Processing method for improving accuracy in high-speed biochemical detection

Abstract

The invention belongs to the technical field, and particularly relates to a treatment method for improving the accuracy of high-speed biochemical detection. The invention discloses a processing method for improving the accuracy of high-speed biochemical detection. The processing method comprises the following steps of sampling window extraction, anomaly recognition and abnormal data screening. Compared with the prior art, the treatment method for improving the accuracy in high-speed biochemical detection has the beneficial effects that (1) the accuracy of a sampling position in high-speed biochemical detection is greatly improved; (2) according to the processing method for improving the accuracy in high-speed biochemical detection, the efficiency of anomaly identification is improved, anomaly sources, anomaly factors and anomaly conditions are classified more finely, and anomaly reasons can be analyzed conveniently; and (3) the processing method for improving the accuracy of high-speed biochemical detection is beneficial to instrument detection and alarm, enlarges the range of a sampling window, and is beneficial to later noise reduction.

Description

Technical Field

[0001] The invention belongs to the technical field of biochemical detection algorithms, and in particular relates to a processing method for improving accuracy of high-speed biochemical detection. Background Art

[0002] At present, most high-speed biochemical detection in vitro diagnostic instruments on the market include a reaction module and a detection module. During the detection, the reaction disk in the reaction module is used as a carrier to drive the reaction cup containing the sample to be detected to rotate, and the position of the light spot is fixed. The reaction disk is driven by a stepper motor, and the detection module synchronously collects data with the stepper motor motion signal to obtain the light transmittance data of each reaction cup, and extracts the data (sampling window) of the light spot falling into the reaction cup, which is processed as the detection result.

[0003] The commonly used implementation method is to obtain the sampling window through mechanical structure calibration and average the data to obtain the test result. This method has no abnormal feedback on the test result and process, such as the reaction disk motor out of step resulting in low sampling position accuracy (such as Figure 1 ), light source fluctuations (such as Figure 2 ), bubble interference (such as Figure 3 )wait.

[0004] In order to eliminate the motion deviation of the reaction disk and to give an alarm for abnormal phenomena, an effective sampling window and an abnormality recognition algorithm are proposed to be identified from the change curve of a single reaction cup. Based on this, the present invention provides a processing method for improving the accuracy of high-speed biochemical detection. Summary of the invention

[0005] The object of the present invention is to provide a method for improving the accuracy of high-speed biochemical detection to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions: A processing method for improving accuracy of high-speed biochemical detection includes the following steps: extracting sampling windows, identifying anomalies, and screening abnormal data; The steps of extracting the sampling window are as follows: (1) Obtain all extreme points by differentially analyzing the light spot reaction cup transmittance change curve data, and extract a peak or trough according to the peak characteristics between the extreme points and the horizontal threshold. The trough corresponds to waveform profile 1, and the peak corresponds to waveform profile 2. The horizontal threshold is: when the transmittance data changes from strong to weak, there is a critical value that makes waveform profile 1 just turn into waveform profile 2. The critical value depends on the absorbance of the light spot at the wall of the reaction cup. The two positions correspond to the inflection point a and the inflection point b of waveform profile 2, respectively. The number of data points between the inflection point a and the inflection point b is the horizontal threshold. (2) Curve peak clipping and valley filling: Using the approximately symmetrical characteristics of the peaks and valleys, the transmittance change curve of the spot reaction cup is clipped and valley filled, so that the fluctuation data at the original peaks and valleys becomes a constant value interval; suppose the data to be calculated is f, the width of the constant value interval is M, the data e with a value of 0 is used as the calculation parameter, x is the intermediate process data, and y is the final output data; Peak shaving: ① ; ② ; Filling the valley: ① ; ② ; (3) performing differential processing on the data y, and searching for a continuous segment of data with a length of M and a value of 0 in the differential result to obtain a constant value interval; taking the average of the initial index and the final index of the constant value interval to obtain the center of the constant value interval, the center of the constant value interval is approximately the center of the peak and the trough; (4) Reduce the approximate influence between the center of the constant value interval and the center of the peak and valley. After each peak shaving and valley filling, determine whether to perform peak shaving and valley filling again based on whether the constant value interval width M is equal to the data e width, so that the e value gradually approaches the M value; (5) Take the center of the constant value interval as the index and select points on both sides of the same index position in the original data; the number of points selected must be less than the maximum valid data length corresponding to the light spot falling into the reaction cup; The steps of anomaly identification are: (1) Classify abnormal factors, including abnormal temperature, mechanical vibration, bubbles passing through the light path, impurity particles passing through the light path, uneven mixing, uneven surface, and false triggering of the optical coupler; (2) Identify abnormalities based on different abnormal factors according to classification; The steps for abnormal data screening are: (1) Get the minimum value Vmin and the maximum value Vmax of the data to be filtered. The minimum difference between two adjacent points is Dmin. Calculate the number of intervals nBins = ; (2) Calculate the number of data points in each interval, determine the interval with the most data points, and determine the boundary of the interval with the most data points; (3) Subtract the light source fluctuation index from the left boundary of the interval and add the light source fluctuation index to the right boundary of the interval to form a new interval boundary; (4) Remove the data outside the boundaries of the new interval.

[0007] Preferably, after the abnormal data is screened, filtering and denoising are performed, and the steps of filtering and denoising are: (1) Convert the actual data into the frequency domain to calculate the noise spectrum density, and integrate the noise spectrum with respect to the frequency to obtain the relationship between frequency and noise; (2) Select the corresponding frequency as the cutoff frequency of the digital filter according to the noise requirement; (3) Designing a digital low-pass filter based on the cutoff frequency of the digital filter; the low-pass digital filter is a finite impulse response filter with a gain of 0 dB at DC; wherein the passband attenuation tends to 0 dB and the stopband attenuation is less than -3 dB.

[0008] Preferably, the abnormal source corresponding to the temperature abnormality is the light source, and the abnormal condition is that the heat dissipation mechanism of the light source is damaged; the abnormal source corresponding to the mechanical vibration is the light source, and the abnormal condition is that the reaction module is impacted by the outside world; the abnormal source corresponding to the bubbles passing through the light path is a water bath, and the abnormal condition is that there are bubbles in the pipeline; the abnormal source corresponding to the impurity particles passing through the light path is a water bath, and the abnormal condition is that there are impurity particles in the pipeline; the abnormal source corresponding to the uneven mixing is the mixing action, and the abnormal condition is uneven mixing; the abnormal source corresponding to the surface unevenness is the reaction cup, and the abnormal condition is that the reaction cup is aging and wear; the abnormal source corresponding to the optical coupler false triggering is the reaction disk, and the abnormal condition is that the optical coupler is interfered with or aged.

[0009] Preferably, the method for identifying temperature anomalies is as follows: in waveform profile 1 and waveform profile 2, the waveform at the gap where the light spot passes is fixed and can be regarded as a reference for full light transmittance of the light intensity; when the intensity of direct light is detected to drop by more than the light source fluctuation threshold at the same cup position gap for five consecutive times, it is judged as a temperature anomaly; the light source fluctuation threshold is the maximum fluctuation range of the light source itself.

[0010] Preferably, the method for identifying mechanical vibration is: take the mean and extreme value of the data in the sampling window, observe whether there are values ​​above or below the mean at the same time, and exceed the light source fluctuation threshold, where the light source fluctuation threshold is the maximum fluctuation range of the light source itself; if so, it is determined to be mechanical vibration interference.

[0011] Preferably, the method for identifying bubbles passing through the light path is: differentiate the light spot reaction cup transmittance change data curve to obtain all maximum and minimum points, and each time adjacent maximum and minimum values ​​exceed the light source fluctuation threshold, it is used as a new curve; observe whether there are new curves at both ends that drop and then rise; if so, it is determined that there are bubbles.

[0012] Preferably, the method for identifying uneven mixing is: when the light spot is in the concentrated area of ​​the sample solution, the transmittance data is relatively large, and when passing through the dilute area, the transmittance data is relatively small; when the light spot passes through the light path several times repeatedly, the above situation occurs each time, and it is determined to be uneven mixing.

[0013] Preferably, the method for identifying the uneven surface is as follows: pure water is placed in the reaction cup, and when the data fluctuation of the light transmission data exceeds the light source fluctuation threshold, it is determined that the surface of the reaction cup is uneven.

[0014] Preferably, the method for identifying false triggering of the optocoupler is as follows: the reaction disk is used as a carrier for the reaction cup, and a yardstick is designed at the bottom. The yardstick is designed for counting the reaction cups, and a typical value is maintained between the trigger position and the sampling window position. If it exceeds 20, it is determined to be a false triggering of the optocoupler.

[0015] Working principle: The complete light transmission curve of a single reaction cup passing through the light spot is as follows: Figure 4 As shown in the figure, the sampling window is at the trough or peak due to physics. The optical coupler trigger position reflects the possible deviation of mechanical calibration when the reaction disk rotates. The data in the sampling window reflects the detection value and abnormal phenomenon of the reaction cup. The data in the gap of the same cup reflects the change of the light source.

[0016] Compared with the prior art, the present invention has the following beneficial effects: (1) The present invention provides a method for improving the accuracy of high-speed biochemical detection, which greatly improves the accuracy of high-speed biochemical detection sampling position; (2) The present invention provides a method for improving the accuracy of high-speed biochemical testing, which increases the efficiency of abnormality identification, divides abnormal sources, abnormal factors and abnormal conditions into smaller parts, and facilitates the analysis of abnormal causes; (3) The present invention provides a method for improving the accuracy of high-speed biochemical detection, which is beneficial to instrument detection alarm, increases the sampling window range, and is beneficial to reducing noise in the later stage. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 : Schematic diagram of the detection results when the reaction disk motor loses step, resulting in low sampling position accuracy; Figure 2 : Schematic diagram of detection results when the light source fluctuates; Figure 3 : Schematic diagram of the detection results during bubble disturbance; Figure 4 : The waveform profile 1 of the complete light transmission data change curve of a single reaction cup passing through the light spot when the absorbance is strong; Figure 5 : The waveform profile 2 of the complete light transmission data change curve of a single reaction cup passing through the light spot when the absorbance is weak; Figure 6 : Extract the peak clipping effect diagram shown in step 2 of the sampling window, the part above the dotted line is eliminated, and the dotted line part presents a constant value interval; Figure 7 : Extract the valley filling effect shown in step 2 of the sampling window, the part below the dotted line is eliminated, and the dotted line part shows a constant value interval; Figure 8 : Extract the trend graph of the gradually approaching peak shown in step 4 of the sampling window; Fig. 9: Extract the trend graph of the gradually approaching trough shown in step 4 of the sampling window; Fig.10 : Specific flow chart for extracting sampling window; Fig.11 : Schematic diagram of the interval boundary with the most data points; Fig.12 : Specific flow chart of the temperature anomaly identification method; Fig.13 : Specific flow chart of the mechanical vibration identification method; Fig.14 : A specific flow chart of the method for identifying the light path through which bubbles pass; Fig.15 : Specific flow chart of the uneven mixing identification method; Fig.16 : Specific flow chart of the surface unevenness identification method; Fig.17 : When the heat dissipation mechanism of the light source is damaged and the light intensity slowly decreases, the trend diagram of the relationship between light intensity and temperature; Fig.18 : Schematic diagram of periodic fluctuations in light intensity caused by subtle mechanical vibrations; Fig.19 : Schematic diagram of the change curve when the bubble passes through the light spot, where the bubble passing through the light spot causes a downward spike, and the slope of the falling edge is smaller than the slope of the rising edge; Fig. 20 : When the reaction cup contains pure water, the transmitted light passes through the uneven surface of the reaction cup, and the light intensity curve shows obvious fluctuations; Fig.21 : The waveform between the two red dots in the embodiment is a schematic diagram of the extracted peak; Fig. 22 : Schematic diagram of the process of peak shaving and valley filling and approximation in the embodiment; Fig.23 : Schematic diagram of taking the central value in the embodiment; Fig.24 : Comparison diagram before and after screening in the value-taking process of the embodiment. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] See also Figure 1-Figure 24 The present invention provides a technical solution for a high-speed biochemical detection method to improve accuracy: A processing method for improving accuracy of high-speed biochemical detection includes the following steps: extracting sampling windows, identifying anomalies, and screening abnormal data; The steps of extracting the sampling window are as follows: (1) Obtain all extreme points by differentially analyzing the light spot reaction cup transmittance change curve data, and extract a peak or trough according to the peak characteristics between the extreme points and the horizontal threshold. The trough corresponds to waveform profile 1, and the peak corresponds to waveform profile 2. The horizontal threshold is: when the transmittance data changes from strong to weak, there is a critical value that makes waveform profile 1 just turn into waveform profile 2. The critical value depends on the absorbance of the light spot at the wall of the reaction cup. The two positions correspond to the inflection point a and the inflection point b of waveform profile 2, respectively. The number of data points between the inflection point a and the inflection point b is the horizontal threshold. (2) Curve peak clipping and valley filling: Using the approximately symmetrical characteristics of the peaks and valleys, the transmittance change curve of the spot reaction cup is clipped and valley filled, so that the fluctuation data at the original peaks and valleys becomes a constant value interval; suppose the data to be calculated is f, the width of the constant value interval is M, the data e with a value of 0 is used as the calculation parameter, x is the intermediate process data, and y is the final output data; Peak shaving: ① ; ② ; Filling valley: ① ; ② ; (3) performing differential processing on the data y, and searching for a continuous segment of data with a length of M and a value of 0 in the differential result to obtain a constant value interval; taking the average of the initial index and the final index of the constant value interval to obtain the center of the constant value interval, the center of the constant value interval is approximately the center of the peak and the trough; (4) Reduce the approximate influence between the center of the constant value interval and the center of the peak and valley. After each peak shaving and valley filling, determine whether to perform peak shaving and valley filling again based on whether the constant value interval width M is equal to the data e width, so that the e value gradually approaches the M value; (5) Take the center of the constant value interval as the index and select points on both sides of the same index position in the original data; the number of points selected must be less than the maximum valid data length corresponding to the light spot falling into the reaction cup; The steps of anomaly identification are: (1) Classify abnormal factors, including abnormal temperature, mechanical vibration, bubbles passing through the light path, impurity particles passing through the light path, uneven mixing, uneven surface, and false triggering of the optical coupler; (2) Identify abnormalities based on different abnormal factors according to classification; The steps for abnormal data screening are: (1) Get the minimum value Vmin and the maximum value Vmax of the data to be filtered. The maximum difference between two adjacent points is Dmin. Calculate the number of intervals nBins = ; (2) Calculate the number of data points in each interval, determine the interval with the most data points, and determine the boundary of the interval with the most data points; (3) Subtract the light source fluctuation index from the left boundary of the interval and add the light source fluctuation index to the right boundary of the interval to form a new interval boundary; (4) Remove the data outside the boundaries of the new interval.

[0020] Based on measured data 45909,45901,45899,45903,45903,45896,45891,45893,45896,45904,45912,45912,45910,45904,45904,45905,45905,45906,45906,45908,45905,45908,45906,45904,45903,45906,45908,45907,45905,45907,45908,45901,45899,45899,45902,45907,45908,459 07,45906,45902,45895,45892,45897,45899,45887,45873,45760,45234,44032,42119,39607,36737,33756,30761,27891,25334,23123,21083,19243,17627,16075,14703,13705,13019,12525,11890,10890,9854,9101,8535,8005,7419,6779,6297,6162,6336,6670 ,6977,7145,7205,7214,7217,7217,7225,7231,7233,7230,7224,7226,7226,7224,7223,7222,7221,7218,7217,7217,7215,7215,7210,7211,7212,7213,7213,7216,7221,7222,7221,7225,7229,7229,7230,7229,7232,7233,7235,7236,7235,7233,7235,7233,7234, 7234,7232,7224,7221,7219,7216,7211,7208,7203,7199,7192,7157,7049,6854,6597,6404,6444,6772,7319,7958,8580,9237,9969,10716,11485,12252,13091,14241,15652,17046,18485,20169,21891,23583,25456,27549,29838,32405, the horizontal threshold is 48, and the sampling window extraction process is described in detail: (1) Extracting peaks The extraction results are as follows Fig.21 , the waveform between the two red dots is the extracted peak; (2) Peak shaving and valley filling and approximation Fig. 22 The red curve in the middle is the first peak clipping, the yellow is the second, the green-purple is the third, and the green is the fourth. It can be seen that the peak center of the fourth approximation is the best; (3) Take the center value Difference the peak clipping data, and average the index values ​​with a difference value of 0 to get the center value, see Fig.23 ; (4) Value The data screening process is explained in detail using the measured data of bubbles passing through the light path as an example: 47319,47319,47319,47320,47321,47319,47319,47320,47318,47317,47315,47316,47316,47314,47313,47316,47315,47313,47310,47303,47295,47284,47273,47259,47245,4 7232,47218,47199,47181,47163,47145,47129,47113,47098,47083,47066,47053,47037,47025,47013,47000,46989,46980,46968,46962,46954,46950,46943,46936,46933,469 30,46924,46923,46919,46917,46916,46913,46914,46916,46920,46923,46930,46933,46942,46948,46952,46960,46968,46974,46982,46993,47001,47010,47022,47034,4704 5,47059,47073,47088,47102,47115,47132,47148,47163,47177,47192,47207,47219,47230,47239,47249,47256,47263,47267,47273,47278,47282,47288,47290,47292,47296.

[0021] Take nBins = , the corresponding number of numbers falling into each Bin is as follows: (14,8,6,5,4,4,3,4,2,4,3,4,3,6,9,51).

[0022] 51 corresponds to the interval (47298, 47323), plus the light source fluctuation range, the update interval is (47248, 47283), and the data outside this interval is filtered out. Fig.24 .

[0023] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for improving the accuracy of high-speed biochemical detection, characterized in that: The following steps are involved: Extract sampling windows, identify anomalies, and filter abnormal data; The steps of extracting the sampling window are as follows: (1) Obtain all extreme points by differentially analyzing the light spot reaction cup transmittance change curve data, and extract a peak or trough according to the peak characteristics between the extreme points and the horizontal threshold. The trough corresponds to waveform profile 1, and the peak corresponds to waveform profile 2. The horizontal threshold is: when the transmittance data changes from strong to weak, there is a critical value that makes waveform profile 1 just turn into waveform profile 2. The critical value depends on the absorbance of the light spot at the wall of the reaction cup. The two positions correspond to the inflection point a and the inflection point b of waveform profile 2, respectively. The number of data points between the inflection point a and the inflection point b is the horizontal threshold. (2) Curve peak clipping and valley filling: Using the approximately symmetrical characteristics of the peaks and valleys, the transmittance change curve of the spot reaction cup is clipped and valley filled, so that the fluctuation data at the original peaks and valleys becomes a constant value interval; suppose the data to be calculated is f, the width of the constant value interval is M, the data e with a value of 0 is used as the calculation parameter, x is the intermediate process data, and y is the final output data; Peak shaving: ① ; ② ； Filling valley: ① ; ② ； (3) performing differential processing on the data y, and searching for a continuous segment of data with a length of M and a value of 0 in the differential result to obtain a constant value interval; taking the average of the initial index and the final index of the constant value interval to obtain the center of the constant value interval, the center of the constant value interval is approximately the center of the peak and the trough; (4) Reduce the approximate influence between the center of the constant value interval and the center of the peak and valley. After each peak shaving and valley filling, determine whether to perform peak shaving and valley filling again based on whether the constant value interval width M is equal to the data e width, so that the e value gradually approaches the M value; (5) Using the center of the constant value interval as the index, select points on both sides of the same index position in the original data; the number of points selected must be less than the maximum valid data length corresponding to the light spot falling into the reaction cup.

2. A method for improving accuracy of high-speed biochemical detection according to claim 1, characterized in that: After the abnormal data is screened, filtering and denoising are performed. The steps of filtering and denoising are as follows: (1) Convert the actual data into the frequency domain to calculate the noise spectrum density, and integrate the noise spectrum with respect to the frequency to obtain the relationship between frequency and noise; (2) Select the corresponding frequency as the cutoff frequency of the digital filter according to the noise requirement; (3) Designing a digital low-pass filter based on the cutoff frequency of the digital filter; the low-pass digital filter is a finite impulse response filter with a gain of 0 dB at DC; wherein the passband attenuation tends to 0 dB and the stopband attenuation is less than -3 dB.

3. A method for improving accuracy of high-speed biochemical detection according to claim 1, characterized in that: The abnormal source corresponding to the temperature abnormality is the light source, and the abnormal condition is that the heat dissipation mechanism of the light source is damaged; the abnormal source corresponding to the mechanical vibration is the light source, and the abnormal condition is that the reaction module is impacted by the outside world; the abnormal source corresponding to the bubbles passing through the light path is the water bath, and the abnormal condition is that there are bubbles in the pipeline; the abnormal source corresponding to the impurity particles passing through the light path is the water bath, and the abnormal condition is that there are impurity particles in the pipeline; the abnormal source corresponding to the uneven mixing is the mixing action, and the abnormal condition is uneven mixing; the abnormal source corresponding to the surface unevenness is the reaction cup, and the abnormal condition is that the reaction cup is aging and wear; the abnormal source corresponding to the optical coupler false triggering is the reaction disk, and the abnormal condition is that the optical coupler is interfered with or aged.

4. A method for improving accuracy of high-speed biochemical detection according to claim 1, characterized in that: The method for identifying temperature anomalies is as follows: in waveform profile 1 and waveform profile 2, the waveform at the gap where the light spot passes is fixed and can be regarded as a reference for full light intensity transmittance; when the intensity of direct light is detected to drop by more than the light source fluctuation threshold at the same cup gap for 5 consecutive times, it is judged as a temperature anomaly; the light source fluctuation threshold is the maximum fluctuation range of the light source itself.

5. A method for improving accuracy of high-speed biochemical detection according to claim 1, characterized in that: The method for identifying mechanical vibration is: take the mean and extreme value of the data in the sampling window, observe whether there are values ​​above or below the mean at the same time, and whether they exceed the light source fluctuation threshold, where the light source fluctuation threshold is the maximum fluctuation range of the light source itself; if so, it is determined to be mechanical vibration interference.

6. A method for improving accuracy of high-speed biochemical detection according to claim 1, characterized in that: The method for identifying the light path through which the bubble passes is as follows: the light spot reaction cup light transmittance change data curve is differentiated to obtain all the maximum and minimum points, and each time the adjacent maximum and minimum values ​​exceed the light source fluctuation threshold, it is used as a new curve; Observe whether there are new curves at both ends that drop and then rise; if so, it is determined that there are bubbles.

7. A method for improving accuracy of high-speed biochemical detection according to claim 1, characterized in that: The identification method of uneven mixing is: when the light spot is in the concentrated area of ​​the sample solution, the transmittance data is relatively large, and when passing through the dilute area, the transmittance data is relatively small; when the light spot passes through the light path several times repeatedly, the above situation occurs each time, and it is judged as uneven mixing.

8. A method for improving accuracy of high-speed biochemical detection according to claim 1, characterized in that: The method for identifying the uneven surface is as follows: pure water is placed in a reaction cup, and when the data fluctuation of the light transmission data exceeds the light source fluctuation threshold, it is determined that the surface of the reaction cup is uneven.

9. A method for improving accuracy of high-speed biochemical detection according to claim 1, characterized in that: The method for identifying false triggering of optocouplers is as follows: the reaction disk is used as a carrier for reaction cups, and a yardstick design is provided at the bottom. The yardstick design is used for counting reaction cups, and a typical value is maintained between the trigger position and the sampling window position. If the value exceeds 20, it is determined to be a false triggering of the optocoupler.

10. A method for improving accuracy of high-speed biochemical detection according to claim 1, characterized in that: The steps of anomaly identification are: (1) Classify abnormal factors, including abnormal temperature, mechanical vibration, bubbles passing through the light path, impurity particles passing through the light path, uneven mixing, uneven surface, and false triggering of the optical coupler; (2) Identify abnormalities based on different abnormal factors according to classification; The steps for abnormal data screening are: (1) Get the minimum value Vmin and the maximum value Vmax of the data to be filtered. The minimum difference between two adjacent points is Dmin. Calculate the number of intervals nBins = ; (2) Calculate the number of data points in each interval, determine the interval with the most data points, and determine the boundary of the interval with the most data points; (3) Subtract the light source fluctuation index from the left boundary of the interval and add the light source fluctuation index to the right boundary of the interval to form a new interval boundary; (4) Remove the data outside the boundaries of the new interval.

Images