A method and system for detecting cell viability using dual-channel differential dynamic adhesion impedance analysis

Abstract

This invention discloses a method and system for cell viability detection using dual-channel differential dynamic adhesion impedance analysis. The system utilizes a microfluidic chip integrating detection and reference channels to synchronously apply fluid disturbances to both channels, inducing a dynamic process of "adhesion → detachment → re-adhesion" in cells. During this process, the differential impedance curve (after background subtraction) is acquired in real time, and six dynamic feature parameters, including the amplitude and rate of decrease and recovery, are extracted. A ridge regression algorithm is used to construct a cell viability prediction model, and the six dynamic feature parameters are input into the model to obtain cell viability values. This invention effectively eliminates environmental and fluid common-mode noise through differential structure and achieves label-free, in-situ, highly sensitive measurement of cell viability using impedance analysis. It has advantages such as accurate results, good stability, and high automation, and is suitable for fields such as cytotoxicity testing and drug screening.

Description

Technical Field

[0001] This invention relates to the field of bio-intelligent sensing, specifically to a method and system for detecting cell viability based on dual-channel differential dynamic adhesion impedance analysis. Background Technology

[0002] Traditional biochemical methods for detecting cell viability, such as trypan blue and MTT assays, are endpoint assays and cannot provide rapid, real-time analysis of cell viability. Furthermore, these methods only obtain metabolic indicators and are insufficient to reflect changes in cell adhesion behavior and mechanical adaptability in fluid environments.

[0003] Impedance analysis is a novel label-free method for detecting cell viability in recent years. For example, Chinese patent application CN201810448547.6 discloses a cell viability assessment device and method based on an interdigital electrode impedance sensor. This method analyzes cell viability and state information by detecting the impedance value output by the interdigital electrode sensor. However, this method only characterizes the adhesion strength and impedance of cells in a static environment. It cannot simulate the adhesion loss caused by fluid shear forces in the in vivo environment, nor can it analyze the active recovery behavior of cells after the fluid disappears, resulting in low accuracy in cell viability assessment.

[0004] Furthermore, existing microfluidic impedance detection technologies mostly employ a single-channel direct measurement mode. In dynamic detection processes involving fluid flow, the impedance signal is highly susceptible to interference from microfluidic environmental factors, such as instability at the electrode-solution interface due to continuous fluid flow and subtle changes in the conductivity of the culture medium. These non-cell-specific environmental noises (common-mode interference) often superimpose on the weak cell impedance signal, making it difficult to accurately distinguish minute dynamic changes in cell adhesion, thus limiting the signal-to-noise ratio and reproducibility of the detection system.

[0005] Therefore, there is an urgent need for a cell activity detection system that can simultaneously simulate the in vivo fluid environment and effectively eliminate environmental common-mode interference. Summary of the Invention

[0006] To address the aforementioned issues, this invention proposes a cell viability detection method and system based on dual-channel differential dynamic adhesion impedance analysis. This system constructs a differential detection architecture by introducing a reference channel, effectively eliminating common-mode interference caused by fluid flow and environmental factors. Under fluid disturbance, the differential impedance method monitors the entire cell adhesion process in real time. The complete "impact-recovery" dynamic characterization enriches the detection dimensions of cell state and improves the accuracy of cell viability assessment.

[0007] The technical solution adopted in this invention is as follows:

[0008] A cell viability detection system based on dual-channel differential dynamic adhesion impedance analysis includes:

[0009] The dual-channel microfluidic chip is composed of a lower substrate and an upper cover plate bonded together. The middle layer integrates two parallel microchannels, namely the reference channel and the detection channel. The dual-channel microfluidic chip integrates interdigitated electrodes, which are placed on the surface of the glass substrate and are respectively connected to the bottom of the reference channel and the detection channel.

[0010] The differential electrical signal acquisition unit is connected to the interdigital electrode signal to acquire the impedance signal of the detection channel cells during the entire process of adhesion, partial detachment and re-adhesion, as well as the impedance signal of the reference channel during the same process.

[0011] The synchronous stable fluid disturbance providing unit is connected to the same end of the reference channel and the detection channel through pipelines, so as to simultaneously form controllable fluid disturbances with consistent parameters in the reference channel and the detection channel;

[0012] A data processing unit connected to the differential electrical signal acquisition unit performs calculations on the acquired impedance signal to obtain the dual-channel differential dynamic adhesion impedance change curve for the entire process. Based on the dual-channel differential dynamic adhesion impedance change curve, the key points of the adhesion state change are extracted, and the dynamic adhesion characteristic parameters are calculated.

[0013] A model calculation unit is connected to the data processing unit by signal. The model calculation unit has a pre-stored cell activity prediction model, and the model outputs cell activity measurement results based on dynamic adhesion characteristic parameters.

[0014] Furthermore, the interdigitated portion of the interdigitated electrode is specifically disposed on the glass substrate surface at the bottom of each channel, and its ends are respectively led out to the chip edge to form a reference electrode port, a common ground port and a detection electrode port, which are connected to the differential electrical signal acquisition unit.

[0015] Furthermore, the reference channel and the detection channel are connected at one end to form a common liquid inlet, which is connected to the synchronous stable fluid disturbance providing unit; the other end of each is connected to the waste liquid cylinder.

[0016] Furthermore, the differential electrical signal acquisition unit employs an impedance meter.

[0017] A cell viability detection method based on dual-channel differential dynamic adhesion impedance analysis, based on the aforementioned cell viability detection system based on dual-channel differential dynamic adhesion impedance analysis, comprises the following steps:

[0018] Step 1: Prepare the cell suspension to be tested and inject it into the detection channel of the microfluidic chip for static culture, while keeping the reference channel in a blank state at all times;

[0019] Step 2: Simultaneously inject controllable fluid disturbances with consistent parameters into the reference channel and the detection channel, causing adherent cells in the detection channel to undergo a dynamic adhesion process;

[0020] Step 3: Before, during, and after the fluid disturbance is applied, the cell impedance signal Z located in the detection channel is acquired simultaneously. sense and the background impedance signal Z located in the reference channel ref ;

[0021] Step 4: Use the differential algorithm to process the synchronously acquired data to obtain the differential impedance data Z. diff =Z sense -Z ref Thus, a complete dual-channel differential dynamic adhesion impedance variation curve is obtained;

[0022] Step 5: Extract four key points characterizing the changes in the adhesion state from the dual-channel differential dynamic adhesion impedance change curve, and calculate six dynamic adhesion characteristic parameters based on the coordinates of the key points: impedance drop amplitude Z_drop, impedance drop rate K_drop, impedance recovery amplitude Z_recover, impedance recovery rate K_recover, impedance change amplitude Z_flow during the fluid disturbance stage, and impedance change slope K_flow during the fluid disturbance stage.

[0023] Step 6: After preprocessing the six dynamic adhesion feature parameters, input them into the cell viability prediction model and output the cell viability prediction results.

[0024] Furthermore, the calculated differential impedance data are recorded and stored in chronological order to form a differential data sequence reflecting the impedance change over time during cell dynamic adhesion, thus obtaining a complete dual-channel differential dynamic adhesion impedance change curve.

[0025] Furthermore, in step 5, the differential impedance data is first subjected to Savitzky-Golay smoothing filtering, and then the smoothed curve is subjected to baseline normalization.

[0026] Furthermore, the selection criteria for the four key points characterizing the changes in the adhesion state are as follows: the impedance point within the stable range where the impedance change rate is lower than a preset threshold before the fluid disturbance is applied is identified as key point A; the impedance point where the curve drops rapidly and reaches a local minimum after the fluid disturbance begins is identified as key point B; the starting point where the fluid disturbance ends and the impedance changes from a low plateau to a continuous rise is identified as key point C; and the local maximum point corresponding to the subsequent impedance recovery and return to stability is identified as key point D.

[0027] Furthermore, the cell viability prediction model in step 6 is constructed using the ridge regression algorithm.

[0028] Furthermore, the preprocessing in step 6 involves organizing the six obtained dynamic adhesion feature parameters, constructing the feature vector of the sample to be tested, and performing standardization processing.

[0029] Compared with the prior art, the present invention has the following advantages:

[0030] 1. This invention constructs a novel dual-channel differential microfluidic detection architecture, effectively improving the accuracy and stability of dynamic detection. By integrating a reference channel and a detection channel into the microfluidic chip and simultaneously applying fluid disturbances, the system can effectively cancel common-mode interference caused by fluid flow and fluctuations in culture medium conductivity in real time through differential operations. This design completely solves the problems of severe baseline drift and difficulty in extracting weak cell signals in traditional single-channel microfluidic impedance detection under dynamic fluid environments, greatly improving the reliability of detection data.

[0031] 2. This invention proposes a novel method for dynamic adhesion characterization based on the "impact-recovery" mechanism. Changes in cell adhesion state under fluid disturbance are important characteristics reflecting their physiological activity and mechanical homeostasis. This invention induces cells to undergo a complete "full adhesion → partial detachment → re-adhesion" process in a microfluidic environment, transforming the abstract cell adhesion behavior into a quantifiable differential electrical signal. This achieves an objective characterization of cell activity state, avoiding the limitations of relying solely on a single metabolic or endpoint indicator.

[0032] 3. This invention extracts multiple dynamic characteristic parameters reflecting cell adhesion loss, recovery ability, and steady-state adaptive behavior based on high-fidelity differential impedance response curves. By modeling and training multiple sets of samples, a cell activity prediction model is established, transforming the assessment of cell activity from qualitative or semi-quantitative analysis to a quantitative prediction method based on multi-feature comprehensive judgment, effectively improving the accuracy and comprehensiveness of cell activity assessment.

[0033] 4. This invention uses high-frequency alternating current as the detection signal, eliminating the need for cell staining or labeling. It does not damage cell structure or physiological state during the detection process, enabling continuous and real-time monitoring of the same batch of cells. Compared with traditional staining methods, it avoids reagent consumption and errors introduced by human operation. The detection process is highly automated, has good repeatability, and is suitable for standardized operation. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of a cell viability detection system based on dual-channel differential dynamic adhesion impedance analysis as described in this invention.

[0035] Figure 2 This is a schematic diagram of the structure of the dual-channel microfluidic chip designed in this invention, where a is a top view and b is a side view.

[0036] Figure 3 This is a schematic diagram of the dynamic cell adhesion process.

[0037] Figure 4 It shows the dynamic adhesion impedance curve after dual-channel differential operation and four key points A, B, C, and D.

[0038] Figure 5 This is a graph showing the relationship between the actual and predicted values ​​of the cell activity prediction model established in this invention.

[0039] Figure 6 yes Figure 1 The flowchart of the device shown.

[0040] The following are the serial numbers and names of the components or processes in the attached diagram: 1. Centrifuge tube; 3. 96-well plate; 4. Interdigitated electrode; 5. Dual-channel microfluidic chip; 6. Impedance meter; 8. Syringe pump; 9. Computer; 10. CCK8; 11. Reference electrode port; 12. Common ground port; 13. Detection electrode port; 14. Substrate; 15. Cover plate; 16. Reference channel; 17. Detection channel; 18. Adherent cells; 19. Interdigitated portion of the electrode; 20. Complete cell adhesion stage; 21. Cell adhesion begins to weaken stage; 22. Cell adhesion further weakens stage; 23. Weakest cell adhesion stage. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention.

[0042] like Figure 1 , 2 As shown, this invention designs a cell activity detection system based on dual-channel differential dynamic adhesion impedance analysis. This system constructs a dual-channel differential microfluidic-impedance integrated structure, applies synchronous and controllable fluid disturbances to adherent cells within a microfluidic chip, records the dynamic changes in differential impedance generated by cells during adhesion, partial detachment, and re-adhesion in real time, and calculates six dynamic characteristic parameters characterizing cell dynamic adhesion behavior based on the differential impedance response curve. Then, it combines a machine learning model to quantitatively predict cell activity.

[0043] The system specifically includes:

[0044] The dual-channel microfluidic chip 5, which integrates both reference and detection channels, has the following specific structure: Figure 2As shown in Figures a and b, the dual-channel microfluidic chip 5 is formed by bonding a lower substrate 14 and an upper cover plate 15. Two parallel microchannels are integrated in the intermediate layer between the substrate 14 and the cover plate 15, namely a reference channel 16 and a detection channel 17. The dual-channel microfluidic chip 5 integrates interdigitated electrodes 4. The interdigitated portions 19 of the interdigitated electrodes 4 are specifically disposed on the surface of the glass substrate 14 at the bottom of each channel, with their ends leading out to the chip edge to form a reference electrode port 11, a common ground port 12, and a detection electrode port 13, respectively.

[0045] The differential electrical signal acquisition unit 6 is connected to the interdigitated electrode 4 to acquire the impedance signal throughout the process; in this embodiment, an impedance meter 6 is used.

[0046] The synchronous stable fluid disturbance providing unit 8 is connected to the same side of the reference channel 16 and the detection channel 17 via pipelines, and simultaneously injects complete culture medium into the reference channel 16 and the detection channel 17, thereby simultaneously forming controllable fluid disturbances with consistent parameters in the reference channel 16 and the detection channel 17. In this embodiment, an injection pump 8 is used.

[0047] The data processing unit is connected to the differential electrical signal acquisition unit. It performs calculations on the acquired impedance signal to obtain the dual-channel differential dynamic adhesion impedance change curve for the entire process. Based on the dual-channel differential dynamic adhesion impedance change curve, it extracts the key points of the adhesion state change and calculates the dynamic adhesion characteristic parameters.

[0048] The model calculation unit pre-stores a cell viability prediction model and outputs cell viability measurement results based on dynamic adhesion characteristic parameters.

[0049] In this embodiment, the reference channel 16 and the detection channel 17 are connected at the same side to form a common liquid inlet 12, which can be connected to the injection pump 8 to maintain synchronous injection.

[0050] In this embodiment, the outlet ends of the reference channel 16 and the detection channel 17 are connected to the waste liquid cylinder 7 through pipelines for centralized recycling of waste liquid.

[0051] In this embodiment, the substrate 14 is a rigid insulating substrate, such as glass; the cover plate 15 typically has good biocompatibility and elastic sealing properties, such as PDMS.

[0052] In this embodiment, the system also includes a cell preparation unit for preparing a cell suspension to be tested. In this embodiment, centrifuge tube 1 is used.

[0053] Based on the aforementioned cell viability detection system based on dual-channel differential dynamic adhesion impedance analysis, this invention can also realize a cell viability detection method based on dual-channel differential dynamic adhesion impedance analysis, combined with... Figure 6 The specific steps of this method are as follows:

[0054] Step 1: Prepare the cell suspension to be tested and inject it into the detection channel 17 of the microfluidic chip 5 for static culture.

[0055] In this embodiment, the cells were first digested, centrifuged, and then resuspended in complete culture medium in centrifuge tube 1, and the cell density was adjusted to approximately 1×10^6 cells / mL.

[0056] During the experiment, the obtained cell suspension was injected into the detection channel 17 of the microfluidic chip 5 using a syringe.

[0057] After injection, the microfluidic chip 5 was placed in an incubator for static incubation (37 ℃, 5% CO2) for 8 hours. During this period, the test cells 18 in the detection channel 17 gradually settled and naturally adhered to the surface of the electrode interdigital portion 19 at the bottom, forming a stable initial adhesion state, corresponding to... Figure 3 The cell adhesion integrity stage 20 is completed; while the reference channel 16 remains blank because no cells are added, and is used to provide a fluid and environmental background reference in subsequent steps.

[0058] Step 2: Connect the common inlet of the dual-channel microfluidic chip 5 to the syringe pump 8, and connect the outlets of the reference channel 16 and the detection channel 17 to the waste liquid tank via conduits. By controlling the operating parameters (flow rate and time) of the syringe pump 8, complete culture medium is simultaneously injected into the dual-channel microfluidic chip 5. At this time, controllable fluid disturbances with consistent parameters are simultaneously formed in the reference channel 16 and the detection channel 17.

[0059] Under the action of fluid shear force, the adherent cells 18 located in the detection channel 17 undergo a dynamic adhesion process. For example... Figure 3 As shown: When no fluid is applied, the cell is in the initial stage of complete cell adhesion 20; after fluid disturbance is applied, the cell deforms under force and enters the stage of weakening cell adhesion 21; as the fluid disturbance continues, the cell further shrinks, the contact area decreases, and the cell adhesion further weakens 22; when the fluid disturbance reaches the set intensity and stabilizes, the cell exhibits the stage of weakest cell adhesion 23. Figure 3 The solid arrow in the middle reflects this process of force release.

[0060] When the infusion pump stops working and the fluid disturbance is removed, the cells, relying on the restorative force of their own cytoskeleton, spread out and adhere back to the electrode surface, showing a change in their adhesion state that is the opposite of what was mentioned above (i.e., gradually recovering from stage 23 to stage 20). Figure 3 The dashed arrows indicate the sequence of changes. These stages together constitute the dynamic adhesion process of cells under fluid disturbance within detection channel 17.

[0061] Step 3: Before, during and after the fluid disturbance is applied, the impedance signal on the interdigital electrode 4 of the microfluidic chip 5 is continuously acquired by the impedance meter 6.

[0062] In this embodiment, the impedance meter 6 is electrically connected to the three interdigitated electrodes 4 on the microfluidic chip 5 (specifically corresponding to the reference electrode port 11, the common ground port 12, and the detection electrode port 13) via wires. The impedance meter 6 uses 1kHz AC current as the excitation signal to synchronously acquire the cell impedance signal (denoted as Z) located in the detection channel 17. sense ) and the background impedance signal located in reference channel 16 (denoted as Z) ref ).

[0063] Step 4: The two acquired impedance signals are transmitted to computer 9 in real time via a data interface. On computer 9, a differential algorithm is used to process the synchronously acquired data (Z). diff =Z sense -Z ref This is to deduct common-mode interference caused by fluid flow and fluctuations in culture medium conductivity. The computer records and stores the calculated differential impedance data in chronological order to form a differential data sequence reflecting the impedance changes over time during cell dynamic adhesion, obtaining a complete dual-channel differential dynamic adhesion impedance change curve, such as... Figure 4 As shown, the downward arrows represent the moment when fluid is applied, the upward arrows represent the moment when fluid is removed, and the gray area represents the continuous phase of fluid application.

[0064] During signal acquisition, the impedance data acquisition time range covers the stages of stable cell adhesion, fluid disturbance, and adhesion change after disturbance removal, thereby obtaining a complete dual-channel differential dynamic adhesion impedance change curve, providing high signal-to-noise ratio basic data for subsequent signal processing and feature parameter extraction.

[0065] Step 5: Automatic signal preprocessing is performed on the dual-channel differential dynamic adhesion impedance change curve in computer 9. Then, based on the trend of the differential impedance curve over time, four key points A, B, C, and D characterizing the adhesion state change are identified. Six dynamic adhesion characteristic parameters are calculated based on the coordinates of these key points: impedance drop amplitude Z_drop, impedance drop rate K_drop, impedance recovery amplitude Z_recover, impedance recovery rate K_recover, impedance change amplitude Z_flow during the fluid disturbance stage, and impedance change slope K_flow during the fluid disturbance stage. Savitzky-Golay smoothing filtering is applied to the differential impedance data to further reduce the impact of environmental noise and instantaneous fluctuations on the impedance curve. Baseline normalization is then performed on the smoothed curve, i.e., the differential impedance values ​​at all time points are divided by the differential impedance value at the initial moment, ensuring comparability of impedance curves obtained from different experimental groups.

[0066] In this embodiment, the selection criteria for the four key points characterizing the changes in the adhesion state are as follows: the impedance point corresponding to the stable interval where the rate of change of impedance is lower than a preset threshold before the fluid disturbance is applied is identified as key point A (corresponding to...). Figure 3 The complete cell adhesion stage 20); the impedance point where the curve drops rapidly and reaches a local minimum after the start of fluid disturbance is identified as key point B (corresponding to Figure 3 The weakest stage of cell adhesion in the process (23); the starting point where the fluid disturbance ends and the impedance changes from a low plateau to a continuous rise is identified as key point C; and the local maximum point corresponding to the subsequent impedance recovery and re-stabilization is identified as key point D.

[0067] Point A corresponds to the stable adhesion state before the fluid disturbance is applied, and its coordinates are set as (t). A Z A Point B corresponds to the state after the rapid descent of the adhesive layer, and its coordinates are set as (t). B Z B Point C corresponds to the state at the end of the fluid disturbance, and its coordinates are set as (t). C Z C Point D corresponds to the state after adhesion recovery, and its coordinates are set as (t). D Z D ).

[0068] Based on the above key points, the following six dynamic adhesion characteristic parameters are calculated from the difference curve: impedance drop amplitude Z_drop, impedance drop rate K_drop, impedance recovery amplitude Z_recover, impedance recovery rate K_recover, impedance change amplitude Z_flow during the fluid disturbance stage, and impedance change slope K_flow during the fluid disturbance stage. The calculation formula is:

[0069]

[0070] Step 6: After preprocessing the six dynamic adhesion feature parameters obtained in Step 5, use them as input variables for the cell viability prediction model, and output the cell viability prediction results from the pre-stored cell viability prediction model in Computer 9.

[0071] In this embodiment, six dynamic adhesion feature parameters obtained from the dual-channel differential dynamic adhesion experiment of the test cells are used as inputs, and the established cell activity prediction model is used to predict the activity of the test cells.

[0072] In this embodiment, the six dynamic adhesion feature parameters obtained by the cells to be tested in steps 1 to 5 are first organized to construct a feature vector of the sample. The feature vector is then standardized to ensure that each feature parameter has a uniform scale range, thereby eliminating the influence of differences in the dimensions of different feature parameters on the estimation results of the regression coefficients. This yields the input vector X_test used for prediction.

[0073] Subsequently, the input vector X_test is input into the trained ridge regression model, and the regression coefficient vector β* determined in the model is used to calculate the input vector, outputting the corresponding predicted cell activity value ŷ; wherein, the predicted value ŷ is a continuous value, used to quantitatively characterize the relative activity level of the cells under the current experimental conditions.

[0074] In practical applications such as drug screening, the predicted cell activity value can be directly used to compare the activity levels between different test samples, or to analyze the activity change trend of the same cell under different treatment conditions (such as different fluid disturbance intensities, different treatment times, or different external stimuli).

[0075] More specifically, the cell viability prediction model based on the ridge regression algorithm and its training process are as follows:

[0076] In this embodiment, to construct a modeling sample with a continuous activity distribution, a gradient thermal damage method was used to treat cells from the same source. By heating the cells at different temperatures for a fixed time, cell samples with different activity levels were obtained. Subsequently, the cell activity of each group of cells was measured in a 96-well plate using the cell activity detection reagent CCK-8 10, and the results were used as cell activity reference values.

[0077] For cell samples with different activity levels, the dual-channel differential dynamic adhesion experiment described in steps 1 to 5 was performed. Dynamic adhesion impedance change curves were acquired under the same fluid disturbance conditions and detection parameters, and six dynamic adhesion characteristic parameters were calculated from these curves. Thus, each cell sample was represented as a feature vector containing the six characteristic parameters, and each vector was paired with a corresponding cell activity reference value.

[0078] In computer 9, the dynamic adhesion feature vectors corresponding to all samples are summarized in the order of the samples to construct a feature matrix X, where each row corresponds to a cell sample and each column corresponds to a dynamic adhesion feature parameter; at the same time, the cell activity reference values ​​obtained by the cell activity detection reagent are constructed as a target vector y, and it is made to correspond one-to-one with the feature matrix X according to the samples.

[0079] Before model training, the feature matrix X is standardized to ensure that each feature parameter has a uniform scale range, thereby eliminating the influence of differences in the dimensions of different feature parameters on the regression coefficient estimation results.

[0080] Based on the feature matrix X and the target vector y, a cell activity prediction model is established using the ridge regression method. The regression coefficient vector β is solved by minimizing the objective function containing the regularization term. The regularization term is controlled by the regularization parameter λ, which is used to constrain the amplitude of the regression coefficients to reduce the impact of the correlation between the dynamic adhesion feature parameters on the model stability, thereby avoiding model overfitting and improving the model's prediction robustness under different activity sample conditions.

[0081] After training the ridge regression model, its performance was analyzed and evaluated. The model was validated using the GroupKFold cross-validation strategy. The ridge regression model achieved R²=0.955 and RMSE=7.15% on the validation set, indicating that the model has high fitting accuracy and low prediction error.

[0082] The cell viability value ŷ predicted by the ridge regression model was correlated with the cell viability reference value y measured by the CCK-8 gold standard method. The results showed that the Pearson correlation coefficient reached r=0.992 and passed the significance test (p<0.001), indicating that the predicted results and the gold standard detection results are highly consistent.

[0083] Furthermore, the predicted cell viability value ŷ output by the model was compared and quantitatively evaluated with the corresponding reference cell viability value y. The results are as follows: Figure 5As shown, the model-predicted values ​​and actual activity values ​​are closely distributed around y=x throughout the entire activity range, indicating that the established cell activity prediction model can stably and reliably characterize the quantitative mapping relationship between differential dynamic adhesion feature parameters and cell activity.

[0084] Therefore, the ridge regression model is determined to be successfully trained and used as an effective model for subsequent prediction of cell viability in the test samples.

[0085] This invention enables label-free, in-situ, and quantitative assessment of cell viability without the need for staining or other destructive treatments of the cells to be tested. Furthermore, since the prediction model is based on dual-channel differential dynamic adhesion impedance characteristics, this method can effectively reduce the interference of dead cell residues or non-specific adhesion on the detection results, and can also eliminate common-mode noise in the fluid environment at the hardware level, thereby significantly improving the stability and reliability of cell viability detection results.

[0086] The above embodiments are only used to illustrate the design concept and features of the present invention, and their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. The protection scope of the present invention is not limited to the above embodiments. Therefore, all equivalent changes or modifications made based on the principles and design ideas disclosed in the present invention are within the protection scope of the present invention.

Claims

1. A cell viability detection system based on dual-channel differential dynamic adhesion impedance analysis, characterized in that, include: The dual-channel microfluidic chip (5) is formed by bonding a lower substrate (14) and an upper cover plate (15). The middle layer integrates two parallel microchannels, namely a reference channel (16) and a detection channel (17). The dual-channel microfluidic chip (5) integrates interdigitated electrodes (4), which are placed on the surface of the glass substrate (14) and connected to the bottom of the reference channel (16) and the detection channel (17) respectively. The differential electrical signal acquisition unit is connected to the interdigital electrode (4) to acquire the impedance signal of the cell in the detection channel (17) during the entire process of adhesion, partial detachment and re-adhesion, as well as the impedance signal of the reference channel (16) during the same process. The synchronous stable fluid disturbance providing unit is connected to the same end of the reference channel (16) and the detection channel (17) through pipelines, and simultaneously forms controllable fluid disturbances with consistent parameters in the reference channel (16) and the detection channel (17); A data processing unit connected to the differential electrical signal acquisition unit performs calculations on the acquired impedance signal to obtain the dual-channel differential dynamic adhesion impedance change curve for the entire process. Based on the dual-channel differential dynamic adhesion impedance change curve, the key points of the adhesion state change are extracted, and the dynamic adhesion characteristic parameters are calculated. A model calculation unit is connected to the data processing unit by signal. The model calculation unit has a pre-stored cell activity prediction model, and the model outputs cell activity measurement results based on dynamic adhesion characteristic parameters.

2. The cell viability detection system based on dual-channel differential dynamic adhesion impedance analysis according to claim 1, characterized in that, The interdigitated portion (19) of the interdigitated electrode (4) is disposed on the surface of the substrate (14) at the bottom of each channel, and its ends are led out to the edge of the chip to form a reference electrode port (11), a common ground port (12) and a detection electrode port (13), which are connected to the differential electrical signal acquisition unit.

3. The cell viability detection system based on dual-channel differential dynamic adhesion impedance analysis according to claim 1, characterized in that, The reference channel (16) and the detection channel (17) are connected at one end to form a common liquid inlet (12), which is connected to the synchronous stable fluid disturbance providing unit; the other end is connected to the waste liquid cylinder.

4. The cell viability detection system based on dual-channel differential dynamic adhesion impedance analysis according to claim 1, characterized in that, The differential electrical signal acquisition unit uses an impedance meter (6).

5. A method for detecting cell viability based on dual-channel differential dynamic adhesion impedance analysis, characterized in that, Based on the cell viability detection system according to claim 1, the method steps are as follows: Step 1: Prepare the cell suspension to be tested and inject it into the detection channel (17) of the microfluidic chip (5) for static culture. The reference channel (16) remains blank at all times. Step 2: Simultaneously inject controllable fluid disturbances with consistent parameters into the reference channel (16) and the detection channel (17), and the adherent cells (18) in the detection channel (17) undergo a dynamic adhesion process; Step 3: Before, during, and after the fluid disturbance is applied, the cell impedance signal Z in the detection channel (17) is simultaneously acquired. sense and the background impedance signal Z located in the reference channel (16) ref ; Step 4: Use the differential algorithm to process the synchronously acquired data to obtain the differential impedance data Z. diff =Z sense -Z ref Thus, a complete dual-channel differential dynamic adhesion impedance variation curve is obtained; Step 5: Extract four key points characterizing the changes in the adhesion state from the dual-channel differential dynamic adhesion impedance change curve, and calculate six dynamic adhesion characteristic parameters based on the coordinates of the key points: impedance drop amplitude Z_drop, impedance drop rate K_drop, impedance recovery amplitude Z_recover, impedance recovery rate K_recover, impedance change amplitude Z_flow during the fluid disturbance stage, and impedance change slope K_flow during the fluid disturbance stage. Step 6: After preprocessing the six dynamic adhesion feature parameters, input them into the cell viability prediction model and output the cell viability prediction results.

6. The cell viability detection method based on dual-channel differential dynamic adhesion impedance analysis according to claim 5, characterized in that, The calculated differential impedance data are recorded and stored in chronological order to form a differential data sequence reflecting the impedance change over time during cell dynamic adhesion, thus obtaining a complete dual-channel differential dynamic adhesion impedance change curve.

7. The cell viability detection method based on dual-channel differential dynamic adhesion impedance analysis according to claim 5, characterized in that, In step 5, the differential impedance data is first smoothed by Savitzky-Golay filtering, and then the smoothed curve is normalized.

8. The cell viability detection method based on dual-channel differential dynamic adhesion impedance analysis according to claim 5, characterized in that, The selection criteria for the four key points characterizing the changes in the adhesion state are as follows: the impedance point within the stable range where the impedance change rate is lower than the preset threshold before the fluid disturbance is applied is identified as key point A; the impedance point where the curve drops rapidly and reaches a local minimum after the fluid disturbance begins is identified as key point B; the starting point where the fluid disturbance ends and the impedance changes from a low plateau to a continuous rise is identified as key point C; and the local maximum point corresponding to the subsequent recovery of impedance and re-stability is identified as key point D.

9. The cell viability detection method based on dual-channel differential dynamic adhesion impedance analysis according to claim 5, characterized in that, In step 6, the cell viability prediction model is constructed using the ridge regression algorithm.

10. The cell viability detection method based on dual-channel differential dynamic adhesion impedance analysis according to claim 5, characterized in that, The preprocessing in step 6 is as follows: the six obtained dynamic adhesion feature parameters are sorted out, the feature vector of the sample to be tested is constructed, and the standardization process is performed.

Images