Construction and application of instant ginseng near-infrared quantitative correction model for maintaining eyesight

Abstract

The invention relates to the field of infrared detection, in particular to a construction method and application of a near-infrared quantitative correction model of instant ginseng for maintaining eyesight. The method specifically comprises the following steps: selecting instant ginseng with known ginsenoside Rg1, Rb1, Re and moisture content; collecting near infrared spectrum data of the instant ginseng; spectral information under the characteristic wave band is selected to be associated with ginsenosides Rg1, Rb1 and Re of the instant ginseng and the moisture content, and a quantitative standard model is established by adopting a partial least square method and a cross-validation method; and carrying out near infrared spectrum scanning on an unknown instant ginseng sample, and importing the unknown instant ginseng sample into the established quantitative correction model to obtain ginsenosides Rg1, Rb1 and Re and the moisture content of the unknown instant ginseng sample. The near-infrared prediction model is used for detecting ginsenoside Rg1, Rb1, Re and moisture content in the instant ginseng, has the advantages of being easy to operate, high in analysis speed, high in prediction result accuracy and the like, provides an experimental basis for inspection of the quality of the instant ginseng, and has remarkable application value.

Description

Technical Field

[0001] The invention relates to the field of infrared detection, and in particular to the construction and application of a near-infrared quantitative correction model of instant ginseng for maintaining vision. Background Art

[0002] An instant ginseng product for maintaining vision is a product that is developed for people with vision problems. The main ingredient of ginseng is ginsenosides. Currently, the most commonly used method for detecting ginsenosides is high performance liquid chromatography, which has the disadvantages of long experimental cycle, cumbersome operation, high cost, and environmental pollution. Therefore, it is of great practical significance to find a new detection method that is fast, simple, safe, environmentally friendly, and non-destructive.

[0003] As a fast, non-destructive, green analytical technology, near-infrared (NIR) spectroscopy has the characteristics of rapid analysis, simple sample processing, and no need to consume reagents. In recent years, as an indirect analytical technology, near-infrared spectroscopy has been gradually applied to food, medicine and other fields. The fifth-generation acousto-optic tunable (AOTF) near-infrared spectrometer, this new type of near-infrared spectrometer has the characteristics of small size, easy to carry, stable performance, strong environmental adaptability, etc. It can be expanded to online production and outdoor sites, and can meet the application needs of comprehensive quality control of food production processes. Therefore, near-infrared spectroscopy technology is applied to the rapid quality detection of an instant ginseng for maintaining vision, so as to ensure the quality of the product and achieve the purpose of rapid and efficient quality control. Summary of the invention

[0004] In view of this, the present invention provides a construction and application of a near-infrared quantitative calibration model for instant ginseng for maintaining vision. The near-infrared determination method is used to quickly determine the content of ginsenosides Rg1, Rb1, Re and water in instant ginseng. The method is simple to operate, has a fast analysis speed, does not require reagents, is green and environmentally friendly, and has a high accuracy in prediction results. It can provide a theoretical basis and analytical means for the rapid detection and evaluation of instant ginseng ingredients and product quality control.

[0005] In order to achieve the above-mentioned object of the invention, the present invention provides the following technical solutions:

[0006] The present invention provides a method for constructing a near-infrared quantitative correction model of instant ginseng for maintaining eyesight. Instant ginseng is taken, ginsenosides Rg1, Rb1, Re and water are taken as key quality control indicators, and high performance liquid chromatography detection and near-infrared spectrum detection are performed to obtain corresponding high performance liquid chromatography detection results and near-infrared spectrum results. A near-infrared quantitative correction model of ginsenosides Rg1, Rb1, Re and water in instant ginseng is obtained by partial least squares method and cross-validation method.

[0007] In some specific embodiments of the present invention, a pulverizer is used for pulverization, and the sieve mesh number is 65-120 meshes.

[0008] In some specific embodiments of the present invention, ginsenosides Rg1, Rb1, and Re are determined by high performance liquid chromatography, and the conditions of the high performance liquid chromatography are:

[0009] The chromatographic column is ACE-C18 (250mm*4.6mm, 5μm); octadecylsilane bonded silica gel is used as the filler; acetonitrile is used as the mobile phase A, water is used as the mobile phase B, and the gradient elution is 0-35min, 19% (A), 35-55min, 19% (A)-29% (A), 55-70min, 29% (A), 70-85min, 29% (A)-40% (A); the detection wavelength is 200-210nm; the flow rate is 0.5-1.5mL / min; the column temperature is 20-40℃; the preferred detection wavelength is 203nm; the flow rate is 1mL / min; the column temperature is 30℃; the theoretical plate number calculated based on the ginsenoside Rg1 peak should be no less than 6000.

[0010] In some specific embodiments of the present invention, the method for preparing the reference solution in the HPLC detection is as follows: taking ginsenoside Rg1 reference, ginsenoside Re reference and ginsenoside Rb1 reference, and adding methanol to prepare a reference solution. Preferably, accurately weigh ginsenoside Rg1 reference, ginsenoside Re reference and ginsenoside Rb1 reference, add methanol to prepare a mixed solution containing 0.2 mg of each per 1 mL, and shake well to obtain.

[0011] In some specific embodiments of the present invention, the preparation method of the test solution in the high performance liquid chromatography detection is: take about 1g of the powder of the product (passed through a No. 4 sieve), accurately weigh it, place it in a Soxhlet extractor, add chloroform and heat to reflux for 2 to 4 hours, discard the chloroform liquid, evaporate the solvent from the residue, move it into a 100mL conical flask together with the filter paper tube, accurately add 30 to 70mL of saturated n-butanol with water, plug it, leave it overnight, ultrasonically treat it, filter it, discard the initial filtrate, accurately measure 20 to 50mL of the subsequent filtrate, place it in an evaporating dish and evaporate it to dryness, dissolve the residue in methanol and transfer it to a 2 to 10mL volumetric flask, add methanol to dilute to the scale, shake it well, filter it, and take the subsequent filtrate to obtain.

[0012] In some specific embodiments of the present invention, the step of adding chloroform and heating under reflux is performed for 2 to 4 hours, preferably, adding chloroform and heating under reflux is performed for 3 hours.

[0013] In some specific embodiments of the present invention, the precisely added water-saturated n-butanol is 30 to 70 mL, preferably, the precisely added water-saturated n-butanol is 50 mL.

[0014] In some specific embodiments of the present invention, the ultrasonic treatment is carried out at 250W, 50kHz for 20 to 40 minutes, preferably, the ultrasonic treatment is carried out for 30 minutes.

[0015] In some specific embodiments of the present invention, the near-infrared spectroscopy detection includes the steps of spectral acquisition and spectral preprocessing;

[0016] The spectrum acquisition is as follows: instant ginseng is taken, crushed and sieved, near-infrared spectrum is collected at 25°C±5°C using a diffuse reflection built-in light source, air is used as a reference background, the number of scans is 200-400, preferably 300, the scanning spectrum range is 1100-2300nm, the wavelength increment is 1-3nm, preferably, the wavelength increment is 2nm, the sample thickness is 1-2cm, preferably, the sample thickness is 1cm;

[0017] In some specific embodiments of the present invention, the spectrum preprocessing is: converting the format of reflectance to absorbance, and then performing first-order differential 9-point smoothing method processing;

[0018] In some specific embodiments of the present invention, the near-infrared quantitative calibration model for instant ginseng is obtained by using partial least squares method and cross-validation method, using The Unscrambler quantitative analysis software, associating the preprocessed spectral data with the sample content data, and establishing a calibration model for each indicator.

[0019] The method of obtaining the instant ginseng near-infrared quantitative correction model also includes using the spectral influence value Leverage and chemical value error Residual statistical test to eliminate abnormal values ​​of the NIR spectrum and the measurement results to determine the final NIR quantitative correction model.

[0020] In some specific embodiments of the present invention, the spectral influence value Leverage and the chemical value error Residual are used for statistical testing, the abnormal values ​​of the NIR spectrum and the measurement results of each index are eliminated, and the main parameters of the model are: the internal and external verification determination coefficient (R 2 ), root mean square error of correction (RMSEC), and root mean square error of prediction (RMSEP) were used to obtain the near-infrared quantitative correction model of each index.

[0021] In some specific embodiments of the present invention, in the NIR model of ginsenoside Rg1 in instant ginseng, the internal validation linear equation is Y=0.9889X+0.0026, and the external validation linear equation is Y=0.9121X+0.0214, where X is the assay value, Y is the predicted value, and the internal validation determination coefficient R 2 =0.9924, external validation determination coefficient R 2=0.9646, RMSEC=0.0082, RMSEP=0.0208;

[0022] In some specific embodiments of the present invention, in the NIR model of ginsenoside Rb1 in instant ginseng, the internal validation linear equation is Y=0.9845X+0.0069, and the external validation linear equation is Y=0.9316X+0.0235, where X is the assay value, Y is the predicted value, and the internal validation determination coefficient R 2 =0.9922, external validation determination coefficient R 2 =0.9691, RMSEC=0.0653, RMSEP=0.1297;

[0023] In some specific embodiments of the present invention, in the ReNIR model of ginsenosides in instant ginseng, the internal validation linear equation is Y=0.9860X+0.0059, and the external validation linear equation is Y=0.9227X+0.0294, where X is the assay value, Y is the predicted value, and the internal validation determination coefficient R 2 =0.9930, external validation determination coefficient R 2 =0.9731, RMSEC=0.0398, RMSEP=0.0781;

[0024] In some specific embodiments of the present invention, in the NIR model of water content in instant ginseng, the internal validation linear equation is Y=0.9661X+0.3272, and the external validation linear equation is Y=0.9064X+0.8908, where X is the assay value, Y is the predicted value, and the internal validation determination coefficient R 2 =0.9829, external validation determination coefficient R 2 =0.9581, RMSEC=0.3268, RMSEP=0.5094;

[0025] In some specific embodiments of the present invention, the ginsenosides Rg1, Rb1, Re and moisture content involved in the modeling have a good correlation with their NIR spectra, and the established NIR detection model has good performance.

[0026] On the basis of the above technical scheme, the present invention also provides an application of an instant ginseng near-infrared quantitative correction model for maintaining vision. The instant ginseng near-infrared quantitative correction model is obtained according to the construction method, and the near-infrared spectrum data of the instant ginseng sample to be tested is imported into the instant ginseng near-infrared quantitative correction model to obtain the content of ginsenosides Rg1, Rb1, Re and water in the sample to be tested.

[0027] The present invention discloses a method for constructing a near-infrared quantitative calibration model for instant ginseng for maintaining vision and its application. The method is specifically as follows: (1) selecting instant ginseng with known ginsenosides Rg1, Rb1, Re and water content for standby use; (2) collecting near-infrared spectral data of instant ginseng; (3) selecting spectral information under characteristic bands and correlating them with the known ginsenosides Rg1, Rb1, Re and water content of the instant ginseng, and using partial least squares method and cross-validation method to establish a quantitative standard model; (4) performing near-infrared spectral scanning on unknown instant ginseng samples, selecting spectral information under characteristic bands, and importing it into the established quantitative calibration model to obtain ginsenosides Rg1, Rb1, Re and water content of unknown instant ginseng samples. The near-infrared prediction model is used for the detection of ginsenosides Rg1, Rb1, Re and water content in instant ginseng, which has the advantages of simple operation, fast analysis speed, high accuracy of prediction results, etc. The present invention provides an experimental basis for the inspection of the quality of instant ginseng and has significant application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below.

[0029] Figure 1 HPLC charts of ginsenoside Rg1, Rb1, and Re reference substances are shown;

[0030] Figure 2 Shows the HPLC chart of the instant ginseng test product for maintaining vision;

[0031] Figure 3 The original NIR spectrum of instant ginseng for maintaining vision is shown;

[0032] Figure 4 The NIR pre-processed spectrum of instant ginseng for maintaining vision is shown;

[0033] Figure 5 The results of NIR modeling of ginsenoside Rg1, an instant ginseng for maintaining vision, are shown;

[0034] Figure 6 The results of NIR modeling of ginsenoside Rb1 of instant ginseng for maintaining vision are shown;

[0035] Figure 7 The results of ReNIR modeling of instant ginseng-ginsenosides for maintaining vision are shown;

[0036] Figure 8 Shows the NIR modeling results of instant ginseng-moisture for maintaining vision.

[0037] For the modeling results:

[0038] The pink color is the verification value predicted by the model that does not include this point, and the blue color is the correction value predicted by the model built with all the samples in the correction set for each sample in the correction set. Both are simulations of the actual prediction ability of the model during the interactive verification process, but the pink verification value is closer to the external prediction. When actually optimizing the model, the external determination coefficient is mainly used. DETAILED DESCRIPTION

[0039] The present invention discloses the construction and application of an instant ginseng near-infrared quantitative correction model for maintaining vision. Those skilled in the art can refer to the content of this article and appropriately improve the process parameters to achieve it. It is particularly important to point out that all similar substitutions and modifications are obvious to those skilled in the art, and they are all considered to be included in the present invention. The methods and applications of the present invention have been described through preferred embodiments, and relevant personnel can obviously modify or appropriately change and combine the methods and applications described herein without departing from the content, spirit and scope of the present invention to implement and apply the technology of the present invention.

[0040] The present invention provides a construction and application method of an instant ginseng near-infrared quantitative correction model for maintaining vision, which adopts the following steps:

[0041] (1) Collecting the near-infrared spectrum of instant ginseng: Under room temperature conditions, crush and sieve the powder of instant ginseng and put it into the sample tank. The sample thickness is required to be 1 to 2 cm. With air as the background, the near-infrared spectrum is collected in diffuse reflection and ratio mode. The spectral range is 1100 to 2300 nm; the wavelength averaging times are 200 to 400, and the wavelength increment is 1 to 3 nm.

[0042] (2) Detection of ginsenosides Rg1, Rb1 and Re by high performance liquid chromatography: Accurately weigh ginsenoside Rg1 reference substance, ginsenoside Re reference substance and ginsenoside Rb1 reference substance, add methanol to make a mixed solution containing 0.2 mg of each per 1 mL, shake well, and obtain. Take about 1 g of the powder of this product (passed through a No. 4 sieve), accurately weigh it, put it in a Soxhlet extractor, add chloroform and heat under reflux for 2 to 4 hours, discard the chloroform solution, evaporate the solvent from the residue, transfer it together with the filter paper tube into a 100 mL conical flask, accurately add 30 to 70 mL of saturated n-butanol with water, plug it, leave it overnight, ultrasonically treat it, filter it, discard the initial filtrate, accurately measure 20 to 50 mL of the subsequent filtrate, evaporate it to dryness in an evaporating dish, dissolve the residue in methanol and transfer it to a 2 to 10 mL volumetric flask, add methanol to dilute to the mark, shake well, filter it, and take the subsequent filtrate to obtain. The prepared samples were used to determine the contents of ginsenosides Rg1, Rb1, and Re by HPLC to obtain the chemical values ​​of the contents of ginsenosides Rg1, Rb1, and Re in the samples.

[0043] Determination of moisture content: Determine according to the moisture determination method (General Rule 0832 Method 2 Drying Method).

[0044] (3) The original spectra were preprocessed using the spectral processing program in the SNAP software and the first-order derivative 9-point smoothing method (Savitzky-Golay) to eliminate the spectral baseline offset and drift caused by factors such as color differences. The preprocessed spectral data were associated with the sample HPLC content data using the Unscrambler quantitative analysis software, using the partial least squares method and cross-validation method to establish a correction model for each indicator. Statistical tests such as spectral leverage and chemical value error residual were used to eliminate outliers in the NIR spectrum and measurement results, and the NIR detection model was determined by screening and optimizing the main parameters of the model.

[0045] (4) performing near infrared spectroscopy scanning on the unknown instant ginseng sample according to the method described in step (3), and importing the obtained characteristic spectrum information into the established quantitative calibration model to obtain the ginsenosides Rg1, Rb1, Re and water content of the unknown instant ginseng.

[0046] Preferably, the main parameter in step (3) is: internal and external validation determination coefficient (R 2 ), prediction standard deviation (RMSEP).

[0047] Preferably, the sample set contains no less than 50 samples.

[0048] The raw materials, reagents and auxiliary materials used in the construction and application of the instant ginseng quality rapid detection method provided by the present invention can all be purchased from the market.

[0049] The present invention will be further described below in conjunction with embodiments:

[0050] Example 1

[0051] (1) Determination of ginsenosides Rg1, Rb1, Re and moisture content in instant ginseng

[0052] HPLC detection of ginsenoside Rg1, Rb1, Re content determination:

[0053] The chromatographic column is ACE-C18 (250mm*4.6mm, 5μm); octadecylsilane bonded silica gel is used as the filler; acetonitrile is used as the mobile phase A, water is used as the mobile phase B, and the gradient elution is 0-35min, 19% (A), 35-55min, 19% (A)-29% (A), 55-70min, 29% (A), 70-85min, 29% (A)-40% (A); the detection wavelength is 203nm; the flow rate is 1.0mL / min; the column temperature is 30°C; the theoretical plate number calculated based on the ginsenoside Rg1 peak should not be less than 6000.

[0054] Accurately weigh ginsenoside Rg1 reference substance, ginsenoside Re reference substance and ginsenoside Rb1 reference substance, add methanol to make a mixed solution containing 0.2 mg of each substance per 1 mL, shake well, and obtain.

[0055] Take about 1g of the powder of this product (passed through a No. 4 sieve), weigh it accurately, put it in a Soxhlet extractor, add chloroform and heat to reflux for 3 hours, discard the chloroform solution, evaporate the solvent from the residue, move it into a 100mL conical flask together with the filter paper tube, accurately add 50mL of saturated n-butanol with water, plug it tightly, leave it overnight, treat it with ultrasound, filter it, discard the initial filtrate, accurately measure 25mL of the subsequent filtrate, put it in an evaporating dish and evaporate it to dryness, dissolve the residue in methanol and transfer it to a 5mL volumetric flask, add methanol to dilute to the scale, shake it well, filter it, and take the subsequent filtrate to obtain it. HPLC was used to determine the content of ginsenosides Rg1, Rb1, and Re, and the chemical values ​​of the content of ginsenosides Rg1, Rb1, and Re in instant ginseng were obtained.

[0056] Moisture content determination:

[0057] The chemical value of the moisture content in instant ginseng is obtained by measuring it according to the moisture determination method (General Rule 0832 Method 2 Drying Method).

[0058] The HPLC spectrum of the reference substance is as follows Figure 1 As shown; the HPLC spectrum of the test sample and the determination results are shown Figure 2 , as shown in Table 1.

[0059] Table 1 Results of various indicators of instant ginseng for maintaining vision

[0060]

[0061]

[0062]

[0063] (2) Collecting the near-infrared spectrum of instant ginseng: Under room temperature, instant ginseng was crushed and the powder passed through an 80-mesh sieve was placed in a sample slot. The sample loading thickness was required to be 1 cm. With air as the background, the near-infrared spectrum was collected in diffuse reflection mode and ratio mode. The spectral range was 1100-2300 nm. The wavelength average number was 300, the wavelength increment was 2 nm, and each sample was measured 3 times to obtain the instant ginseng spectrum. Figure 3 shown.

[0064] (3) Using the spectrum processing program in the SNAP software, the original spectrum was preprocessed using the first-order derivative 9-point smoothing method (Savitzky-Golay) to eliminate the spectral baseline shift and drift caused by factors such as color differences. Figure 4As shown. Unscrambler quantitative analysis software, partial least squares method and cross-validation method were used to associate the pre-processed spectral data with the sample content data to establish the correction model for each index. Statistical tests such as spectral influence value Leverage and chemical value error Residual were used to eliminate the abnormal values ​​of NIR spectra and measurement results, and the main parameters of the model were screened and optimized to determine the NIR detection model. Figures 5 to 8 shown.

[0065] Internal and external validation determination coefficient (R 2 ), calibration standard deviation (RMSEC), prediction standard deviation (RMSEP), etc., to evaluate the model quality. The specific calculation formula is as follows:

[0066]

[0067] In the formula, y i,actual is the measured value of the reference method for the i-th sample; is the average value of the reference method determination of all samples in the calibration set or validation set; i,predicted is the predicted value of the i-th sample in the calibration set or validation set prediction process; n is the number of samples in the calibration set or validation set. The closer R is to 1, the better the regression should be.

[0068]

[0069] In the formula, y i,actual is the measured value of the reference method for the i-th sample; y i,predicted is the predicted value of the i-th sample in the calibration set using the established model; n is the number of samples in the calibration set. The smaller the RMSEC value, the better.

[0070]

[0071] In the formula, y i,actual is the measured value of the reference method for the i-th sample; y i,predicted is the spectral method prediction value of the i-th sample in the validation set prediction process; m is the number of samples in the validation set. The smaller the RMSEP value, the better, and RMSEC should be smaller than the RMSEP value.

[0072] Results: In the ginsenoside Rg1NIR model, the internal validation linear equation was Y = 0.9889X + 0.0026, and the external validation linear equation was Y = 0.9121X + 0.0214, where X was the test value, Y was the predicted value, and the internal validation determination coefficient R 2 =0.9924, external validation determination coefficient R 2=0.9646, RMSEC=0.0082, RMSEP=0.0208; in the ginsenoside Rb1NIR model, the internal validation linear equation is Y=0.9845X+0.0069, the external validation linear equation is Y=0.9316X+0.0235, and the internal validation determination coefficient R 2 =0.9922, external validation determination coefficient R 2 =0.9691, RMSEC=0.0653, RMSEP=0.1297; in the ginsenoside ReNIR model, the internal validation linear equation is Y=0.9860X+0.0059, the external validation linear equation is Y=0.9227X+0.0294, and the internal validation determination coefficient R 2 =0.9930, external validation determination coefficient R 2 =0.9731, RMSEC=0.0398, RMSEP=0.0781; in the moisture NIR model, the internal validation straight line method is Y=0.9661X+0.3272, the external validation straight line equation is Y=0.9064X+0.8908, and the internal validation determination coefficient R 2 =0.9829, external validation determination coefficient R 2 =0.9581, RMSEC=0.3268, RMSEP=0.5094.

[0073] The above parameters show that the contents of various indicators of instant ginseng for maintaining vision involved in the modeling are well correlated with its NIR spectrum, and the established NIR detection model has good performance.

[0074] (4) Perform near infrared spectroscopy scanning on the unknown instant ginseng sample according to the methods described in steps (2) and (3), and import the obtained characteristic spectrum information into the established quantitative calibration model to obtain the ginsenosides Rg1, Rb1, Re and water content of the unknown instant ginseng. Verify the consistency of the spectrum acquisition parameter settings of the sample and the modeling sample, obtain the prediction results of each index, calculate the deviation between the predicted value and the measured value, verify the accuracy of the model, and the verification results are shown in Table 2. The deviation value calculation formula is as follows:

[0075]

[0076] Table 2 Verification results of the NIR model for maintaining vision with instant ginseng

[0077]

[0078]

[0079] As shown in Table 2, the average deviations between the predicted values ​​and the HPLC measured values ​​of ginsenosides Rg1, Rb1, and Re in the five batches of verification samples were 7.02%, 7.80%, and 9.39%, respectively, and the mean deviation of each indicator was less than 10%; the average deviation between the predicted value and the measured value of water content was 2.61%, and the mean deviation was less than 5%.

[0080] It can be seen that there is a good correlation between the predicted values ​​of each indicator NIR detection model and the measured values ​​of the standard detection method, and the measurement accuracy is high. The NIR spectral model can be used to quickly detect and analyze ginsenosides Rg1, Rb1, Re and water content in instant ginseng for maintaining vision.

[0081] (5) The established quantitative models of ginsenosides Rg1, Rb1, Re and water were externally validated, and the results showed that the prediction results of NIR technology were basically consistent with the results of traditional HPLC, with no significant deviation. NIR rapid detection has high accuracy and can replace liquid phase analysis. A comparison was also made in terms of application and related consumables, as shown in Table 3.

[0082] Table 3 Comparison of detection technologies

[0083]

[0084] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for constructing a near-infrared quantitative calibration model of instant ginseng for maintaining vision, characterized in that: Instant ginseng was taken, and ginsenosides Rg1, Rb1, Re and water were used as key quality control indicators. The corresponding HPLC test results and near-infrared spectroscopy results were obtained by HPLC and near-infrared spectroscopy. The near-infrared quantitative calibration model of ginsenosides Rg1, Rb1, Re and water in instant ginseng was obtained by partial least squares method and cross-validation method. Ginsenosides Rg1, Rb1 and Re were determined by high performance liquid chromatography, and the conditions of the high performance liquid chromatography were as follows: The chromatographic column is ACE-C18 (250mm*4.6mm, 5μm); octadecylsilane bonded silica gel is used as the filler; acetonitrile-water is used as the mobile phase; the detection wavelength is 200-210nm; the column temperature is 20-40℃; the flow rate is 0.5-1.5mL / min; the theoretical plate number calculated based on the ginsenoside Rg1 peak should be no less than 6000. The method for preparing the reference solution in the high performance liquid chromatography detection is as follows: taking ginsenoside Rg1 reference substance, ginsenoside Re reference substance and ginsenoside Rb1 reference substance, and adding methanol to prepare the reference solution. The preparation method of the test solution in the high performance liquid chromatography detection is as follows: take about 1g of the powder of the product (passed through a No. 4 sieve), accurately weigh it, place it in a Soxhlet extractor, add chloroform and heat to reflux for 2 to 4 hours, discard the chloroform liquid, evaporate the solvent from the residue, move it into a 100mL conical flask together with the filter paper tube, accurately add 30 to 70mL of saturated n-butanol with water, plug it, leave it overnight, ultrasonically treat it, filter it, discard the initial filtrate, accurately measure 20 to 50mL of the subsequent filtrate, place it in an evaporating dish and evaporate it to dryness, dissolve the residue in methanol and transfer it to a 2 to 10mL volumetric flask, add methanol to dilute to the scale, shake it well, filter it, and take the subsequent filtrate to obtain it. The moisture determination adopts the second method (drying method) of the moisture determination method in Part 0832 of the 2020 edition of the Chinese Pharmacopoeia. The near-infrared spectrum detection includes the steps of spectrum acquisition and spectrum preprocessing; The spectrum acquisition is as follows: instant ginseng is taken, crushed and sieved, and near-infrared spectrum is acquired at 25°C±5°C using a diffuse reflection built-in light source, with air as a reference background, the number of scans is 200-400, the scanning spectrum range is 1100-2300nm, the wavelength increment is 1-3nm, and the sample thickness is 1-2cm; The spectrum preprocessing is: converting the original spectrum collected by the spectrum into a format of reflectance to absorbance, and then performing a first-order differential 9-point smoothing method; The method of obtaining the near-infrared quantitative calibration model of instant ginseng is specifically as follows: using The Unscrambler quantitative analysis software, using partial least squares method and cross-validation method, the pre-processed spectral data is associated with the sample content data, and the calibration model of each index is established; The method of obtaining the instant ginseng near-infrared quantitative correction model also includes using the spectral influence value Leverage and chemical value error Residual statistical test to eliminate abnormal values ​​of the NIR spectrum and the measurement results to determine the final NIR quantitative correction model.

2. An application of an instant ginseng near-infrared quantitative correction model for maintaining vision, characterized in that: A near-infrared quantitative calibration model for instant ginseng is obtained according to the construction method as described in claim 1, and the near-infrared spectrum data of the instant ginseng sample to be tested is imported into the instant ginseng near-infrared quantitative calibration model to obtain the content of ginsenosides Rg1, Rb1, Re and water in the sample to be tested.

Images