Automatic sample distribution method for flow detection

Abstract

The invention provides an automatic sample distribution method for flow cytometry, which belongs to the technical field of flow cytometry and comprises the following steps: acquiring basic information of a sample; the method comprises the following steps: pre-analyzing a sample through a micro-fluidic chip to obtain sample density, sample viscosity and sample ion concentration; measuring the micro-channel diameter and the micro-channel length of the micro-fluidic chip by adopting a laser ranging system, and recording the characteristic parameters of the micro-channel material; a target distribution volume is set, a high-precision injection pump system is started for pre-injection, and an initial distribution rate is obtained through measurement of a flow sensor; calculating the initial distribution rate by using a fluid dynamic compensation equation set to obtain a corrected distribution rate; controlling the high-precision injection pump system to perform sample distribution according to the corrected distribution rate; performing real-time monitoring on a sample liquid level by using an ultrasonic sensor, and synchronously acquiring a real-time distribution rate; and sample shunting is performed through the micro-fluidic chip, so that multi-channel parallel distribution is realized.

Description

Technical Field

[0001] The invention belongs to the technical field of flow cell detection, and in particular, relates to an automatic sample distribution method for flow cell detection. Background Art

[0002] In the process of sample flow detection, accurate and stable sample distribution is a key step to ensure the accuracy of analysis results. However, there are some urgent problems to be solved in the existing flow detection sample distribution technology: First, there are large deviations and uncertainties in the sample distribution process. Since the flow of fluid in the microchannel is affected by many factors, such as the physical and chemical properties of the fluid, the geometric dimensions of the microchannel, the flow rate, etc., it is difficult to accurately predict the volume and rate of sample distribution. Traditional sample distribution methods usually adopt a constant flow rate or constant pressure method, which cannot dynamically compensate for these interference factors and cannot ensure high accuracy and repeatability of distribution. Secondly, the automation level of sample distribution is low. At present, most flow detection systems still rely on manual operation for sample loading and distribution, which is inefficient and has large human errors. Although there are some automated sample distribution systems, most of them are limited to specific application scenarios, and it is difficult to be compatible with sample containers of different types and specifications, and the flexibility is poor. Furthermore, there is a lack of real-time monitoring and feedback mechanism for the distribution process. In the prior art, the sample distribution process is generally carried out in a black box manner, that is, it is impossible to know the changes in key parameters in the distribution process in real time, and it is impossible to detect and correct abnormal situations in time, resulting in instability of analysis results.

[0003] In summary, the existing sample distribution technology for flow cytometry has the technical problem of insufficient distribution accuracy. Summary of the invention

[0004] In view of this, the present invention provides a method for automatically distributing samples for flow cytometry, which can solve the technical problem of insufficient distribution accuracy in the existing sample distribution technology for flow cytometry.

[0005] The present invention is achieved in that:

[0006] The present invention provides a method for automatically distributing samples for flow detection, comprising the following steps:

[0007] S10. Obtain basic information of the sample, including sample number, type, and collection;

[0008] S20, pre-analyzing the sample through a microfluidic chip to obtain sample density, sample viscosity, and sample ion concentration;

[0009] S30, using a laser ranging system to measure the microchannel diameter and microchannel length of the microfluidic chip, and recording the characteristic parameters of the microchannel material;

[0010] S40. Set the target dispensing volume, start the high-precision injection pump system for pre-injection, and obtain the initial dispensing rate through measurement by the flow sensor;

[0011] S50. Calculate the initial dispensing rate using the hydrodynamic compensation equation set to obtain the corrected dispensing rate;

[0012] S60. Control the high-precision injection pump system to perform sample dispensing according to the corrected dispensing rate;

[0013] S70. Use the ultrasonic sensor to monitor the sample liquid level in real time and synchronously collect the real-time dispensing rate;

[0014] S80. Perform sample shunting through the microfluidic chip to achieve multi-channel parallel dispensing, and the multi-channel parallel dispensing adopts laminar flow control.

[0015] Optionally, the step S10 specifically includes:

[0016] Step 101. Place the sample container within the acquisition field of view of the machine vision system;

[0017] Step 102. Start the image acquisition device of the machine vision system to acquire a high-definition image of the sample container;

[0018] Step 103. Preprocess the high-definition image, including image enhancement, noise elimination, and edge extraction;

[0019] Step 104. Use a deep learning model to perform target detection and segmentation on the sample container;

[0020] Step 105. Perform barcode scanning within a preset area of the sample container;

[0021] Step 106. Analyze the barcode information and extract the sample number, sample type, and collection time;

[0022] Step 107. Write the sample number, the sample type, and the collection time into the database.

[0023] The hydrodynamic compensation equation set includes a fluid resistance equation, a surface tension equation, a pressure loss equation, and a flow rate correction equation;

[0024] The fluid resistance equation is used to calculate the fluid motion resistance coefficient in the microfluidic channel. The inputs include the sample viscosity, the microchannel diameter, and the initial dispensing rate, and the output is the fluid motion resistance coefficient;

[0025] The surface tension equation is used to calculate the surface tension correction coefficient. The inputs include the sample density, the sample ion concentration, and the microchannel material characteristic parameters, and the output is the surface tension correction coefficient;

[0026] The pressure loss equation is used to calculate the friction pressure loss value. The inputs include the fluid motion resistance coefficient, the microchannel length, and the initial distribution rate, and the output is the friction pressure loss value;

[0027] The flow rate correction equation is used to calculate the corrected distribution rate. The inputs include the friction pressure loss value, the surface tension correction coefficient, and the target distribution volume, and the output is the corrected distribution rate.

[0028] The following provides a detailed explanation of the relevant equations or mathematical models:

[0029] 1. The fluid resistance equation is specifically expressed as follows:

[0030]

[0031] In the formula, f r is the fluid motion resistance coefficient; μ is the sample viscosity; v is the initial distribution rate; D is the microchannel diameter; ρ is the sample density; α1, α2, α3 are undetermined coefficients; ε1 is the error term.

[0032] Parameter acquisition method: μ is obtained through pre-analysis of the microfluidic chip; v is measured by a flow sensor; D is measured by a laser ranging system; ρ is obtained through pre-analysis of the microfluidic chip; α1, α2, α3 are obtained by least squares fitting of experimental data; the range of ε1 is 0.001 to 0.01. This equation considers the combined effects of laminar resistance (the first term), turbulent resistance (the second term), and viscous dissipation (the third term). The power-law relationship is adopted because there is a non-linear relationship between fluid resistance and velocity and geometric dimensions.

[0033] 2. The surface tension equation is specifically expressed as follows:

[0034]

[0035] In the formula, σ c is the surface tension correction coefficient; γ is the surface tension coefficient; c is the sample ion concentration; c0 is the reference ion concentration; θ is the contact angle; β1, β2, β3 are undetermined coefficients; ε2 is the error term.

[0036] Parameter acquisition method: γ is obtained by the sessile drop method: Step 1: Drop the sample on the surface of the microchannel material; Step 2: Use a high-speed camera to capture the shape of the droplet; Step 3: Obtain the surface tension coefficient through image processing. c is obtained through pre-analysis of the microfluidic chip; c0 takes the standard physiological saline concentration of 0.9%; θ is obtained by measuring with a contact angle meter; β1, β2, β3 are obtained through multiple regression analysis; the range of ε2 is 0.01 - 0.05. This equation takes into account the relationship between surface tension and density (the first term), the influence of ion concentration (the second term), and the influence of material wettability (the third term). The square root relationship is used because there is a classical square root relationship between surface tension and density.

[0037] 3. The pressure loss equation is specifically expressed as follows:

[0038]

[0039] In the formula, ΔP is the value of the frictional pressure loss; L is the length of the microchannel; n is the number of local loss points; ξ i is the local loss coefficient; λ1, λ2 are undetermined coefficients; ε3 is the error term.

[0040] Parameter acquisition method: L is obtained by a laser ranging system; ξ i is obtained through experimental calibration: Step 1: Measure the pressure loss at different flow rates; Step 2: Obtain the local loss coefficient through numerical fitting; λ1, λ2 are obtained through least squares fitting; the range of ε3 is 1 - 5 Pa. This equation is based on the Darcy - Weisbach equation, taking into account the frictional loss (the first term) and the local loss (the second term). The quadratic relationship is used because the pressure loss is proportional to the square of the flow rate.

[0041] 4. The flow rate correction equation is specifically expressed as follows:

[0042]

[0043] In the formula, v c is the corrected dispensing rate; P0 is the standard atmospheric pressure; σ0 is the surface tension of water under standard conditions; V t is the target dispensing volume; V c is the currently dispensed volume; η1, η2, η3 are undetermined coefficients; ε4 is the error term.

[0044] Parameter acquisition method: P0 takes 101325 Pa; σ0 takes 72.75 mN / m (20 °C); V t is the set value; V c is obtained through integrated flow rate: η1, η2, η3 are obtained through non-linear optimization; the range of ε4 is 0.01 - 0.05 mm / s. This equation takes into account the influence of pressure loss (the first term), the influence of surface tension (the second term), and volume compensation (the third term). The square relationship is used to enhance the correction effect for large pressure losses.

[0045] 5. The laminar flow control equation for multi-channel parallel distribution is specifically expressed as follows:

[0046]

[0047] In the formula, v i is the flow velocity of the i-th channel; Q i is the target flow rate of the i-th channel; k ij is an element of the flow rate distribution matrix; ε 5i is the error term of each channel.

[0048] Parameter acquisition method: Q i is set according to the distribution requirements; k ij is obtained through fluid network analysis: Step 1: Establish a fluid network model; Step 2: Solve the node equations; Step 3: Calibrate the matrix parameters; ε 5i has a range of 0.001 - 0.01 mm / s. This matrix equation describes the flow rate coupling relationship between multiple channels, takes into account the mutual influence between channels, and the matrix form is convenient for describing complex fluid networks.

[0049] Compared with the prior art, an automatic sample distribution method for flow cytometry provided by the present invention achieves high precision and high repeatability of sample distribution. This method uses a microfluidic chip to pre-analyze properties such as the density, viscosity, and ion concentration of the sample, and combines laser ranging technology to measure the geometric dimensions of the microchannels to establish a fine hydrodynamic model. On this basis, through means such as initial distribution rate measurement and flow velocity correction algorithm, it dynamically compensates for the influence of factors such as fluid motion resistance and surface tension, and can accurately control the distribution rate and volume, which is much better than the traditional constant flow velocity / pressure distribution method. At the same time, a real-time monitoring and feedback mechanism for the distribution process is established. This method uses an ultrasonic sensor to real-time monitor the change in the sample liquid level height, calculates the real-time distribution rate and volume, and compares them with the preset target values. When it is found that the deviation exceeds the threshold, it can timely adjust the operating parameters of the injection pump to ensure that the distribution result meets the requirements. This real-time feedback control greatly improves the stability and reliability of the analysis results. It solves the technical problem of insufficient distribution accuracy existing in the existing sample distribution technology for flow cytometry. Brief Description of the Drawings

[0050] Figure 1 is a flowchart of the method provided by the present invention;

[0051] Figure 2 It is a distribution diagram of sample property parameters, including three sub - graphs. From left to right, they are sub - graphs of density, dynamic viscosity, and ion concentration of three flies in sequence;

[0052] Figure 3 It is a three - dimensional contour map of the microchannel;

[0053] Figure 4 It is a relationship diagram of hydrodynamic parameters, including two sub - graphs. The left - hand sub - graph shows the relationship between pressure loss and flow velocity, and the right - hand sub - graph shows the relationship between surface tension coefficient and ion concentration;

[0054] Figure 5 It is a real - time monitoring diagram of the dispensing rate;

[0055] Figure 6 It is a heat map of multi - channel flow distribution. Specific implementation mode

[0056] To make the purpose, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0057] As Figure 1 shown, it is a flow chart of an automatic sample dispensing method for flow cytometry detection provided by the present invention. This method includes the following steps

[0058] S10. Obtain basic sample information, including sample number, type, and collection;

[0059] S20. Pre - analyze the sample through a microfluidic chip to obtain sample density, sample viscosity, and sample ion concentration;

[0060] S30. Use a laser ranging system to measure the diameter and length of the microchannel of the microfluidic chip, and record the material characteristic parameters of the microchannel;

[0061] S40. Set the target dispensing volume, start the high - precision injection pump system for pre - injection, and measure the initial dispensing rate through a flow sensor;

[0062] S50. Calculate the initial dispensing rate using the hydrodynamic compensation equations to obtain the corrected dispensing rate;

[0063] S60. Control the high - precision injection pump system for sample dispensing according to the corrected dispensing rate;

[0064] S70. Use an ultrasonic sensor to perform real - time monitoring of the sample liquid level and synchronously collect the real - time dispensing rate;

[0065] S80. Perform sample splitting through the microfluidic chip to achieve multi - channel parallel dispensing. The multi - channel parallel dispensing adopts laminar flow control.

[0066] The specific implementation manners of the above steps are described in detail as follows:

[0067] The specific implementation manner of step S10 is to directly obtain the basic sample information such as the sample number, type, and collection time through the sample statement, or to obtain it quickly by using the machine vision method. First, place the sample container within the acquisition field of view of the machine vision system. Then start the image acquisition device of the machine vision system to acquire a high-definition image of the sample container. Next, preprocess the acquired high-definition image, including operations such as image enhancement, noise elimination, and edge extraction. After that, use a deep learning model to perform object detection and segmentation on the sample container to determine the position of the sample container. Perform barcode scanning within the preset area of the sample container, parse the barcode information, and extract the basic sample information such as the sample number, sample type, and collection time. Finally, write this basic sample information into the database to provide basic support for subsequent sample flow tracking and data management.

[0068] The specific implementation manner of step S20 is to first inject the sample into the injection port of the microfluidic chip. Then control the electrode array in the microfluidic chip to generate an electric field driving force to make the sample flow in the microchannel. During this process, measure the flow time of the sample in the microchannel and calculate the density of the sample according to the pre-calibrated calibration curve. At the same time, collect the pressure change data of the sample in the microchannel and calculate the viscosity of the sample based on the rheology model. In addition, use a conductivity sensor to measure the conductivity of the sample, and then calculate the ion concentration of the sample according to the standard curve of conductivity and ion concentration. Finally, store the obtained parameters such as the sample density, viscosity, and ion concentration in the control system to provide a basis for subsequent automatic distribution control.

[0069] The specific implementation manner of step S30 is to first control the laser ranging system to emit a laser beam. Then collect the reflection signal of the laser beam on the surface of the microfluidic chip, calculate the lateral dimension of the microchannel according to the reflection signal to obtain the microchannel diameter. Next, perform multi-point scanning along the axial direction of the microchannel, reconstruct the three-dimensional contour of the microchannel according to the multi-point scanning data, and calculate the axial length of the microchannel to obtain the length of the microchannel. Finally, read the characteristic parameters such as the surface energy, contact angle, and roughness of the microchannel material from the material database to provide the necessary geometric and material parameters for subsequent hydrodynamic calculations.

[0070] The specific implementation of step S40 is as follows: First, input the target dispensing volume into the control system. Then, calculate the theoretical injection time based on the target dispensing volume and set the injection parameters of the high-precision injection pump system, including the injection rate and injection volume. Next, control the high-precision injection pump system to perform a pre-injection while collecting the real-time data of the flow sensor. Filter the real-time data, calculate the average value of the real-time data, and obtain the initial dispensing rate. The purpose of this step is to obtain the initial dispensing rate through the actual pre-injection and provide basic data for subsequent hydrodynamic compensation.

[0071] The specific implementation of step S50 is as follows: First, substitute parameters such as the sample density, sample viscosity, microchannel diameter, and initial dispensing rate into the fluid resistance equation to solve for the fluid motion resistance coefficient. This fluid resistance equation comprehensively considers factors such as laminar resistance, turbulent resistance, and viscous dissipation, and is described by a power relationship, which can accurately predict the fluid motion resistance. Next, substitute the sample density, sample ion concentration, and microchannel material property parameters into the surface tension equation to solve for the surface tension correction coefficient. This surface tension equation considers the relationship between surface tension and factors such as density, ion concentration, and wettability, and is described by a square root relationship, which can better predict the influence of surface tension. Then, substitute the fluid motion resistance coefficient, microchannel length, and initial dispensing rate into the pressure loss equation to solve for the value of the pressure loss along the way. This pressure loss equation is based on the Darcy-Weisbach equation, comprehensively considers the loss along the way and local loss, and is described by a quadratic relationship, which can more accurately predict the pressure loss. Finally, substitute the value of the pressure loss along the way, the surface tension correction coefficient, and the target dispensing volume into the flow rate correction equation to solve for the corrected dispensing rate. This flow rate correction equation considers the influence of factors such as pressure loss, surface tension, and volume compensation, and is described by a square relationship, which can effectively correct the dispensing rate.

[0072] The specific implementation of step S60 is as follows: First, convert the corrected dispensing rate into a control instruction for the injection pump and send this instruction to the high-precision injection pump system. Then, control the high-precision injection pump system to perform sample dispensing according to the corrected dispensing rate. During this process, monitor the operating state of the high-precision injection pump system in real time and record the actually dispensed volume. Next, calculate the error between the actually dispensed volume and the target dispensing volume. If the error exceeds the preset threshold (such as 5%), then trigger an alarm signal to prompt that the dispensing parameters need to be adjusted. The purpose of this step is to ensure that the actual dispensing result meets the expected goal and optimize the dispensing control through feedback adjustment.

[0073] The specific implementation of step S70 is as follows: First, control the ultrasonic sensor to emit ultrasonic waves towards the sample liquid level. Then, receive the ultrasonic wave signal reflected by the sample liquid level, and calculate the height of the sample liquid level based on the time difference. Next, calculate the real-time volume change according to the change in the height of the sample liquid level, so as to obtain the real-time dispensing rate. Finally, compare this real-time dispensing rate with the previously corrected dispensing rate. If the deviation between the two exceeds the preset range (for example, 10%), trigger the rate adjustment mechanism to re-adjust the operating parameters of the high-precision syringe pump. The purpose of this step is to monitor the rate change in the dispensing process in real time, perform feedback control in a timely manner, and ensure the accuracy and stability of the dispensing result.

[0074] The specific implementation of step S80 is as follows: First, determine the number of channels of the microfluidic chip according to the actual dispensing requirements. Then, establish a fluid network model of the microfluidic chip and calculate the flow distribution coefficients of each channel. Next, set the opening degrees of the flow control valves of each channel according to these flow distribution coefficients and start the flow splitting mechanism of the microfluidic chip. During the dispensing process, monitor the flow data of each channel in real time and dynamically adjust the opening degrees of the flow control valves to maintain a stable laminar flow state. The laminar flow control equation for multi-channel parallel dispensing describes the flow coupling relationship between channels in matrix form and can better handle complex fluid networks. The purpose of this step is to achieve uniform dispensing of the sample among multiple channels and ensure consistent analysis conditions for each test channel.

[0075] Specifically, the principle of the present invention is: establish a closed-loop dynamic control system, and through collecting the physical and chemical property parameters of the sample and the geometric dimension parameters of the microchannels, and combining with a fine fluid dynamics model, achieve precise control and real-time feedback of the sample dispensing process.

[0076] First, before sample dispensing, the method uses machine vision technology to identify the sample container and read barcode information to obtain basic information of the sample, such as sample number, type, collection time, etc. This provides basic support for subsequent sample flow tracking and data management.

[0077] Secondly, pre-analyze the sample through the microfluidic chip to measure physical and chemical property parameters of the sample, such as density, viscosity, and ion concentration. These parameters are key inputs for fluid dynamics modeling. At the same time, use laser ranging technology to measure the geometric dimensions of the microchannels, including diameter and length, as well as characteristics such as the surface energy and contact angle of the microchannel material. These geometric and material science parameters are also necessary inputs for fluid dynamics calculations.

[0078] Based on the obtained sample property parameters and microchannel parameters, this method establishes a hydrodynamic compensation equation set including a fluid resistance equation, a surface tension equation, a pressure loss equation, and a flow rate correction equation. This equation set can accurately predict the flow behavior of the sample in the microchannel and dynamically compensate for the influence of factors such as fluid motion resistance and surface tension. By solving these equations, the corrected optimal dispensing rate can be obtained, providing a basis for subsequent precise dispensing.

[0079] Next, this method uses a high-precision injection pump system to dispense samples according to the corrected dispensing rate. At the same time, the ultrasonic sensor is used to monitor the change of the sample liquid level in real time, calculate the real-time dispensing rate and volume, and compare them with the target values. If the deviation is found to exceed the preset range, the rate adjustment mechanism will be triggered to reset the operating parameters of the injection pump to ensure that the dispensing result meets the requirements.

[0080] In addition, in order to achieve uniform parallel dispensing of samples among multiple channels, this method adopts a laminar flow control strategy. By establishing a fluid network model of the microfluidic chip, the flow rate distribution coefficient of each channel is calculated, and the opening degree of the flow control valve is dynamically adjusted to maintain a stable laminar flow state.

[0081] Next, a specific embodiment 1 of this method is provided. The specific implementation methods of each step in this embodiment 1 are described in detail as follows: First, in step S10, this method uses a machine vision system to identify the sample container and read the sample barcode information. The specific implementation process is as follows:

[0082] In the first step, the sample container is placed within the acquisition field of view of the machine vision system.

[0083] In the second step, the image acquisition device of the machine vision system is started to acquire a high-definition image of the sample container. The image acquisition process can be described as:

[0084] I = f capture (t);

[0085] In the formula, I represents the acquired image, f capture represents the image acquisition function, and t represents the acquisition time.

[0086] In the third step, preprocessing is performed on the acquired high-definition image, including operations such as image enhancement, noise elimination, and edge extraction. The image preprocessing process can be described as:

[0087] I pre = f preprocess (I);

[0088] In the formula, I pre represents the preprocessed image, and f preprocess represents the image preprocessing function.

[0089] Step 4: Use a deep learning model to perform object detection and segmentation on the sample container. The object detection and segmentation process can be described as:

[0090] O = f detect (I pre );

[0091] In the formula, O represents the sample container area obtained from detection and segmentation, and f detect represents the object detection and segmentation function.

[0092] Step 5: Perform barcode scanning within the preset area of the sample container, parse the barcode information, and extract basic sample information such as the sample number, sample type, and collection time. The barcode parsing process can be described as:

[0093] {S id , S type , S time} = f decode (O);

[0094] In the formula, S id represents the sample number, S type represents the sample type, S time represents the collection time, and f decode represents the barcode decoding function.

[0095] Step 6: Write the extracted basic sample information into the database to provide basic support for subsequent sample flow tracking and data management. The database writing process can be described as:

[0096] {S id , S type , S time} → Database;

[0097] In the formula, Database represents the sample information database.

[0098] In step S20, the method pre-analyzes the sample through a microfluidic chip to obtain parameters such as the sample density, viscosity, and ion concentration. The specific implementation process is as follows:

[0099] Step 1: Inject the sample into the inlet of the microfluidic chip. The sample injection process can be described as:

[0100] S → Chip;

[0101] In the formula, S represents the sample to be analyzed.

[0102] Step 2: Control the electrode array in the microfluidic chip to generate an electric field driving force to make the sample flow in the microchannel. The electric field driving process can be described as:

[0103]

[0104] Wherein, represents the electric field driving force, q represents the charge of the sample ion, represents the electric field strength in the microfluidic chip.

[0105] In the third step, measure the flow time of the sample in the microchannel, and calculate the density of the sample according to the pre-calibrated calibration curve. The density calculation process can be described as:

[0106] ρ = f ρ (t);

[0107] Wherein, ρ represents the sample density, t represents the flow time of the sample in the microchannel, and f ρ represents the density calculation function.

[0108] In the fourth step, collect the pressure change data of the sample in the microchannel, and calculate the viscosity of the sample based on the rheological model. The viscosity calculation process can be described as:

[0109] μ = f μ (ΔP);

[0110] Wherein, μ represents the sample viscosity, ΔP represents the pressure change of the sample in the microchannel, and f μ represents the viscosity calculation function.

[0111] In the fifth step, measure the conductivity of the sample using a conductivity sensor, and calculate the ion concentration of the sample according to the standard curve of conductivity and ion concentration. The ion concentration calculation process can be described as:

[0112] c = f c (κ);

[0113] Wherein, c represents the sample ion concentration, κ represents the sample conductivity, and f c represents the ion concentration calculation function.

[0114] In the sixth step, store the obtained parameters such as the sample density ρ, viscosity μ, and ion concentration c into the control system to provide a basis for subsequent automatic distribution control. The parameter storage process can be described as:

[0115] {ρ, μ, c} → ControlSystem;

[0116] In step S30, the method uses a laser ranging system to measure the diameter and length of the microchannel of the microfluidic chip, and records the microchannel material property parameters. The specific implementation process is as follows:

[0117] In the first step, control the laser ranging system to emit a laser beam. The laser emission process can be described as:

[0118] L = f laser (t);

[0119] where L represents the laser beam, f laser represents the laser emission function, and t represents the emission time.

[0120] In the second step, collect the reflection signal of the laser beam on the surface of the microfluidic chip. The reflection signal collection process can be described as:

[0121] R = f detect (L, Chip);

[0122] where R represents the reflection signal, f detect represents the reflection signal detection function.

[0123] In the third step, calculate the lateral dimension of the microchannel based on the reflection signal to obtain the microchannel diameter D. The diameter calculation process can be described as:

[0124] D = f D (R);

[0125] where f D represents the diameter calculation function.

[0126] In the fourth step, perform multi-point scanning along the axial direction of the microchannel, and reconstruct the three-dimensional profile of the microchannel based on the multi-point scanning data. The three-dimensional profile reconstruction process can be described as:

[0127] P = f profile (R);

[0128] where P represents the three-dimensional profile, f profile represents the profile reconstruction function.

[0129] In the fifth step, calculate the axial projection length of the three-dimensional profile to obtain the microchannel length L. The length calculation process can be described as:

[0130] L = f L (P);

[0131] where f L represents the length calculation function.

[0132] In the sixth step, read the characteristic parameters such as the surface energy γ, contact angle θ, and roughness ε of the microchannel material from the material database. The material parameter reading process can be described as:

[0133] {γ, θ, ε} = f material (Chip);

[0134] where f material represents the material parameter reading function.

[0135] In step S40, the method sets the target dispensing volume, starts the high-precision injection pump system for pre-injection, and measures the initial dispensing rate through a flow sensor. The specific implementation process is as follows:

[0136] First step, input the target dispensing volume V into the control system target .

[0137] Second step, calculate the theoretical injection time t according to the target dispensing volume theory :

[0138]

[0139] In the formula, v init represents the initial dispensing rate.

[0140] Third step, set the injection parameters of the high-precision injection pump system, including the injection rate v pump and the injection volume V pump .

[0141] Fourth step, control the high-precision injection pump system to perform pre-injection. The pre-injection process can be described as:

[0142]

[0143] In the formula, V pre represents the pre-injection volume, and t pre represents the pre-injection time.

[0144] Fifth step, collect the real-time data Q(t) of the flow sensor.

[0145] Sixth step, perform filtering processing on the real-time data to obtain the filtered data Q filtered (t). The filtering process can be described as:

[0146] Q filtered (t) = f filter (Q(t));

[0147] In the formula, f filter represents the filtering function.

[0148] Seventh step, calculate the average value of the filtered data to obtain the initial dispensing rate v init :

[0149]

[0150] In step S50, the method calculates the initial dispensing rate using the hydrodynamic compensation equation set to obtain the corrected dispensing rate. The specific implementation process is as follows:

[0151] In the first step, substitute the sample density ρ, sample viscosity μ, microchannel diameter D, and initial dispensing rate v init into the fluid resistance equation to solve for the fluid motion resistance coefficient f r :

[0152]

[0153] where α1, α2, α3 are undetermined coefficients and ε1 is an error term.

[0154] In the second step, substitute the sample density ρ, sample ion concentration c, and microchannel material characteristic parameters {γ, θ, ε} into the surface tension equation to solve for the surface tension correction coefficient σ c :

[0155]

[0156] where β1, β2, β3 are undetermined coefficients, ε2 is an error term, and c0 is the reference ion concentration.

[0157] In the third step, substitute the fluid motion resistance coefficient f r , microchannel length L, and initial dispensing rate v init into the pressure loss equation to solve for the friction pressure loss value ΔP:

[0158]

[0159] where λ1, λ2 are undetermined coefficients, ξ i is the i-th local loss coefficient, ε3 is an error term, and n is the number of local loss points.

[0160] In the fourth step, substitute the friction pressure loss value ΔP, surface tension correction coefficient σ c , and target dispensing volume V target into the flow rate correction equation to solve for the corrected dispensing rate v c :

[0161]

[0162] where P0 is the standard atmospheric pressure, σ0 is the surface tension of water under standard conditions, V c is the currently dispensed volume, η1, η2, η3 are undetermined coefficients, and ε4 is an error term.

[0163] In step S60, the method controls the high-precision syringe pump system for sample dispensing according to the corrected dispensing rate. The specific implementation process is as follows:

[0164] In the first step, convert the corrected dispensing rate v c into a control command for the syringe pump.

[0165] In the second step, send the control instruction to the high-precision injection pump system.

[0166] In the third step, control the high-precision injection pump system to perform sample dispensing according to the corrected dispensing rate v c The dispensing process can be described as follows:

[0167]

[0168] where, V a represents the actual dispensed volume, and t a represents the dispensing time.

[0169] In the fourth step, monitor the operating state of the high-precision injection pump system in real time.

[0170] In the fifth step, record the actual dispensed volume V of the high-precision injection pump system a .

[0171] In the sixth step, calculate the error ε between the actual dispensed volume V a and the target dispensed volume V target : V

[0172]

[0173] In the seventh step, if the error ε V exceeds the preset threshold (such as 5%), trigger an alarm signal.

[0174] In step S70, the method uses an ultrasonic sensor to monitor the sample liquid level in real time and synchronously collect the real-time dispensing rate. The specific implementation process is as follows:

[0175] In the first step, control the ultrasonic sensor to emit ultrasonic waves towards the sample liquid level. The ultrasonic wave emission process can be described as:

[0176] U = f ultrasonic (t);

[0177] where, U represents the ultrasonic wave signal, and f ultrasoic represents the ultrasonic wave emission function.

[0178] In the second step, receive the ultrasonic wave signal reflected from the sample liquid level. The reflected signal reception process can be described as:

[0179] R U = f detect (U, S);

[0180] where, R U represents the reflected signal, and f detect represents the reflected signal detection function.

[0181] In the third step, calculate the sample liquid level height h based on the time difference of the reflected signals:

[0182]

[0183] where c U is the propagation speed of ultrasonic waves in the sample medium, and Δt is the time difference between the transmitted and received signals.

[0184] In the fourth step, calculate the real-time volume change ΔV based on the change in the sample liquid level height h:

[0185] ΔV = A·Δh;

[0186] where A is the cross-sectional area of the sample container.

[0187] In the fifth step, calculate the real-time dispensing rate v based on the real-time volume change ΔV r :

[0188]

[0189] In the sixth step, compare the real-time dispensing rate v r with the corrected dispensing rate v c for comparison.

[0190] In the seventh step, if |v r - v c | > ε v , then trigger the rate adjustment mechanism to re-adjust the parameters of the syringe pump.

[0191] In step S80, the method uses laminar flow control to achieve multi-channel parallel dispensing of the microfluidic chip. The specific implementation process is as follows:

[0192] In the first step, determine the number of channels m according to the dispensing requirements.

[0193] In the second step, establish a fluid network model of the microfluidic chip and calculate the flow distribution coefficient k of each channel ij :

[0194]

[0195] where v i is the flow velocity of the i-th channel, Q i is the target flow of the i-th channel, and ε 5i is the error term of the i-th channel.

[0196] In the third step, set the opening degree of the flow control valve of each channel according to the flow distribution coefficient k ij of each channel.

[0197] In the fourth step, start the shunt mechanism of the microfluidic chip.

[0198] Step 5: Monitor the flow rate data v of each channel in real time i 。

[0199] Step 6: Dynamically adjust the valve opening according to the flow rate data to maintain a stable laminar flow state.

[0200] To better understand and implement the present invention, Example 2 of a specific application scenario of the present invention is provided below: When a certain testing enterprise conducts a disease diagnosis and testing service, it adopts the automatic sample allocation method for flow cytometry testing proposed by the present invention.

[0201] First, before the sample enters the testing process, the staff places the sample container within the acquisition range of the machine vision system. The system then activates the image acquisition device and captures a high-definition image of the sample container. Through preprocessing, including operations such as image enhancement, noise elimination, and edge extraction, the quality and analyzability of the image are improved.

[0202] After that, the system uses a deep learning object detection model to analyze the image, quickly locates and segments the area of the sample container. Within this preset area, the system activates the barcode scanning module to read the QR code information on the sample container. Through decoding, basic information such as the sample number, type, and collection time is extracted and written into the central sample information database, laying a foundation for subsequent sample flow tracking and data management.

[0203] After the sample information is collected, the staff injects the sample into a pre-prepared microfluidic chip. The chip is equipped with an electrode array that can generate an electric field driving force to promote the flow of the sample in the microchannel. During the flow of the sample, the system collects a series of key parameters:

[0204] First, by measuring the flow time of the sample in the microchannel and combining it with the previously established density-flow time calibration curve, the density of the sample is calculated. The test results show that the average density of a certain batch of nucleic acid testing samples for a certain disease is 1.005 g / cm3, and the standard deviation is 0.002 g / cm3.

[0205] Secondly, the system collects the pressure change data of the sample in the microchannel and analyzes and calculates it based on a rheological model to obtain the dynamic viscosity of the sample. The test results show that the average dynamic viscosity of the sample is 1.002 mPa·s, and the standard deviation is 0.005 mPa·s.

[0206] In addition, the system also measures the conductivity of the sample through a conductivity sensor and converts it to obtain the ion concentration with reference to the standard curve. The test results show that the average ion concentration of the sample is 149 mmol / L, and the standard deviation is 1 mmol / L.

[0207] The above sample property parameters provide the necessary input data for subsequent hydrodynamic modeling.

[0208] Figure 2 It is a distribution diagram of sample property parameters, showing the distribution of three key physical parameters of the sample: density, dynamic viscosity, and ion concentration. The distribution characteristics of 100 samples are presented in the form of a histogram, and the average value line is marked.

[0209] Meanwhile, the center also used a laser ranging system to conduct a detailed measurement of the microchannels of the microfluidic chip. First, the system emitted a laser beam to irradiate the chip surface and collected the reflected signals. By analyzing these reflected signals, the lateral dimension of the microchannel was calculated, and the microchannel diameter was obtained as 50 μm with a standard deviation of 1 μm.

[0210] Next, the system performed multi-point scanning along the axial direction of the microchannel to obtain the three-dimensional contour data of the microchannel. After data processing and reconstruction, the length of the microchannel was determined to be 20 mm with a standard deviation of 0.1 mm.

[0211] In addition, it was read from the material database that the surface energy of the microfluidic chip is 72 mN / m, the contact angle is 45°, and the surface roughness is 50 nm. These microchannel geometry and material parameters will also serve as inputs for hydrodynamic modeling.

[0212] Figure 3 It is a three-dimensional contour diagram of the microchannel, using a three-dimensional surface map to show the precise contour of the microchannel, including the subtle variations in surface roughness.

[0213] With the sample property parameters and microchannel parameters, the system can next establish a hydrodynamic compensation equation set to perform fine modeling and dynamic control of the sample dispensing process.

[0214] First, parameters such as sample density, viscosity, microchannel diameter, and initial dispensing rate are substituted into the fluid resistance equation to calculate the fluid motion resistance coefficient. Through experimental calibration, the undetermined coefficients of the fluid resistance equation were determined as α1 = 96, α2 = 1.2, α3 = 0.15, and the range of the error term ε1 is between 0.001 and 0.01.

[0215] Second, sample density, ion concentration, and microchannel material parameters are substituted into the surface tension equation to solve for the surface tension correction coefficient. Through multiple regression analysis, the undetermined coefficients of the surface tension equation were determined as β1 = 0.08, β2 = 0.12, β3 = 0.05, and the range of the error term ε2 is between 0.01 and 0.05. Referring to standard normal saline, the center set the reference ion concentration c0 as 0.9%.

[0216] Next, parameters such as the fluid resistance coefficient, microchannel length, and initial dispensing rate are substituted into the pressure loss equation to calculate the value of the pressure loss along the way. Through experimental measurement and numerical fitting, the undetermined coefficients of the pressure loss equation are determined as λ1 = 1.2, λ2 = 0.8, and the local loss coefficients ξ i are 0.5, 0.3, 0.2 respectively, and the range of the error term ε3 is between 1 and 5 Pa.

[0217] Finally, the value of the pressure loss along the way, the surface tension correction coefficient, and the target dispensing volume are substituted into the flow rate correction equation to solve for the corrected optimal dispensing rate. Through nonlinear optimization, the undetermined coefficients of the flow rate correction equation are determined as η1 = 0.6, η2 = 0.4, η3 = 0.3, and the range of the error term ε4 is between 0.01 and 0.05 mm / s.

[0218] With the above hydrodynamic compensation equations, the system can dynamically calculate the optimal dispensing rate based on the actually measured sample property parameters and microchannel parameters, providing a basis for subsequent precise dispensing.

[0219] Figure 4 It is a diagram of hydrodynamic parameter relationships, showing two important diagrams: the relationship between pressure loss and flow rate, and the relationship between surface tension coefficient and ion concentration.

[0220] Next, the staff input a target dispensing volume of 50 μL into the control system. Based on this target volume, the system calculates the theoretical injection time as 50 s. Then, the system sets the injection rate of the high-precision injection pump to 1 μL / s and the injection volume to 50 μL for a pre-injection test.

[0221] During the pre-injection process, the system collects the real-time data of the flow sensor and filters it. Finally, the initial dispensing rate is calculated as 0.998 μL / s, and the standard deviation is 0.005 μL / s.

[0222] With the initial dispensing rate, combined with the previously established hydrodynamic compensation equations, the system calculates the corrected optimal dispensing rate as 1.002 μL / s.

[0223] The system immediately converts this corrected dispensing rate into a control command for the injection pump and sends it to the high-precision injection pump. The injection pump performs sample dispensing according to this command. During the dispensing process, the system monitors the operating state of the injection pump in real time and records the actually dispensed volume.

[0224] Meanwhile, the system also uses ultrasonic sensors to measure the height of the sample liquid surface in real time. According to the change in the liquid surface height, the real-time dispensing rate and volume are calculated. By comparing the corrected target dispensing rate with the measured real-time dispensing rate, it is found that the deviation between the two is always controlled within 5%, meeting the experimental requirements.

[0225] Figure 5 This is a real-time monitoring graph of the dispensing rate, showing the real-time comparison between the target dispensing rate and the actual dispensing rate during the 50-second dispensing process, including the error range band of ±5%.

[0226] In addition to single-channel dispensing, the automatic dispensing system at this center also supports multi-channel parallel dispensing. For 100 COVID-19 nucleic acid test samples on the same day, the system determined a plan to use 10 parallel test channels.

[0227] First, the engineers at this center established a fluid network model of the microfluidic chip and calculated the flow distribution coefficient matrix for each channel:

[0228]

[0229] According to this flow distribution coefficient matrix, the system set the opening degrees of the flow control valves for each test channel to ensure that the flow ratio of each channel meets the requirements and maintains a stable laminar flow state.

[0230] Figure 6 This is a heat map of multi-channel flow distribution, visually showing the 10×10 flow distribution coefficient matrix in the form of a heat map and clearly displaying the flow distribution relationship between each channel.

[0231] During the actual dispensing process, the system monitored the flow data of each test channel in real time and adjusted the valve opening degrees in a timely manner to ensure that the 100 samples were evenly and orderly dispensed among the 10 channels. The entire dispensing process took only 15 minutes, greatly improving the sample processing efficiency.

[0232] The testing personnel then carried out subsequent nucleic acid amplification and fluorescence detection on the dispensed samples. The detection results of all samples showed good conditions and met the quality control requirements. Compared with the previous manual dispensing method, this automatic dispensing system increased the detection efficiency by nearly 50%, greatly reducing human errors and providing strong support for the COVID-19 diagnosis work of testing enterprises.

[0233] Generally speaking, when conducting the diagnostic test for a certain disease, the testing enterprise fully applies the automatic sample allocation method for flow cytometry testing proposed in the present invention. By means of machine vision technology for sample information acquisition, microfluidic technology for sample property analysis, laser ranging technology for measuring microchannel parameters, and ultrasonic sensors for real-time monitoring, a closed-loop dynamic control system is established. This system can accurately predict the flow behavior of samples in the microchannels and accordingly control a high-precision injection pump for optimized sample allocation. Meanwhile, the system supports multi-channel parallel allocation, which can significantly improve the throughput of sample testing.

[0234] In implementation, the allocation accuracy of this automatic allocation system can reach within 5%, which is significantly better than the traditional constant flow rate / pressure allocation method. Meanwhile, the system realizes fully automated operation throughout the process, greatly improving work efficiency and reducing human errors. In addition, through real-time monitoring and feedback control, the stability and reliability of the analysis results are ensured.

[0235] It should be noted that the interpretation of the variable parameters involved in the present invention is shown in Table 1 below.

[0236] Table 1 Interpretation Table of Variable Parameters

[0237]

[0238]

[0239] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention.

Claims

1. A method for automatically distributing samples for flow cytometry, characterized in that: The following steps are involved: S10. Obtain basic information of the sample, including sample number, type, and collection; S20, pre-analyzing the sample through a microfluidic chip to obtain sample density, sample viscosity, and sample ion concentration; S30, using a laser ranging system to measure the microchannel diameter and microchannel length of the microfluidic chip, and recording the characteristic parameters of the microchannel material; S40, setting the target dispensing volume, starting the high-precision injection pump system for pre-injection, and obtaining the initial dispensing rate through measurement by the flow sensor; S50, calculating the initial distribution rate using a group of fluid dynamics compensation equations to obtain a corrected distribution rate; S60, controlling the high-precision syringe pump system to perform sample distribution according to the corrected distribution rate; S70, using an ultrasonic sensor to monitor the sample liquid level in real time and synchronously collect the real-time distribution rate; S80, performing sample diversion through the microfluidic chip to achieve multi-channel parallel distribution, wherein the multi-channel parallel distribution adopts laminar flow control.

2. The automatic sample distribution method for flow detection according to claim 1, characterized in that: The fluid dynamics compensation equation group includes a fluid resistance equation, a surface tension equation, a pressure loss equation, and a flow rate correction equation; The fluid resistance equation is used to calculate the fluid motion resistance coefficient in the microfluidic channel, the input includes the sample viscosity, the microchannel diameter, the initial distribution rate, and the output is the fluid motion resistance coefficient; The surface tension equation is used to calculate the surface tension correction coefficient, the input includes the sample density, the sample ion concentration, and the microchannel material characteristic parameters, and the output is the surface tension correction coefficient; The pressure loss equation is used to calculate the pressure loss value along the way, the input includes the fluid motion resistance coefficient, the microchannel length, and the initial distribution rate, and the output is the pressure loss value along the way; The flow rate correction equation is used to calculate the corrected distribution rate, and the input includes the pressure loss value along the process, the surface tension correction coefficient, and the target distribution volume, and the output is the corrected distribution rate.

3. The automatic sample distribution method for flow detection according to claim 1, characterized in that: The step S20 specifically includes: Step 201, injecting the sample into the injection port of the microfluidic chip; Step 202: controlling the electrode array in the microfluidic chip to generate an electric field driving force; Step 203, measuring the flow time of the sample in the microchannel, and calculating the sample density according to a pre-calibrated curve; Step 204, collecting pressure change data of the sample in the microchannel, and calculating the viscosity of the sample according to a rheological model; Step 205: measuring the conductivity of the sample using a conductivity sensor; Step 206, calculating the sample ion concentration according to the standard curve of conductivity and ion concentration; Step 207: Store the sample density, the sample viscosity, and the sample ion concentration into a control system.

4. The automatic sample distribution method for flow detection according to claim 1, characterized in that: The step S30 specifically includes: Step 301, controlling the laser ranging system to emit a laser beam; Step 302: collecting the reflection signal of the laser beam on the surface of the microfluidic chip; Step 303: Calculate the transverse dimension of the microchannel according to the reflection signal to obtain the diameter of the microchannel; Step 304, performing multi-point scanning along the axial direction of the microchannel; Step 305, reconstructing the three-dimensional profile of the microchannel according to the multi-point scanning data; Step 306, calculating the axial projection length of the three-dimensional profile to obtain the microchannel length; Step 307: Read the characteristic parameters of the microchannel material from the material database, including surface energy, contact angle, and roughness.

5. The automatic sample distribution method for flow detection according to claim 1, characterized in that: The step S40 specifically includes: Step 401, inputting the target dispensing volume into the control system; Step 402, calculating the theoretical injection time according to the target distribution volume; Step 403, setting the injection parameters of the high-precision injection pump system, including injection rate and injection volume; Step 404, controlling the high-precision injection pump system to perform pre-injection; Step 405: collecting real-time data of the flow sensor; Step 406: filtering the real-time data; Step 407: Calculate the average value of the real-time data to obtain the initial allocation rate.

6. The automatic sample distribution method for flow detection according to claim 1, characterized in that: The step S50 specifically includes: Step 501, substituting the sample density, the sample viscosity, the microchannel diameter, and the initial distribution rate into the fluid resistance equation; Step 502, solving the fluid resistance equation to obtain the fluid motion resistance coefficient; Step 503, substituting the sample density, the sample ion concentration, and the microchannel material characteristic parameters into the surface tension equation; Step 504, solving the surface tension equation to obtain the surface tension correction coefficient; Step 505, substituting the fluid motion resistance coefficient, the microchannel length, and the initial distribution rate into the pressure loss equation; Step 506, solving the pressure loss equation to obtain the pressure loss value along the route; Step 507, substituting the pressure loss value along the way, the surface tension correction coefficient, and the target distribution volume into a flow rate correction equation; Step 508: Solve the flow rate correction equation to obtain the corrected distribution rate.

7. The automatic sample distribution method for flow detection according to claim 1, characterized in that: The step S60 specifically includes: Step 601, converting the corrected dispensing rate into a syringe pump control instruction; Step 602: Send the syringe pump control instruction to the high-precision syringe pump system; Step 603, controlling the high-precision syringe pump system to operate according to the corrected dispensing rate; Step 604: monitor the operating status of the high-precision injection pump system in real time; Step 605, recording the actual dispensed volume of the high-precision syringe pump system; Step 606, calculating the error between the actual dispensing volume and the target dispensing volume; Step 607: When the error exceeds a preset threshold, an alarm signal is triggered.

8. The automatic sample distribution method for flow detection according to claim 1, characterized in that: The step S70 specifically includes: Step 701, controlling the ultrasonic sensor to emit ultrasonic waves toward the sample liquid surface; Step 702, receiving the ultrasonic signal reflected by the sample liquid surface; Step 703, calculating the sample liquid level according to the time difference of the ultrasonic signal; Step 704, calculating the real-time volume change according to the change in the sample liquid level; Step 705: Calculate the real-time distribution rate according to the real-time volume change; Step 706: compare the real-time allocation rate with the corrected allocation rate; Step 707: When the real-time allocation rate deviates from the corrected allocation rate by more than a preset range, a rate adjustment mechanism is triggered.

9. The automatic sample distribution method for flow detection according to claim 1, characterized in that: The step S80 specifically includes: Step 801, determining the number of channels according to allocation requirements; Step 802: establishing a fluid network model of the microfluidic chip; Step 803, calculating the flow distribution coefficient of each channel; Step 804, setting the flow control valve opening of each channel according to the flow distribution coefficient; Step 805, starting the flow diversion mechanism of the microfluidic chip; Step 806: monitor the flow data of each channel in real time; Step 807: dynamically adjust the flow control valve opening according to the flow data to maintain a laminar flow state.

Images