A single-cell high-throughput detection system and method based on vertical three-dimensional focusing

Abstract

The present invention discloses a single-cell high-throughput detection system and method based on vertical three-dimensional focusing, which belongs to the field of cell detection technology. It includes a sample reaction module, a vertical three-dimensional sheath flow focusing module, a fluid control module and a single-cell multi-frequency impedance detection module; the present invention carries out a design based on a vertical three-dimensional sheath flow channel, so that the flow direction of the sample flow in the detection area is in the same straight line with the gravity direction, reducing the cell loss during the detection process; using a high-speed sample flow to flow in a hose, it is difficult for the cells to settle, further reducing the cell loss during the detection process; by constructing a combination of positive pressure and negative pressure systems, a complete set of automatic injection, detection, and cleaning processing procedures are designed, realizing the automation of three-dimensional focused single-cell detection, and improving the maximum throughput of the actual operation of the system.

Description

Technical Field

[0001] The present invention belongs to the technical field of cell detection, and in particular relates to a single-cell high-throughput detection system and method based on vertical three-dimensional focusing. Background Art

[0002] White blood cells are an important part of the human immune system. White blood cells in peripheral blood can be classified based on morphology and function. Accurate classification and counting of different types of white blood cells is of great significance in clinical diagnosis. The gold standard for white blood cell classification and counting is microscopic examination. Usually, the blood smear is stained first, and then trained professionals use counting plates to count the stained blood smear under a microscope. Although this method is still the mainstream in many clinical hospitals, it is heavily dependent on the proficiency of technicians, and the time and labor costs are relatively high. With the development of artificial intelligence this year, automated blood smear detection instruments such as slide pushers have gradually been used to improve the classification performance of microscopic examinations and reduce the demand for labor costs, but they still do not solve the problem of low microscopic examination throughput.

[0003] In response to the clinical demand for rapid classification and typing of white blood cells, detection principles such as electrical impedance and light scattering are applied to the classification and counting of various types of cells in peripheral blood. Currently, commercial blood cell analyzers using this principle have been widely used in clinical testing in hospitals. When classifying white blood cells, a single cell quickly passes through the detection area, causing the scattered light or electrical impedance waveform to change and be collected. A scatter plot is then drawn to determine the threshold for counting and classifying cells. Commercial blood cell analyzers generally have a high throughput, but in terms of electrical detection, the impedance signal changes at the single-cell level are weak, the sensitivity is low, and the accuracy of white blood cell identification and classification is limited. They can only be used as a rapid initial screening tool and cannot replace the "gold standard" of microscopy.

[0004] Since microfluidic technology has good adaptability to detection at the single-cell scale, the microfluidic detection platform based on micromachining technology improves the performance of existing single-cell detection methods. The high-precision and high-consistency microchannels and microelectrodes made by micromachining technology improve the sensitivity and consistency of single-cell impedance detection. However, in terms of electrical detection, due to the small number of electric field lines blocked by the measured cells, it is impossible to obtain high-sensitivity impedance changes of single cells. In response to the problems existing in the existing microfluidic single-cell detection platform, a single-cell detection platform based on a three-dimensional virtual compression channel was proposed. By using a non-conductive viscous sheath flow solution to focus the conductive sample solution in the microchannel, a virtual compression microchannel close to the cell's geometric size is formed. When the measured cell is forced to compress through the virtual compression channel, the electric field lines are effectively blocked, and the uniform electric field area at the center of the electric field increases the electrical detection sensitivity and alleviates the problem of electrical detection position dependence.

[0005] In signal processing, through reasonable construction, deep neural networks can extract massive features from original signals, effectively improve classification accuracy, and are widely used for feature extraction and classification of one-dimensional sequence signals. Among them, recurrent neural networks are very effective in processing sequence characteristics due to their high-quality feature extraction and classification of time series information in sequence signals. Summary of the invention

[0006] The present invention combines microfluidic technology to invent a single-cell high-throughput detection system and method based on vertical three-dimensional focusing. Compared with the existing technology, there are four main improvement goals:

[0007] (1) In the existing single-cell detection technology based on three-dimensional sheath flow focusing, the cell flow direction is perpendicular to the gravity direction of the cells and the flow rate in the channel is slow, resulting in serious cell sedimentation in the hose, microchannel and other locations. To address this shortcoming, the present invention places the microfluidic chip vertically so that the cell flow direction and the gravity direction of the cells are in a straight line. At the same time, a sample flow with a higher flow rate is used to pass through the detection area to reduce the occurrence of cell sedimentation.

[0008] (2) In the existing single-cell detection technology based on three-dimensional sheath flow focusing, the liquid path from sample injection to detection is directly connected to the microfluidic chip by a hose, which has many dead zones and serious cell loss, seriously affecting the accuracy of cell classification and counting. In view of this shortcoming, the present invention designs a sample delivery liquid path that combines positive pressure and negative pressure systems. The sample is drawn from the reaction pool into the hose by negative pressure, and the sample flow is quickly pushed into the detection area by positive pressure, which reduces the dead zone in the liquid path, thereby reducing the cell loss of sample delivery and improving the accuracy of white blood cell detection.

[0009] (3) In the existing single-cell detection technology based on three-dimensional sheath flow focusing, manual settings and adjustments are required from blood collection, sample injection to detection, and there is a lack of automated processes, which makes it difficult to achieve the theoretical maximum throughput. In view of this shortcoming, the present invention designs a set of automated sample injection, detection and cleaning processes, which reduces manual operations during the experiment and improves the actual maximum throughput.

[0010] (4) In the existing microfluidic single-cell flow cytometry detection platform, the extracted feature parameters rely on the corresponding mathematical model calculation, the number of features is small, and the cell classification effect is poor. To address this shortcoming, the present invention uses a recurrent neural network to process single-cell electrical impedance signals and mines the characteristic single-cell information in the electrical impedance data to achieve high-precision classification of the tested cells.

[0011] The technical solution of the present invention is as follows:

[0012] A single-cell high-throughput detection system based on vertical three-dimensional focusing, comprising a sample reaction module, a vertical three-dimensional sheath flow focusing module, a fluid control module and a single-cell multi-frequency impedance detection module;

[0013] The sample reaction module comprises a reaction pool and a reaction liquid channel, the reaction pool opening is a sample inlet, and the middle and lower part is connected to the reaction liquid channel;

[0014] The vertical three-dimensional sheath flow focusing module comprises a sample flow channel, a sheath flow channel and a three-dimensional focusing channel. After the sample flow channel converges with the sheath flow channel, it is connected with the three-dimensional focusing channel. The starting end of the sample flow channel is the sample flow inlet, the middle opening of the sample flow channel is the detection electrode inlet, the starting end of the sheath flow channel is the sheath flow inlet, and the ending end of the three-dimensional focusing channel is the mixed fluid outlet;

[0015] The fluid control module includes a reaction liquid control module, a sample flow control module and a sheath flow control module, the reaction liquid control module includes a reaction liquid pool and a positive pressure interface, the positive pressure interface is connected to the reaction liquid pool, and the reaction liquid pool is connected to the reaction pool through the reaction liquid channel; the sample flow control module includes a positive pressure interface, a negative pressure interface, a buffer solution pool, a waste liquid pool, a sample flow hose and a chip connector, the positive pressure interface is connected to the buffer solution pool and then connected to the sample flow hose for pushing the sample liquid, and is connected to the reaction pool through a tee, and is connected to the vertical three-dimensional focusing module with the chip connector through another tee, the end of the sample flow hose is connected to the waste liquid pool, and the outlet of the waste liquid pool is connected to the negative pressure interface for controlling the discharge of waste liquid; the sheath flow control module includes a positive pressure interface, a sheath flow solution pool and a sheath flow connector, the positive pressure interface is connected to the sheath flow solution pool, and is connected to the sheath flow inlet through a liquid guide hose and the sheath flow connector, and the positive pressure interface is used to control the flow rate of the sheath flow;

[0016] The single-cell multi-frequency impedance detection module includes a phase-locked amplifier and a data acquisition card connected thereto for collecting data. The phase-locked amplifier is respectively connected to the detection electrodes at the middle opening of the sample flow injection channel and the outlet of the mixed fluid through electrode lines.

[0017] In the above technical solution, the cross section of the sheath flow channel is circular with a radius of 50 microns.

[0018] In the above technical solution, the cross-section of the sheath flow injection channel is square, with a length of 600 microns and a width of 300 microns.

[0019] In the above technical solution, the sample flow direction in the sample flow channel is on the same straight line as the gravity direction.

[0020] In the above technical solution, the pressure control range of the pressure interface is between -1 bar and 8 bar.

[0021] In the above technical solution, the output frequency of the phase-locked amplifier is 865,800 sampling points / second.

[0022] In the above technical solution, the detection system also includes a single cell recognition and classification module, which analyzes the extracted electrical impedance data through the recurrent neural network model and classifies and counts the white blood cells.

[0023] A single cell high-throughput detection method based on vertical three-dimensional focusing includes the following liquid control steps:

[0024] Step 1.1, inject 0.25 ml of lysis solution into the reaction pool, add quantitative blood sample and react for a certain period of time, then flush sufficient buffer into the reaction solution inlet to terminate the reaction;

[0025] Step 1.2, open the solenoid valves at the outlet of the reaction pool and the inlet of the waste liquid pool, keep the solenoid valves at the outlet of the buffer solution pool and the inlet of the sample flow closed, and use negative pressure to draw the sample into the liquid guide hose at high speed;

[0026] Step 1.3, open the solenoid valves at the outlet of the buffer solution pool and the inlet of the sample flow, close the solenoid valves at the outlet of the reaction pool and the inlet of the waste liquid pool, use positive pressure to push the sample flow into the detection area at high speed through the liquid guide hose, and start the detection after the flow stabilizes;

[0027] Step 1.4: After the test is completed, continue to push for a certain period of time to clean the test area. At the same time, use the reaction liquid inlet to push the buffer solution to clean the inner wall of the reaction pool, then close the solenoid valves at the buffer solution pool outlet and the sample flow inlet, open the solenoid valves at the reaction pool outlet and the waste liquid pool inlet, and while pushing the buffer solution into the reaction liquid inlet, use negative pressure to extract the liquid to further clean the reaction pool and the liquid guide hose;

[0028] Step 1.5, close the solenoid valves at the reaction pool outlet and the waste liquid pool inlet, push 0.25 ml of lysis solution into the reaction liquid inlet, repeat steps 1.1-1.4, and the automated sampling test can be realized.

[0029] A single cell high-throughput detection method based on vertical three-dimensional focusing comprises the following steps:

[0030] Step 2.1, using the fluid control module to control the reaction liquid to enter the reaction pool to react with the sample added by the injection needle, and open the solenoid valve after a certain period of time;

[0031] Step 2.2, using the fluid control module, negative pressure draws the sample into the sample flow hose until the sample flow fills the entire sample flow hose, and then closes the corresponding solenoid valve;

[0032] Step 2.3, using the fluid control module, open the corresponding solenoid valve, use the positive pressure interface to push the sample flow, and open the sheath flow channel at the same time, so that the high conductivity sample flow carrying the cells and the low conductivity sheath flow pass through the sample flow inlet and the sheath flow inlet respectively, and then enter the microfluidic chip;

[0033] Step 2.4, using a single-cell multi-frequency impedance detection module to detect four-frequency impedance amplitude and phase data between two inserted electrode lines when cells pass through and when no cells pass through;

[0034] Step 2.5: Use the single-cell neural network recognition and classification module based on the recurrent neural network to build and train a neural network model for cell classification based on the four-frequency impedance amplitude and phase data obtained in step 2.4.

[0035] In the above technical solution, the construction and training process of the neural network in step 2.5 is as follows: the four-frequency impedance amplitude and phase data of each single cell are matched one by one with the image video data, 70% of the impedance data set is taken as the training set to enter the recurrent neural network for training, and 15% of the data set is used as the validation set to test the training results, and 15% of the completely isolated data set is used as the integrated learning test set to test the training results.

[0036] Beneficial effects:

[0037] (1) The present invention is designed based on a vertical three-dimensional sheath flow channel, so that the flow direction of the sample flow in the detection area is in the same straight line as the gravity direction, thereby reducing cell loss during the detection process.

[0038] (2) The present invention uses a high-speed sample flow to flow in a hose, making it difficult for cells to settle, further reducing cell loss during the detection process.

[0039] (3) The present invention designs a complete set of automatic sampling, detection, and cleaning processing procedures by constructing a combination of positive pressure and negative pressure systems, thereby realizing the automation of three-dimensional focused single-cell detection and improving the maximum throughput of the actual operation of the system.

[0040] (4) The present invention uses a recurrent neural network model to process single-cell impedance signals. More specifically, a long short-term memory neural network in the recurrent neural network is used to process single-cell multi-frequency impedance signals to achieve high-precision classification of the measured cells. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is the system device and module diagram;

[0042] Figure 2 Schematic diagram of vertical three-dimensional sheath flow focusing module;

[0043] Figure 3 It is the flow chart of liquid circuit control;

[0044] Figure 4 Schematic diagram of the construction and training process of a recurrent neural network. DETAILED DESCRIPTION

[0045] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. However, the following embodiments are limited to explaining the present invention, and the protection scope of the present invention should include the entire contents of the claims, and through the description of the following embodiments, those skilled in the art can fully implement the entire contents of the claims of the present invention.

[0046] Example

[0047] This embodiment is mainly divided into hardware devices and experimental procedures.

[0048] (1) Hardware device

[0049] The hardware devices required by the present invention are as follows Figure 1 As shown, the system consists of: a sample reaction module, which is used to realize automated sampling and processing of blood samples; a vertical three-dimensional sheath flow focusing module, which is used to realize integrated three-dimensional sheath flow focusing and reduce cell sedimentation; a fluid control module, which is used to realize flow rate control of the reaction liquid, sample flow, and sheath flow channel to reduce cell loss; a multi-frequency impedance detection module, which is used to collect the multi-frequency impedance of single cells in the focused detection area and process and analyze it.

[0050] The fluid control module is used to achieve flow rate control of the reaction liquid, sample flow, and sheath flow channels, including a microfluidic pressure controller and its supporting system, four solenoid valves, four sets of liquid guide hoses, gas guide hoses, and a set of liquid guide hoses for sample flow. The microfluidic pressure controller and its supporting system can output the required flow rate by manually setting, and can provide pressure control in the range of -1bar~8bar. A centrifuge tube with a capacity of 5mL is used as a solution pool, and the liquid guide tube and gas guide tube are respectively connected to the liquid circuit and pressure interface to form a fluid pressure control system, providing the sample reaction module and the vertical three-dimensional sheath flow focusing module with adjustable flow rates of reaction liquid, sample flow, and sheath flow. The four solenoid valves are respectively located at the sample flow outlet of the sample reaction module, the sample flow inlet of the vertical three-dimensional sheath flow focusing module, the buffer solution pool of the fluid control module, and the waste liquid pool, which are used to control the isolation between different areas during automated sampling and detection to prevent mutual influence between liquids.

[0051] The vertical 3D focusing module is used to achieve integrated 3D sheath flow focusing. The main part is an insulating resin material microfluidic chip made using high-precision 3D printing technology, with a minimum layer thickness of 10 microns. The overall structure of the microfluidic chip is shown in the attached figure. Figure 2As shown, the integrated three-dimensional sheath flow focusing channel mainly includes a sample flow inlet, a sheath flow inlet, a three-dimensional sheath flow focusing area, a detection area, a detection electrode inlet, and a mixed fluid outlet. When used specifically, the sheath flow solution is a deionized water insulating solution in which sucrose is dissolved, and the sample flow solution and the solution in the buffer solution pool are phosphate buffered saline (PBS for short).

[0052] The single-cell multi-frequency impedance detection module is used to collect the single-cell multi-frequency impedance in the focus detection area and process and analyze it. It includes a phase-locked amplifier and a data acquisition card. According to the needs of the embodiment, it can accurately detect the presence or absence of impedance changes, with an output frequency of 865,800 sampling points / second, and connect the two electrode inlets of the aforementioned three-dimensional sheath flow focusing module to detect the electrode inlet and the mixed fluid outlet by inserting an electrode line.

[0053] The system also includes a single-cell recognition and classification module based on a recurrent neural network. In this module, the long short-term memory (LSTM) network in the recurrent neural network is used to realize feature extraction and recognition and classification of 8-dimensional impedance sequence signals, thereby realizing single-cell classification and counting.

[0054] (2) Experimental process

[0055] In the experimental operation of this embodiment, the sample reaction module, the vertical three-dimensional sheath flow focusing module, the fluid control module and the multi-frequency impedance detection module are first connected. The fluid control module is connected to the sample reaction module outlet, the reaction liquid inlet, the vertical three-dimensional sheath flow focusing module sample flow inlet, the sheath flow inlet and the mixed fluid outlet through a liquid guide hose; the multi-frequency impedance detection module is connected to the vertical three-dimensional sheath flow focusing module detection electrode inlet and the mixed fluid outlet through two detection electrodes.

[0056] Next, the fluid in the system is controlled by the fluid control module. The control process is as shown in the attached figure. Figure 3 As shown, the specific process is as follows:

[0057] Step 1.1, inject 0.25 ml of lysis solution into the reaction pool, add quantitative blood sample and react for a certain period of time, then flush sufficient buffer into the reaction solution inlet to terminate the reaction;

[0058] Step 1.2, open the solenoid valves at the outlet of the reaction pool and the inlet of the waste liquid pool, keep the solenoid valves at the outlet of the buffer solution pool and the inlet of the sample flow closed, and use negative pressure to draw the sample into the liquid guide hose at high speed;

[0059] Step 1.3, open the solenoid valves at the outlet of the buffer solution pool and the inlet of the sample flow, close the solenoid valves at the outlet of the reaction pool and the inlet of the waste liquid pool, use positive pressure to push the sample flow into the detection area at high speed through the liquid guide hose, and start the detection after the flow stabilizes;

[0060] Step 1.4: After the test is completed, continue to push for a certain period of time to clean the test area. At the same time, use the reaction liquid inlet to push the buffer solution to clean the inner wall of the reaction pool, then close the solenoid valves at the buffer solution pool outlet and the sample flow inlet, open the solenoid valves at the reaction pool outlet and the waste liquid pool inlet, and while pushing the buffer solution into the reaction liquid inlet, use negative pressure to extract the liquid to further clean the reaction pool and the liquid guide hose;

[0061] Step 1.5, close the solenoid valves at the reaction pool outlet and the waste liquid pool inlet, push 0.25 ml of lysis solution into the reaction liquid inlet, repeat steps 1.1-1.4, and the automated sampling test can be realized.

[0062] During the experiment, the single cell detection method includes the following experimental steps:

[0063] Step 2.1, using the fluid control module to control the reaction liquid to enter the reaction pool to react with the sample added by the injection needle, and open the solenoid valve after a certain period of time;

[0064] Step 2.2, using the fluid control module, negative pressure draws the sample into the sample flow hose until the sample flow fills the entire sample flow hose, and then closes the corresponding solenoid valve;

[0065] Step 2.3, using the fluid control module, open the corresponding solenoid valve, use the positive pressure interface to push the sample flow, and open the sheath flow channel at the same time, so that the high conductivity sample flow carrying the cells and the low conductivity sheath flow pass through the sample flow inlet and the sheath flow inlet respectively, and then enter the microfluidic chip;

[0066] Step 2.4, using a single-cell multi-frequency impedance detection module to detect four-frequency impedance amplitude and phase data between two inserted electrode lines when cells pass through and when no cells pass through;

[0067] Step 2.5: Use the single-cell neural network recognition and classification module based on the recurrent neural network to build and train a neural network model for cell classification based on the four-frequency impedance amplitude and phase data obtained in step 2.4.

[0068] The neural network construction and training process described in step 2.5 is as shown in the attached figure. Figure 4 70% of the impedance data set is used as a training set to enter the recurrent neural network for training, 15% of the data set is used as a validation set to test the training results, and 15% of the completely isolated data set is used as a test set to test the model training results.

[0069] In this embodiment, the sheath flow is an insulating solution of deionized water in which sucrose is dissolved, and a low-conductivity solution that is miscible with the water system does not affect the realization of basic functions; the sample flow is a phosphate buffered saline (PBS for short), and a conductive solution such as iodixanol prepared with physiological saline, cell culture medium, and PBS does not affect the realization of basic functions.

[0070] In this embodiment, the three-dimensional sheath flow focusing module is formed by 3D printing using a photosensitive resin material. The three-dimensional fluid focusing module can also be formed using different 3D printers and printing materials, and femtosecond laser quartz processing technology.

[0071] In this embodiment, the direction of cell flow is opposite to the direction of gravity. The fact that the direction of cell flow is the same as the direction of gravity does not affect the realization of basic functions.

[0072] In this embodiment, the shape and size of the interface between the sample inlet, sheath inlet, detection electrode inlet, mixed solution outlet and fluid control module in the three-dimensional sheath flow focusing module can be changed without affecting the basic function.

[0073] The above description is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest range consistent with the principles and novel features applied for herein.

Claims

1. A single-cell high-throughput detection system based on vertical three-dimensional focusing, characterized in that: It includes a sample reaction module, a vertical three-dimensional sheath flow focusing module, a fluid control module and a single cell multi-frequency impedance detection module; The sample reaction module comprises a reaction pool and a reaction liquid channel, the opening of the reaction pool is an injection port, and the middle and lower parts are connected to the reaction liquid channel; The vertical three-dimensional sheath flow focusing module comprises a sample flow channel, a sheath flow channel and a three-dimensional focusing channel. After the sample flow channel converges with the sheath flow channel, it is connected with the three-dimensional focusing channel. The starting end of the sample flow channel is the sample flow inlet, the middle opening of the sample flow channel is the detection electrode inlet, the starting end of the sheath flow channel is the sheath flow inlet, and the ending end of the three-dimensional focusing channel is the mixed fluid outlet; The fluid control module includes a reaction liquid control module, a sample flow control module and a sheath flow control module, the reaction liquid control module includes a reaction liquid pool and a positive pressure interface, the positive pressure interface is connected to the reaction liquid pool, and the reaction liquid pool is connected to the reaction pool through the reaction liquid channel; the sample flow control module includes a positive pressure interface, a negative pressure interface, a buffer solution pool, a waste liquid pool, a sample flow hose and a chip connector, the positive pressure interface is connected to the buffer solution pool and then connected to the sample flow hose for pushing the sample liquid, and is connected to the reaction pool through a tee, and is connected to the vertical three-dimensional sheath flow focusing module with the chip connector through another tee, the end of the sample flow hose is connected to the waste liquid pool, and the outlet of the waste liquid pool is connected to the negative pressure interface for controlling the discharge of waste liquid; the sheath flow control module includes a positive pressure interface, a sheath flow solution pool and a sheath flow connector, the positive pressure interface is connected to the sheath flow solution pool, and is connected to the sheath flow inlet through a liquid guide hose and the sheath flow connector, and the positive pressure interface is used to control the flow rate of the sheath flow; The single-cell multi-frequency impedance detection module includes a phase-locked amplifier and a data acquisition card connected thereto for collecting data. The phase-locked amplifier is respectively connected to the detection electrodes at the middle opening of the sample flow injection channel and the outlet of the mixed fluid through electrode lines.

2. A single-cell high-throughput detection system based on vertical three-dimensional focusing according to claim 1, characterized in that: The cross section of the sheath flow channel is circular, and the radius is 50 microns.

3. The single-cell high-throughput detection system based on vertical three-dimensional focusing according to claim 1, characterized in that: The sheath flow injection channel has a square cross section with a length of 600 microns and a width of 300 microns.

4. The single-cell high-throughput detection system based on vertical three-dimensional focusing according to claim 1, characterized in that: The sample flow direction in the sample flow channel is in the same straight line as the gravity direction.

5. The single-cell high-throughput detection system based on vertical three-dimensional focusing according to claim 1, characterized in that: The pressure control range of the positive pressure interface is between -1bar and 8bar.

6. The single-cell high-throughput detection system based on vertical three-dimensional focusing according to claim 1, characterized in that: The output frequency of the lock-in amplifier is 865,800 samples per second.

7. The single-cell high-throughput detection system based on vertical three-dimensional focusing according to claim 1, characterized in that: The detection system also includes a single cell recognition and classification module, which analyzes the extracted electrical impedance data and classifies and counts the white blood cells through a recurrent neural network model.

8. A single-cell high-throughput detection method based on vertical three-dimensional focusing, characterized in that: Using a single cell high-throughput detection system based on vertical three-dimensional focusing as claimed in any one of claims 1 to 7, comprising the following liquid control steps: Step 1.1, inject 0.25 ml of lysis solution into the reaction pool, add quantitative blood sample and react for a certain period of time, then flush sufficient buffer through the injection port to terminate the reaction; Step 1.2, open the solenoid valves at the outlet of the reaction pool and the inlet of the waste liquid pool, keep the solenoid valves at the outlet of the buffer solution pool and the inlet of the sample flow closed, and use negative pressure to draw the sample into the liquid guide hose at high speed; Step 1.3, open the solenoid valves at the outlet of the buffer solution pool and the inlet of the sample flow, close the solenoid valves at the outlet of the reaction pool and the inlet of the waste liquid pool, use positive pressure to push the sample flow into the detection area at high speed through the liquid guide hose, and start the detection after the flow stabilizes; Step 1.4: After the test is completed, continue to push for a certain period of time to clean the test area. At the same time, use the injection port to push the buffer solution to clean the inner wall of the reaction pool. Then close the solenoid valves at the outlet of the buffer solution pool and the sample flow inlet, open the solenoid valves at the outlet of the reaction pool and the inlet of the waste liquid pool, and while pushing the buffer solution into the injection port, use negative pressure to extract the liquid to further clean the reaction pool and the liquid guide hose; Step 1.5, close the solenoid valves at the reaction pool outlet and the waste liquid pool inlet, push 0.25 ml of lysis solution into the injection port, and repeat steps 1.1-1.4 to achieve automated injection detection.

9. A single-cell high-throughput detection method based on vertical three-dimensional focusing, characterized in that: Using a single cell high-throughput detection system based on vertical three-dimensional focusing as claimed in any one of claims 1 to 7 comprises the following steps: Step 2.1, using the fluid control module to control the reaction liquid to enter the reaction pool and react with the sample added to the injection port, and open the solenoid valve after a certain period of time; Step 2.2, using the fluid control module, negative pressure draws the sample into the sample flow hose until the sample flow fills the entire sample flow hose, and then closes the corresponding solenoid valve; Step 2.3, using the fluid control module, open the corresponding solenoid valve, use the positive pressure interface to push the sample flow, and open the sheath flow channel at the same time, so that the high conductivity sample flow carrying the cells and the low conductivity sheath flow pass through the sample flow inlet and the sheath flow inlet respectively, and then enter the microfluidic chip; Step 2.4, using a single-cell multi-frequency impedance detection module to detect four-frequency impedance amplitude and phase data between two inserted electrode lines when cells pass through and when no cells pass through; Step 2.5: Use the single-cell neural network recognition and classification module based on the recurrent neural network to build and train a neural network model for cell classification based on the four-frequency impedance amplitude and phase data obtained in step 2.4.

Images