Fluorometer Peak Finding Algorithm

Abstract

This invention discloses a peak-finding algorithm for a fluorometer, which can achieve fast peak finding with high accuracy and can eliminate the influence of interference signals. The steps are as follows: plotting a preliminary waveform curve; processing the values ​​Y at the sampling points... N Perform smoothing and plot the smoothed waveform curve; find the sampling points within the peak-finding interval that conform to formula b, along with their positions N and values ​​Y. N The peak point is denoted as . Find the minimum values ​​and their positions P and M to the left and right of the peak point on the smooth waveform curve, and take the point with the smaller distance as the baseline valley point of the peak point. Calculate the peak height value. Calculate the peak height values ​​and peak point positions of peak T and peak C in sequence, and confirm the allowable range deviation.

Description

Technical Field

[0001] This invention relates to the field of fluorescence immunoassay technology, and in particular to a peak-finding algorithm for a fluorescence analyzer. Background Technology

[0002] The principle of immunochromatography is to use specific antibodies to bind to the analyte and detect the presence or absence of the target substance by using labeled fluorescent dyes. It can be used to detect small and very small amounts of electron-emitting substances, giving fluorescent immunochromatography advantages such as high sensitivity, high specificity, speed and simplicity.

[0003] The principle of fluorescence detection by a fluorescence immunoassay analyzer is to focus a specific wavelength excitation light source through a lens, collimate the optical path, and form a narrow slit to form a line light source, which illuminates the fluorescent material and excites it into a visible light signal. After the visible light signal enters the detection system, it is converted into an electrical signal and amplified. Thus, a complete waveform curve is formed by gradually moving the reagent card and scanning through the detection system.

[0004] During fluorescence detection, the fluorescence signal is easily affected by environmental factors (such as temperature, light, and pollutants). These factors may cause changes in the intensity of the fluorescence signal, affecting the accuracy of the detection results. Especially when the concentration of the analyte is low, the upper part of the reagent card may be chromatographically retained, causing the substrate signal generated by the detection system to be superimposed on the fluorescence intensity data of the detection area. This results in a large interference signal superimposed on the fluorescence background signal, which has a significant impact on the measurement accuracy and can lead to misjudgment. Traditional peak-finding algorithms mainly find the maximum value within a set interval, and record the peak value based on the decreasing trend of the corresponding values ​​on both sides of the maximum value. Then, they search for valley values ​​within the set interval on both sides of the peak, taking the lowest value as the valley value, and using the difference between the peak and the valley as the final relative peak height. Traditional peak-finding algorithms are mainly suitable for detection items with relatively flat baselines. However, due to the background effect that occurs during reagent strip chromatography, the baseline may not be flat and may show an increasing, decreasing, or wavy trend. In addition, there may be some assembly errors in the instrument and reagent strip assembly process. If the lowest value within a set interval is used as the valley point value, it will result in a large inter-stage and inter-strip error in the final relative peak height value, which will have a significant impact on the accuracy of low-concentration sample judgment. Summary of the Invention

[0005] The purpose of this invention is to provide a peak finding algorithm for fluorometers, which solves the problems existing in the prior art, can achieve fast peak finding with high accuracy, and can eliminate the influence of interference signals.

[0006] To achieve the above objectives, the solution of the present invention is:

[0007] A peak-finding algorithm for a fluorometer includes the following steps:

[0008] Step 1: The detection system moves and scans the reagent card, recording the location N of all sampling points and their corresponding values ​​Y. N And plot the initial waveform curve;

[0009] Step 2: Use formula a to process the values ​​Y at each sampling point in the initial waveform curve. N The smoothed value Y′ is obtained. N And plot the smooth waveform curve, formula a is as follows:

[0010] Y′ N =(Y N-2 +Y N-1 +Y N +Y N+1 +Y N+2 ) / 5

[0011] Step 3: Set the peak finding interval for T-peak. Within the peak finding interval for T-peak, sequentially select sampling points that conform to formula b. Formula b is as follows:

[0012] Y′ N-3 <Y′ N-2 <Y′ N-1 ≤Y′ N ≥Y′ N++ >Y′ N+2 >Y′ N+3

[0013] Find the sampling points within the peak-finding interval of T that conform to formula b, along with their locations N and values ​​Y. N This is recorded as the peak point;

[0014] Step 4: For the peak points recorded in Step 3, on the smooth waveform curve, find the minimum value to the left of the peak point in the valley-finding interval [NX, N] on the left side, and record its corresponding position as P. Find the minimum value to the right of the peak point in the valley-finding interval [N, N+X] on the right side, and record its corresponding position as M. Where X is the position point occupied by half of the peak.

[0015] Step 5: Based on the minimum values ​​on both sides of peak N found in the interval [NX,N] and [N,N+X] in Step 4, compare the absolute values ​​of the positional distances NP and NM. Take the positional distance with the smaller absolute value as the baseline valley point of the peak point, and record the fluorescence value Y′ corresponding to this baseline valley point. G The peak height T at this peak point N =Y′ N -Y′ G ;

[0016] Step 6: Repeat steps 4 and 5 until the peak height values ​​of all peak points recorded in step 3 are calculated. Take the largest peak height value as the calculated peak value T, and obtain the position of peak point T.

[0017] Step 7: Set the range of C-peak finding interval, and calculate the C-peak value and C-peak position of the smooth waveform curve according to the process of steps 3 to 6; confirm the allowable range deviation of the distance from the T-peak position to the C-peak position.

[0018] In step 1, a stepper motor is used to drive the reagent card to move at a constant speed, with a step distance of 0.02 mm / step. A fluorescence signal is collected once for each step.

[0019] After adopting the above technical solution, the present invention has the following technical effects:

[0020] This invention ensures that the corresponding peak value is within the set peak-finding interval, without needing to consider whether the valley point is within the peak-finding interval. It can find the fluorescence intensity of the relative peak height with uniqueness, and does not have overly strict requirements on the instrument's inter-stage assembly position deviation and the clamping strip assembly position deviation, resulting in higher compatibility. Furthermore, regarding the peak-finding interval setting, when the selected interval shifts left or right, as long as the peak point is within the selected interval, a unique and unchanging relative peak height fluorescence value can be calculated. This invention utilizes the method of first finding all peak values ​​that conform to the peak-shaving pattern, and then finding the corresponding minimum value on both sides of each peak within a specified range of points. The point corresponding to this minimum value has the shortest horizontal distance to the point corresponding to the peak; this point is the valley point of the corresponding peak. The fluorescence difference between each peak and valley is calculated, and the value with the largest difference is the relative T-peak or C-peak value, meaning the slope from the peak to the valley point must be the largest. This ensures the uniqueness of peak and valley finding, resulting in higher accuracy and stronger adaptability. It avoids excessive deviation in results due to baseline fluctuations, increasing or decreasing trends. Detailed Implementation

[0021] To further explain the technical solution of the present invention, the present invention will be described in detail below through specific embodiments.

[0022] This invention discloses a peak-finding algorithm for a fluorometer, comprising the following steps:

[0023] Step 1: The detection system moves and scans the reagent card, recording the location N of all sampling points and their corresponding values ​​Y. N And plot the initial waveform curve;

[0024] Step 2: Due to the discrete fluctuations of the sampling points, it is necessary to perform moving average filtering on the values ​​of all sampling points using a smoothing method. Therefore, in Step 2, formula a is used to process the values ​​Y of each sampling point in the initial waveform curve. N The smoothed value Y′ is obtained. N And plot the smooth waveform curve, formula a is as follows:

[0025] Y′ N =(Y N-2 +Y N-1 +Y N +Y N+1 +Y N+2 ) / 5

[0026] Step 3: Considering the impact of reagent card assembly errors and instrument-to-instrument assembly errors in the detection system, set the T-peak search interval range for each reagent card item (the range for each peak search can be expanded to ±2mm). Within the T-peak search interval range, sequentially select sampling points that conform to formula b, as follows:

[0027] Y′ N-3 <Y′ N-2 <Y′ N-1 ≤Y′ N ≥Y′ N+1 >Y′ N+2 >Y′ N+3

[0028] Find the sampling points within the peak-finding interval of T that conform to formula b, along with their locations N and values ​​Y. N This is recorded as the peak point;

[0029] Step 4: For the peak points recorded in Step 3, find X points forward and backward on the smooth waveform curve. That is, find the minimum value to the left of the peak point in the valley-finding interval [NX, N] on the left and record its corresponding position as P. Find the minimum value to the right of the peak point in the valley-finding interval [N, N+X] on the right and record its corresponding position as M. Here, X is the number of position points occupied by half a peak. Its value depends on the sum of the detection line width of the reagent card and the line light source width. When the design of the detection line width of the reagent card and the line light source width of the instrument are determined, this value of X can be set to a fixed value.

[0030] Step 5: Based on the minimum values ​​on both sides of peak N found in the interval [NX,N] and [N,N+X] in Step 4, compare the absolute values ​​of the positional distances NP and NM. Take the positional distance with the smaller absolute value as the baseline valley point of the peak point, and record the fluorescence value Y′ corresponding to this baseline valley point. G Then the peak height T of the peak point N =Y′N -Y′ G ;

[0031] Step 6: Repeat steps 4 and 5 until the peak height values ​​of all peak points recorded in step 3 are calculated. Take the largest peak height value as the calculated peak value T, and obtain the position of peak point T.

[0032] Step 7: Set the C-peak search interval range. Calculate the C-peak value and its position on the smooth waveform curve following the process in steps 3 to 6 (i.e., replace the T-peak search interval range with the C-peak search interval range). Confirm the allowable range deviation of the distance between the T-peak position and the C-peak position (the allowable range deviation is determined by the reagent card item). Since the T-line and C-line are generally sprayed using a spraying technique during reagent card manufacturing, their relative positions are relatively fixed. By introducing the relative distance between the T-peak and C-peak to confirm the allowable range deviation, the problem of inaccurate test results caused by dirt adhering to the reagent card surface can be solved. The total number of points from the T-peak to the C-peak represents the distance from the center of the T-line to the center of its C-line on the reagent card. The allowable deviation setting for this distance depends on the production control requirements of the reagent card.

[0033] In step 1 above, a stepper motor is used to drive the reagent card to move at a constant speed with a step distance of 0.02 mm / step. A fluorescence signal is collected once for each step. The fluorescence collection of the entire reagent card is completed by collecting 600 fluorescence signals, and the total moving distance is 12 mm.

[0034] To accommodate reagent cards from different manufacturers, users can set the starting position, step size, half-peak width, and total number of collection points according to the actual needs of the reagent card; to accommodate different reagent projects, users can set the peak finding interval for T peak and the peak finding interval for C peak according to the needs of different projects.

[0035] Through the above scheme, this invention can find the fluorescence intensity of the relative peak height as long as the corresponding peak value is within the set peak-finding interval, without considering whether the valley point is within the peak-finding interval. This method is unique and does not have strict requirements on the inter-instrument assembly position deviation and the assembly position deviation of the clips, resulting in higher compatibility. Furthermore, regarding the peak-finding interval setting, when the selected interval shifts left or right, as long as the peak point is within the selected interval, a unique and unchanging relative peak height fluorescence value can be calculated. This invention utilizes the method of first finding all peak values ​​that conform to the peak-shaped change, and then finding the corresponding minimum value on both sides of each peak within a specified range of points. The point corresponding to the minimum value has the shortest horizontal distance to the point corresponding to the peak. This point is the valley point of the corresponding peak. The fluorescence difference between each peak and valley is calculated, and the value with the largest difference is the relative T-peak or C-peak value, meaning the slope from the peak to the valley point must be the largest. This ensures the uniqueness of peak and valley finding, resulting in higher accuracy and stronger adaptability. It also prevents excessive deviation in results due to baseline fluctuations, increases, or decreases.

[0036] The above embodiments are not intended to limit the product form and style of the present invention. Any appropriate changes or modifications made by those skilled in the art should be considered as not departing from the patent scope of the present invention.

Claims

1. A peak-finding algorithm for a fluorometer, characterized in that... Includes the following steps: Step 1: The detection system moves and scans the reagent card, recording the position N of all sampling points and their corresponding values ​​Y. N And plot the initial waveform curve; Step 2: Use formula a to process the values ​​Y at each sampling point in the initial waveform curve. N The smoothed value Y′ is obtained. N And plot the smooth waveform curve, formula a is as follows: AND' N =(And N-2 +And N-1 +And N +And N+1 +And N+2 ) / 5 Step 3: Set the peak finding interval for T-peak. Within the peak finding interval for T-peak, sequentially select sampling points that conform to formula b. Formula b is as follows: AND' N-3 <Y′ N-2 <Y′ N-1 ≤Y′ N ≥Y′ N+1 >Y′ N+2 >Y′ N+3 Find the sampling points within the peak-finding interval of T that conform to formula b, along with their locations N and values ​​Y. N This is recorded as the peak point; Step 4: For the peak points recorded in Step 3, on the smooth waveform curve, find the minimum value to the left of the peak point in the valley-finding interval [NX, N] on the left side, and record its corresponding position as P. Find the minimum value to the right of the peak point in the valley-finding interval [N, N+X] on the right side, and record its corresponding position as M. Where X is the position point occupied by half of the peak. Step 5: Based on the minimum values ​​on both sides of peak N found in the interval [NX, N] and [N, N+X] in Step 4, compare the absolute values ​​of the position distances NP and NM. Take the position distance with the smaller absolute value as the baseline valley point of the peak point, and record the fluorescence value Y′ corresponding to the baseline valley point. G The peak height T at this peak point N =Y′ N -Y′ G ; Step 6: Repeat steps 4 and 5 until the peak height values ​​of all peak points recorded in step 3 are calculated. Take the largest peak height value as the calculated peak value T, and obtain the position of peak point T. Step 7: Set the range of C-peak finding interval, and calculate the C-peak value and C-peak position of the smooth waveform curve according to the process of steps 3 to 6; confirm the allowable range deviation of the distance from the T-peak position to the C-peak position.

2. The fluorescence spectrometer peak-finding algorithm as described in claim 1, characterized in that: In step 1, a stepper motor is used to drive the reagent card to move at a constant speed, with a step distance of 0.02 mm / step. A fluorescence signal is collected once for each step.