Fluorescence calibration method

Abstract

The invention provides a fluorescence calibration method, which comprises the following steps of: 1, preparing a plurality of groups of same fluorescent dyes with different concentrations, respectively injecting the fluorescent dyes into test holes of a test chip, and irradiating the test chip filled with the fluorescent dyes with different concentrations by using exciting light; step 2, acquiring excitation fluorescence of the test chip by using optical acquisition equipment to obtain an acquired fluorescence grayscale image of the test chip, processing the fluorescence grayscale image, and calculating an average pixel value of the fluorescence grayscale image as an acquired fluorescence value; carrying out image processing on the fluorescent grayscale images of the test chips filled with the fluorescent dyes with different concentrations to obtain a plurality of groups of collected fluorescent values; 3, the calibration fluorescence value is the product of the collected fluorescence value and the calibration coefficient, a model for solving the minimum error between the variation coefficient of the calibration fluorescence values of the fluorescent dyes with different concentrations and the expected variation coefficient is built, and the model is solved to obtain the calibration coefficient. The method has the advantages of simplicity in operation and high accuracy.

Description

Technical Field

[0001] This invention belongs to the field of biomolecular detection technology, and in particular relates to a fluorescence calibration method. Background Technology

[0002] Polymerase chain reaction (PCR) is a molecular biology technique used to amplify specific DNA fragments. Real-time quantitative PCR (qPCR) utilizes fluorescence signals to achieve accurate, rapid, and comprehensive detection of large amounts of information about cells, proteins, nuclei, and other biological components.

[0003] Traditional fluorescence PCR detection systems require a coaxial arrangement between the excitation source and the fluorescence receiving system. This vertical excitation necessitates a collimation system that is more expensive and larger in size to filter out the emitted excitation light entering the fluorescence filter while allowing fluorescence to pass through. To address the limitations in device layout caused by these issues, using high-reflectivity silicon-based chips instead of traditional polypropylene PCR tubes for fluorescence experiments has become a new research direction. A laser source is used to illuminate the silicon-based chip at a specific angle. Since silicon does not exhibit fluorescence, only the test area on the silicon-based chip is excited to fluoresce. The excited fluorescence can then be received and tested by an image receiving system positioned above the silicon-based chip.

[0004] However, since the excitation light is irradiated onto the silicon chip at a certain angle, the excitation light will not be uniformly irradiated onto every area of ​​the silicon chip. This leads to uneven irradiation of the excitation light, resulting in large differences in fluorescence values ​​in different areas of the silicon chip. Consequently, the linear curve obtained when plotting the fluorescence intensity is poor.

[0005] Therefore, it is necessary to design a fluorescence correction method that can effectively compensate for the problem of inaccurate fluorescence value measurement caused by uneven excitation light irradiation. Summary of the Invention

[0006] This invention provides a fluorescence calibration method that can effectively calibrate the measured fluorescence values ​​so that the measured fluorescence values ​​meet the requirements.

[0007] Other objects and advantages of the present invention can be further understood from the technical features disclosed herein.

[0008] To achieve one, some, or all of the above objectives or other objectives, the present invention provides a fluorescence calibration method comprising the following steps: Step 1: Preparing multiple concentrations of the same fluorescent dye and injecting them into the test holes of a test chip, and irradiating the test chips containing different concentrations of fluorescent dye with excitation light; Step 2: Using an optical acquisition device to acquire the excitation fluorescence of the test chip, obtaining the acquired fluorescence grayscale image of the test chip, processing the fluorescence grayscale image, and calculating the average pixel value of the fluorescence grayscale image as the acquired fluorescence value; performing image processing on the fluorescence grayscale images of the test chips containing different concentrations of the same fluorescent dye respectively to obtain multiple sets of acquired fluorescence values; Step 3: The calibration fluorescence value is the product of the acquired fluorescence value and the calibration coefficient, constructing a model to solve for the minimum error between the coefficient of variation and the expected coefficient of variation of the calibration fluorescence value of different concentrations of fluorescent dye, and solving the model to obtain the calibration coefficient. The beneficial effects of this technical solution are as follows: This invention processes the fluorescence grayscale image acquired in a fluorescence experiment, calculates the average pixel value of the acquired fluorescence grayscale image to obtain the acquired fluorescence value, constructs a model minimizing the error between the coefficient of variation of the corrected acquired fluorescence value and the expected coefficient of variation, solves this model to obtain the correction coefficient, and uses this correction coefficient to calibrate the fluorescence value acquired by the excitation light irradiating the chip, which can effectively compensate for the acquired fluorescence value, ensuring that the acquired fluorescence value meets the requirements of the fluorescence experiment. The method of this invention has the advantages of simple operation and high accuracy.

[0009] The fluorescent dyes consist of at least three groups.

[0010] The concentrations of the multiple groups of fluorescent dyes were diluted in a gradient.

[0011] The test chip is a silicon-based test chip, and the silicon-based test chip is provided with a plurality of test holes, which are evenly arranged on the test chip.

[0012] Step 2, the fluorescence grayscale image processing, further includes dividing the fluorescence grayscale image into several regions, calculating the average pixel value of the fluorescence grayscale image in each region as the collected fluorescence value of each region; constructing a model to solve for the minimum error between the coefficient of variation and the expected coefficient of variation of the calibration fluorescence value at different concentrations in each region, and solving the model to obtain the calibration coefficient for each region.

[0013] The method for dividing the fluorescence grayscale image into several regions in step 2 is as follows: several regions of equal size are selected on the test chip, and each region contains one test hole.

[0014] The average pixel value of the fluorescent grayscale image in each region is calculated using an image processing algorithm.

[0015] The constructed model for minimizing the error between the coefficient of variation and the expected coefficient of variation when solving for calibration fluorescence values ​​at different concentrations is as follows:

[0016]

[0017] K represents the amount of fluorescent dye at different concentrations, cv n Let be the coefficient of variation for the nth concentration of fluorescent dye, and e be the expected coefficient of variation. The derivation formula for cv is:

[0018]

[0019] Data nj Here are the fluorescence values ​​after calibration for the nth concentration of fluorescent dye; std is the standard deviation, mean is the mean, and Data... nj The derivation formula is as follows:

[0020]

[0021] Where x m a represents the calibration coefficient of the m-th test hole on the test chip. nm This represents the fluorescence value collected in the m-th test well of the n-th concentration fluorescent dye, where m represents the number of test wells on the test chip.

[0022] The expected coefficient of variation range is: 2% ≤ e ≤ 3%.

[0023] The model for minimizing the error between the coefficient of variation and the expected coefficient of variation of calibrated fluorescence values ​​at different concentrations is solved using the quasi-Newton method.

[0024] Compared with existing technologies, the main advantages of this invention include: This invention acquires fluorescence grayscale images, processes the pixel values ​​of the acquired fluorescence grayscale images to obtain the acquired fluorescence values, and then calibrates the acquired fluorescence values, offering the advantage of a simple calibration method. Furthermore, by constructing a model for calibration, each area on the test chip can be calibrated, ensuring the accuracy of the fluorescence experimental data from the test chip.

[0025] To make the above and other objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the specific embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 This is a schematic diagram of the fluorescence experimental detection system of the present invention.

[0028] Figure 2 This is a flowchart of the fluorescence calibration coefficient calculation in Embodiment 1 of the present invention.

[0029] Figure 3 The image shows the fluorescence intensity after injection of different concentrations of fluorescent dye in Example 1.

[0030] In the diagram: 1. Temperature control module; 2. Test chip; 3. Imaging lens; 4. Fluorescent filter; 5. CMOS image sensor; 6. FPGA image acquisition board; 7. Excitation light source; 8. Collimating lens; 9. Dichroic mirror; 10. LED light source. Detailed Implementation

[0031] The foregoing and other technical contents, features, and effects of the present invention will be clearly presented in the following detailed description of a preferred embodiment with reference to the accompanying drawings. The directional terms mentioned in the following embodiments, such as up, down, left, right, front, or back, are merely for reference to the accompanying drawings. Therefore, the directional terms used are for illustrative purposes and not for limiting the present invention.

[0032] Example 1

[0033] To better illustrate the present invention, Example 1 uses three concentrations of fluorescent dyes to conduct fluorescence experiments and calculate the corresponding fluorescence values. The fluorescent dyes selected in Example 1 are FAM, ROX, HEX, and CY5 dyes, and their names and characteristics are shown in Table 1 below.

[0034] Table 1: Four types of fluorescent dyes and their properties

[0035]

[0036] See Figure 2 The specific steps for fluorescence value calibration are as follows:

[0037] Step 1: Prepare three concentrations of fluorescent dye. The three concentrations of fluorescent dye are diluted in a gradient. In Example 1, the three concentrations of fluorescent dye are high-concentration (0.5 L mol / L), medium-concentration (0.25 L mol / L), and low-concentration (0.125 L mol / L). The fluorescent dyes in Example 1 are diluted in a gradient of 2 times. Example 1 uses only three concentrations of fluorescent dye. For fluorescence calibration results, the more different concentrations of fluorescent dye used, the more accurate the calculated correction coefficient will be.

[0038] Three sets of fluorescent dyes were injected into the test chip, the structure of which can be found in [reference needed]. Figure 3 The test chip is a silicon-based chip with 16 serpentine regions, which are test holes formed after etching. During fluorescence experiments, the test chip needs to be heated to excite the fluorescent reagent at a suitable temperature. The chip temperature is maintained between 50℃ and 55℃, and heating is achieved using a temperature control module attached to the chip. Different concentrations of fluorescent dye are injected; since silicon itself does not fluoresce, only the serpentine regions shown in the image exhibit fluorescence after excitation. For the detailed excitation process, please refer to [link to relevant documentation]. Figure 1 In the figure, the excitation light source 7 is tilted at a certain angle to illuminate the test chip 2. When the excitation light source illuminates the test chip containing different concentrations of fluorescent dye during the fluorescence experiment, the angle at which the excitation light illuminates the test chip is the same.

[0039] The excitation light source 7 includes a collimating lens 8 located at the light source outlet and an LED light source 10 disposed on the excitation light source 7. The LED light source 10 includes four types of LED light sources. In Example 1, in order to be compatible with the fluorescent dyes of the four fluorescent channels mentioned above, a four-wavelength integrated light source module was selected, which integrates the LED light sources with wavelengths of 480nm (FAM), 580nm (ROX), 525nm (HEX), and 638nm (CY5) into a single light outlet using a dichroic mirror 9.

[0040] Step 2: Above the test chip is an image acquisition device, which includes an imaging lens 3, a fluorescence filter 4, a CMOS image sensor 5, and an FPGA image acquisition board 6. The CMOS image sensor 5 is used to sense images, and the FPGA image acquisition board 6 acquires images from the CMOS image sensor 5 and performs subsequent processing. This subsequent processing includes generating a fluorescence grayscale image and segmenting the fluorescence grayscale image. The segmentation method is based on the region of the serpentine test holes on the test chip 2, using rectangular frames to completely enclose the serpentine test hole region while ensuring that the rectangular frames on the test chip 2 are uniformly arranged. The collected fluorescence grayscale images can be found in [reference needed]. Figure 3 , Figure 3 The differences in fluorescence intensity at three different concentrations are presented; higher concentrations of the fluorescent dye result in higher fluorescence intensity. The acquired fluorescence grayscale images were processed, and the collected data are as follows:

[0041]

[0042] In Equation 1, Data1 represents high-concentration data, Data2 represents intermediate-concentration data, and Data3 represents low-concentration data. It is assumed that the calibration coefficient to be solved for this channel is X = {x1, x2, ..., x...}. 16 Then, the calibrated fluorescence values ​​for each concentration can be obtained as shown in Equation 2:

[0043]

[0044] The derivation formulas for cv at high, medium, and low concentrations, calculated using formula (2), are shown in formula (3):

[0045]

[0046] Where std is the standard deviation and mean is the mean;

[0047] Assuming the expected cv is e, then the calibration coefficient model for the k-th channel can be derived from the formula for calculating the coefficient of variation.

[0048] f(x) is the objective function, X is the constraint set and X = R. m That is, X is an m-dimensional real space.

[0049] The expected coefficient of variation (cv) is required to be less than or equal to 5% in the polymerase chain reaction analyzer standard. In order to improve the calibration effect in the experiment, the expected coefficient of variation is required to be 2% ≤ e ≤ 3%. In this example 1, 3% is selected and the model is solved.

[0050] Step 3: Obtain the optimization model for the k-th channel. For the unconstrained optimization model of this channel, the quasi-Newton method is used to solve the model. Specifically, the Matlab optimization function fminunc is used with the optimization parameter quasi-newton set to find the minimum value of the model to obtain the calibration coefficients of each hole in the k-th channel. Thus, the hole calibration coefficient matrix M for this channel is obtained. m×1 .

[0051] Step 4: Repeat steps 2 and 3 to calculate the calibration coefficient matrix for each hole in the other channels, finally obtaining the hole calibration coefficient matrix for all channels.

[0052] M = [M1 M2 … M] n (5)

[0053] Among them, M k It is an m-row, 1-column matrix, k∈[1,n].

[0054] To facilitate rapid computation and processing of image computing models, the built-in function fminunc in the scientific analysis software MATLAB R2014b can be used to calculate the calibration coefficients. This software can be burned into the FPGA image acquisition board to improve computational efficiency.

[0055] The following is a data analysis of fluorescence experiments conducted using different concentrations of fluorescent dyes in Example 1. To evaluate the accuracy of the fluorescence calibration method, four fluorescence channels were used in the experiments. The four channels used FAM, ROX, HEX, and CY5 fluorescent dyes, respectively. The 16 serpentine test wells on the silicon-based chip were arranged and numbered sequentially from bottom to top and from left to right. The fluorescence values ​​of each test well under each fluorescence channel obtained by the image acquisition module during the experiment are shown in Tables 2-4 below.

[0056] Table 2: Fluorescence data collected from high-concentration fluorescent dyes

[0057]

[0058]

[0059]

[0060]

[0061] Table 3: Fluorescence data collected from medium-concentration fluorescent dyes

[0062]

[0063]

[0064]

[0065]

[0066] Table 4: Fluorescence data collected from low-concentration fluorescent dyes

[0067]

[0068]

[0069]

[0070]

[0071] The 10 data points measured in each test well were obtained by repeatedly acquiring grayscale images of the excitation fluorescence 10 times under the same fluorescence channel and concentration, and calculating the fluorescence image of the grayscale image in that test well.

[0072] The collected fluorescence values ​​at each concentration in each fluorescence channel were used to calculate the calibration coefficients using scientific computing analysis software. The calibration coefficients for each well are shown in Table 5 below.

[0073] Table 5. Calibration coefficients for each well in each fluorescence channel.

[0074]

[0075]

[0076] The correction coefficients for each well in Table 5 are calculated by averaging the 10 fluorescence values ​​collected from each well in Tables 2-4.

[0077] The calibration fluorescence value is obtained by multiplying the correction coefficient of each well obtained in Table 5 by the fluorescence value collected in Tables 2-4 by the calibration coefficient. The coefficient of variation of the calibration fluorescence value and the coefficient of variation of the uncalibrated fluorescence values ​​in Tables 2-4 are calculated respectively, and the data in Table 6 below are obtained.

[0078] Table 6. Calibration coefficients for each well in each fluorescence channel.

[0079]

[0080]

[0081] Table 6 shows that the calibrated coefficient of variation meets the requirement of less than 5%. At the same time, the dispersion of the coefficient of variation of each test well is significantly reduced, which weakens the influence of uneven light source distribution on fluorescence acquisition.

[0082] To evaluate whether the calibration coefficient affects fluorescence linearity, fluorescence linearity needs to be calculated. Higher fluorescence linearity indicates better acquired fluorescence data; generally, a linearity of at least 99% is required. Fluorescence linearity is calculated by calibrating fluorescence values ​​at different concentrations using the calibration coefficient, and then using the linear regression coefficient r = ∑(Xi-X)(Yi-Y / sqrt[∑(Xi-X)]). 2 ×∑(Yi-Y) 2 To verify whether it affects linearity, Xi is the calibrated fluorescence value for each concentration (average of three sample data for each concentration), X is the average calibrated fluorescence value for each concentration, YI is the concentration of the tested fluorescent dye, and Y is the average value of the tested fluorescent dye at each concentration. The experimental results are shown in Table 7 below.

[0083] Table 7 Fluorescence linearity before and after calibration

[0084]

[0085]

[0086] As shown in Table 7, the fluorescence calibration method of the present invention does not affect the original fluorescence linearity.

[0087] The fluorescence value calibration method provided by this invention has been described in detail above. Specific examples have been used to illustrate the structure and working principle of this invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this invention. It should be noted that those skilled in the art can make several improvements and modifications to this invention without departing from the principles of this invention, and these improvements and modifications also fall within the scope of protection of the claims of this invention.

Claims

1. A fluorescence calibration method, characterized in that, Includes the following steps: Step 1: Prepare multiple concentrations of the same fluorescent dye and inject them into the test wells of the test chip. Use an excitation light source to irradiate the test chip containing different concentrations of fluorescent dye. Step 2: Use an optical acquisition device to acquire the excitation fluorescence of the test chip to obtain the acquired fluorescence grayscale image of the test chip. Process the fluorescence grayscale image and calculate the average pixel value of the fluorescence grayscale image as the acquired fluorescence value. Image processing was performed on the fluorescence grayscale images of test chips containing different concentrations of the same fluorescent dye to obtain multiple sets of collected fluorescence values; Step 3: The calibration fluorescence value is the product of the collected fluorescence value and the calibration coefficient. Construct a model to minimize the error between the coefficient of variation and the expected coefficient of variation of the calibration fluorescence value for different concentrations of fluorescent dyes, and solve the model to obtain the calibration coefficient.

2. The fluorescence calibration method according to claim 1, characterized in that: The fluorescent dyes consist of at least three groups.

3. The fluorescence calibration method according to claim 2, characterized in that: The concentrations of the multiple groups of fluorescent dyes were diluted in a gradient.

4. The fluorescence calibration method according to claim 3, characterized in that: The test chip is a silicon-based test chip, and the silicon-based test chip is provided with a plurality of test holes, which are evenly arranged on the test chip.

5. The fluorescence calibration method according to claim 4, characterized in that: Step 2, the fluorescence grayscale image processing, further includes dividing the fluorescence grayscale image into several regions, calculating the average pixel value of the fluorescence grayscale image in each region, and using it as the collected fluorescence value for each region. A model is constructed to minimize the error between the coefficient of variation and the expected coefficient of variation for calibration fluorescence values ​​at different concentrations in each region. Solving this model yields the calibration coefficient for each region.

6. The fluorescence calibration method according to claim 5, characterized in that: The method for dividing the fluorescence grayscale image into several regions in step 2 is as follows: several regions of equal size are selected on the test chip, and each region contains one test hole.

7. The fluorescence calibration method according to claim 5, characterized in that: The average pixel value of the fluorescent grayscale image in each region is calculated using an image processing algorithm.

8. The fluorescence calibration method according to claim 7, characterized in that: The constructed model for minimizing the error between the coefficient of variation and the expected coefficient of variation when solving for calibration fluorescence values ​​at different concentrations is as follows: K represents the amount of fluorescent dye at different concentrations, cv n Let be the coefficient of variation for the nth concentration of fluorescent dye, and e be the expected coefficient of variation. The derivation formula for cv is: Data nj Here are the fluorescence values ​​after calibration for the nth concentration of fluorescent dye; std is the standard deviation, mean is the mean, and Data... nj The derivation formula is as follows: Where x m a represents the calibration coefficient of the m-th test hole on the test chip. nm This represents the fluorescence value collected in the m-th test well of the n-th concentration fluorescent dye, where m represents the number of test wells on the test chip.

9. A fluorescence calibration method according to claim 8, characterized in that: The expected coefficient of variation range is: 2% ≤ e ≤ 3%.

10. A fluorescence calibration method according to claim 9, characterized in that: The model for minimizing the error between the coefficient of variation and the expected coefficient of variation of calibrated fluorescence values ​​at different concentrations is solved using the quasi-Newton method.

Images