CN105277509A - Near infrared nondestructive test method for aspartate aminotransferase activity in serum - Google Patents

Abstract

The invention provides a rapid and nondestructive detection method for aspartate aminotransferase (AST) activity in human serum, which belongs to the field of clinical testing. The method comprises the following steps: collection of blood samples; preparation of serum samples; reference value of AST activity in the determination of serum samples; measurement of near-infrared transmission spectrum of serum samples by Fourier transform near-infrared spectrometer; The data is processed, and the optimal preprocessing method is screened out; a reasonable near-infrared spectral modeling area is selected; the processed spectral data is associated with the reference value of AST activity, and a quantitative prediction model for AST activity in serum is established, and the The established model was verified; the near-infrared transmission spectrum of the serum sample to be tested was collected and processed in the same way, and the model was used to predict the activity of AST in the serum to be tested. The method is rapid, non-destructive, easy to operate and high in accuracy.

Description

Near-infrared non-destructive detection method for aspartate aminotransferase activity in serum

technical field

The invention relates to a rapid and non-destructive analysis method for aspartate aminotransferase (aspartateaminotransferase, AST) activity in human serum, specifically relates to the rapid quantitative analysis of AST activity in serum by using near-infrared spectroscopy combined with chemometrics technology, which belongs to clinical Inspection field.

Background technique

Aspartate aminotransferase (aspartateaminotransferase, AST) is an intracellular functional enzyme, which is widely distributed in the heart, liver, skeletal muscle and kidney tissue cells of the human body, among which the heart muscle and liver have the highest content. AST can catalyze the transfer of an amino group from a specific amino acid (L-glutamic acid or L-aspartic acid) to the corresponding ketoacids (α-ketoglutarate and oxaloacetate). Under normal circumstances, the extracellular AST content is extremely low, and the serum content is generally only one thousandth of the intracellular content. When the heart, liver and other organs are damaged, the content of AST in the serum increases significantly, indicating the generation of inflammation and cell necrosis in the body. At present, the magnitude of AST activity in serum is mainly used to judge the damage of the liver. Since AST has the highest content in the myocardium, AST activity in serum can also be used as a sensitive indicator for judging the effects of drug toxicity on cardiac function. At present, medical institutions such as hospitals and blood collection stations usually use biochemical methods assisted by automatic biochemical analyzers to detect the activity of AST in serum. The incubation process and pretreatment process of this method are cumbersome and time-consuming, which affects doctors' timely diagnosis and treatment of critically ill patients; secondly, this method uses destructive detection technology, and serum samples can no longer be used for the determination of other clinical test indicators; in addition, this method requires the use of Biochemical reagents cause a certain degree of biochemical pollution to the environment. Therefore, inventing a rapid, non-destructive, non-polluting, and low-cost method for detecting AST activity in serum is of great significance for clinical testing.

Near infrared spectroscopy (near infrared spectroscopy, NIRS) characterizes the double frequency and combined frequency absorption of hydrogen-containing groups in molecular vibrational transitions, which carries a wealth of sample information. NIRS combined with chemometric techniques (chemometrictechniques) can extract the characteristic information of samples, and realize the rapid, non-destructive, non-polluting and low-cost analysis of complex systems. Near-infrared spectroscopy is a convenient analytical method.

The invention combines NIRS and chemometric techniques to establish a rapid non-destructive detection method for AST in human serum.

Contents of the invention

The purpose of the present invention is to provide a rapid, non-destructive detection method for AST activity in human serum, which mainly includes the following steps:

(1) Collection of blood samples;

(2) Preparation of serum samples;

(3) Determination of the reference value of AST activity in serum samples;

(4) Measure the near-infrared transmission spectrum of the serum sample with a Fourier transform near-infrared spectrometer;

(5) Process the data of the measured spectrum with chemometrics technology, and screen out the optimal preprocessing method;

(6) Select a reasonable near-infrared spectrum modeling area;

(7) correlate the processed spectral data with the reference value of AST activity, establish a quantitative prediction model for AST activity in serum, and verify the established model;

(8) Collect and process the near-infrared transmission spectrum of the serum sample to be tested in the same way, and use the established model to predict the activity of AST in the serum to be tested.

The preparation method of the serum sample in the step (1) is as follows: in a place where the medical institution meets the requirements, the medical staff collects an appropriate amount of venous blood according to the regulations, puts it in a blood collection tube containing a coagulant, and waits for the blood to coagulate and stratify at room temperature. , take the supernatant as the serum sample.

The method for determining the reference value of AST activity in the step (3) is a method commonly used in medical institutions at present, including but not limited to detection by a fully automatic biochemical analyzer.

In the step (4), the near-infrared transmission spectrum measurement method is determined with reference to the operating procedures of the instrument. Spectrometers include but not limited to Fourier transform near-infrared spectrometers; measurement modes include but not limited to transmission mode; measurement parameters include but not limited to spectral measurement range, resolution and scan times.

The preprocessing method of near-infrared spectral data in the described step (5) can be but not limited to one or in unprocessed, first derivative, second derivative, Norris smoothing, Savitzky-Golay smoothing, mean centralization, calibration Various.

The software used for the selection of reasonable modeling spectral regions in the step (6) includes but not limited to TQAnalyst8.0 software; the selection methods include but not limited to software automatic selection, manual selection, the combination of software automatic selection and manual selection.

The software that is used for data processing in described step (7) includes but not limited to TQAnalyst software and MATLAB software; Quantitative modeling algorithm includes but not limited to partial least square method (PLS), principal component regression method (PCR) and artificial neural network One or more of the methods (ANN).

When using the built calibration model in the step (8) to predict the AST activity in the sample to be tested, the spectral measurement method, spectral measurement parameters, spectral data preprocessing method and spectral modeling area used by the sample to be tested should be consistent with the calibration. The set of samples is consistent.

This method is suitable for the quantitative analysis of AST activity in serum. The detection process does not require complex sample pretreatment and reagents, which can greatly shorten the analysis time. It is a convenient, rapid and non-destructive analysis technique, which is worth promoting.

Description of drawings

Figure 196 Fourier transform near-infrared transmission original spectrum of serum samples

Figure 2 The linear correlation diagram of the optimal PLS model reference value and predicted value for the determination of AST activity in serum samples

detailed description

The present invention will be further described below in conjunction with the accompanying drawings and examples, which should not be construed as limiting the present invention.

Example

1. Instrument

AntarisII Fourier transform near-infrared spectrometer (Thermo Company, USA), the sampling device is a transmission analysis module, the signal acquisition software is RESULT3.0, and the data analysis and modeling software is TQAnalyst8.0. The enzyme activity assay instrument is BS-800M automatic biochemical analyzer (Shenzhen Mindray Company).

2. Serum samples

96 serum samples, 54 blood samples were collected from women and 42 blood samples were collected from men. The preparation method of the serum sample is to collect an appropriate amount of venous blood, place it in a blood collection tube containing a coagulant, wait for the blood to coagulate and stratify at room temperature, and take the supernatant to be the serum sample.

3. Determination of AST activity reference value in serum samples

The AST activity of serum samples was measured by BS-800M automatic biochemical analyzer. The range of AST activity of 96 serum samples was 12.30-79.83U/L.

4. Acquisition of near-infrared transmission spectra

Spectrum acquisition method: Take an appropriate amount of serum sample and inject it into the clean liquid sample tube equipped with the instrument. The injection volume is about 2/3 of its volume, and put it into the liquid sample holder correctly. The spectral signal acquisition software is RESULT3.0.

Spectral measurement parameters: the spectral scanning range is 10000-4000cm ^-1 , the resolution is 8cm ^-1 , and the number of scanning is 32 times. Before each scan of the sample, scan with the same parameters and subtract the background.

The original spectra of Fourier transform near-infrared transmission of 96 serum samples are shown in Figure 1.

5. Preprocessing of spectral data

The spectral processing was completed by TQAnalyst8.0 software, and the processing methods included mean centering, first derivative, second derivative, 7-point Savitzky-Golay smoothing and 5-point Norris smoothing.

6. Select a reasonable near-infrared spectrum modeling area

The spectral range used for modeling was automatically screened by TQAnalyst8.0 software into two segments of 7185-7023cm ^-1 and 7020-6869cm ^-1 .

7. Establishment of Quantitative Model

In this example, the partial least squares (PLS) method was used to establish a quantitative model of AST activity in serum samples. The 96 samples were divided into calibration set and validation set. The AST activity reference value range of 72 calibration set samples was 12.30-79.83U/L, and the AST activity reference value range of the remaining 24 validation set samples was 14.54-68.60U/L.

The model performance is evaluated by the following indicators: calibration set correlation coefficient (R _c ), cross-validation set correlation coefficient (R _cv ), validation set correlation coefficient (R _p ), calibration set root mean square error (RMSEC), cross-validation set mean Root Square Error (RMSECV) and Validation Set Root Mean Square Error (RMSEP).

The preprocessing method of the PLS optimal quantitative model is the combination of mean centering, second derivative and 5-point Norris smoothing. The number of principal factors is 7, and the selection method is the number of principal factors corresponding to the minimum RMSECV by leave-one-out cross-validation. The linear relationship between the reference value of the calibration set and the real value of the best model obtained was significant (R _c =0.903), robustness (R _cv =0.859) and prediction effect (R _p =0.931) were good, and RMSEC (7.01), Both RMSECV (8.39) and RMSEP (5.98) are small. The graphical effect of the correlation between the reference value and the predicted value of the calibration set and prediction set is shown in Figure 2. The optimal PLS model established by Fourier transform near-infrared spectroscopy can be used for accurate and rapid quantitative detection of AST activity in serum.

8. Conclusion

The invention establishes a quantitative model by combining Fourier transform near-infrared spectroscopy with chemometrics technology, which can be used for accurate and rapid quantitative detection of AST activity in serum. Compared with traditional methods, this method does not require complex sample pretreatment and reagents, is fast, non-destructive, pollution-free, low-cost, easy to operate, and has high accuracy. It is suitable for the determination of AST activity in human serum in medical institutions.

Claims (8)

1. human serum Mid-Heaven Gate winter histidine amino group transferase (AST) is active quick, a lossless detection method, mainly comprise the following steps:

(1) collection of blood sample;

(2) preparation of blood serum sample;

(3) reference value of AST activity in blood serum sample is measured;

(4) NIR transmittance spectroscopy of blood serum sample is measured with ft-nir spectrometer;

(5) process by the data of chemometric techniques to institute's light-metering spectrum, filter out optimum preprocess method;

(6) rational near infrared spectrum modeling region is chosen;

(7) spectroscopic data after process is associated with the reference value of AST activity, sets up the Quantitative Prediction Model of AST activity in serum, and institute's established model is verified;

(8) NIR transmittance spectroscopy of acquisition and processing test serum sample in the same way, by the activity of AST in institute's established model prediction test serum.

2. the method for claim 1, it is characterized in that: in described step (1), the preparation method of blood serum sample is as follows: in the satisfactory place of medical institutions, venous blood is gathered by regulation appropriate by medical personnel, be placed in the heparin tube containing set accelerator, treat blood clotting layering under room temperature, get supernatant and be blood serum sample.

3. the method for claim 1, is characterized in that: in described step (3), the assay method of the active reference value of AST is the method that current medical institutions commonly use, and includes but not limited to adopt automatic clinical chemistry analyzer to detect.

4. the method for claim 1, is characterized in that: in described step (4), NIR transmittance spectroscopy measuring method measures with reference to the working specification of instrument.Spectrometer includes but not limited to ft-nir spectrometer; Measurement pattern includes but not limited to transmission mode; Measurement parameter includes but not limited to spectral measurement ranges, resolution and scanning times.

5. the method for claim 1, is characterized in that: in described step (5) preprocess method of near infrared spectrum data can be but be not limited to untreated, first order derivative, second derivative, Norris are level and smooth, Savitzky-Golay is level and smooth, average centralization, calibrate in one or more.

6. the method for claim 1, is characterized in that: in described step (6), the software of the choice for use of Rational Model SPECTRAL REGION includes but not limited to TQAnalyst8.0 software; Selection mode includes but not limited to that software is selected automatically, artificial selection, software are selected to combine with artificial selection automatically.

7. the method for claim 1, is characterized in that: include but not limited to TQAnalyst software and MATLAB software for the software of data processing in described step (7); Quantitative modeling algorithm include but not limited in partial least square method (PLS), principal component regression method (PCR) and artificial neural network method (ANN) one or more.

8. the method for claim 1, it is characterized in that: the middle institute of described step (8) positive model for school building is to when in testing sample, AST activity is predicted, the preprocess method of testing sample spectral measurement method used, spectral measurement parameter, spectroscopic data all should be consistent with calibration set sample with spectrum modeling region.