Multi-parameter fusion single-cell flow thermal system and multi-parameter fusion method

Abstract

The invention provides a multi-parameter fusion single-cell flow thermal system and a multi-parameter fusion method. The differential symmetric structure can directly measure the product and heat production change before and after metabolism of the micro-channels on the two sides of the single-cell micro-fluidic chip, the attribute of a single cell can be known, and the development rule of cell metabolism heat along with growth, division and apoptosis can be accurately revealed. According to the microfluidic vacuum cavity structure design, the stability of a temperature field in the calorimetric cavity is ensured, and the influence of the external environment on cell metabolism is avoided. Through the microfluid driving module and the electron microscope microscopic cell measurement module, single cell motion control and extracellular microenvironment accurate control are realized, and the problem that the existing research lacks dynamic monitoring and quantitative analysis of the cell size is solved. By means of the method, multi-dimensional and multi-parameter research on the cell metabolism process can be achieved, and meanwhile, the invention provides a method for measuring cell metabolites and heat production in a non-invasive mode.

Description

Technical Field

[0001] The present invention belongs to the field of single-cell metabolic heat detection, and particularly relates to a multi-parameter fusion single-cell flow calorimetry system and a multi-parameter fusion method. Background Art

[0002] Abnormal cell energy metabolism is a common cause of various diseases including obesity, diabetes, and cancer. Therefore, monitoring the cell metabolism process is of great significance for basic biomedical science research and the development of clinical diagnosis and treatment technologies for metabolic diseases. To study cell metabolism, current techniques include measuring oxygen consumption rate (OCR), extracellular acidification (ECA), and cell metabolic heat.

[0003] The oxygen consumption rate of cell metabolism is one of the important factors that can reflect the functional state of cells (including cell differentiation, growth, and mitochondrial energy metabolism). Therefore, monitoring the oxygen consumption rate during the cell metabolism process is of great significance. In 1985, Vanderkooi et al. developed an oxygen probe molecule based on phosphorescence quenching and applied it to the detection of oxygen in biological tissues. The core of the oxygen nanosensor is a highly sensitive oxygen probe molecule. During the detection process, molecular oxygen penetrates through the matrix pores to the probe molecule and collides with it, resulting in fluorescence quenching. According to the quantitative relationship between the luminescence intensity and the oxygen concentration, the oxygen concentration in the surrounding environment can be deduced.

[0004] Extracellular acidification is mainly measured by the pH value of the cell fluid, and a pH meter is usually used. The pH meter determines the acidity or alkalinity of the solution by measuring the concentration of hydrogen ions in the solution, thereby representing the pH value. The specific principle is as follows: When a reversible hydrogen ion indicator electrode and a reference electrode are simultaneously immersed in the solution to be measured, and the potential of the reference electrode remains constant, a potential difference will be generated between the two electrodes at a certain temperature. This potential difference is related to the activity of hydrogen ions in the solution. Therefore, the change in the pH value of the solution to be measured can be directly represented as the change in the electromotive force of the battery formed by it:

[0005]

[0006] where R is the gas constant 8.3143 J·℃ -1 ·g -1 , F is the Faraday constant 96487.0 C / mol, T is the absolute temperature, and E 0 is the standard electrode potential. It can be seen from the above formula that after measuring the electromotive force between the two electrodes, the pH value of the cell fluid can be calculated.

[0007] Meanwhile, the metabolism of cells also depends on cell size, and this feedback between size and metabolism provides a mechanism for controlling growth and cell size. Cell size is usually in the micrometer range, so a special micrometer is needed to observe the size changes of individual cells before and after metabolism under a microscope. A microscopic micrometer consists of an ocular micrometer and a stage micrometer. The ocular micrometer can be directly used to measure cell size, but it needs to be calibrated with the stage micrometer before use to determine the actual length represented by each small grid of the ocular micrometer under a certain magnification of the eyepiece and objective lens of the microscope. Then, based on the number of grids of the ocular micrometer equivalent to the cell, the size of the cell can be calculated.

[0008] In addition, accurate measurement of cell metabolic heat is of great basic scientific and clinical application value for revealing the regulatory mechanisms of cell metabolism, differentiation, growth, and apoptosis, exploring new paths for the evolution of tumor cells, conducting drug screening and toxicity testing, and assisting in the diagnosis and treatment of metabolic diseases.

[0009] Calorimetry is the science of heat measurement, which has advantages such as direct measurement, real-time, and in-situ compared to indirect methods such as oxygen consumption rate and metabolite analysis, and has a long history and important position in the measurement of metabolic heat effects. Chinese Patent with Publication No. CN118209583A discloses a differential power compensation single-cell flow calorimetry system, which proposes to use chip calorimetry technology and a differential structure to compensate for the heat loss caused by cell metabolism in the microfluidic channel, thereby measuring the metabolic heat of single cells. Although this system can already reflect the metabolic situation of cells by measuring cell metabolic heat, it is impossible to grasp the mutual relationship between multiple parameters and obtain more accurate single-cell metabolic processes by observing a single parameter. At the same time, due to the limitations of the accuracy and sensitivity of existing single-cell calorimetry systems, combined with the lack of methods for multi-dimensional metabolic information fusion processing such as heat production and product component analysis, the research on biological thermodynamics and metabolic regulation mechanisms at the single-cell level has not been systematically carried out.

[0010] Combining the simultaneously measured extracellular acidification rate (ECAR), oxygen consumption rate (OCR) with real-time metabolite flux analysis can significantly improve the interpretive value of isolated measurements made in these complex metabolic systems, and thus it is of great research significance to fuse and process multi-dimensional metabolic information. Simply put, the data set fusion processing of ECAR, OCR, cell size, and metabolic heat is a process of effectively integrating, reshaping, transforming, and analyzing various types of metabolic data and other related biological information. Since there are certain relationships between the data sets, multivariate statistical analysis techniques can be used to deeply study the interactions and correlations between multiple variables and provide a more comprehensive description of cell metabolism characteristics. The fusion processing of multi-dimensional metabolic information can reveal the potential associations and patterns between different metabolic indicators, provide richer input variables for the construction of metabolic models, and provide important clues for the construction of metabolic networks and the study of metabolic regulation mechanisms.

[0011] In summary, based on the design of a differential power compensation single-cell calorimeter, the present invention proposes a multi-parameter fusion single-cell flow calorimetry system and a multi-parameter fusion method. The system fuses multi-dimensional information such as cell metabolic heat production and product components, studies the biothermodynamic behavior of processes such as metabolism and stress response at the single-cell level, and explores the internal relationship between cell metabolism and internal physiological mechanisms such as differentiation, proliferation, and apoptosis and macroscopic thermodynamics; it provides a new technical means for studying processes such as metabolism and stress response at the single-cell level and exploring the relationship between physiological phenomena such as cell metabolism and differentiation, proliferation, and apoptosis, and is expected to play an important role in the fields of basic biomedical science research and clinical diagnosis. Summary of the Invention

[0012] Aiming at the problem of the lack of multi-dimensional metabolic information fusion processing in the existing research mentioned in the background technology, the present invention provides a multi-parameter fusion single-cell flow calorimetry system and a multi-parameter fusion method.

[0013] On the one hand, the present invention provides a multi-parameter fusion single-cell flow calorimetry system, including a microfluidic vacuum calorimeter, an on-line control and data acquisition system, and a sensor device.

[0014] The microfluidic vacuum calorimeter includes a microfluidic chip, a single-crystalline silicon thermopile, and a heating wire.

[0015] The on-line control and data acquisition system includes a parameter measurement module, a temperature measurement module, and a power compensation module.

[0016] The microfluidic chip adopts a differential structure and includes a sample microchannel and a reference microchannel; the monocrystalline silicon thermopile is respectively placed at the inlet and outlet of the microchannel to measure the temperature difference between the two microchannels; cell micro-wells are arranged inside the microchannel, and cells are captured by an electron microscope, and the cell size is measured by a micrometer; the heating wire is embedded at the bottom of the cell micro-well, and power compensation is performed by heating, so as to keep the temperature difference between the two microchannels at the inlet equal to the temperature difference between the two microchannels at the outlet, and the applied electric power for power compensation is the heat generated by the metabolism of a single cell.

[0017] The sensor device is placed at the front end of the liquid inlet of the microfluidic chip; wherein, the parameter measurement module processes the data collected by the sensor device.

[0018] On the other hand, the present invention also provides a multi-parameter fusion method applying the multi-parameter fusion single-cell flow calorimetry system as described above, which is specifically as follows:

[0019] Perform data preprocessing on the obtained dataset of multiple variables:

[0020] Perform data cleaning on it, delete or correct the error data; perform standardization processing on the cleaned dataset.

[0021] Perform fusion processing on the standardized dataset:

[0022] Use principal component analysis technology to reduce the variables to a few principal components; group the cell samples according to the metabolite characteristics, distinguish normal data points and abnormal data points, so as to screen and simplify the dataset; establish a multiple regression model to analyze how the metabolites jointly affect the biological characteristics of cells.

[0023] Present the analysis results in an intuitive way through visualization technology:

[0024] Explain the interactions and associations between different metabolites, perform correlation analysis on the fused data, and reveal the degree of association between different variables.

[0025] The beneficial effects of the present invention:

[0026] (1) The differential symmetric structure described in the present invention directly measures the changes in products and heat generation before and after metabolism in the microchannels on both sides of a single-cell microfluidic chip, understands the properties of a single cell, and accurately reveals the development law of cell metabolic heat with growth, division and apoptosis.

[0027] (2) The present invention realizes the multi-dimensional and multi-parameter research on the cell metabolism process, and provides a non-invasive method for measuring cell metabolites and heat generation.

[0028] (3) The microfluidic vacuum chamber structure design of the present invention ensures the stability of the temperature field inside the calorimetric chamber, avoiding the influence of the external environment on cell metabolism.

[0029] (4) The present invention realizes the manipulation of single-cell movement and the precise control of the extracellular microenvironment through the microfluidic driving module and the electron microscope microscopic cell measurement module, solving the problem that existing research lacks dynamic monitoring and quantitative analysis of cell size. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a structural diagram of a multi-parameter fusion single-cell flow calorimetry system of the present invention;

[0031] Figure 2 It is a three-dimensional structural diagram of the microfluidic vacuum calorimeter device of the present invention;

[0032] Figure 3 It is a schematic diagram of the composition of the sensor device of the present invention;

[0033] Figure 4 It is a schematic diagram of the composition of the dissolved oxygen sensor of the present invention;

[0034] Figure 5 It is a schematic diagram of the composition of the parameter measurement module;

[0035] Figure 6 It is a schematic diagram of the composition of the microfluidic driving module of the present invention;

[0036] Figure 7 It is a schematic diagram of the electron microscope microscopic cell measurement module of the present invention;

[0037] Figure 8 It is a flow chart of the cell image processing algorithm of the present invention;

[0038] Figure 9 It is a flow chart of a multi-parameter fusion method of the present invention;

[0039] Figure 10 It is a flow chart of the principal component analysis calculation of the present invention;

[0040] Figure 11 It is a flow chart of the cluster analysis of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0042] The present invention will be further described below with reference to the drawings:

[0043] As Figure 1 shown, the embodiment of the present application provides a multi-parameter fusion single-cell flow calorimetry system, which includes a microfluidic vacuum calorimeter 1, an on-line control and data acquisition system 2, a sensor device 3, a computer 2-1, a microfluidic driving module 2-3, and an electron microscopy measurement module 2-4.

[0044] As Figure 2 shown, the microfluidic vacuum calorimeter 1 includes a vacuum calorimeter chamber 1-1, an electron microscope 1-2, a microfluidic chip 1-3, a sample inlet tube 1-4-a, a reference inlet tube 1-4-b, a sample outlet tube 1-5-a, a reference outlet tube 1-5-b, an optical seal feedthrough 1-6, a microfluidic fixture 1-7, single-crystalline silicon thermopiles 1-8-a and 1-8-b, a heating wire 1-9, and a microfluidic driving module 2-3.

[0045] The interior of the vacuum calorimeter chamber 1-1 is a vacuum environment with three layers of nested shielding, which can significantly reduce heat loss caused by air conduction, convection, and radiation, etc., help maintain the stability of experimental conditions, improve the accuracy and sensitivity of experiments, and ensure the repeatability of experimental data. A micro window is provided below the calorimeter chamber body. The electron microscopy measurement module 2-4 uses the electron microscope 1-2 to observe the fluid and cell movement in the calorimeter chamber in real time through the micro window, and a micrometer is installed on the microscope to measure the size of the imprisoned cells.

[0046] The sample inlet tube 1-4-a, the reference inlet tube 1-4-b, the sample outlet tube 1-5-a, and the reference outlet tube 1-5-b are respectively connected to the inlets and outlets of the microchannels of the microfluidic chip 1-3 for introducing and discharging the fluids required for the experiment. At the same time, the sample inlet tube 1-4-a and the reference inlet tube 1-4-b are respectively connected to the sensor device 3.

[0047] The microfluidic chip 1-3 has a differential structure and includes two microchannels, namely a sample microchannel and a reference microchannel. The microchannels are two microfluidic channels with exactly the same size and equal flow rates. Among them, live cells are placed in the sample microchannel, and inactivated cells or empty samples are placed in the reference microchannel as a reference. Cell micro-wells are processed inside the microchannels for capturing cells. The single-crystalline silicon thermopile is divided into an inlet thermopile and an outlet thermopile. The inlet thermopile measures the temperature difference between the two microchannels at the inlet, and the outlet thermopile measures the temperature difference between the two microchannels at the outlet. Thin-film heating wires are embedded at the bottoms of the cell micro-wells of the two microchannels, including a sample-side heating wire and a reference-side heating wire, and power compensation is performed through heating to keep the temperature difference between the two microchannels at the inlet and the temperature difference between the two microchannels at the outlet equal.

[0048] Fluids enter from the sample side and the reference side respectively. Cells are captured by an electron microscope at the cell micro-wells. At this time, living cells generate heat during metabolic activities, while the reference side (inactivated cells or empty samples) does not generate heat. Due to the heat generated by cell metabolic activities, the temperature of the sample microchannel will be slightly higher than that of the reference microchannel, which will result in unequal temperature differences between the two microchannels at the inlet and the two microchannels at the outlet. The temperature difference between the two microchannels is measured by a thermopile, and then the power compensation module automatically adjusts the PID parameters according to the measured temperature difference to output current to heat the heating wire, making the temperature differences at the two places equal. The heating power is the metabolic heat of the cells. After the measurement is completed, a pulsed flow is applied by a microfluidic driving pump to discharge the fluid in the microchannel.

[0049] The optical seal feedthroughs 1-6 are used to transmit optical signals while maintaining a vacuum environment. The electron microscope 1-2 can observe the microchannel through the lens on the optical seal feedthrough 1-6. The imaging system based on an inverted microscope helps to track the activity of cell specimens. The microfluidic fixture 1-7 is used to fix the microchannel and ensure an effective vacuum seal is formed between the microchannel and the calorimetric chamber to prevent temperature disturbances caused by fluid evaporation. The inlet pipe is further connected to the external microfluidic driving module 2-3 through the microfluidic fixture 1-7. The single-crystalline silicon thermopile is divided into the inlet thermopile 1-8-a and the outlet thermopile 1-8-b, which are fabricated by MEMS and can detect extremely small temperature changes. They are respectively placed at the inlet and outlet of the microchannel. Among them, the inlet thermopile 1-8-a measures the temperature difference between the two microchannels at the inlet, and the outlet thermopile 1-8-b measures the temperature difference between the two microchannels at the outlet. KΩ-level thin-film heating wires 1-9 are embedded at the bottom of the cell micro-wells of the two microchannels, which are fabricated by MEMS and include a sample-side heating wire and a reference-side heating wire. Power compensation is carried out through heating to keep the temperature differences between the two microchannels at the inlet and the two microchannels at the outlet equal.

[0050] Refer to Figure 3 , the sensor device 3 consists of a sample reservoir 3-1-a, a pH measurement device body 3-2-a, a sample liquid outlet 3-3-a, a reference reservoir 3-1-b, a pH measurement device body 3-2-b, and a reference liquid outlet 3-3-b. The microfluidic driving module 2-3 transports fluids to the sample reservoir 3-1-a and the reference reservoir 3-1-b for sealed storage at a constant flow rate. The hydrostatic pressure caused by gravity injects the solution from the reservoir into the inlet of the microfluidic channel, and the flow rate is adjusted by changing the height of the reservoir to ensure stable fluid transportation. The pH measurement device body measures the pH value of the injected fluid.

[0051] Furthermore, as Figure 4As shown, 3-4-a and 3-4-b are both dissolved oxygen sensors, which are patch-type and can be integrated into a microfluidic channel to measure the dissolved oxygen concentration in the fluid by immersion. The dissolved oxygen sensor used is based on the fluorescence probe method, and the concentration of dissolved oxygen is calculated by measuring the phase difference between the excitation fluorescence and the reference light. There is a relationship between the phase difference and the fluorescence lifetime, so the phase difference can be measured by a phase-locked amplifier and other technologies. Then, the oxygen concentration value is obtained according to the established phase difference-dissolved oxygen concentration calibration curve.

[0052] Further, the pH value and dissolved oxygen concentration measured by the sensor device 3 are processed by the parameter measurement module 2-2. Figure 5 , the specific process is:

[0053] Firstly, the signal conditioning circuit constructed by the operational amplifier performs preliminary amplification and linearization processing on the weak voltage signals generated by the pH meter and the dissolved oxygen sensor, and then a low-noise amplifier is introduced to further amplify the signal.

[0054] Secondly, the amplified signal is low-pass filtered to filter out the input noise of the amplifier and the additional noise generated by itself;

[0055] Next, the filtered signal is converted into a digital signal through an analog-to-digital converter, and the signal at this time can be transmitted to a microprocessor for data processing;

[0056] Finally, communicate with the computer to upload the collected parameter measurement results.

[0057] like Figure 6 As shown, the microfluidic drive module 2-3 is the key to achieve fine fluid control, which includes a microfluidic drive pump 2-3-1-a, a sample liquid reservoir 2-3-2-a, a microfluidic drive pump 2-3-1-b, a reference liquid reservoir 2-3-2-b, a flow sensor 2-3-3 and a waste liquid pool 2-3-4. The microfluidic drive pump uses the pressure difference between the inlet and outlet of the sealed liquid reservoir to send the fluid into the microfluidic channel at a constant flow rate. The operating parameters of the microfluidic drive pump, including flow rate, delivery volume and running time, can be set by the host computer software of the computer 2-1. The sample liquid reservoir 2-3-2-a and the reference liquid reservoir 2-3-2-b are used to seal and store cell culture fluid and are connected to the microfluidic drive pump 2-3-1 through a fluid passage. During the movement of the fluid, the flow sensor 2-3-3 monitors the flow rate in real time, realizes feedback control of the flow rate, and ensures that the fluid delivery process is stable and meets the preset conditions. The fluid first passes through the sensor device 3 to measure pH value and oxygen concentration, and then flows into the microfluidic chip 1-3. The waste liquid pool 2-3-4 is used to receive the fluid flushed out after the measurement to avoid contamination.

[0058] The electron microscope microscopic cell measurement module 2-4 observes the microchannel by designing a micro window under the cavity of the microfluidic vacuum calorimeter and using advanced microscopic imaging technology. Refer to Figure 7 , when the fluid containing cells passes through the microchannel, the electron microscope continuously captures cell images. These images are transmitted to a computer, and single-cell automatic recognition is achieved through analysis by an image processing algorithm. An ocular micrometer is loaded on the eyepiece of the electron microscope, and an objective micrometer is loaded on the objective lens. Size calibration is performed before measurement. After a single cell is recognized, its size is measured, and the measured data is uploaded to the computer. During the whole process, the microfluidic driving module 2-3 continuously and stably conveys the fluid, providing a necessary flow environment for single-cell analysis.

[0059] The temperature measurement module 2-5 processes the weak voltage signal generated by the single-crystal silicon thermopile through a high-precision thermoelectric signal acquisition circuit. First, the voltage signal is amplified by a signal amplification circuit, then the noise in the signal is removed by filtering to ensure the clarity of the signal. The amplified analog signal is converted into a digital signal by an analog-to-digital converter (ADC), and then the temperature is calculated by a microprocessor (MCU) (based on the characteristic curve and calibration data of the thermopile), and communicated with the computer through a communication module. In the embodiment of the present application, a switch is used to upload the temperature data.

[0060] As Figure 8 shown, the cell images need to be processed in the computer to achieve the recognition of single cells. The specific process of the image processing algorithm is as follows:

[0061] First, the cell images continuously captured by the electron microscope are preprocessed to remove noise and interference factors and improve the image quality:

[0062] Since grayscale images are easier to process and can retain sufficient image information, the color images are converted into grayscale images and normalized.

[0063] The salt-and-pepper noise and Gaussian noise in the cell images are removed by median filtering and Gaussian filtering algorithms respectively to preserve the edge information of the images.

[0064] The grayscale histogram of the image is adjusted by histogram equalization to make the distribution of the grayscale values of the image more uniform, thereby enhancing the contrast of the image and making the cell structure clearer.

[0065] Then, edge detection is used to segment the preprocessed cell images to separate single cells from the background.

[0066] Finally, feature extraction is performed on the segmented cell images according to the histogram of oriented gradients method:

[0067] Calculate the gradient magnitude and direction of each pixel in the image to capture the contour information of the cells;

[0068] Divide the image into small cell units, calculate the histogram of gradient directions within each cell unit, combine multiple cell units into a block, and normalize the histogram of gradient directions within the block;

[0069] Collect the orientation gradient feature vectors of all blocks in the image to form the final feature vector, and perform quantitative analysis on the extracted features to evaluate the cell morphology.

[0070] As Figure 9 shown, the embodiment of the present application also provides a multi-parameter fusion method applying a multi-parameter fusion single-cell flow calorimetry system as described above, including the following steps:

[0071] First, after obtaining the dataset of multiple variables, perform data preprocessing:

[0072] To ensure the integrity and accuracy of the measured pH value, dissolved oxygen concentration, cell size, and metabolic heat data, clean the data, deleting or correcting incorrect data;

[0073] Furthermore, since the units and dimensions of different datasets are different, standardize the cleaned dataset in the computer to make it comparable.

[0074] Then, perform fusion processing on the standardized dataset:

[0075] Use the principal component analysis (PCA) technique to reduce the dimensionality of the four variables of pH value, oxygen consumption rate, metabolic heat, and cell size to a few principal components, which can summarize most of the information of the original data and reduce data redundancy;

[0076] Adopt a clustering algorithm (such as K-means clustering) to group cell samples according to metabolite characteristics (including pH value, oxygen consumption rate, metabolic heat, cell size), which can distinguish normal data points from abnormal data points, achieving the effect of screening and simplifying the dataset;

[0077] Establish a multiple regression model to predict the relationship between one or more dependent variables (such as cell viability, metabolic rate, etc.) and independent variables (pH value, oxygen consumption rate, metabolic heat, cell size), which helps to analyze how these metabolites jointly affect the metabolic process of cells.

[0078] Finally, according to the results of PCA, clustering analysis, and regression analysis, present the analysis results in an intuitive way through visualization techniques:

[0079] Explain the interactions and associations among different metabolites. By using the Pearson correlation coefficient, perform a correlation analysis on the fused data to reveal the degree of association between different variables.

[0080] As Figure 10 shown, due to the high data dimension, it is necessary to perform data dimensionality reduction and feature extraction through principal component analysis. The specific calculation process is as follows:

[0081] First, organize the data of each pH value, oxygen consumption rate, metabolic heat, and cell size after preprocessing into an n×4 data matrix, where n represents the number of samples and the columns represent the four variables.

[0082] Calculate the 4×4 covariance matrix. The element in the i-th row and j-th column of the matrix represents the covariance between variable i and variable j, and the diagonal elements of the matrix represent the covariance of each variable with itself.

[0083] Perform eigen decomposition on the covariance matrix, that is, solve the eigenvalues and eigenvectors of the covariance matrix.

[0084] Sort the eigenvalues from largest to smallest, and select the eigenvectors corresponding to the top k largest eigenvalues as the principal components.

[0085] Construct a new feature space based on the selected principal components, and project the original four-variable data into this new space. Project the original data onto the selected principal components to obtain the scores of each sample on the principal components, and these scores can be used for subsequent analysis and visualization.

[0086] Furthermore, since the amount of data to be fused is large, the K-means clustering method is selected for clustering analysis. The process is as Figure 11 shown, specifically as follows:

[0087] First, select the number of clusters K value through the silhouette coefficient method. The silhouette coefficient can measure the similarity of a point to the points within its cluster and the difference from the points in other clusters. Its value ranges from -1 to 1, and the larger the value, the better the clustering effect.

[0088] Select a series of possible K values, and run the K-means algorithm on the data set for each K value to perform clustering, obtaining k clusters and the clustering labels of each sample point.

[0089] For each sample point, calculate its average distance from other sample points within the same cluster (i.e., the cohesion a_i), and the minimum value of its average distance from sample points in other clusters (i.e., the separation b_i).

[0090] Calculate the silhouette coefficient of each sample point through s_i = (b_i - a_i) / max(a_i, b_i).

[0091] Calculate the average silhouette coefficient for all sample points to obtain the average silhouette coefficient at this k value.

[0092] After all K values have been calculated as described above, select the K value with the largest average silhouette coefficient as the optimal number of clusters.

[0093] Then, select K samples as the initial cluster centers, calculate the Euclidean distance from each sample to the K cluster centers, and assign each sample to the nearest cluster center.

[0094] Finally, recalculate the mean of all samples in the cluster as the new cluster center, and repeat the above process until the cluster centers no longer change.

[0095] The working process of one embodiment of this application:

[0096] 1. Prepare fluid samples

[0097] Before starting the experiment, use an appropriate culture medium for cell culture as the sample to be tested, inactivate the cells through chemical treatment as the reference, and ensure that its pH, ionic strength, temperature, etc. match those of the sample to be tested.

[0098] 2. Load fluid into the microfluidic driving pump

[0099] Load the fluid sample into the liquid storage pool of the microfluidic driving pump. This step needs to be carried out under sterile conditions to prevent sample contamination, ensure that the connection between the liquid storage pool and the microfluidic driving pump is firm and sealed, and avoid fluid leakage or air entering the system.

[0100] 3. Set operating parameters

[0101] Set the operating parameters of the microfluidic driving pump through the upper computer software of the computer, including flow rate, delivery volume, operating time, etc. The cross-sectional area of the microfluidic channel is 35×50μm 2 , considering the cross-sectional area of the microchannel, inject at a fluid velocity of 0.1mm s -1 .

[0102] 4. Start the driving pump

[0103] After confirming that all settings are correct, start the driving pump. The driving pump starts to deliver the fluid in the liquid storage pool at a constant flow rate through the inlet pipe to the microchannel of the microfluidic chip. During the fluid movement, the system monitors the flow rate in real time through the flow sensor to achieve feedback control of the flow rate and ensure that the fluid delivery process is stable and meets the preset conditions.

[0104] 5. Cell capture and size measurement

[0105] The movement states of the fluid and cells in the microfluidic channel of the microfluidic chip are observed in real time through the electron microscope microscopic cell measurement module. The flow rate is adjusted by cooperating with the microfluidic driving pump module, and the cells are confined in the cell microwells. The automatic recognition of single cells is realized by combining the image processing algorithm, and the size change during the metabolic process of single cells is measured by a micrometer.

[0106] 6. Parameter measurement and power compensation

[0107] After the cells are stabilized in the cell microwells, the weak voltage signal collected by the single-crystal silicon thermopile is processed by the temperature measurement module, and the voltage signals of all parameters collected by the sensors are processed by the parameter measurement module. The data is uploaded to the computer through the microprocessor and the switch. The power compensation module automatically adjusts the PID parameters according to the temperature data uploaded by the temperature measurement module to heat the output current of the heating wire, and the applied electric power for power compensation is the heat generated by the metabolism of single cells.

[0108] 7. End the experiment

[0109] After the measurement is completed, a pulsed flow is applied by the microfluidic driving pump to discharge the fluid in the microfluidic channel, and necessary cleaning and disinfection are carried out for use in the next experiment.

[0110] In summary, the present invention has good applicability in the multi-dimensional research of the cell metabolic process, and provides a non-invasive method for measuring cell metabolites and heat production.

[0111] The above specific description further details the purpose, technical solution and beneficial effects of the invention. It should be understood that the above is only a specific embodiment of the present invention and is not used to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A multi-parameter fusion single-cell flow calorimetry system, characterized in that: It includes a microfluidic vacuum calorimeter, an online control and data acquisition system, and a sensor device; The microfluidic vacuum calorimeter comprises a microfluidic chip, a single crystal silicon thermopile and a heating wire; The online control and data acquisition system includes a parameter measurement module, a temperature measurement module and a power compensation module; The microfluidic chip adopts a differential structure, including a sample microchannel and a reference microchannel; the single crystal silicon thermopiles are respectively placed at the inlet and outlet of the microchannel, and are used to measure the temperature difference between the two microchannels; a cell microwell is arranged inside the microchannel, and cells are captured by an electron microscope microscopic cell measurement module; a heating wire is embedded in the bottom of the cell microwell, and is used to perform power compensation by heating, so as to keep the temperature difference between the two microchannels at the inlet and the temperature difference between the two microchannels at the outlet equal, and the electric power applied for power compensation is the heat generated by the metabolism of a single cell; The sensor device is placed at the front end of the liquid inlet of the microfluidic chip; The parameter measurement module is used to process the data collected by the sensor device.

2. A multi-parameter fusion single-cell flow calorimetry system according to claim 1, characterized in that: The microfluidic chip is placed in a vacuum calorimetry chamber, the interior of the vacuum calorimetry chamber is a vacuum environment, and has multiple layers of nested shielding to reduce heat loss, maintain stability, and ensure data repeatability.

3. A multi-parameter fusion single-cell flow calorimetry system according to claim 1 or 2, characterized in that: A micro window is also provided below the vacuum calorimetry chamber body, and the electron microscope microscopic cell measurement module observes the fluid and cell movement in the calorimetry chamber in real time through the micro window.

4. The multi-parameter fusion single-cell flow calorimetry system according to claim 1, characterized in that: The sensor device includes a pH measurement unit and an oxygen concentration measurement unit, which are placed at the inlet and outlet of the liquid inlet pipe and are used to measure the changes in pH concentration and oxygen concentration at the two places; the liquid inlet pipe is connected to the microchannel inlet of the microfluidic chip and is used to introduce fluid.

5. A multi-parameter fusion single-cell flow calorimetry system according to claim 4, characterized in that: The oxygen concentration measurement unit is integrated in a microfluidic channel and is used for immersion measurement of dissolved oxygen concentration in a fluid.

6. A multi-parameter fusion single-cell flow calorimetry system according to claim 1 or 4, characterized in that: The parameter measurement module is used to process the pH value and dissolved oxygen concentration measured by the sensor device, specifically: First, a signal conditioning circuit constructed by an operational amplifier performs preliminary amplification and linearization processing on the weak voltage signal generated by the sensor device, and then a low-noise amplifier is introduced to further amplify the signal; Secondly, the amplified signal is low-pass filtered to filter out the input noise of the amplifier and the additional noise generated by itself; Next, the filtered signal is converted into a digital signal through an analog-to-digital converter, and the signal at this time can be transmitted to a microprocessor for data processing; Finally, communicate with the computer to upload the collected parameter measurement results.

7. The multi-parameter fusion single-cell flow calorimetry system according to claim 1, characterized in that: It also includes a microfluidic driving module, which adjusts the flow rate by changing the height of the liquid storage tank and is used to monitor the flow rate in real time during the movement of the fluid to ensure stable fluid delivery.

8. The multi-parameter fusion single-cell flow calorimetry system according to claim 1, characterized in that: The electron microscope microscopic cell measurement module is used to continuously capture cell images and realize automatic identification of single cells through image processing algorithm analysis; at the same time, a micrometer is loaded to measure the size of the imprisoned cells, and size calibration is performed before measurement, and size measurement is performed after the single cell is identified.

9. The multi-parameter fusion single-cell flow calorimetry system according to claim 1, characterized in that: The temperature measurement module processes the weak voltage signal generated by the single crystal silicon thermopile through a high-precision thermoelectric signal acquisition circuit. Specifically, the voltage signal is first amplified by a signal amplification circuit, and then the noise in the signal is removed by filtering to ensure the clarity of the signal. The amplified analog signal is converted into a digital signal, and then the temperature is calculated by a microprocessor to complete the upload of temperature data.

10. A multi-parameter fusion method using a multi-parameter fusion single-cell flow calorimetry system according to any one of claims 1 to 9, characterized in that: The following steps are involved: Perform data preprocessing on the obtained data sets of multiple variables: Perform data cleaning, delete or correct erroneous data; standardize the cleaned data set; Perform fusion processing on the standardized data set: Use principal component analysis techniques to reduce the variables to a few principal components; Grouping cell samples according to metabolite characteristics to distinguish normal data points from abnormal data points to screen and simplify data sets; establishing a multivariate regression model to analyze how the metabolites jointly affect the metabolic process of cells; The analysis results are presented in an intuitive way through visualization technology: Interpret the interactions and associations between different metabolites, and perform correlation analysis on the fused data to reveal the degree of association between different variables.

Images