Method and system for rapidly determining blood BHB threshold values of different dairy cow groups

By establishing Lasso regression model and second-order derivative model, the blood BHB concentration prediction of mid-infrared spectroscopy is optimized, and the problems of low diagnostic sensitivity and specificity in the prior art are solved, achieving more efficient monitoring of ketone disease in dairy cattle herd.

CN118981608APending Publication Date: 2024-11-19SHANDONG OX LIVESTOCK BREEDING CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411168264.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-23
Publication Date
2024-11-19

AI Technical Summary

Technical Problem

The existing mid-infrared spectrometry is used to detect blood BHB concentration in dairy cows with low diagnostic sensitivity and specificity, making it difficult to effectively monitor subclinical ketosis, affecting animal health and productivity.

Method used

By obtaining milk sample data and blood data, a Lasso regression model was established, and pre-processing and feature selection was combined with the second-order derivative model, the optimal combination model was constructed, blood BHB concentration prediction was optimized, and the subject's working characteristic curve was drawn based on the prediction results. Determine the optimal blood BHB threshold to improve diagnostic accuracy.

Benefits of technology

The sensitivity and specificity of blood BHB concentration prediction by mid-infrared spectroscopy is significantly improved, providing a more efficient subclinical ketosis diagnosis method, which can more accurately monitor the ketosis condition of dairy cow population.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118981608A_ABST
    Figure CN118981608A_ABST
Patent Text Reader

Abstract

The invention discloses a method and system for rapidly determining blood BHB threshold values of different dairy cow groups, and belongs to the technical field of blood detection.The method comprises the steps that milk sample data and blood data in conventional production performance determination of sample dairy cows are obtained, an intermediate infrared spectrum is obtained based on the milk sample data, and the blood BHB concentration is obtained based on the blood data; establishing an optimal model for predicting the BHB concentration of the blood based on the mid-infrared spectrum and the BHB concentration of the blood, and training and optimizing the optimal model to obtain a trained optimal model; performing blood BHB concentration prediction according to the trained optimal model, and obtaining a subject working characteristic curve based on a prediction result; and determining a blood BHB threshold value according to the working characteristic curve of the subject, and monitoring the subclinical ketosis condition of the dairy cow group according to the blood BHB threshold value. The problem that a general SCK positive diagnosis threshold value, namely the blood BHB concentration, is low in sensitivity in mid-infrared spectroscopy diagnosis is solved, and the sensitivity and specificity of mid-infrared spectroscopy diagnosis are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of blood detection, and in particular relates to a method and a system for quickly determining blood BHB thresholds of different dairy cow groups. Background Art

[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art.

[0003] Ketosis is a common metabolic disease in dairy cows, among which subclinical ketosis (SCK) is often overlooked because it is difficult to detect. A high incidence of SCK can easily have a negative impact on animal health, production and profitability. Therefore, it is extremely important to establish a subclinical ketosis monitoring system, in which ketone body content can be detected in blood, urine and milk, which can be used to diagnose SCK, and early treatment can improve production losses in dairy cows.

[0004] Currently, beta-hydroxybutyrate (BHB) in the blood has been widely used for the detection of SCK and is used as the gold standard. However, the biochemical method or electronic blood ketone meter commonly used for the determination of blood BHB concentration is not suitable as a routine method for SCK diagnosis and monitoring in large cattle farms due to labor costs and other limitations.

[0005] Through milk sampling in the routine production performance measurement (DHI) of dairy cows, the use of Fourier transform mid-infrared spectroscopy can provide a more efficient and practical blood BHB prediction strategy, thereby quickly, accurately and in real time monitoring the SCK status of dairy cow herds. However, compared with direct blood BHB concentration determination, the prediction of blood BHB by mid-infrared spectroscopy through milk sampling only shows a moderate level of diagnostic sensitivity and specificity. Based on the previously recommended SCK positive diagnostic threshold, that is, BHB concentration ≥1.2mmol / L, the sensitivity of mid-infrared spectroscopy to this threshold is only 69%-70%, and the diagnostic results are not ideal. Summary of the invention

[0006] In order to overcome the deficiencies of the above-mentioned prior art, the present invention provides a method and system for quickly determining the blood BHB thresholds of different dairy cow groups, which solves the problem of low sensitivity of the universal SCK positive diagnostic threshold, i.e., blood BHB concentration, in mid-infrared spectroscopy diagnosis, and improves the sensitivity and specificity of mid-infrared spectroscopy diagnosis.

[0007] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:

[0008] In a first aspect, the present invention provides a method for quickly determining the blood BHB thresholds of different dairy cow groups, comprising:

[0009] Obtaining milk sample data and blood data from routine production performance measurements of sample dairy cows, and obtaining a mid-infrared spectrum based on the milk sample data, and obtaining a blood BHB concentration based on the blood data;

[0010] Establishing an optimal model for predicting blood BHB concentration based on the mid-infrared spectrum and blood BHB concentration, and training and optimizing the optimal model to obtain a trained optimal model;

[0011] The blood BHB concentration is predicted according to the trained optimal model, and the receiver operating characteristic curve is obtained based on the prediction results;

[0012] The blood BHB threshold is determined according to the receiver operating characteristic curve, and the subclinical ketosis condition of the dairy cow group is monitored by the blood BHB threshold.

[0013] A further technical solution is to pre-process the mid-infrared spectrum and select an effective spectral region for analysis, wherein the pre-processing is to use a second-order derivative to eliminate interference in the spectrum.

[0014] According to a further technical solution, the optimal model is a Lasso regression model.

[0015] In a further technical solution, the fitting of the Lasso regression model is expressed as:

[0016]

[0017] Among them, y i is the response variable, x i is the predictor variable, β is the coefficient variable, ∥β∥1 represents the L1 norm of the coefficient vector, is the regularization parameter, which controls the intensity of the penalty on the coefficients.

[0018] In a further technical solution, the Lasso regression model is expressed as:

[0019]

[0020] In the formula, is the blood BHB concentration, β0 is the intercept term, N is the total number of wave numbers, β i is the regression coefficient of the ith wave number, X i is the spectral value of the ith wave number, and ε is the error term.

[0021] A further technical solution is to obtain a receiver operating characteristic curve based on the prediction results: the true positive rate and the false positive rate are calculated according to the prediction results, and the receiver operating characteristic curve is plotted with the false positive rate as the horizontal axis and the true positive rate as the vertical axis.

[0022] In a further technical solution, the blood BHB threshold is determined based on the concentration that maximizes the true positive rate and the false positive rate.

[0023] In a second aspect, the present invention provides a system for rapidly determining blood BHB thresholds of different dairy cow groups, comprising:

[0024] A data acquisition module, which is configured to: acquire milk sample data and blood data in routine production performance measurement of sample dairy cows, and obtain a mid-infrared spectrum based on the milk sample data, and obtain a blood BHB concentration based on the blood data;

[0025] A model building module, which is configured to: establish an optimal model for predicting blood BHB concentration based on the mid-infrared spectrum and blood BHB concentration, and train and optimize the optimal model to obtain a trained optimal model;

[0026] A prediction drawing module is configured to: predict blood BHB concentration according to the trained optimal model, and obtain a receiver operating characteristic curve based on the prediction result;

[0027] The threshold determination module is configured to determine the blood BHB threshold according to the receiver operating characteristic curve, and monitor the subclinical ketosis condition of the dairy cow group with the blood BHB threshold.

[0028] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in a method for rapidly determining blood BHB thresholds of different cow groups as described in the first aspect.

[0029] In a fourth aspect, the present invention provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in a method for rapidly determining blood BHB thresholds of different dairy cow groups as described in the first aspect are implemented.

[0030] One or more of the above technical solutions have the following beneficial effects:

[0031] The present invention establishes an optimal model for predicting blood BHB concentration using DHI milk samples through mid-infrared spectra and blood BHB concentration of sample cow milk sample data, draws a receiver operating characteristic curve (ROC) based on the prediction result of the trained optimal model, and determines the optimal blood BHB threshold of the corresponding dairy cow population using the receiver operating characteristic curve (ROC). When diagnosing SCK using mid-infrared spectroscopy based on milk sample data, the optimal blood BHB threshold is used for diagnosis, which can greatly improve the sensitivity of diagnosis.

[0032] The present invention combines the second-order derivative (SD) model and the minimum absolute value selection and shrinkage operator (Lasso) regression model to construct an optimal combined model structure from milk sample spectrum to BHB prediction, thereby improving the accuracy of blood BHB concentration prediction and providing a more efficient diagnosis method for subclinical ketosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0034] Figure 1 This is a flow chart of a method for quickly determining blood BHB thresholds of different dairy cow groups according to an embodiment of the present invention;

[0035] Figure 2 is a receiver operating characteristic (ROC) curve generated by predicting blood BHB concentration according to an embodiment of the present invention. DETAILED DESCRIPTION

[0036] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0037] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0038] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0039] In order to solve the problems of low sensitivity and specificity mentioned in the background technology, for the prediction of blood BHB concentration by mid-infrared spectroscopy of DHI milk samples, it is necessary to redefine the diagnostic threshold of SCK-positive gold standard BHB in different dairy cow populations, so as to further improve the sensitivity and specificity.

[0040] Embodiment 1

[0041] like Figure 1 As shown, this embodiment discloses a method for quickly determining the blood BHB thresholds of different dairy cow groups, the method comprising the following steps:

[0042] S1: obtaining milk sample data and blood data of a sample dairy cow in a routine production performance measurement, and obtaining a mid-infrared spectrum based on the milk sample data, and obtaining a blood BHB concentration based on the blood data;

[0043] In this embodiment, by investigating the basic conditions of the dairy farm, feeding methods, milking methods, feeding management measures, etc., dairy cows that meet the DHI measurement requirements are selected, and milk sample data and blood data in the routine production performance measurement (DHI) of dairy cows are collected according to the standard process of DHI measurement.

[0044] The processed milk sample data was spectrally measured using the milk composition analyzer Milkoscan TM FT+milk (Foss Electric A / S, Hiller, Denmark) of the Danish FOSS company to obtain the mid-infrared spectrum of the milk sample data. When collecting blood samples, a suitable blood sampling site was selected, such as the jugular vein or the tail vein. The blood data, i.e., the blood BHB concentration in the blood sample, was quantitatively measured and analyzed based on the enzyme colorimetric β-hydroxybutyrate determination kit, and the measurement result was the true value of BHB.

[0045] It should be noted that the obtained mid-infrared spectrum and blood BHB concentration are one-to-one corresponding, that is, the infrared spectrum of milk sample data and the blood BHB concentration of blood sample data are collected from a cow. Labeling can be performed according to the label information, such as cow number, sampling date, sampling time, etc., to avoid data confusion, resulting in data errors affecting the prediction results.

[0046] In this embodiment, 541 Chinese Holstein dairy cows with a lactation period of less than 50 days were selected from a dairy farm. Blood samples of the tail venous blood of all the dairy cows were collected from 6:00 to 10:00 in the morning for measuring the BHB blood concentration; milk samples were collected from 12:00 to 4:00 in the afternoon of the same day for mid-infrared spectroscopy analysis.

[0047] S2: establishing an optimal model for predicting blood BHB concentration based on the mid-infrared spectrum and blood BHB concentration, and training and optimizing the optimal model to obtain a trained optimal model;

[0048] In this embodiment, the optimal model for predicting blood BHB concentration is a model for predicting blood BHB concentration using DHI milk sample data. By training and optimizing the optimal model, the prediction of blood BHB concentration can achieve the best effect.

[0049] The mid-infrared spectrum and the corresponding blood BHB concentration are used as a data set, and the data set is divided into a training set, a validation set and a test set; the training set is used to train the optimal model, the validation set is used to adjust the parameters of the optimal model and optimize the optimal model, and the test set is used to evaluate the performance of the optimal model and determine the blood BHB threshold.

[0050] S201: Preprocess the mid-infrared spectrum, including denoising, line correction, normalization, etc., to improve data quality, and select effective spectral regions for analysis to ensure the quality of subsequent modeling results.

[0051] In order to ensure the quality of subsequent modeling, the spectral data should be properly processed. -1 between 3,025 cm and OH stretching region (above 3,025 cm -1 ) will lead to noise-like areas. Water absorption areas and non-information areas are excluded. This embodiment selects 2964.4-2833.4cm -1 、1807.9-1784.8cm -1 、1761.7-1734.7cm -1 and 1576.6-971.46cm -1 The spectral region with a total of 208 wavenumbers was used for subsequent analysis.

[0052] The second derivative (SD) can effectively eliminate the interference of various backgrounds and baselines in the spectrum and improve the quality of spectral data. The formula is as follows:

[0053]

[0054] In the formula, x i is the absorbance at the current wave number i, and g is the window width.

[0055] S202: Check whether the blood BHB concentration data has missing values ​​or abnormal values, and perform corresponding processing, such as filling or deleting missing values, eliminating abnormal values, etc.

[0056] S203: Establish an optimal model for predicting blood BHB concentration based on mid-infrared spectrum and blood BHB concentration. The optimal model is the Lasso regression model. Data fitting is performed when the optimal model is established. The Lasso regression model performs feature selection while fitting the data. Specifically, the minimum absolute value selection and the shrinkage operator (Lasso) regression are suitable for processing situations where there is a strong correlation between variables; and the L1 penalty (adding the L1 norm as a regularization term in the loss function of the model) will make some coefficients completely zero, thereby achieving feature selection. The formula is as follows:

[0057]

[0058] In the formula, y i is the response variable, x i is the predictor variable, β is the coefficient variable, ∥β∥1 represents the L1 norm of the coefficient vector, is the regularization parameter, which controls the intensity of the penalty on the coefficients.

[0059] In this embodiment, As the prior information penalty term, η is used as a balance parameter to balance the relative importance of data and prior information, and when it is infinitely small, the prior Lasso model is equivalent to the ordinary Lasso model.

[0060] By integrating the above prior information (such as 2964.4-2833.4cm -1 、1807.9-1784.8cm -1 、1761.7-1734.7cm -1 and 1576.6-971.46cm -1 spectral region) to obtain pseudo-response variables Combined with the prior information penalty term Obtain the penalized coefficient variable β, which is used to predict the blood BHB concentration of individuals with known spectral values The formula is as follows:

[0061]

[0062] In the formula, is the blood BHB concentration, β0 is the intercept term, N is the total number of wave numbers (in this embodiment, the wave number i has a total of 208 wave numbers), β i is the regression coefficient of the ith wave number, X i is the spectral value of the ith wave number, and ε is the error term.

[0063] The optimal combined model structure from milk spectrum to BHB prediction is constructed by combining the second derivative (SD) model and the least absolute value selection and shrinkage operator (Lasso) regression model. The Lasso regression model is trained using the training set data. During the training process, the regularization parameter λ is adjusted to find the optimal complexity of the model. This can be achieved through cross-validation such as K-fold cross-validation to minimize the prediction error on the validation set.

[0064] Parameter optimization of the optimal model: According to the performance on the validation set, adjust the parameters of the Lasso regression model, mainly the regularization parameters, to find the optimal model parameters and obtain the trained optimal model.

[0065] S3: Predicting blood BHB concentration according to the trained optimal model, and obtaining a receiver operating characteristic curve based on the prediction results;

[0066] The performance of the trained optimal model was evaluated using the test set data and the blood BHB threshold was obtained by analysis. Specifically, the mid-infrared spectral data obtained from the milk sample data in the test set was input into the trained optimal model to predict the blood BHB concentration of the corresponding individual. The predicted blood BHB concentration was calculated with the blood BHB concentration measured by the corresponding individual, and the receiver operating characteristic curve was calculated by discriminant model analysis.

[0067] The three indicators for evaluating the accuracy of blood BHB prediction include: correlation coefficient (R 2 ), root mean square error (RMSE) and residual prediction deviation (RPD), where the correlation coefficient (R 2 ) and residual prediction deviation (RPD), the larger the prediction accuracy, while the smaller the root mean square error (RMSE), the more accurate the prediction.

[0068] In this embodiment, the prediction accuracy of the SD-Lasso combined model structure was compared with that of other combined model structures. The results showed that the prediction result of the SD-Lasso combined model structure proposed in this embodiment was the best. The correlation coefficient (R2), root mean square error (RMSE) and residual prediction deviation (RPD) of the three indicators of BHB prediction accuracy were 0.663, 0.281 and 1.724, respectively, which can be used to predict the concentration of blood BHB by mid-infrared spectroscopy.

[0069] Table 1. BHB prediction accuracy (R 2 , RMSE and RPD values)

[0070]

[0071] Note: FD refers to the first-order derivative model; SD refers to the second-order derivative model; RFR refers to the random regression forest model; PLSR refers to the least squares regression model; Lasso refers to the least absolute value selection and shrinkage operator model; Ridge refers to the ridge regression model.

[0072] The receiver operating characteristic (ROC) curve was calculated based on the prediction results of the optimal model through discriminant model analysis, specifically:

[0073] True Positive Rate (TPR): The ratio of the number of samples that are actually positive and are correctly judged as positive to the number of samples that are actually positive (that is, they are really true). TPR is also called Sensitivity.

[0074] False Positive Rate (FPR): The ratio of the number of samples that are actually negative but are mistakenly judged as positive to the number of samples that are actually negative (i.e., the false positives are mistakenly considered to be true and are falsely reported). FPR is equal to 1-Specificity.

[0075] Among them, true positives (TP) are the number of true positives, false positives (FP) are the number of false positives, true negatives (TN) are the number of true negatives, and false negatives (FN) are the number of false negatives. Then calculate the accuracy, sensitivity, specificity, precision and F-test score:

[0076] Accuracy = (TP+TN) / (TP+TN+FP+FN)

[0077] Sensitivity = TP / (TP+FN) = TPR

[0078] Specificity = TN / (TN+FP)

[0079] Accuracy = TP / (TP+FP)

[0080] Finally, the ROC curve is actually a curve drawn with FPR as the horizontal axis and TPR as the vertical axis. This is because the assumed cost of a false negative is greater than the cost of further testing and treatment associated with a false positive.

[0081] S4: Determine a blood BHB threshold value according to a receiver operating characteristic curve, and use the blood BHB threshold value to monitor subclinical ketosis in the dairy cow population.

[0082] The Youden Index, also known as the correct index, is a commonly used method when assuming that the harm of false negatives (missed diagnosis rate) and false positives (misdiagnosis rate) is of equal significance, which reflects the total ability of real patients and non-patients. Youden Index = Sensitivity (i.e., true positive rate TPR) + (specificity-1) (i.e., the Y-axis minus the X-axis of the ROC curve, i.e., the false positive rate FPR). The Youden Index is a statistical measure commonly used to select the best classification threshold. The value range of the Youden Index is between -1 and 1. The closer its value is to 1, the better the performance of the model. In general, selecting a classification threshold that maximizes the Youden Index can achieve the best performance of the model while balancing sensitivity and specificity. Based on the Youden Index selection method, the true positive rate TPR and the false positive rate FPR are obtained, and the receiver operating characteristic (ROC) curve is generated. The optimal blood BHB threshold is determined based on the concentration that maximizes the sensitivity and specificity of predicting SCK. The optimal predicted blood BHB threshold is determined based on maximizing the concentration that maximizes the sensitivity and specificity of predicting subclinical ketosis (SCK).

[0083] For the example data in this embodiment, Figure 2 It can be inferred that when the sensitivity and specificity are 0.930 and 0.899 respectively, the predicted blood BHB concentration is the largest. At this time, the optimal blood BHB threshold is 1.082mmol / L.

[0084] Further, in order to evaluate the potential diagnostic ability of blood BHB threshold as a single test, the present embodiment also calculates the area under the curve (AUC) of the ROC curve. The discriminant model analysis is to calculate the receiver operating characteristic curve, and the accuracy of the discriminant model is evaluated by generating and calculating the area under the ROC (receiver operating characteristic) curve (AUC). The optimal cutoff value of each detection variable is defined as the point where the sum of sensitivity and specificity is the largest (i.e., the weights of false positive and false negative test results are equal). Since high specificity may also cause samples to be judged as false positives, the present embodiment also analyzes and verifies the results with PPV (Positive predictive value) and NPV (Negative predictive value) as considerations, and the verification results are shown in Table 2.

[0085] Table 2. Sensitivity, specificity, global accuracy, PPV, NPV and AUC results obtained by the discriminant model

[0086]

[0087] Note: PPV = Positive Predictive Value, NPV = Negative Predictive Value, AUC = Area Under the Curve

[0088] Furthermore, based on all the results, the final blood BHB threshold was determined to obtain the optimal threshold. The diagnostic performance of the blood BHB threshold was judged by all the obtained results of sensitivity, specificity, global accuracy, PPV, NPV and AUC. The optimal blood BHB threshold was determined with sensitivity ≥ 90% as the diagnostic performance standard, which further improved the predictive ability of mid-infrared spectroscopy for blood BHB in this cattle farm.

[0089] For the example data in this embodiment, as shown in Table 2, when the sensitivity and specificity are approximately 93% and 90% respectively, it indicates that 1.082mmol / L is the optimal blood BHB threshold, which can accurately identify samples that are actually SCK positive, and the proportion of predicted results as SCK positive is approximately 93%. The global accuracy (i.e., the total proportion of correctly classified records) reached 90%. Therefore, the discriminant model can classify dairy cows with reasonable accuracy. The value of the area under the ROC curve AUC is 0.914, indicating that the detection ability of the model is high, further confirming that the model has high accuracy in classifying dairy cows. In addition, the PPV is 52%, which means that 48% of dairy cows are misdiagnosed as hyperketonemia. The NPV reached 99%, and the prediction value index was high. This shows that the detection ability of the model is good. The use of the optimal blood BHB threshold of 1.082mmol / L further improves the predictive ability of mid-infrared spectroscopy for blood BHB in this cattle farm. In this embodiment, the optimal blood BHB threshold of 1.082 mmol / L is only an example, and is a monitoring threshold for a group of 541 Chinese Holstein cows with a lactation period of less than 50 days.

[0090] It should be noted that the optimal blood BHB thresholds corresponding to different dairy cow groups, i.e., different dairy farms or dairy cow groups in different areas of the same dairy farm are all different. The determined optimal blood BHB threshold is specific. Different dairy cow groups need to determine their own optimal blood BHB thresholds to monitor the SCK status of their respective dairy cow groups, which can improve the sensitivity of milk sample data in monitoring blood BHB concentration and improve the accuracy of diagnostic results.

[0091] Embodiment 2

[0092] This embodiment provides a system for quickly determining blood BHB thresholds of different dairy cow groups, including:

[0093] A data acquisition module, which is configured to: acquire milk sample data and blood data in routine production performance measurement of sample dairy cows, and obtain a mid-infrared spectrum based on the milk sample data, and obtain a blood BHB concentration based on the blood data;

[0094] A model building module, which is configured to: establish an optimal model for predicting blood BHB concentration based on the mid-infrared spectrum and blood BHB concentration, and train and optimize the optimal model to obtain a trained optimal model;

[0095] A prediction drawing module is configured to: predict blood BHB concentration according to the trained optimal model, and obtain a receiver operating characteristic curve based on the prediction result;

[0096] The threshold determination module is configured to determine the blood BHB threshold according to the receiver operating characteristic curve, and monitor the subclinical ketosis condition of the dairy cow group with the blood BHB threshold.

[0097] Embodiment 3

[0098] The purpose of this embodiment is to provide a computing device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method of embodiment 1 when executing the program.

[0099] Embodiment 4

[0100] The purpose of this embodiment is to provide a computer-readable storage medium, a computer-readable storage medium having a computer program stored thereon, and when the program is executed by a processor, the steps of the method of embodiment 1 are performed.

[0101] The steps involved in the apparatus of the above embodiments 3 and 4 correspond to the method embodiment 1, and the specific implementation method can refer to the relevant description part of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method in the present invention.

[0102] Those skilled in the art should understand that the modules or steps of the present invention described above can be implemented by a general-purpose computer device, or alternatively, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0103] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0104] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. A method for quickly determining the blood BHB threshold of different dairy cow groups, characterized in that: include: Obtaining milk sample data and blood data from routine production performance measurements of sample dairy cows, and obtaining a mid-infrared spectrum based on the milk sample data, and obtaining a blood BHB concentration based on the blood data; Establishing an optimal model for predicting blood BHB concentration based on the mid-infrared spectrum and blood BHB concentration, and training and optimizing the optimal model to obtain a trained optimal model; The blood BHB concentration is predicted according to the trained optimal model, and the receiver operating characteristic curve is obtained based on the prediction results; The blood BHB threshold is determined according to the receiver operating characteristic curve, and the subclinical ketosis condition of the dairy cow group is monitored by the blood BHB threshold.

2. A method for rapidly determining blood BHB thresholds of different dairy cow groups as claimed in claim 1, characterized in that: The mid-infrared spectrum is preprocessed, and an effective spectral region is selected for analysis. The preprocessing is to use a second-order derivative to eliminate interference in the spectrum.

3. A method for rapidly determining blood BHB thresholds of different dairy cow groups as claimed in claim 1, characterized in that: The optimal model is the Lasso regression model.

4. A method for rapidly determining blood BHB thresholds of different dairy cow groups as claimed in claim 3, characterized in that: The fitting of the Lasso regression model is expressed as: Among them, y i is the response variable, x i is the predictor variable, β is the coefficient variable, ∥β∥1 represents the L1 norm of the coefficient vector, is the regularization parameter, which controls the intensity of the penalty on the coefficients.

5. A method for rapidly determining blood BHB thresholds of different dairy cow groups as claimed in claim 3, characterized in that: The Lasso regression model is expressed as: In the formula, is the blood BHB concentration, β0 is the intercept term, N is the total number of wave numbers, β i is the regression coefficient of the ith wave number, X i is the spectral value of the ith wave number, and ε is the error term.

6. A method for rapidly determining blood BHB thresholds of different dairy cow groups as claimed in claim 1, characterized in that: The receiver operating characteristic curve based on the prediction results is specifically as follows: the true positive rate and the false positive rate are calculated according to the prediction results, and the receiver operating characteristic curve is plotted with the false positive rate as the horizontal axis and the true positive rate as the vertical axis.

7. A method for rapidly determining blood BHB thresholds of different dairy cow groups as claimed in claim 6, characterized in that: The blood BHB threshold is determined based on the concentration that maximizes the true positive rate and the false positive rate.

8. A system for rapidly determining the blood BHB thresholds of different groups of dairy cows, characterized in that: include: A data acquisition module, which is configured to: acquire milk sample data and blood data in routine production performance measurement of sample dairy cows, and obtain a mid-infrared spectrum based on the milk sample data, and obtain a blood BHB concentration based on the blood data; A model building module, which is configured to: establish an optimal model for predicting blood BHB concentration based on the mid-infrared spectrum and blood BHB concentration, and train and optimize the optimal model to obtain a trained optimal model; A prediction drawing module is configured to: predict blood BHB concentration according to the trained optimal model, and obtain a receiver operating characteristic curve based on the prediction result; The threshold determination module is configured to determine the blood BHB threshold according to the receiver operating characteristic curve, and monitor the subclinical ketosis condition of the dairy cow group with the blood BHB threshold.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in a method for rapidly determining blood BHB thresholds of different cow groups as described in any one of claims 1 to 7 are implemented.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps in the method for rapidly determining the blood BHB thresholds of different cow groups as described in any one of claims 1 to 7 are implemented.