Microchannel-based cell elastic modulus acquisition method, device, electronic device and storage medium

Abstract

The present application relates to a method, device, electronic device and storage medium for obtaining cell elastic modulus based on microchannels. The method for obtaining cell elastic modulus based on microchannels includes: obtaining first cell deformation data; the first cell deformation data includes cell deformation characteristic data of elastic deformation of the cell at at least one moment in the process of passing through a variable cross-section pipe; the cross-sectional areas of at least two places of the variable cross-section pipe are different; the inner diameter of the narrowest area of ​​the variable cross-section pipe is greater than the cell diameter and less than twice the cell diameter, so as to avoid direct squeezing of the cell by the variable cross-section pipe in the process of passing through the variable cross-section pipe, and only a single cell is allowed to pass through the narrowest area each time; the first cell deformation data is input into a trained cell elastic modulus prediction model, so that the cell elastic modulus prediction model outputs the cell elastic modulus. The technical solution of the present application can improve the throughput of detecting cell elastic modulus and reduce damage to cells.

Description

Technical Field

[0001] The present invention relates to the technical field of cell detection, and in particular to a method, device, electronic equipment and storage medium for obtaining cell elastic modulus based on a microchannel. Background Art

[0002] In the related technologies, the measurement techniques used to measure the elastic modulus of cells include atomic force microscopy, micropipette, magnetic tweezers, optical tweezers, shear flow, cell stretching instrument, micro-pillar array, variable cross-section cell extrusion microchannel, etc. However, these technical means have problems such as low throughput and great damage to cells. Summary of the invention

[0003] The purpose of the present application is to provide a method, device, electronic device and storage medium for obtaining cell elastic modulus based on a microchannel, which can improve the throughput of detecting cell elastic modulus and reduce damage to cells.

[0004] According to a first aspect of an embodiment of the present application, a method for obtaining a cell elastic modulus based on a microchannel is provided, comprising:

[0005] Acquire first cell deformation data; the first cell deformation data includes cell deformation characteristic data of elastic deformation of the cell at least at a time when the cell passes through the variable cross-section pipeline; the cross-sectional areas of at least two places of the variable cross-section pipeline are different; the inner diameter of the narrowest area of ​​the variable cross-section pipeline is larger than the cell diameter and smaller than twice the cell diameter, so as to avoid the cell being directly squeezed by the variable cross-section pipeline in the process of passing through the variable cross-section pipeline, and only a single cell is allowed to pass through the narrowest area each time;

[0006] The first cell deformation data is input into a trained cell elastic modulus prediction model so that the cell elastic modulus prediction model outputs the cell elastic modulus.

[0007] In one embodiment, the cell elastic modulus prediction model is a fully connected neural network;

[0008] The fully connected neural network includes an input layer, a flattening layer, a fusion layer, at least one hidden layer and an output layer connected in sequence;

[0009] The input layer is used to input the first cell deformation data; the flattening layer is used to flatten the first cell deformation data to obtain the flattened first cell deformation data; the fusion layer is used to fuse the flattened first cell deformation data to obtain fused data; the fusion layer is fully connected to the hidden layer, and the last hidden layer of the at least one hidden layer is connected to the output layer, and the output layer is used to output the cell elastic modulus.

[0010] In one embodiment, the first cell deformation data also includes pressure information at the inlet of the variable cross-section pipe.

[0011] In one embodiment, the cell deformation characteristic data includes at least one of cell contour data, extension index, out-of-roundness, curvature ratio, relative cell size and inverse transit time.

[0012] In one embodiment, when the cell deformation feature data includes cell contour data, the cell contour data is the coordinates of contour points on the cell contour in the cell image; the step of acquiring the cell feature data includes:

[0013] Taking a picture of the cell by a camera device to obtain the cell image;

[0014] Binarizing the cell image to obtain a binarized cell image;

[0015] Performing contour extraction on the binarized cell image to obtain a cell contour;

[0016] determining the centroid of the cell outline;

[0017] Extracting contour points from the cell contour at a specified angle interval to obtain N contour points, where N=360 / α, α is the specified angle interval;

[0018] For each of the contour points, the coordinates of the contour point are obtained; wherein the centroid is the origin.

[0019] In one embodiment, for each moment, the cell contour data includes first contour data of a main view of the cell and second contour data of a top view.

[0020] In one embodiment, the cell elastic modulus prediction model is a convolutional neural network; the first cell deformation data is a cell image; and the cell image includes a front view and a top view of the cell.

[0021] In one embodiment, the cell elastic modulus prediction model is trained by the following method:

[0022] The specified constitutive model and elastic modulus are used to calculate and obtain the deformation data of the second cell;

[0023] The second cell deformation data is used as training data to train the cell elastic modulus prediction model until the cell elastic modulus prediction model meets specified conditions, thereby obtaining a trained cell elastic modulus prediction model.

[0024] In one embodiment, the constitutive model regards the cell as a shell-like structure object, and the inside and outside of the cell are viscous liquid; the constitutive model is the Hookean model, the Mooney-Rivlin model, the Neo-Hookean model, the Skalak model or the Evans & Skalak model.

[0025] According to a second aspect of an embodiment of the present application, a cell elastic modulus device is provided, comprising:

[0026] The acquisition module is configured to acquire first cell deformation data; the first cell deformation data includes cell deformation characteristic data of elastic deformation of the cell at least at a time when the cell passes through the variable cross-section pipe; the cross-sectional areas of at least two places of the variable cross-section pipe are different; the inner diameter of the narrowest area of ​​the variable cross-section pipe is larger than the cell diameter and smaller than twice the cell diameter, so as to avoid the cell being directly squeezed by the variable cross-section pipe during the process of passing through the variable cross-section pipe, and only a single cell is allowed to pass through the narrowest area each time;

[0027] The prediction module is configured to input the first cell deformation data into a trained cell elastic modulus prediction model so that the cell elastic modulus prediction model outputs the cell elastic modulus.

[0028] According to a third aspect of an embodiment of the present application, an electronic device is provided, comprising a memory and a processor, wherein the memory is used to store a computer program executable by the processor; and the processor is used to execute the computer program in the memory to implement the above method.

[0029] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, on which a computer program is stored, characterized in that when the executable computer program in the storage medium is executed by a processor, the above method can be implemented.

[0030] Compared with the prior art, the beneficial effect of the present application is that: by acquiring the first cell deformation data, and inputting the first cell deformation data into the trained cell elastic modulus prediction model, so that the cell elastic modulus prediction model outputs the cell elastic modulus, wherein the first cell deformation data includes the cell deformation characteristic data of elastic deformation of the cell at least at one moment in the process of passing through the variable cross-section pipeline, the cross-sectional areas of the variable cross-section pipeline are different at least at two places, and the inner diameter of the narrowest area of ​​the variable cross-section pipeline is greater than the cell diameter and less than twice the cell diameter, so as to avoid the cell being directly squeezed by the variable cross-section pipeline in the process of passing through the variable cross-section pipeline, and only a single cell is allowed to pass through the narrowest area each time, so that the cell elastic modulus is measured faster than using laboratory measurement tools, the throughput is high, and because no measurement tool is used to apply force to the cell, damage to the cell can be avoided. Therefore, the technical solution provided by the present application can improve the throughput of detecting the cell elastic modulus and reduce damage to the cell. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 The figure is a flow chart of a method for obtaining cell elastic modulus based on a microchannel according to an exemplary embodiment.

[0032] Figure 2 is a schematic diagram showing the deformation of a cell in a pipe with a variable cross-section according to an exemplary embodiment.

[0033] Figure 3 is a cell image showing deformation of a cell at different times according to an exemplary embodiment.

[0034] Figure 4 is a flow chart of a method for obtaining cell elastic modulus based on a microchannel according to another exemplary embodiment.

[0035] Figure 5 is a cell image showing a deformed cell according to an exemplary embodiment.

[0036] Figure 6 is a binarized cell image shown according to an exemplary embodiment.

[0037] Figure 7 is a cell outline image shown according to an exemplary embodiment.

[0038] Figure 8 is a schematic diagram showing how to obtain coordinates of contour points based on a cell contour image according to an exemplary embodiment;

[0039] Fig. 9 It is a schematic structural diagram of a cell elastic modulus prediction model according to an exemplary embodiment.

[0040] Fig.10 The present invention is a flowchart of a method for training a cell elastic modulus prediction model according to an exemplary embodiment.

[0041] Fig.11 It is a block diagram of a cell elastic modulus prediction device according to an exemplary embodiment.

[0042] Fig.12 It is a block diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0043] Unless otherwise defined, the technical terms or scientific terms used in this specification and claims shall have the usual meanings understood by persons with ordinary skills in the technical field to which the invention belongs. The specific embodiments of the present invention will be described below in conjunction with the accompanying drawings. It should be noted that in the specific description of these embodiments, in order to provide a concise description, it is impossible for this specification to provide a detailed description of all the features of the actual embodiments. Without departing from the spirit and scope of the present invention, those skilled in the art may modify and replace the embodiments of the present invention, and the resulting embodiments are also within the scope of protection of the present invention.

[0044] Figure 1 This is a flow chart of a method for obtaining cell elastic modulus based on a microchannel according to an exemplary embodiment. The method for obtaining cell elastic modulus based on a microchannel can be applied to electronic devices such as computers and servers. Figure 1 The microchannel-based cell elastic modulus acquisition method may include the following steps:

[0045] Step 101, obtaining first cell deformation data; the first cell deformation data includes cell deformation characteristic data of elastic deformation of the cell at least at one moment in the process of passing through the variable cross-section pipe; the cross-sectional areas of at least two places of the variable cross-section pipe are different; the inner diameter of the narrowest area of ​​the variable cross-section pipe is larger than the cell diameter and smaller than twice the cell diameter, so as to avoid the cell being directly squeezed by the variable cross-section pipe in the process of passing through the variable cross-section pipe, and only a single cell is allowed to pass through the narrowest area at a time.

[0046] In this embodiment, if Figure 2As shown, the variable cross-section pipeline includes a first pipeline segment 21, a second pipeline segment 22, a third pipeline segment 23, a fourth pipeline segment 24 and a fifth pipeline segment 25 connected in sequence. The cross-sectional areas of the first pipeline segment 21, the third pipeline segment 23 and the fifth pipeline segment 25 are evenly distributed along the length direction of the variable cross-section pipeline. The cross-sectional area of ​​the first pipeline segment 21 is the same as the cross-sectional area of ​​the fifth pipeline segment 25, and the cross-sectional area of ​​the first pipeline segment 21 is greater than the cross-sectional area of ​​the third pipeline segment 23. The second pipeline segment 22 and the fourth pipeline segment 24 are transition pipeline segments. The cross-sectional area of ​​the second pipeline segment 22 near the first pipeline segment 21 is greater than the cross-sectional area of ​​the second pipeline segment 22 near the third pipeline segment 23, and the cross-sectional area of ​​the second pipeline segment 22 gradually decreases in the direction from the first pipeline segment 21 to the third pipeline segment 23. The cross-sectional area of ​​the fourth pipeline segment 24 near the third pipeline segment 23 is less than the cross-sectional area of ​​the fourth pipeline segment 24 near the fifth pipeline segment 25, and the cross-sectional area of ​​the fourth pipeline segment 24 gradually increases in the direction from the first pipeline segment 21 to the third pipeline segment 23.

[0047] like Figure 2 As shown, the third pipe section 23 is the narrowest area of ​​the variable cross-section pipe, and the inner diameter D of the third pipe section 23 is 7 microns, which is larger than the diameter of the undeformed cell 28 and smaller than twice the diameter of the cell 28, so as to prevent the undeformed cell 28 from being directly squeezed by the variable cross-section pipe when passing through the variable cross-section pipe, and only a single cell is allowed to pass through the narrowest area each time. Only a single cell is allowed to pass through the narrowest area each time means that at any position in the narrowest area at the same time, there is only one cell.

[0048] like Figure 2 As shown, in the xyz coordinate system, the variable cross-section pipeline extends along the y-axis direction. The length L1 of the first pipeline segment 21 is the same as the length L5 of the fifth pipeline segment 25, the length L3 of the third pipeline segment 23 is the sum of the length L1 of the first pipeline segment 21 and the length L5 of the fifth pipeline segment 25, and the length L2 of the second pipeline segment 22 is the same as the length L4 of the fourth pipeline segment 24.

[0049] In one embodiment, the length L1 of the first pipe segment 21 is 15 microns, the length L2 of the second pipe segment 22 is 5 microns, the length L3 of the third pipe segment 23 is 30 microns, the length L4 of the fourth pipe segment 24 is 5 microns, and the length L5 of the fifth pipe segment 25 is 15 microns. The difference ΔD between the outer diameter of the first pipe segment 21 and the outer diameter of the third pipe segment 23 is 3.5 microns. Of course, in other embodiments, the sizes of the various parts of the variable cross-section pipe can also be other values.

[0050] like Figure 2As shown, the inlet 26 of the variable cross-section pipe is located on one side of the first pipe segment 21, and the outlet 27 of the variable cross-section pipe is located on one side of the fifth pipe segment 25. The pressure at the inlet 26 in the variable cross-section pipe is greater than the pressure at the outlet 27, so that the cell enters the variable cross-section pipe from the inlet 26 of the variable cross-section pipe and flows out from the outlet 27. After the undeformed cell 28 enters the third pipe segment 23, it is elastically deformed due to being squeezed, and the force-sensitive channel of the elastically deformed cell 29 is opened. After the force-sensitive channel of the cell is opened, the technical means of the jet can be used to transport the macromolecular substance into the cell. However, in order to better set the control parameters of the jet device, it is necessary to understand the elastic modulus of the cell.

[0051] In this embodiment, if Figure 3 As shown, the cells undergo different degrees of elastic deformation at different times during the process of passing through the variable cross-section pipe. For example, at the first moment t1, the cells do not undergo elastic deformation, and the main view P11 and the top view P12 of the cells are both circular. At the second moment t2, the third moment t3, and the fourth moment t4, the cells undergo different degrees of elastic deformation, and the main view P21 and the top view P22 of the cells at the second moment t2, the main view P31 and the top view P32 of the cells at the third moment t3, and the main view P41 and the top view P42 of the cells at the fourth moment t4 are all irregular.

[0052] In this embodiment, the first cell deformation data includes cell deformation characteristic data of elastic deformation of the cell at multiple moments in the process of the cell passing through the variable cross-section pipe. In other embodiments, the first cell deformation data may also only include cell deformation characteristic data of elastic deformation of the cell at a certain moment in the process of the cell passing through the variable cross-section pipe.

[0053] In this embodiment, the cell deformation characteristic data is cell contour data, and the first cell deformation data also includes pressure information at the entrance of the variable cross-section pipe. The cell contour data is the coordinates of the contour points on the cell contour in the cell image. The cell image includes a front view and a top view of the cell. For each moment, the cell contour data includes the first contour data of the front view of the cell and the second contour data of the top view. In other embodiments, the cell image may only include the front view or the top view of the cell.

[0054] In this embodiment, if Figure 4 As shown, for each cell image, the method for acquiring cell contour data may include the following steps:

[0055] Step 401, photographing cells by using a camera to obtain cell images.

[0056] In this embodiment, the coordinates of contour points on the cell contour in the main view of the cell are obtained as an example for description.

[0057] In this step, if Figure 5 As shown, by photographing the front side of the cell with a camera, a cell image 51 including a cell 29 that has undergone elastic deformation can be obtained. The cell image 51 is a front view of the cell.

[0058] Step 402: binarize the cell image to obtain a binarized cell image.

[0059] In this step, if Figure 6 As shown, Figure 5 The cell image 51 shown is binarized to obtain a binarized cell image 61 .

[0060] Step 403: extract the outline of the binarized cell image to obtain the cell outline.

[0061] In this step, if Figure 7 As shown, the boundary following algorithm proposed by Suzuki et al. can be used to detect the cell contour in the binary cell image 61 to obtain a cell contour image 71, which includes a cell contour 72. Other algorithms can also be used to extract the cell contour. Among them, the cell contour is a footprint contour, not a cross-sectional contour.

[0062] Step 404, determine the centroid of the cell outline.

[0063] In this step, the centroid of the cell outline 72 is calculated. Figure 7 As shown, the centroid D is located in the cell outline 72 and is also the centroid of the cell outline 72 .

[0064] Step 405 , starting from the designated contour point on the cell contour, extract contour points in a counterclockwise direction at designated angle intervals, and obtain N contour points, where N=360 / α, and α is the designated angle interval.

[0065] In this embodiment, α is 2° and N is 180°.

[0066] In this step, if Figure 8 As shown, starting from the designated contour point A on the cell contour 72, contour points are extracted counterclockwise at an angle interval of 2° to obtain 180 contour points.

[0067] In this embodiment, 180 contour points are taken from the cell contour 72. These 180 contour points are sufficient to depict the cell contour 72, and can reduce the amount of calculation and improve the calculation speed. Of course, in other embodiments, other numbers of contour points can also be taken.

[0068] Of course, in other embodiments, the method of extracting contour points from the cell contour is not limited to the method in this embodiment. For example, starting from the designated contour point A on the cell contour 72, contour points can be extracted clockwise at an angle interval of 2° to obtain 180 contour points.

[0069] Step 406: For each contour point, a vector obtained by connecting the origin and the contour point is used as the coordinate of the contour point.

[0070] In this step, if Figure 8 As shown, for each contour point 81, the origin O and the contour point 81 are connected to obtain the corresponding vector, which is used as the coordinate of the contour point 81. The combination of these 180 vectors is an array describing the cell contour, which is also a set of input data for the cell elastic modulus prediction model. If the array is recorded as X1(180,2), then the array X1(180,2) is as follows:

[0071] <![CDATA[x1]]> <![CDATA[y1]]> <![CDATA[x2]]> <![CDATA[y2]]> <![CDATA[x3]]> <![CDATA[y3]]> …… …… <![CDATA[x 180 ]]> <![CDATA[y 180 ]]>

[0072] In one embodiment, after obtaining the coordinates of the contour points, the coordinates of the contour points need to be normalized, and normalization can make the value range of the data input to the cell elastic modulus prediction model within a relatively consistent interval. Moving the center of mass D to the origin O can make normalization easier. Moreover, normalizing the data input to the cell elastic modulus prediction model helps to increase the convergence speed of the model, provide the robustness of the model, and avoid differences between features.

[0073] The above describes the method for obtaining the cell contour data. The first contour data of the main view and the second contour data of the top view of the cell can be obtained by using the above method, which will not be introduced one by one here.

[0074] Step 102: input the first cell deformation data into the trained cell elastic modulus prediction model so that the cell elastic modulus prediction model outputs the cell elastic modulus.

[0075] In this embodiment, the model selected for the cell is a shell structure cell model, which considers that the cell membrane is a shell structure, and the inside and outside of the cell are viscous liquids. For the shell structure cell model, the cell membrane constitutive model adopts the Hookean model, the Mooney-Rivlin model, the Neo-Hookean model, the Skalak model or the Evans & Skalak model. For example, in this embodiment, the cell membrane constitutive model can adopt the Mooney-Rivlin model.

[0076] The constitutive model can also be called a constitutive equation. For example, the following equation is a constitutive model, where c1 and c2 are elastic moduli. The elastic modulus can also be called a constitutive parameter.

[0077]

[0078] Where U is the strain energy density, I1, I2, and I3 are the three deformation invariants of the Cauchy-Green tensor, and κ is a material constant used to describe the nonlinear characteristics of cells.

[0079] If it is known that the cell membrane constitutive model adopts the above equation, the elastic modulus thereof can be predicted using a cell elastic modulus prediction model trained on the above equation.

[0080] In this embodiment, the cell elastic modulus prediction model is used to predict the cell elastic modulus for a specified constitutive model. For example, if the cell membrane constitutive model adopts the Mooney-Rivlin model, the cell elastic modulus prediction model is used to predict the cell elastic modulus when the cell membrane constitutive model adopts the Mooney-Rivlin model.

[0081] In this step, the acquired first cell deformation data is input into a trained cell elastic modulus prediction model, and the cell elastic modulus prediction model outputs the cell elastic modulus.

[0082] In this embodiment, if Fig. 9 As shown, the cell elastic modulus prediction model is a fully connected neural network. The fully connected neural network includes an input layer input, a flatten layer flatten, a fusion layer merge, hidden layers hide1, hide2, hide3 and an output layer output which are connected in sequence. The number of hidden layers can be other numbers, not limited to 3 layers.

[0083] The input layer input is used to input the first cell deformation data. The input layer includes 2n+1 input nodes 91, which are used to input the first time t1 to the nth time (t n ) and the pressure information at the entrance of the variable cross-section pipe. Wherein n is a positive integer. The pressure information at the entrance of the variable cross-section pipe is a constant value, which can be recorded as an array P(1). Each input node 91 is used to input an array. For example, at the first moment t1, the cell contour data includes the first contour data of the main view of the cell (recorded as X1(180,2)) and the second contour data of the top view (recorded as X2(180,2)), at the second moment t2, the cell contour data includes the first contour data of the main view of the cell (recorded as X3(180,2)) and the second contour data of the top view (recorded as X4(180,2)), ..., at the nth moment (t n ), the cell outline data includes the first outline data of the main view of the cell (denoted as X 2n-1(180,2)) and the second contour data of the top view (denoted as X 2n The 2n+1 input nodes 91 of the input layer are used to input the arrays X1(180,2), X2(180,2), X3(180,2), X4(180,2), ..., X 2n-1 (180,2),X 2n (180,2) and P(1). The cell contour data can be obtained by processing the cell image through the CV2 library (computer vision library) in Python.

[0084] The flatten layer flatten is used to flatten the first cell deformation data to obtain the flattened first cell deformation data. The flatten layer flatten includes 2n+1 flattening nodes 92, and the 2n+1 flattening nodes 92 are connected to the 2n+1 input nodes 91 in a one-to-one correspondence. The flattening node 92 is used to flatten the data received by the corresponding input node 91.

[0085] The fusion layer merge is used to fuse the flattened first cell deformation data to obtain fused data, wherein the fused data is a tensor. The fusion layer is connected to all flattened nodes 92 in the flatten layer flatten. In this embodiment, the Concatenate function can be used to connect the flattened first cell deformation data in series to obtain fused data, but is not limited thereto. The fusion layer merge is fully connected to the hidden layers hide1, hide2, and hide3.

[0086] The hidden layer hide1 includes 2n+1 neuron nodes 93, and each neuron node 93 in the hidden layer hide1 is respectively connected to the fusion layer merge. The hidden layer hide2 also includes 2n+1 neuron nodes 93, and each neuron node 93 in the hidden layer hide2 is respectively connected to each neuron node 93 in the hidden layer hide1, and is respectively connected to each neuron node 93 in the hidden layer hide3.

[0087] Hidden layer hide3 is connected to the output layer output, and the output layer output is used to output the cell elastic modulus. The output layer output includes output nodes 94, the number of output nodes 94 is the same as the number of cell elastic moduli, and each output node 94 is used to output a cell elastic modulus. For example, when the cell elastic modulus includes c1 and c2, the output layer output includes two output nodes 94, one output node 94 is used to output c1, and the other output node 94 is used to output c2.

[0088] The above describes a method for predicting the cell elastic modulus using a trained cell elastic modulus prediction model. The following describes a method for training the cell elastic modulus prediction model.

[0089] like Fig.10 As shown, the training method of the cell elastic modulus prediction model may include the following steps:

[0090] Step 1001, using the specified constitutive model and elastic modulus to perform calculations to obtain second cell deformation data.

[0091] Step 1002: Use the second cell deformation data as training data to train the cell elastic modulus prediction model until the cell elastic modulus prediction model meets specified conditions, thereby obtaining a trained cell elastic modulus prediction model.

[0092] In this embodiment, the specified constitutive model and elastic modulus can be used for calculation to obtain the corresponding second cell deformation data as training data. For example, the specified constitutive model can be the above equation, and then M sets of c1 and c2 values ​​are prepared, and M can be 3*10 6 , but not limited thereto. The values ​​of each group of c1 and c2 are different. For each group of c1 and c2 values, the corresponding second cell deformation data is calculated using the above equation to describe the shape of the cell after deformation.

[0093] Then, the calculated training data is input into the cell elastic modulus prediction model, and the cell elastic modulus prediction model is trained until the cell elastic modulus prediction model meets the specified conditions, thereby obtaining a trained cell elastic modulus prediction model. The specified conditions are that the training loss and the validation loss reach 10 -5 ~10 -4 , and the accuracy on the test set reaches over 99%.

[0094] By using computational methods to obtain training data, we can quickly obtain massive training data, improve training speed, and save training costs. Of course, training data can also be experimental data.

[0095] In this embodiment, by acquiring the first cell deformation data and inputting the first cell deformation data into the trained cell elastic modulus prediction model, so that the cell elastic modulus prediction model outputs the cell elastic modulus, it is faster and has a higher throughput than using laboratory measurement tools to measure the cell elastic modulus, and because no measurement tool is used to apply force to the cell, damage to the cell can be avoided. Therefore, the technical solution provided by the present application can increase the throughput of detecting the cell elastic modulus and reduce damage to the cell.

[0096] In other embodiments, the cell deformation characteristic data may also include any one or any combination of cell contour data, extension index, out-of-roundness, curvature ratio, relative cell size and inverse of transit time.

[0097] In one embodiment, the calculation formula for the stretch index is:

[0098] D1=(WH) / (W+H)

[0099] Where D1 is the extension index, W is the maximum length of the cell after deformation, and H is the minimum length of the cell after deformation.

[0100] In another embodiment, the calculation formula for the extension index is:

[0101]

[0102] Among them, W is the maximum length of the cell after deformation, and H is the minimum length of the cell after deformation.

[0103] In another embodiment, the calculation formula for the extension index is:

[0104]

[0105] Among them, D t is the deformation of the cell at time t, D t+1 is the deformation of the cell at time t+1, W t is the maximum length of the cell after deformation at time t, H t is the minimum length of the cell after deformation at time t, W t+1 is the maximum length of the cell after deformation at time t+1, H t+1 It is the minimum length of the cell after deformation at time t+1.

[0106] In another embodiment, the calculation formula for the extension index is:

[0107] D1=W / 2R1

[0108] Among them, R1 is the radius of the cell when it is not deformed, and W is the maximum length of the cell after deformation.

[0109] The calculation formula for out-of-roundness is as follows:

[0110] O=1-c

[0111]

[0112] Among them, O is the out-of-roundness, c is the roundness, A is the area of ​​the cell deformed cross section, and p is the perimeter of the cell deformed cross section.

[0113] The curvature ratio is calculated as follows:

[0114] ratio = κ min / κ max

[0115] Among them, ratio is the curvature ratio, κ min is the minimum curvature of the cell deformation section, κ max is the maximum curvature of the cell deformation section.

[0116] The relative cell size was calculated as follows:

[0117] D2=R1 / R2

[0118] Wherein, D2 is the relative cell size, R1 is the radius of the cell when it is not deformed, and R2 is the radius of the inner wall cross section of the third pipe segment 23 .

[0119] The calculation formula for the inverse of time is as follows:

[0120]

[0121] Wherein, D3 is the reciprocal of the passage time, and the passage time is the time length for the cells to pass through the third pipe segment 23 .

[0122] Another exemplary embodiment of the present application also provides a method for obtaining cell elastic modulus based on a microchannel. Different from the above embodiment, in this embodiment, the cell elastic modulus prediction model is a convolutional neural network. The first cell deformation data is a cell image taken of a deformed cell; the cell image includes a front view and a top view of the cell.

[0123] In this embodiment, the collected cell image can be directly input into the trained cell elastic modulus prediction model, and the cell elastic modulus prediction model can directly output the cell elastic modulus. In this way, after the cell image is collected and before it is input into the cell elastic modulus prediction model, the step of converting the cell image into data is saved, which is convenient to use and can improve the acquisition speed of the cell elastic modulus.

[0124] Fig.11 FIG. 1 is a block diagram of a cell elastic modulus device according to an exemplary embodiment. Fig.11 As shown, in this embodiment, the cell elastic modulus device includes:

[0125] The acquisition module 1101 is configured to acquire first cell deformation data; the first cell deformation data includes cell deformation characteristic data of elastic deformation of the cell at least at a time when the cell passes through the variable cross-section pipeline; the cross-sectional areas of at least two places of the variable cross-section pipeline are different; the inner diameter of the narrowest area of ​​the variable cross-section pipeline is larger than the cell diameter and smaller than twice the cell diameter, so as to avoid the cell being directly squeezed by the variable cross-section pipeline during the process of passing through the variable cross-section pipeline, and only a single cell is allowed to pass through the narrowest area each time;

[0126] The prediction module 1102 is configured to input the first cell deformation data into a trained cell elastic modulus prediction model so that the cell elastic modulus prediction model outputs the cell elastic modulus.

[0127] An embodiment of the present application also proposes an electronic device, comprising a processor and a memory; the memory is used to store a computer program executable by the processor; the processor is used to execute the computer program in the memory to implement the microchannel-based cell elastic modulus acquisition method described in any of the above embodiments.

[0128] The embodiments of the present application further provide a computer-readable storage medium. When the executable computer program in the storage medium is executed by a processor, the microchannel-based cell elastic modulus acquisition method described in any of the above embodiments can be implemented.

[0129] Regarding the device in the above embodiment, the specific manner in which the processor performs the operation has been described in detail in the embodiment of the method, and will not be elaborated here.

[0130] Fig.12 1 is a block diagram of an electronic device according to an exemplary embodiment. For example, the electronic device 1200 may be provided as a server. Fig.12 The device 1200 includes a processing component 1222, which further includes one or more processors, and a memory resource represented by a memory 1232 for storing instructions that can be executed by the processing component 1222, such as an application. The application stored in the memory 1232 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 1222 is configured to execute instructions to perform the above-mentioned method for obtaining cell elastic modulus based on microchannels.

[0131] The device 1200 may also include a power component 1226 configured to perform power management of the device 1200, a wired or wireless network interface 1250 configured to connect the device 1200 to a network, and an input / output (I / O) interface 1258. The device 1200 may operate based on an operating system stored in the memory 1232, such as Windows Server™, MacOS X™, Unix™, Linux™, FreeBSD™, or the like.

[0132] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 1232 including instructions, which can be executed by the processing component 1222 of the device 1200 to perform the above method. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0133] In the present invention, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance. The term "plurality" refers to two or more than two, unless otherwise clearly defined.

[0134] The above description of the embodiments is to facilitate those of ordinary skill in the art to understand and apply the present application. It is obvious that those familiar with the art can easily make various modifications to these embodiments and apply the general principles described herein to other embodiments without having to pay creative labor. Therefore, the present application is not limited to the embodiments herein, and improvements and modifications made by those skilled in the art based on the contents disclosed in the present application without departing from the scope and spirit of the present application are within the scope of the present application.

Claims

1. A method for obtaining cell elastic modulus based on microchannel, characterized in that: include: Acquire first cell deformation data; the first cell deformation data includes cell deformation characteristic data of elastic deformation of the cell at least at a time when the cell passes through the variable cross-section pipeline; the cross-sectional areas of at least two places of the variable cross-section pipeline are different; the inner diameter of the narrowest area of ​​the variable cross-section pipeline is larger than the cell diameter and smaller than twice the cell diameter, so as to avoid the cell being directly squeezed by the variable cross-section pipeline during the process of passing through the variable cross-section pipeline, and only a single cell is allowed to pass through the narrowest area each time; Inputting the first cell deformation data into a trained cell elastic modulus prediction model so that the cell elastic modulus prediction model outputs a cell elastic modulus; The cell elastic modulus prediction model is a fully connected neural network; The fully connected neural network includes an input layer, a flattening layer, a fusion layer, at least one hidden layer and an output layer connected in sequence; The input layer is used to input the first cell deformation data; The flattening layer is used to flatten the first cell deformation data to obtain the flattened first cell deformation data; The fusion layer is used to perform fusion processing on the flattened first cell deformation data to obtain fusion data; The fusion layer is fully connected to the hidden layer, the last hidden layer in the at least one hidden layer is connected to the output layer, and the output layer is used to output the cell elastic modulus; When the cell deformation characteristic data includes cell contour data, the cell contour data is the coordinates of contour points on the cell contour in the cell image; Acquiring the cell contour data, comprising: Taking a picture of the cell by a camera device to obtain the cell image; Binarizing the cell image to obtain a binarized cell image; Performing contour extraction on the binarized cell image to obtain a cell contour; determining the centroid of the cell outline; Extract contour points from the cell contour at specified angle intervals to obtain N contour points, , α is the specified angle interval; For each of the contour points, obtaining the coordinates of the contour point; wherein the centroid is the origin; For each moment, the cell contour data includes first contour data of a main view and second contour data of a top view of the cell; The cell elastic modulus prediction model is trained by the following method: The specified constitutive model and elastic modulus are used to calculate and obtain the deformation data of the second cell; Using the second cell deformation data as training data, training the cell elastic modulus prediction model until the cell elastic modulus prediction model meets a specified condition, thereby obtaining a trained cell elastic modulus prediction model; The first cell deformation data includes cell contour data and pressure information at the inlet of the variable cross-section pipe, and the first cell deformation data also includes at least one of an extension index, a non-roundness, a curvature ratio, a relative cell size and an inverse of a passing time.

2. The method for obtaining cell elastic modulus based on microchannel according to claim 1, characterized in that: The cell elastic modulus prediction model is a convolutional neural network; the first cell deformation data is a cell image; the cell image includes a front view and a top view of the cell.

3. The method for obtaining cell elastic modulus based on microchannel according to claim 1, characterized in that: The constitutive model regards the cell as a shell-like structure object, and the inside and outside of the cell are viscous liquids; the constitutive model is the Hookean model, the Mooney-Rivlin model, the Neo-Hookean model, the Skalak model or the Evans & Skalak model.

4. A cell elastic modulus device, characterized in that: include: The acquisition module is configured to acquire first cell deformation data; the first cell deformation data includes cell deformation characteristic data of elastic deformation of the cell at least at a time when the cell passes through the variable cross-section pipe; the cross-sectional areas of at least two places of the variable cross-section pipe are different; the inner diameter of the narrowest area of ​​the variable cross-section pipe is larger than the cell diameter and smaller than twice the cell diameter, so as to avoid the cell being directly squeezed by the variable cross-section pipe during the process of passing through the variable cross-section pipe, and only a single cell is allowed to pass through the narrowest area each time; A prediction module, configured to input the first cell deformation data into a trained cell elastic modulus prediction model, so that the cell elastic modulus prediction model outputs a cell elastic modulus; The cell elastic modulus prediction model is a fully connected neural network; The fully connected neural network includes an input layer, a flattening layer, a fusion layer, at least one hidden layer and an output layer connected in sequence; The input layer is used to input the first cell deformation data; The flattening layer is used to flatten the first cell deformation data to obtain the flattened first cell deformation data; The fusion layer is used to perform fusion processing on the flattened first cell deformation data to obtain fusion data; The fusion layer is fully connected to the hidden layer, the last hidden layer in the at least one hidden layer is connected to the output layer, and the output layer is used to output the cell elastic modulus; When the cell deformation characteristic data includes cell contour data, the cell contour data is the coordinates of contour points on the cell contour in the cell image; Acquiring the cell contour data, comprising: Taking a picture of the cell by a camera device to obtain the cell image; Binarizing the cell image to obtain a binarized cell image; Performing contour extraction on the binarized cell image to obtain a cell contour; determining the centroid of the cell outline; Extract contour points from the cell contour at specified angle intervals to obtain N contour points, , α is the specified angle interval; For each of the contour points, obtaining the coordinates of the contour point; wherein the centroid is the origin; For each moment, the cell contour data includes first contour data of a main view and second contour data of a top view of the cell; The cell elastic modulus prediction model is trained by the following method: The specified constitutive model and elastic modulus are used to calculate and obtain the deformation data of the second cell; Using the second cell deformation data as training data, training the cell elastic modulus prediction model until the cell elastic modulus prediction model meets a specified condition, thereby obtaining a trained cell elastic modulus prediction model; The first cell deformation data includes cell contour data and pressure information at the inlet of the variable cross-section pipe, and the first cell deformation data also includes at least one of an extension index, a non-roundness, a curvature ratio, a relative cell size and an inverse of a passing time.

5. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory is used to store a computer program executable by the processor; and the processor is used to execute the computer program in the memory to implement the method according to any one of claims 1 to 4.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the executable computer program in the storage medium is executed by a processor, the method according to any one of claims 1 to 4 can be implemented.

Images