CN106644902A - Evaluation method of stability of laminar flow of flow cytometer - Google Patents

Abstract

The invention provides a method for evaluating the laminar flow stability of a flow cytometer, the method comprising the following steps: 1) using a high-speed microscopic image acquisition system to detect the 90° Mie scattered light of the microspheres in the flow chamber; 2) using The gray cluster analysis method clusters and analyzes the light intensity/smear length insufficient, normal, diffraction and overlapping in a large number of collected images to obtain standard normal smear images; 3) Use the midpoint method to determine the smear boundary , and calculate the corresponding microsphere flow rate; 4) use the stability of the microsphere flow rate to characterize the stability of the flow cytometer fluid system.

Description

A Method for Assessing Laminar Flow Stability of Flow Cytometry

technical field

The invention relates to the field of statistical analysis of microsphere velocity measurement, in particular to the field of laminar flow stability evaluation in a flow chamber of a flow cytometer.

Background technique

Flow cytometer is a clinical testing and analysis instrument that realizes high-speed one-by-one multi-parameter quantitative analysis of single cells or other particles in suspension in high-speed and linear flow by detecting scattered light signals and (or) fluorescent signals. Among them, the main purpose of the liquid circuit system is to make the sample liquid containing the sample to be tested (cells or microspheres) form a stable laminar flow under the wrapping of the sheath liquid, so as to achieve the purpose of obtaining single-cell flow. The stability of the liquid system will directly affect the position and time of the cells/microspheres passing through the detection area of the flow chamber, thereby affecting the signal intensity and light pulse duration of the corresponding scattered light and fluorescence signals. Evaluating the stability of the liquid circuit system, especially the velocity stability when the cells/microspheres pass through the detection area of the flow chamber, can realize a rapid prediction of the stability of the entire instrument.

At present, the methods for judging the stability of the fluid circuit system of flow cytometer mainly include the pressure method and the pulse signal characteristic analysis method. The pressure method refers to the observation of the most critical sample fluid pressure and sheath fluid pressure in the fluid system, and according to different detection rate requirements, the fluid system can be judged to be stable if the variation range of the two is within a certain range. However, the pressure method detects the gas pressure acting on the sample fluid and the sheath fluid, rather than directly detecting the liquid flow rate, so it is impossible to measure the impact of the subsequent sample injection structure and pipeline on the laminar flow and cell/microsphere velocity. The pulse signal characteristic analysis method refers to detecting the scattered light and fluorescence signals generated when the cells pass through the detection area of the flow chamber through the data acquisition module, and using the stability of the obtained pulse width to characterize the stability of the cell velocity. This method needs to complete a series of operations such as cell/microsphere scattered light excitation, collection, photoelectric conversion, pulse processing and parameter extraction, which involves the optical system and electronic circuit processing system, which increases the uncertainty of the measurement process and cannot be real Reflect the flow of cells/microspheres in the flow chamber.

The high-precision flow field characteristic analysis method is mainly particle image velocimetry (Particl imagevelocimetry, PIV), whose velocity measurement depends on the tracer particles scattered in the flow field, by measuring The displacement can be used to indirectly measure the instantaneous velocity distribution of the flow field, and can provide rich spatial structure and flow characteristics of the flow field. However, the size of the microspheres used by PIV technology for liquid flow field analysis is similar to the size of the sample detected by flow cytometry, so it is impossible to use multiple tracer particles to analyze the flow characteristics of laminar flow and single-cell flow in the flow chamber.

Contents of the invention

In order to solve the above problems, the object of the present invention is to provide a method for evaluating the stability of flow cytometer laminar flow, said method comprising the following steps: 1) using a high-speed microscopic image acquisition system to 90 ° Mie scattering of microspheres in the flow chamber 2) Use the gray clustering analysis method to cluster and analyze the light intensity/smear length insufficient, normal, diffraction and overlapping in a large number of collected images to obtain a standard normal smear image, and the specific implementation steps As follows: if there are n observation objects, m evaluation indicators, and s different gray classes, then each observation object has m characteristic data to be observed, and the obtained sequence is shown in formula (1):

X ₁ =(x ₁ (1),x ₁ (2),...,x ₁ (n))

X ₂ =(x ₂ (1),x ₂ (2),...,x ₂ (n))

...

X _m = (x _m (1), x _m (2), ..., x _m (n)) (1)

Determine the center point λ ₁ , λ ₂ ,...,λ _s of the gray classes 1, 2,...,s, and divide the value range of each index into s gray classes; divide the gray classes in different directions Continuation, consider adding 0 gray class and s+1 gray class, and determine its center points λ ₀ and λ _s+1 , so as to obtain a new center point sequence: λ ₀ , λ ₁ , λ ₂ ,...,λ _s ,λ _s+1 , the connection point (λ _k ,1) and the center point (λ _k-1 ,0) of the k-1th small gray class, the connection point (λ _l ,1) and the l+1th small gray class The center point (λ _l+1 , 0) of the small gray class obtains the trapezoidal whitening weight function of the j index with respect to the k gray class For an observed value x of index j, it can be given by

Calculate the degree of membership of the gray class k (k=1,2,...s) Calculate the comprehensive clustering coefficient of object i (i=1,2,...,n) with respect to gray class k

in, is the whitening weight function of j index k subclass, η _j is the weight of index j in the comprehensive clustering,

Depend on Judgment object i belongs to gray class k*;

3) Use the midpoint method to determine the trailing boundary, and calculate the corresponding microsphere flow rate; 4) Use the stability of the microsphere flow rate to characterize the stability of the flow cytometer fluid system.

Preferably, the index of carrying out cluster analysis in said step 2) is determined as follows:

Sum the gray value of each row of pixels in the four types of images, namely light intensity/smear length insufficient, normal, diffraction and overlapping, respectively, to obtain a horizontal gray-scale sum curve; solve the first-order derivative of the horizontal gray-scale sum curve ;Set positive and negative thresholds, and count the number of extreme points within the threshold range, and the number of effective extreme points is used as an index for cluster analysis;

Sum the gray values of each column of pixels in the four types of images: insufficient light intensity/smear length, normal, diffraction and overlap, respectively, to obtain the longitudinal gray sum curve; solve the first derivative of the longitudinal gray sum curve ;Set the positive and negative thresholds, and count the number of times the curve passes through the positive and negative thresholds, and the number of intersections with the positive and negative thresholds can be used as an index for cluster analysis.

Preferably, the microsphere flow rate in the step 3) is obtained by the formula v=l/t, wherein l is the tail length of the microsphere, and t is the exposure time of the camera.

Preferably, said step 4) also includes by the formula Calculate the standard deviation, characterize the microsphere velocity using the mean value of the tail length, and evaluate the stability of the microsphere velocity using the standard deviation of the tail length.

It should be understood that both the foregoing general description and the following detailed description are exemplary illustrations and explanations, and should not be used as limitations on the claimed content of the present invention.

Description of drawings

With reference to the accompanying drawings, more objects, functions and advantages of the present invention will be clarified through the following description of the embodiments of the present invention, wherein:

Fig. 1 is the schematic diagram of high-speed image acquisition microsphere velocity measurement;

Fig. 2 is microsphere trailing image;

Figure 3 is the horizontal gray sum curve: Figure 3(a) normal gray class; Figure 3(b) insufficient length gray class; Figure 3(c) overlapping gray class;

Figure 4 is the lateral gray sum derivative curve: Figure 4(a) normal gray class; Figure 4(b) insufficient length gray class; Figure 4(c) overlapping gray class;

Figure 5 longitudinal gray sum curve: Figure 5(a) normal gray class; Figure 5(b) insufficient length gray class; Figure 5(c) overlapping gray class;

Figure 6 is the longitudinal gray sum derivative curve: Figure 6(a) normal gray class; Figure 6(b) insufficient length gray class; Figure 6(c) overlapping gray class;

Fig. 7 is a graph showing the rising edge of the gray value of the column element;

FIG. 8 is a graph showing the falling edge of the gray value of a column element.

detailed description

The objects and functions of the present invention and methods for achieving the objects and functions will be clarified by referring to the exemplary embodiments. However, the present invention is not limited to the exemplary embodiments disclosed below; it can be implemented in various forms. The essence of the description is only to help those skilled in the relevant art comprehensively understand the specific details of the present invention.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the drawings, the same reference numerals represent the same or similar components, or the same or similar steps.

The invention provides a method for evaluating the stability of laminar flow in a flow chamber of a flow cytometer. The method uses a high-speed microscopic image acquisition system to detect the 90° Mie scattered light of microspheres in the flow chamber, and uses gray clustering analysis Methods Cluster analysis was carried out on the insufficient light intensity/smear length, normal, diffraction and overlapping in a large number of collected images, and the standard normal smear images were obtained. Then, the trailing boundary was determined using the midpoint method, and the corresponding microsphere flow rate was calculated. Finally, the stability of the flow cytometer liquid system was characterized by the stability of the flow rate of the microspheres.

The invention selects 90° side scattered light as the observation object, which can avoid direct light interference of excitation light source, and remove background light source of traditional microscopic image collection, thereby reducing background light information and improving image contrast. When the microsphere passes through the laser excitation area of the flow chamber, the trailing image of the microsphere can be collected by changing the exposure time of the high-speed camera, and then the flow velocity of the microsphere can be obtained. The schematic diagram of the high-speed image acquisition microsphere velocimetry method based on 90° Mie scattering is shown in Fig. 1 .

Since the flow cytometer can detect tens of thousands of cells per second, and the position of the microspheres in the flow chamber at the moment of exposure by the high-speed camera is random, the acquired trailing images of the microspheres generally include blank, normal, length Insufficient and overlapping 4 situations, as shown in Figure 2.

The present invention adopts the cluster analysis method based on the trapezoidal whitening weight function to classify the microsphere trailing image, and its specific implementation steps are as follows:

There are n observation objects, m evaluation indicators, and s different gray classes. Then each observation object has m characteristic data to be observed, and the sequence can be obtained as shown in formula (1):

X ₁ =(x ₁ (1),x ₁ (2),...,x ₁ (n))

X ₂ =(x ₂ (1),x ₂ (2),...,x ₂ (n))

...

X _m = (x _m (1), x _m (2), ..., x _m (n)) (1)

Determine the center point λ ₁ ,λ ₂ ,...,λ _s of the gray classes 1,2,...,s, and divide the value range of each index into s gray classes accordingly. Extend the gray class to different directions, consider adding 0 gray class and s+1 gray class, and determine its center points λ ₀ and λ _s+1 , so as to obtain a new center point sequence: λ ₀ , λ ₁ , λ ₂ ,...,λ _s ,λ _s+1 , the connection point (λ _k ,1) and the center point (λ _k-1 ,0) of the k-1th small gray class, the connection point (λ _l ,1 ) and the center point (λ _l+1 , 0) of the l+1th small gray class to obtain the trapezoidal whitening weight function of the j index with respect to the k gray class For an observed value x of index j, it can be given by

Calculate the degree of membership of the gray class k (k=1,2,...s) Calculate the comprehensive clustering coefficient of object i (i=1,2,...,n) with respect to gray class k

in, is the whitening weight function of j index k subclass, and η _j is the weight of index j in the comprehensive clustering.

Depend on Judgment object i belongs to gray class k*.

In one embodiment of the present invention, excitation light selects wavelength to be the laser diode of 605nm, and power is 80mW; Microscope objective lens is the SPAHL-50 of Japan Sigma Koki company, and numerical aperture is 0.42, magnification 50, working distance 20.5mm; Image acquisition The system uses Dantec's Q450 high-speed image acquisition system, and the supporting high-speed CMOS camera is Visionresearch's V310, with a maximum resolution of 1280×800, a maximum speed of 500,000 fps, a minimum shutter time of 1us, and a geometric size of a single pixel point of 20μm× 20 μm; the detection sample uses the standard quality control microsphere Flow-Check Pro Fluoro-spheres A69183 of Beckman Coulter Company, and the diameter of the microsphere is 20±1 μm. The exposure time was set to 40 μm, and the sample loading operation was performed using a flow cytometer. After the laminar flow state is stable, the image acquisition starts, the sampling frame rate is set to 3300 frames/s, and the total number of sampled pictures is set to 12000.

The gray values of each row of pixels in the four types of images are summed separately, and the horizontal gray sum curve is obtained, as shown in Figure 3. Set the gray value threshold to 35, and count the number of rows above the threshold as an index for cluster analysis. Due to the gray value gradient process in the vertical direction, the insufficient gray class has a slow rising edge in the horizontal gray sum curve; the normal gray class has no longitudinal gray scale gradient process, so the rising and falling edges change quickly; There is no gradient process in the vertical direction, but there are multiple tails overlapping, so the number of rising edges and falling edges is greater than 1. Solve the first derivative of the curve in Figure 3, as shown in Figure 4. Set the positive threshold to 55 and the negative threshold to -50, and count the number of extreme points within the threshold range. That is, when the amplitude of the extreme point is greater than 55 or less than -55, it will be counted as an effective extreme point. The number of valid extreme points for insufficient, normal, and overlapping is 1, 2, and 4, respectively. The number of effective extreme points can be used as an index for cluster analysis.

If the trailing length of the insufficient gray class is seriously insufficient or the gray value is insufficient, the number of effective extreme points may be 0; there are many possibilities for overlapping gray classes, and the trailing quantity, overlapping position, and overlapping method of overlapping gray classes Uncertain, the number of effective extreme points of overlapping gray classes can be other integers greater than 2.

In the same way, the gray value of each column of pixels in the four types of images is summed respectively, and the vertical gray value sum curve is obtained as shown in Fig. 5 . Set the gray value threshold to 200, and count the number of columns above the threshold as an index for cluster analysis. Insufficient gray class has a gray value gradient process in the vertical direction, and the peak value of the vertical gray value sum is small; the overlapping gray class has a low probability of completely overlapping in the vertical direction, so the number of columns whose gray value is not 0 is more than normal . Solve the first derivative of the curve in Figure 5, as shown in Figure 6. Set the positive threshold to 150 and the negative threshold to -250, and count the number of times the curve passes the positive and negative thresholds. The gray level distribution of the trailing image of the normal gray class is relatively uniform, so the first-order derivative curve of the longitudinal gray level sum has a monotonous increase and decrease characteristic, and there is no jitter near the positive and negative thresholds, and there are four intersections with the positive and negative thresholds. There are gradients or jumps in the gray distribution of trailing images with insufficient and overlapping gray classes, and there are jitters near the positive and negative thresholds, so the number of intersection points is not less than 4. The number of intersections with positive and negative thresholds can be used as an index for cluster analysis.

The velocity of the microsphere is obtained by the formula v=l/t, where l is the length of the microsphere tail, and t is the exposure time of the camera. In order to ensure an accurate assessment of the stability of the fluid system, it is necessary to analyze the rising and falling process of the trailing gray value, reasonably select the tailing boundary, and reduce the calculation error of the trailing length of the microspheres.

The gray value of the normal image is summed in columns, and 5 columns of pixels are symmetrically selected centering on the maximum value of the sum of gray values. The rising process of the gray value of the selected 5 columns of pixels is shown in Figure 7. It can be seen from Fig. 7 that the gray value of the elements in the 5 columns before the 303rd row only slightly jitters, and basically keeps stable. From column 304 to column 309, the gray value rises rapidly and changes linearly. Similarly, the process of decreasing the gray value of pixels in the five columns is shown in FIG. 8 . The gray value of pixels in the selected 5 columns is basically stable before the 420th column, and the gray value drops rapidly from the 421st column to the 428th column, and changes linearly.

Based on the characteristics of rapid linear change of pixel gray values in each column, the present invention uses the midpoint method to determine the trailing boundary. The midpoint method refers to selecting the pixel point whose gray value is closest to the average value of the edge change process as the trailing boundary. Taking Figures 7 and 8 as examples, the average gray values of the pixels in the 5 columns during the ascending process are 36.5, 36.5, 34, 33.1, and 35, and the gray values of the pixels in the 307th row (38, 37, 35, 34, and 35) The closest, so the 307th row is used as the starting boundary point of the trailing; the average gray value of the pixels in the 5 columns during the descent process is 33.2, 31.7, 31.2, 31.4 and 27.4 and the gray value of the pixel in the 307th row (32 , 30, 30, 31 and 27) are the closest, so the 424th row is taken as the starting boundary point of trailing.

Use the midpoint method to determine the tailing boundary of the normal image, and calculate the average number of pixels occupied by the length of the tailing to be 116.9, according to the formula Calculated standard deviation σ = 1.7. Since the true value of the velocity of the microsphere in the flow chamber cannot be obtained, the average value of the tail length can be used to characterize the velocity of the microsphere, and the standard deviation of the tail length can be used to evaluate the stability of the velocity of the microsphere, and then the convection flow method can be completed. Assessment of the laminar flow stability of the cytometer.

Other embodiments of the invention will be apparent to and understood by those skilled in the art from consideration of the specification and practice of the invention disclosed herein. The description and examples are considered exemplary only, with the true scope and spirit of the invention defined by the claims.

Claims (4)

1. A flow cytometer laminar flow stability assessment method, the method may further comprise the steps: 1) Use a high-speed microscopic image acquisition system to detect the 90° Mie scattered light of the microspheres in the flow chamber; 2) Use the gray clustering analysis method to cluster and analyze the light intensity/smear length insufficient, normal, diffraction and overlapping in a large number of collected images to obtain standard normal smear images. The specific implementation steps are as follows: Assuming n observation objects, m evaluation indicators, and s different gray classes, each observation object has m characteristic data to be observed, and the available sequence is shown in formula (1): X ₁ =(x ₁ (1),x ₁ (2),...,x ₁ (n)) X ₂ =(x ₂ (1),x ₂ (2),...,x ₂ (n)) ... X _m = (x _m (1), x _m (2), ..., x _m (n)) (1) Determine the center point λ ₁ , λ ₂ ,...,λ _s of the gray classes 1, 2,...,s, and divide the value range of each index into s gray classes; divide the gray classes in different directions Continuation, consider adding 0 gray class and s+1 gray class, and determine its center points λ ₀ and λ _s+1 , so as to obtain a new center point sequence: λ ₀ , λ ₁ , λ ₂ ,...,λ _s ,λ _s+1 , the connection point (λ _k ,1) and the center point (λ _k-1 ,0) of the k-1th small gray class, the connection point (λ _l ,1) and the l+1th small gray class The center point (λ _l+1 , 0) of the small gray class obtains the trapezoidal whitening weight function of the j index with respect to the k gray class (j=1,2,…,m; k,l=1,2,…,s), for an observed value x of index j, it can be given by ${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}^{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\left\{\begin{array}{cc}\mathrm{<span\; class="notranslate">\; 0</span>}& \mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; \&NotElement;</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; \&rsqb;</span>\\ \frac{\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}}{{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}}& \mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; \&rsqb;</span>\\ \mathrm{<span\; class="notranslate">\; 1</span>}& \mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; \&rsqb;</span>\\ \frac{{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; x</span>}}{{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}}& \mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>\end{array}\right.<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$ Calculate the degree of membership of the gray class k (k=1,2,...s) Calculate the comprehensive clustering coefficient of object i (i=1,2,...,n) with respect to gray class k ${\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}^{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; \&eta;</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$ in, is the whitening weight function of j index k subclass, η _j is the weight of index j in the comprehensive clustering, Depend on Judgment object i belongs to gray class k*; 3) Use the midpoint method to determine the trailing boundary, and calculate the corresponding microsphere flow rate; 4) Use the stability of the microsphere flow rate to characterize the stability of the flow cytometer fluid system. 2. method according to claim 1, wherein said step 2) in the index that cluster analysis is carried out, determine as follows: Sum the gray value of each row of pixels in the four types of images, namely light intensity/smear length insufficient, normal, diffraction and overlapping, to obtain the horizontal gray sum curve; solve the first order derivative for the horizontal gray sum curve ;Set positive and negative thresholds, and count the number of extreme points within the threshold range, and the number of effective extreme points is used as an index for cluster analysis; Sum the gray values of each column of pixels in the four types of images: insufficient light intensity/smear length, normal, diffraction and overlap, respectively, to obtain the longitudinal gray sum curve; solve the first derivative of the longitudinal gray sum curve ;Set the positive and negative thresholds, and count the number of times the curve passes through the positive and negative thresholds, and the number of intersections with the positive and negative thresholds can be used as an index for cluster analysis. 3. method according to claim 1, wherein said step 3) in the microsphere flow rate, obtain by formula v=l/t, and wherein ₁ is microsphere trailing length, and t is camera exposure time. 4. method according to claim 1, wherein said step 4) also comprises by formula Calculate the standard deviation, characterize the microsphere velocity using the mean value of the tail length, and evaluate the stability of the microsphere velocity using the standard deviation of the tail length.