback time interval recommended by the system, Δσ is the current data fluctuation, σ baseline The frequency is adjusted to reduce the feedback frequency under stable conditions and increase the feedback frequency during abnormal fluctuations.

Images