Intelligent POCT Detection Method and System Based on 5G and Microfluidics Technology

Abstract

This invention discloses an intelligent point-of-care testing (POCT) method and system based on 5G and microfluidic technology. The method includes: acquiring initial data of samples collected in real-time using microfluidic technology combined with integrated sensors; preprocessing the initial data and transforming the preprocessed data to obtain transformed data; constructing a mathematical model between sample characteristic changes and target analyte concentrations based on the transformed data; analyzing the frequency domain corresponding to the time-series signal of the initial data and determining high-frequency and low-frequency components in the frequency domain using a set threshold; calculating changes in sample characteristics based on the high-frequency and low-frequency components using the mathematical model and inferring abnormalities or pathogen types present in the sample to obtain inference results; and transmitting the inference results using 5G technology. Implementing this method can improve the real-time performance, accuracy, and overall performance of POCT detection.

Description

Technical Field

[0001] This invention relates to POCT detection methods, and more specifically to intelligent POCT detection methods and systems based on 5G and microfluidic technology. Background Technology

[0002] With the increasing demand for healthcare, POCT (Point-of-Care Testing) technology, as a fast, convenient, and on-site testing technology, has been widely used in clinical diagnosis, disease screening, and health monitoring.

[0003] However, traditional POCT devices have limitations in sample processing speed, data transmission efficiency, and result accuracy. Specifically, traditional POCT devices may not be able to produce results quickly, limiting their immediate response capabilities in field applications. Due to the lack of efficient communication protocols or infrastructure support, traditional POCT devices have slow data upload and download speeds, affecting real-time performance and interactivity in applications such as telemedicine. Some POCT tests may require large sample volumes or be insufficiently sensitive to the identification of specific biomarkers, leading to inaccurate results. Limited by technological and materials science advancements, some POCT devices offer limited quantitative analysis capabilities, failing to meet the high precision standards required for clinical diagnosis. Furthermore, existing POCT technologies often fail to fully utilize multi-source information fusion and advanced data analysis techniques to optimize detection performance.

[0004] Therefore, it is necessary to provide a new method to improve the real-time performance, accuracy, and overall performance of POCT detection. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide an intelligent POCT detection method and system based on 5G and microfluidic technology.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent POCT detection method based on 5G and microfluidic technology, comprising:

[0007] Microfluidic technology combined with integrated sensors is used to acquire initial data of samples in real time;

[0008] The initial data is preprocessed, and the preprocessed data is then transformed to obtain the transformed data.

[0009] A mathematical model is constructed based on the transformed data to relate the changes in sample characteristics to the concentration of the target analyte.

[0010] Analyze the frequency domain corresponding to the time series signal of the initial data, and determine the high-frequency and low-frequency components in the frequency domain by combining a set threshold;

[0011] The mathematical model is used to calculate the changes in sample characteristics based on the high-frequency and low-frequency components, and to infer the abnormalities or pathogen types present in the sample to obtain the inference results.

[0012] The inference results are transmitted using 5G technology.

[0013] The further technical solution is as follows: the initial data of the sample acquired in real time using microfluidic technology combined with integrated sensors includes:

[0014] The sample is introduced into the microfluidic system, and relevant data are collected in real time using integrated sensors to obtain initial data;

[0015] The initial data includes the concentration information of the target substance, the flow rate of the sample liquid, and the operating environment parameters of the equipment.

[0016] The further technical solution is as follows: preprocessing the initial data and transforming the preprocessed data to obtain transformed data includes:

[0017] The concentration information of the target substance and the flow rate of the sample liquid are preprocessed according to the equipment environmental operating parameters to obtain preprocessed data;

[0018] The Fourier transform algorithm is used to convert the time-domain signal corresponding to the preprocessed data into a frequency-domain signal to obtain the transformed data.

[0019] The further technical solution is as follows: the analysis of the frequency domain corresponding to the time series signal of the initial data, and the determination of the high-frequency and low-frequency components in the frequency domain in combination with a set threshold, includes:

[0020] Based on the data in the frequency domain and the initial data, combined with the set threshold, high-frequency components and low-frequency components are identified, and the weighting coefficients corresponding to the high-frequency components and low-frequency components are determined.

[0021] The step of identifying high-frequency and low-frequency components based on the frequency domain data and the initial data combined with a set threshold, and determining the weighting coefficients corresponding to the high-frequency and low-frequency components, includes:

[0022] Spectral analysis is performed on the frequency domain data, and high-frequency and low-frequency components are identified by combining sample characteristic changes, target analyte concentrations, and set thresholds.

[0023] Calculate the proportions of high-frequency and low-frequency components in the total to obtain the weighting coefficients corresponding to the high-frequency components and the low-frequency components.

[0024] The further technical solution is as follows: The mathematical model is used to calculate the changes in sample characteristics based on the high-frequency and low-frequency components, and to infer the abnormalities or pathogen types present in the sample to obtain the inference result, including:

[0025] The sample fluctuation is calculated based on the frequency data corresponding to the high-frequency and low-frequency components.

[0026] The overall sample irregularity is calculated based on the sample fluctuations and the weighting coefficients corresponding to the high-frequency and low-frequency components to obtain the sample irregularity.

[0027] Based on the sample fluctuations and irregularities, the abnormal conditions or pathogen types encountered by the sample are identified, and specific characteristics are classified into specific pathological states to obtain inference results.

[0028] The further technical solution is as follows: Identifying abnormal situations or pathogen types encountered by the sample based on its fluctuations and irregularity, and classifying specific characteristics to specific pathological states to obtain inference results, includes:

[0029] Obtain standard sample data as a reference, under the same conditions as the current sample but without the influence of the tested substances.

[0030] Calculate the specific changes between the preprocessed data and the standard sample data to obtain the change vector;

[0031] The change vector, the sample fluctuation, and the sample irregularity quantification are input into the analysis model to determine the abnormal situation or pathogen type encountered by the sample, and the specific characteristics are classified into specific pathological states.

[0032] The further technical solution is as follows: the mathematical model is C(t)=f(S(t),P(t))+b·V(t)+c· Where C(t) represents the concentration of the target substance, S(t) is the sample transfer speed, P(t) is the equipment pressure, V(t) is the liquid flow velocity, b and c are influence coefficients used to quantify the influence of flow velocity and its rate of change on concentration change; f(S(t),P(t)) in the mathematical model represents the influence of sample transfer speed and equipment pressure on the concentration of the target substance.

[0033] This invention also provides an intelligent POCT detection system based on 5G and microfluidic technology, comprising:

[0034] The initial data acquisition unit is used to acquire the initial data of the sample in real time by using microfluidic technology combined with integrated sensors;

[0035] A preprocessing unit is used to preprocess the initial data and transform the preprocessed data to obtain transformed data;

[0036] A mathematical model building unit is used to build a mathematical model between sample feature changes and target analyte concentrations based on the transformed data.

[0037] The analysis unit is used to analyze the frequency domain corresponding to the time series signal of the initial data, and determine the high-frequency and low-frequency components in the frequency domain by combining a set threshold.

[0038] The inference unit is used to calculate the changes in sample characteristics based on the high-frequency and low-frequency components using the mathematical model, and to infer the abnormalities or pathogen types present in the sample to obtain the inference result.

[0039] A transmitting unit is used to transmit the inference result using 5G technology.

[0040] The present invention also provides a computer device, the computer device including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the above-described method.

[0041] The present invention also provides a storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0042] The advantages of this invention compared to existing technologies are as follows: This invention improves the real-time performance and automation of the detection process by acquiring sample data in real time through microfluidic technology and integrated sensors. Initial data undergoes preprocessing and transformation to ensure data quality and consistency, providing an accurate foundation for subsequent analysis. By establishing a mathematical model, changes in sample characteristics are correlated with the concentration of the target analyte, enhancing the accuracy and reliability of the detection results. Frequency domain analysis of the time-series signal, combined with high- and low-frequency component identification of sample changes, effectively identifies potential anomalies or pathogens. The integration of 5G technology enables rapid data transmission and real-time inference, allowing detection results to be promptly delivered to relevant personnel, enhancing overall detection performance and response speed.

[0043] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0044] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 A flowchart illustrating the intelligent POCT detection method based on 5G and microfluidic technology provided in this embodiment of the invention;

[0046] Figure 2 A schematic block diagram of an intelligent POCT detection system based on 5G and microfluidic technology provided in an embodiment of the present invention;

[0047] Figure 3 A schematic block diagram of a computer device provided for an embodiment of the present invention. Detailed Implementation

[0048] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0050] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0051] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0052] Please see Figure 1 , Figure 1This is a schematic flowchart illustrating an intelligent POCT detection method based on 5G and microfluidics technology provided in an embodiment of the present invention. This intelligent POCT detection method based on 5G and microfluidics technology is applied in a server. The server interacts with the microfluidic system and the display terminal. The method in this embodiment mainly aims to address the limitations of traditional POCT devices in sample processing speed, data transmission efficiency, and result accuracy. By combining microfluidics technology and 5G communication technology, it aims to achieve faster detection speed, higher data transmission rate, more accurate results, and more efficient multi-source information fusion and advanced data analysis capabilities. Specifically, by using a microfluidic chip to manipulate trace amounts of liquid, rapid mixing, separation, and reaction of samples can be achieved, greatly improving the speed and efficiency of sample processing while reducing the required sample volume. Integrated sensors are used to monitor sample changes in real time and acquire initial data. These sensors can be electrochemical, optical, or other types of sensors, capable of providing information about changes in sample characteristics. The collected initial data is cleaned and formatted as necessary, and converted into a form suitable for further analysis. A relationship between changes in sample characteristics and the concentration of the target analyte is established to infer information about pathogen type or anomalies from the data. Transforming time-series signals into the frequency domain allows for the identification of different components (such as high-frequency and low-frequency components), leading to a better understanding of how sample features change over time. 5G technology ensures high-speed data transmission, enabling telemedicine and immediate response, significantly enhancing the application scenarios and practicality of point-of-care testing (POCT). It is particularly suitable for scenarios requiring rapid result assessment, such as emergency rooms, ambulances, medical institutions in remote areas, and on-site quarantine points—scenarios requiring pre-diagnosis of diseases. Furthermore, it is also suitable for situations requiring remote monitoring of patient health.

[0053] Figure 1 This is a flowchart illustrating the intelligent POCT detection method based on 5G and microfluidic technology provided in an embodiment of the present invention. Figure 1 As shown, the method includes the following steps S110 to S160.

[0054] S110: Employs microfluidic technology combined with integrated sensors to acquire initial data of samples collected in real time.

[0055] In this embodiment, the sample is introduced into the microfluidic system, and relevant data is collected in real time using integrated sensors to obtain initial data; wherein, the initial data includes the concentration information of the target substance, the flow rate of the sample liquid, and the operating environment parameters of the device.

[0056] In this embodiment, microfluidics utilizes a micrometer-scale channel network to control liquid flow, enabling precise manipulation of extremely small amounts of liquid. Furthermore, it allows for the integration of multiple functions on a single chip, such as mixing, separation, and reaction. This not only reduces sample and reagent consumption but also improves the automation of experiments.

[0057] An integrated sensor combines multiple sensing elements to simultaneously monitor several physical or chemical parameters. For example, in this embodiment, it can acquire real-time information on the concentration of the target substance, the flow rate of the sample liquid, and parameters of the device's operating environment. This data is crucial for subsequent data analysis.

[0058] Specifically, the sample to be tested is added to a specially designed microfluidic chip. Channels within the microfluidic system guide the sample flow, while integrated sensors begin collecting various relevant information in real time, including but not limited to the concentration of the target substance, the sample flow rate, and device operating environment parameters. The target substance refers to the specific component to be detected or analyzed; it can be a chemical substance, a biomolecule (such as DNA or protein), a cell, or other specific particles. Concentration information refers to the level of these target substances in the sample, usually expressed as mass per unit volume (e.g., mg / L) or moles (e.g., μM). Device operating environment parameters include, but are not limited to, factors that may affect the experimental results, such as temperature, humidity, and pressure. In some cases, other factors such as pH value and light intensity may also be involved.

[0059] The small size of microfluidic chips allows for more compact and lightweight POCT devices, making them easier to carry and operate, especially suitable for on-site testing in remote areas or emergency situations. Various types of sensors can be installed at key points, such as biosensors at the inlet, pressure sensors at the outlet, and temperature sensors arranged around the entire channel. Multiple functional modules are integrated onto a single platform, reducing external connecting parts and enhancing system stability and reliability.

[0060] S120. The initial data is preprocessed, and the preprocessed data is transformed to obtain the transformed data.

[0061] In one embodiment, step S120 described above may include steps S121 to S122.

[0062] S121. The concentration information of the target substance and the flow rate of the sample liquid are preprocessed according to the equipment environment operation parameters to obtain preprocessed data.

[0063] In this embodiment, this stage primarily addresses the data deviation problem caused by changes in the equipment's operating environment (such as temperature, humidity, and pressure). By adjusting these parameters, the concentration and flow rate of the target substance under actual conditions can be more accurately reflected. Specifically, the operating environment parameters of the equipment (such as temperature, humidity, and pressure) are analyzed, and the target substance concentration information and sample liquid flow rate are adjusted accordingly to eliminate the influence of environmental factors on the measurement results.

[0064] Effective data preprocessing using equipment operating environment parameters is a crucial step in ensuring the accuracy of measurement results and the stability of system behavior. In this embodiment, for temperature-sensitive sensors (such as enzyme activity detection), calibration curves or lookup tables are established at different temperatures. During the preprocessing stage, sensor readings are adjusted according to the current temperature to compensate for the effects of temperature changes. For pH-sensitive chemical reactions, the variation of sensor response under different pH conditions is determined experimentally, and corresponding correction strategies are then developed.

[0065] If substances in the environment may interfere with sensor readings (such as dissolved gases, suspended particles, etc.), these impurities can be removed using physical or chemical methods, or specialized algorithms can be developed to identify and eliminate their effects. For electromagnetic interference, hardware shielding measures can be implemented, combined with software algorithms to suppress noise, such as using filters to reduce high-frequency interference signals.

[0066] Considering the effects of temperature and pressure on liquid viscosity, an appropriate correction factor is introduced when calculating the flow velocity of the sample liquid to ensure that the flow velocity measurement reflects the true situation. This correction factor can be determined using self-learning historical data; when actual environmental conditions deviate from the flow meter's design standards, the flow meter is recalibrated according to the new environmental parameters to ensure its accuracy remains unaffected.

[0067] S122. Use the Fourier transform algorithm to convert the time-domain signal corresponding to the preprocessed data into a frequency-domain signal to obtain the transformed data.

[0068] In this embodiment, the Fourier transform algorithm is applied to convert the preprocessed time-domain signal (i.e., data that changes over time) into a frequency-domain signal. This step helps to identify hidden time-independent characteristics in the original data, such as frequency components, thereby enhancing the data analysis capabilities.

[0069] This preprocessing step reduces the impact of external environmental changes on measurement results, improving data quality and reliability. Using Fourier transform, data can be examined from a new perspective, revealing features that are not easily noticeable in the time domain, thus supporting more precise analysis.

[0070] S130. Construct a mathematical model between sample characteristic changes and target analyte concentration based on the converted data.

[0071] In this embodiment, the mathematical model is: Where C(t) represents the concentration of the target substance, S(t) is the sample transfer speed, P(t) is the equipment pressure, V(t) is the liquid flow velocity, b and c are influence coefficients used to quantify the influence of flow velocity and its rate of change on concentration change; f(S(t),P(t)) in the mathematical model represents the influence of sample transfer speed and equipment pressure on the concentration of the target substance.

[0072] In this embodiment, the sample transfer speed S(t) and the equipment pressure P(t) are the main influencing factors, while the liquid flow velocity V(t) and its rate of change are also taken into consideration. This also affects the concentration. Therefore, we can assume that there exists a function f(S(t), P(t)) to describe the effect of transport velocity and pressure on concentration, and two other coefficients b and c to quantify the degree of influence of flow velocity and its rate of change on concentration change.

[0073] Determining the influence coefficients b and c in a mathematical model typically involves multiple steps, including experimental design, data collection, selection of parameter estimation methods, and model validation. The detailed process is as follows:

[0074] In order to accurately capture the liquid flow velocity V(t) and its rate of change To determine the effect on the concentration C(t) of the target substance, a series of carefully designed experimental conditions are needed to ensure that all possible combinations of variables are covered. For example, the sample transfer speed S(t) and the equipment pressure P(t) can be varied, while recording the liquid flow rate and its changes under different conditions.

[0075] Under each set condition, high-precision sensors are used to monitor and record in real time the concentration C(t) of the target substance, the sample transfer speed S(t), the equipment pressure P(t), the liquid flow velocity V(t), and the rate of change of the flow velocity. Key parameters, etc.

[0076] Based on the obtained dataset, appropriate statistical or machine learning methods are used to estimate the unknown parameters b and c in the model. Commonly used methods include, but are not limited to, the following:

[0077] If we assume that b and c have a linear relationship, we can directly apply the least squares method or other forms of linear regression algorithms to solve for these two parameters by fitting known data points.

[0078] When there is a complex (e.g., nonlinear) relationship between b and c, nonlinear optimization techniques such as gradient descent and genetic algorithms can be used to find the optimal solution.

[0079] For more complex systems, advanced machine learning models such as neural networks and support vector machines can be used for parameter estimation. These models can automatically learn the mapping relationship between input features and output during training and adjust their internal parameters accordingly to minimize prediction error.

[0080] The collected data is divided into training and test sets. The model is trained using the training set and the values ​​of b and c are calculated. The model's performance is then evaluated on the test set to ensure that it has good generalization ability.

[0081] Examining the model's sensitivity to parameters b and c—that is, whether small changes to these parameter values ​​lead to significantly different output results—helps in understanding the model's stability and reliability.

[0082] Based on the results of the above verification process, if necessary, the original assumptions or model structure may be adjusted appropriately until the parameter configuration that best reflects the actual situation is found.

[0083] Once the optimal estimates of b and c are determined, they can be applied to real-time data analysis to help infer changes in sample characteristics and identify potential anomalies or pathogen types. As more new data accumulates, parameters b and c should be periodically re-evaluated and updated to ensure the model remains optimal and adapts to evolving real-world application scenarios.

[0084] Furthermore, for f(S(t),P(t)), first define the range of variation of S(t) and P(t), which should cover all possible situations in actual operation.

[0085] Using statistical experimental design methods, such as full factorial design, partial factorial design, or response surface methodology, a series of experimental condition combinations are planned to systematically explore the effects of S(t) and P(t) on C(t).

[0086] In each experiment, try to keep other conditions consistent except for S(t) and P(t), such as environmental factors like temperature and humidity, and liquid flow velocity V(t), in order to reduce the impact of external interference on the results.

[0087] The values ​​of C(t), S(t), and P(t) are recorded in real time using precise sensors to ensure data quality and accuracy. Each experimental condition is repeated at least three times to calculate the mean and standard deviation, thus improving data reliability.

[0088] Based on theoretical knowledge and preliminary data analysis, an appropriate mathematical expression is chosen to represent f(S(t), P(t)). For example, if a linear relationship is assumed, f(S(t), P(t)) = a1 S(t) + a2 P(t); if a nonlinear relationship exists, f(S(t), P(t)) is determined through fitting. Using methods such as least squares, maximum likelihood estimation, or machine learning algorithms, the parameters of the selected model are estimated based on the experimental dataset to obtain the specific values ​​of a1 and a2.

[0089] A portion of the data is used as the training set for model training, and another portion is used as the validation set for evaluating model performance. Techniques such as cross-validation are used to check the good agreement between the model's predictions and actual observations. A new dataset, independent of the modeling process, is then used to further test the model's generalization ability and stability.

[0090] Analyze the model's sensitivity to different input parameters to identify which parameters have a significant impact on the output, thus guiding subsequent research priorities. Perform statistical tests on key assumptions in the model, such as linearity and additive effects, and consider introducing interaction terms or other advanced features when necessary. Based on the problems discovered during validation, continuously adjust the model structure and parameters until a form that best reflects reality is found.

[0091] Once the model has been fully validated and deemed reliable, it can be applied to real-world production and monitoring environments. As more new data accumulates, the model's performance should be reassessed periodically, and parameter configurations updated as needed to ensure its long-term effectiveness.

[0092] S140. Analyze the frequency domain corresponding to the time series signal of the initial data, and determine the high-frequency and low-frequency components in the frequency domain by combining the set threshold.

[0093] In this embodiment, high-frequency and low-frequency components are identified based on the data in the frequency domain and the initial data, combined with a set threshold, and the weighting coefficients corresponding to the high-frequency and low-frequency components are determined.

[0094] In this embodiment, high-frequency components refer to components with higher frequencies, which typically correspond to rapidly changing or oscillating portions of the data. In intelligent point-of-care testing (POCT) methods, high-frequency components may represent transient fluctuations, short-term changes, or rapid reaction events in the sample, such as instantaneous concentration changes due to biochemical reactions, turbulence effects in sample flow, or other forms of rapid disturbances. For medical diagnostics, high-frequency components can provide information about the presence and activity level of pathogens, or reveal the presence of certain acute pathological conditions.

[0095] Low-frequency components refer to those with lower frequencies, typically corresponding to slower changes or stable states in the data. In intelligent point-of-care testing (POCT) methods, low-frequency components may reflect long-term trends, gradual changes, or stable baseline levels of sample characteristics, such as a continuous increase or decrease in the concentration of the target analyte, or a gradual change in sample flow rate. Low-frequency components help in understanding the overall behavior and contextual information of the sample, can be used to assess the progression of chronic diseases or monitor treatment effectiveness, and can serve as one of the foundations for building mathematical models.

[0096] By identifying and distinguishing between high-frequency and low-frequency components, this method can analyze sample feature variations and their underlying biological mechanisms in greater detail. For example, after preprocessing the initial data and converting it into a frequency domain signal, a set threshold is used to determine which components are high-frequency and which are low-frequency. Then, a pre-constructed mathematical model is used to calculate the changes in sample features based on these components, thereby inferring the anomalies or pathogen types present in the sample.

[0097] In short, high-frequency and low-frequency components each carry information about different aspects of the sample, working together to achieve more accurate diagnostic results. This method not only improves detection accuracy but also enhances the system's robustness and adaptability, providing strong support for real-time detection.

[0098] In one embodiment, step S140 described above may include steps S141 to S142.

[0099] S141. Perform spectral analysis on the frequency domain data, and identify high-frequency and low-frequency components by combining sample characteristic changes, target analyte concentrations, and set thresholds.

[0100] In this embodiment, a spectrum is generated based on frequency domain data, displaying the amplitude at different frequencies. Characteristic frequency components associated with the target analyte are identified from the spectrum. For example, peaks at certain frequencies may correspond to vibrational modes or reaction rates specific to the target substance. The spectral differences between known samples (positive controls, negative controls) and unknown samples are compared to determine which frequency component changes are associated with the target analyte concentration. If sample characteristics change over time, such as due to reaction progress or environmental factors, the spectral analysis model needs to be dynamically updated to adapt to these changes.

[0101] Based on laboratory validation experiments or historical data, one or more thresholds are set for each characteristic frequency to distinguish between fluctuations within the normal range and abnormal situations. When a frequency component exceeds the preset threshold, an alarm is triggered or the presence of a specific concentration of the target analyte is confirmed.

[0102] Based on the spectral analysis results, frequency components are divided into high-frequency components (usually associated with rapidly changing processes) and low-frequency components (more inclined to reflect slowly changing processes). The importance of these two types of information may vary depending on the application scenario.

[0103] S142. Calculate the proportions of high-frequency components and low-frequency components in the total to obtain the weighting coefficients corresponding to the high-frequency components and the low-frequency components.

[0104] To quantify the importance of high-frequency and low-frequency components, their respective proportions in the overall signal are calculated. These proportions, used as weighting coefficients (w1, w2), represent the degree to which high-frequency and low-frequency components contribute to the total signal. These weighting coefficients help understand the impact of different frequency components on the overall signal, thus better interpreting and predicting the relationship between sample characteristic variations and target analyte concentrations.

[0105] By distinguishing and weighting high-frequency and low-frequency components, the subtle correlation between changes in sample characteristics and the concentration of target analytes can be captured more accurately, thereby improving the accuracy of diagnostic results. In practical applications, external factors such as temperature fluctuations and equipment noise may introduce additional frequency components. By setting appropriate thresholds and accurately identifying high-frequency and low-frequency components, these interferences can be effectively filtered out, ensuring the stability of the detection system. Understanding the respective roles of high-frequency and low-frequency components allows for adjustments to detection strategies based on specific circumstances, such as increasing focus on specific frequency ranges or reducing unnecessary monitoring items, thereby saving energy and time. Based on the characteristics and weights of high-frequency and low-frequency components, a more reliable basis can be provided for inferring abnormalities or pathogen types in samples, contributing to early warning and precision medicine.

[0106] S150. Using the mathematical model, calculate the changes in sample characteristics based on the high-frequency and low-frequency components, and infer the abnormalities or pathogen types present in the sample to obtain the inference results.

[0107] In this embodiment, the inference result refers to whether there are any abnormalities in the sample. If there are abnormalities, the possible pathogen types will also be given.

[0108] In one embodiment, step S150 described above may include steps S151 to S153.

[0109] S151. Calculate the sample fluctuation based on the frequency data corresponding to the high-frequency and low-frequency components.

[0110] In this embodiment, the sample fluctuation includes information derived from changes in both high-frequency and low-frequency component signals.

[0111] In this step, the main focus is on the information derived from signal variations within different frequency ranges in the sample:

[0112] High-frequency components: This part reflects rapidly changing characteristics in the sample, often associated with acute infection or rapidly multiplying pathogens. For example, in a POCT testing environment, this could be a sudden increase in the number of viral particles or a sharp change in bacterial metabolic activity. To quantify these rapidly changing characteristics, the root mean square (RMS) value corresponding to the high-frequency components is calculated. The RMS value is a statistical measure that effectively represents the average level of signal intensity or amplitude. By calculating the RMS of the high-frequency components, the presence and intensity of acute or sudden events in the sample can be assessed.

[0113] Low-frequency components: In contrast, low-frequency components tend to reflect chronic changes or long-term effects in the sample, such as chronic inflammation or persistent pathogen infection. To capture trends in these long-term changes, we calculate the cumulative value of the low-frequency components. The cumulative value can be obtained by summing all measurement points over a specific time period, thus depicting the overall trend or state change of the sample over time.

[0114] S152. Calculate the overall sample irregularity based on the sample fluctuation and the weighting coefficients corresponding to the high-frequency and low-frequency components to obtain the sample irregularity.

[0115] In this embodiment, the comprehensive sample irregularity refers to an indicator that integrates information from high-frequency and low-frequency components to generate an indicator that can reflect the overall health or pathological state of the sample.

[0116] Specifically, a weighted summation method is used, where features within each frequency range are assigned different weights. These weights are chosen based on their diagnostic importance and how best they reflect the behavioral patterns of potential pathogens. For example, if the acute phase of a pathogen is particularly critical for diagnosis, high-frequency components are given higher weights; conversely, for chronic diseases, low-frequency components are given greater weights. The resulting comprehensive sample irregularity is a numerical indicator used to determine whether the sample deviates from a normal healthy state.

[0117] S153. Identify the abnormal situation or pathogen type encountered by the sample based on the sample fluctuation and the irregularity of the sample, and classify the specific characteristics into a specific pathological state to obtain the inference result.

[0118] In one embodiment, step S153 described above may include steps S1531 to S1533.

[0119] S1531. Obtain standard sample data under the same conditions as the current sample but without the influence of the tested substances as a reference.

[0120] In this embodiment, a set of standard sample data under the same conditions as the current test sample is obtained. This data should be unaffected by any test substances to ensure it represents a "normal" state. For example, in medical testing, blood, urine, or other bodily fluid samples from healthy individuals are selected as references. The selection of standard samples should match the environmental conditions of the test sample (such as temperature, humidity, collection time, etc.) as closely as possible to minimize the influence of external variables on the results. These standard samples will be used as the basis for subsequent comparisons.

[0121] S1532. Calculate the specific changes between the preprocessed data and the standard sample data to obtain the change vector.

[0122] In this embodiment, after obtaining preprocessed test sample data and standard sample data, the next step is to calculate the difference between the two, i.e., the change value. Specifically, for each feature dimension i, the sample feature value X is calculated. test (i) and the corresponding standard sample feature value X reference The difference ΔX between (i) i =X test (i)-X reference (i) thus forming a change vector ΔX that contains all the differences in characteristics. This change vector directly reflects the degree of deviation of the test sample from the standard sample, providing important quantitative basis for subsequent analysis.

[0123] S1533. Input the change vector, the sample fluctuation, and the sample irregularity quantification into the analysis model to determine the abnormal situation or pathogen type encountered by the sample, and classify the specific characteristics into a specific pathological state.

[0124] The previously obtained change vector and sample fluctuation σ test The irregularity metric Complexity is integrated into a single feature vector F. input =(ΔX, σ) testThe model evaluates the input feature vector (complexity) to determine if there are any abnormalities or specific types of pathogen infections, and categorizes the specific features into known pathological states. Specifically, if a sample is detected as abnormal, the model will classify it into different pathogen types (such as bacteria, viruses, fungi, etc.) or specific pathological states (such as acute diseases, chronic diseases, etc.) based on feature matching; alternatively, it can infer the possible abnormal type or pathogen in the sample based on the model's output and known pathological states. For example, a large difference in value and an abnormally high fluctuation frequency may indicate a viral infection; large changes in feature values ​​without obvious periodicity may indicate an acute disease.

[0125] The analysis model is trained using the following method:

[0126] First, a large labeled dataset needs to be collected, which should include samples from various known pathological states or pathogen types. Each sample must undergo detailed feature extraction and preprocessing to ensure it meets the requirements of the model input. Furthermore, the true label of each sample (i.e., its actual category) needs to be recorded for supervised learning.

[0127] Before actual training, the feature vectors need to be further optimized. This may involve feature selection (picking the features that best distinguish different categories), feature combination (creating new features to enhance the model's performance), or applying dimensionality reduction techniques (such as principal component analysis, PCA) to simplify the problem space without losing important information.

[0128] Depending on the nature of the problem and the available data types, a suitable machine learning algorithm or deep learning architecture can be selected. Common choices include, but are not limited to, Support Vector Machines (SVM), Random Forests (RF), K-Nearest Neighbors (KNN), Neural Networks (NN), and deep learning models such as Convolutional Neural Networks (CNN) or Recurrent Neural Networks (RNN). Each model has its unique advantages and disadvantages, and is suitable for different types of data and tasks.

[0129] The selected model is trained using a labeled dataset. During this process, the model continuously adjusts its internal parameters to minimize the error between the predicted output and the true label. Cross-validation is typically used to evaluate model performance and prevent overfitting. After training, the model should be able to make accurate predictions on new, unseen data.

[0130] After training, perform a final evaluation of the model using an independent test set to check its generalization ability. If the model performs poorly, you can go back to the previous steps to make adjustments, such as changing the model structure, adding more data, or improving feature engineering, until satisfactory performance is achieved.

[0131] In this embodiment, the inference results may include not only whether the sample is abnormal, the possible pathogen type or pathological state (e.g., "suspected viral infection", "bacterial infection", "no abnormality", etc.), but also clinical recommendations, such as: if a pathogen or abnormality is detected, further examination (e.g., PCR testing, blood culture, etc.) is recommended. If no abnormality is detected, continued observation or regular health checkups are recommended.

[0132] By simultaneously analyzing high-frequency and low-frequency components, the system can detect not only acute events (such as sudden outbreaks of infection) but also chronic changes (such as chronic inflammation or infection). This multi-dimensional analysis improves the accuracy and reliability of detection. Sensitive monitoring of high-frequency components allows the system to capture changes in signals in the early stages of disease, enabling early warning, which is crucial for rapid response to acute illnesses. Taking into account each individual's historical health record and personal differences, this approach provides personalized health management plans for each patient, rather than a one-size-fits-all approach.

[0133] Optimizing resource allocation: Early prediction of disease progression helps medical institutions better plan resources, rationally arrange examinations and treatments, and improve the efficiency of medical services. By comparing standard sample data, anomalies in actual samples can be more clearly identified, increasing the reliability of diagnosis and helping to rule out the possibility of misdiagnosis. The entire process relies heavily on data analysis and mathematical models, reducing the need for human intervention, improving the speed and consistency of diagnosis, and also reducing costs.

[0134] In addition, features can be extracted from the preprocessed data; historical data of the sample can be retrieved, and the sample fluctuation, sample irregularity and features corresponding to the historical data can be extracted to obtain historical information; the historical information, the features, the sample fluctuation and the sample irregularity can be input into the development trend prediction model to predict the development trend to obtain the prediction result; the prediction result can be matched with the known pathological state model stored in the database to obtain the corresponding specific pathological state and form the inference result.

[0135] Specifically, when analyzing the preprocessed data, the first step is to extract meaningful features from it. These features can be statistics directly calculated from the raw data (such as mean and variance), or new features obtained through complex algorithmic transformations (such as obtaining frequency domain features through Fourier transform, or capturing multi-scale characteristics using wavelet transform). The goal of feature extraction is to transform the raw data into a set of numerical values ​​or vectors that best describe the properties of the sample, enabling more effective analysis in subsequent steps. Next, the system retrieves historical data for the current sample. The importance of this step lies in its ability to allow us to understand the currently observed data within a time series context, rather than simply viewing a single measurement result in isolation. By analyzing historical data, we can extract trends and patterns in the sample's changes over time, such as its volatility (i.e., the magnitude of data variation) and irregularity (a measure of the randomness and complexity of the data distribution). This long-term perspective helps identify subtle changes or trends that might be overlooked in a single measurement. Once we have the features of the current sample and its historical information, the next step is to input all this information into a specially designed trend prediction model. The task of this model is to predict future data points, that is, to estimate what might happen if the current trend continues. To achieve this, predictive models are typically built based on machine learning or deep learning techniques, trained on a large amount of similar data to learn the development patterns of different types of samples. The inputs to the model include not only the latest features and current fluctuations and irregularities, but also the previously mentioned historical information to ensure the prediction is as accurate as possible. Finally, based on the output of the predictive model, the system searches the database for the closest known pathological state model. This is a matching process aimed at finding the pathological pattern most similar to the predicted result, thereby inferring the specific pathological state corresponding to the current sample. The key to this step is having a comprehensive and accurately labeled database of pathological states, containing typical data representations of various disease states. Once the best match is found, the system will generate an inference result, informing the user which pathological state the current sample is most likely to belong to, and providing corresponding explanations and supporting information.

[0136] By combining real-time and historical data for comprehensive analysis, early signs or trends of disease can be captured more accurately, helping doctors make more precise diagnoses. Using trend prediction models, potential health risks can be warned before symptoms become apparent, making preventative treatment possible. Considering each individual's historical health record, this approach can provide personalized health management plans for each patient, rather than a one-size-fits-all approach.

[0137] S160. The inference result is transmitted using 5G technology.

[0138] In this embodiment, using 5G technology to send inference results can greatly improve the efficiency, response speed, and accessibility of medical services. With its high bandwidth, low latency, and large connection capacity, 5G networks are ideally suited for real-time transmission of large amounts of data, including but not limited to high-definition medical images, video streams, and complex analysis results. Before transmission, the inference results need to be encrypted before being sent to the display terminal.

[0139] The aforementioned intelligent POCT detection method based on 5G and microfluidics technology improves the real-time performance and automation of the detection process by acquiring sample data in real time through microfluidics and integrated sensors. Initial data undergoes preprocessing and transformation to ensure data quality and consistency, providing an accurate foundation for subsequent analysis. By establishing a mathematical model, changes in sample characteristics are correlated with the concentration of target analytes, enhancing the accuracy and reliability of the detection results. Frequency domain analysis of the time-series signal, combined with high- and low-frequency component analysis, identifies sample changes, effectively detecting potential anomalies or pathogens. The integration of 5G technology enables rapid data transmission and real-time inference, allowing detection results to be promptly delivered to relevant personnel, enhancing overall detection performance and response speed.

[0140] Figure 2 This is a schematic block diagram of an intelligent POCT detection system 300 based on 5G and microfluidic technology provided in an embodiment of the present invention. Figure 2 As shown, corresponding to the above-described intelligent POCT detection method based on 5G and microfluidics technology, the present invention also provides an intelligent POCT detection system 300 based on 5G and microfluidics technology. This intelligent POCT detection system 300 includes units for executing the above-described intelligent POCT detection method based on 5G and microfluidics technology, and the system can be configured in a server. Specifically, please refer to... Figure 2 The intelligent POCT detection system 300 based on 5G and microfluidics technology includes an initial data acquisition unit 301, a preprocessing unit 302, a mathematical model construction unit 303, an analysis unit 304, an inference unit 305, and a transmission unit 306.

[0141] The system comprises: an initial data acquisition unit 301, used to acquire initial data of samples collected in real time using microfluidic technology combined with integrated sensors; a preprocessing unit 302, used to preprocess the initial data and transform the preprocessed data to obtain transformed data; a mathematical model construction unit 303, used to construct a mathematical model between sample characteristic changes and target analyte concentration based on the transformed data; an analysis unit 304, used to analyze the frequency domain corresponding to the time series signal of the initial data and determine the high-frequency and low-frequency components in the frequency domain by combining a set threshold; an inference unit 305, used to calculate the changes in sample characteristics based on the high-frequency and low-frequency components using the mathematical model and infer the abnormalities or pathogen types present in the sample to obtain inference results; and a transmission unit 306, used to transmit the inference results using 5G technology.

[0142] In one embodiment, the initial data acquisition unit 301 is configured to:

[0143] The sample is introduced into the microfluidic system, and relevant data are collected in real time using integrated sensors to obtain initial data; wherein, the initial data includes the concentration information of the target substance, the flow rate of the sample liquid, and the operating environment parameters of the device.

[0144] In one embodiment, the preprocessing unit 302 is configured to:

[0145] The concentration information of the target substance and the flow rate of the sample liquid are preprocessed according to the equipment environmental operating parameters to obtain preprocessed data; the time domain signal corresponding to the preprocessed data is converted into a frequency domain signal using a Fourier transform algorithm to obtain the converted data.

[0146] In one embodiment, the analysis unit 304 is configured to:

[0147] The process involves identifying high-frequency and low-frequency components based on the frequency domain data and the initial data, combined with a set threshold, and determining the weighting coefficients corresponding to the high-frequency and low-frequency components. This includes: performing spectral analysis on the frequency domain data, combining sample characteristic changes, target analyte concentration, and a set threshold to identify high-frequency and low-frequency components; and calculating the proportions of the high-frequency and low-frequency components in the total to obtain the weighting coefficients corresponding to the high-frequency and low-frequency components, respectively.

[0148] In one embodiment, the inference unit 305 is used to calculate the sample fluctuation based on the frequency data corresponding to the high-frequency and low-frequency components; calculate the comprehensive sample irregularity based on the sample fluctuation and the weight coefficients corresponding to the high-frequency and low-frequency components to obtain the sample irregularity; identify the abnormal situation or pathogen type encountered by the sample based on the sample fluctuation and the sample irregularity, and classify the specific characteristics into a specific pathological state to obtain the inference result.

[0149] In one embodiment, the inference unit 305 is further configured to:

[0150] Obtain standard sample data under the same conditions as the current sample but without the influence of tested substances as a reference; calculate specific change values ​​between the preprocessed data and the standard sample data to obtain a change vector; input the change vector, the sample fluctuation, and the sample irregularity quantification into the analysis model to determine the abnormal situation or pathogen type encountered by the sample, and classify the specific characteristics into a specific pathological state.

[0151] It should be noted that those skilled in the art can clearly understand that the specific implementation process of the above-mentioned intelligent POCT detection system 300 based on 5G and microfluidic technology and its various units can be referred to the corresponding descriptions in the foregoing method embodiments. For the sake of convenience and brevity, these details will not be repeated here.

[0152] The aforementioned intelligent POCT detection system 300 based on 5G and microfluidics technology can be implemented as a computer program, which can be used in, for example... Figure 3 It runs on the computer device shown.

[0153] Please see Figure 3 , Figure 3 This is a schematic block diagram of a computer device provided in an embodiment of this application. The computer device can be a server, wherein the server can be a standalone server or a server cluster composed of multiple servers.

[0154] See Figure 3 The computer device 500 includes a processor 502, a memory, and a network interface 505 connected via a system bus 501. The memory may include a non-volatile storage medium 503 and internal memory 504.

[0155] The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032 includes program instructions that, when executed, cause the processor 502 to perform an intelligent POCT detection method based on 5G and microfluidic technology.

[0156] The processor 502 provides computing and control capabilities to support the operation of the entire computer device 500.

[0157] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 can execute an intelligent POCT detection method based on 5G and microfluidic technology.

[0158] This network interface 505 is used for network communication with other devices. Those skilled in the art will understand that... Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device 500 to which the present application is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0159] The processor 502 is used to run a computer program 5032 stored in the memory to perform the following steps:

[0160] Microfluidic technology combined with integrated sensors is used to acquire initial data of samples collected in real time; the initial data is preprocessed and transformed to obtain transformed data; a mathematical model is constructed based on the transformed data to relate sample characteristic changes to target analyte concentrations; the frequency domain corresponding to the time series signal of the initial data is analyzed, and high-frequency and low-frequency components in the frequency domain are determined by combining a set threshold; the mathematical model is used to calculate the changes in sample characteristics based on the high-frequency and low-frequency components, and the abnormalities or pathogen types present in the sample are inferred to obtain the inference results; the inference results are transmitted using 5G technology.

[0161] It should be understood that in the embodiments of this application, the processor 502 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0162] It will be understood by those skilled in the art that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program includes program instructions and can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the embodiments of the above methods.

[0163] Therefore, the present invention also provides a storage medium. This storage medium can be a computer-readable storage medium. The storage medium stores a computer program, wherein when executed by a processor, the computer program causes the processor to perform the following steps:

[0164] Microfluidic technology combined with integrated sensors is used to acquire initial data of samples collected in real time; the initial data is preprocessed and transformed to obtain transformed data; a mathematical model is constructed based on the transformed data to relate sample characteristic changes to target analyte concentrations; the frequency domain corresponding to the time series signal of the initial data is analyzed, and high-frequency and low-frequency components in the frequency domain are determined by combining a set threshold; the mathematical model is used to calculate the changes in sample characteristics based on the high-frequency and low-frequency components, and the abnormalities or pathogen types present in the sample are inferred to obtain the inference results; the inference results are transmitted using 5G technology.

[0165] The storage medium can be any computer-readable storage medium capable of storing program code, such as a USB flash drive, portable hard drive, read-only memory (ROM), magnetic disk, or optical disk.

[0166] It should be noted that the functions or steps that the storage medium or computer device can achieve are described in the relevant descriptions in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0167] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0168] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the division of each unit is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0169] The steps in the method of this invention can be adjusted, merged, or reduced in order according to actual needs. The units in the system of this invention can be merged, divided, or reduced according to actual needs. Furthermore, the functional units in the various embodiments of this invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0170] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a terminal, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0171] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An intelligent POCT detection method based on 5G and microfluidic technology, characterized in that, The application relates to a method for analyzing a sample, comprising: acquiring initial data of a sample collected in real time by using microfluidic technology in combination with an integrated sensor; preprocessing the initial data and converting the preprocessed data to obtain converted data; constructing a mathematical model between sample feature variation and target analyte concentration according to the converted data; analyzing a frequency domain corresponding to a time series signal of the initial data and determining high-frequency components and low-frequency components in the frequency domain in combination with a set threshold value; calculating sample feature variation according to the high-frequency components and the low-frequency components by using the mathematical model, and inferring abnormal conditions or pathogen types existing in the sample to obtain an inference result; sending the inference result by using 5G technology; the analysis of the frequency domain corresponding to the time series signal of the initial data and the determination of the high-frequency components and the low-frequency components in the frequency domain in combination with the set threshold value, comprising: identifying the high-frequency components and the low-frequency components according to the data of the frequency domain and the initial data in combination with the set threshold value, and determining weight coefficients corresponding to the high-frequency components and the low-frequency components; the identification of the high-frequency components and the low-frequency components according to the data of the frequency domain and the initial data in combination with the set threshold value, and the determination of the weight coefficients corresponding to the high-frequency components and the low-frequency components, comprising: performing frequency spectrum analysis on the data of the frequency domain, identifying the high-frequency components and the low-frequency components in combination with sample feature variation, target analyte concentration and the set threshold value, and calculating proportions of the high-frequency components and the low-frequency components in the total to obtain the weight coefficient corresponding to the high-frequency components and the weight coefficient corresponding to the low-frequency components. the acquisition of the initial data of the sample collected in real time by using microfluidic technology in combination with an integrated sensor, comprising: 2.The 5G and microfluidic technology-based intelligent POCT detection method according to claim 1, characterized in that, introducing the sample into a microfluidic system and collecting relevant data in real time by using an integrated sensor to obtain initial data; wherein the initial data comprises concentration information of a target substance, sample liquid flow speed and device operation environment parameters. the preprocessing of the initial data and the conversion of the preprocessed data to obtain converted data, comprising: 3.The 5G and microfluidic technology-based intelligent POCT detection method of claim 1, wherein, preprocessing the concentration information of the target substance and the sample liquid flow speed according to device environment operation parameters to obtain preprocessed data; using a Fourier transform algorithm to convert time domain signals corresponding to the preprocessed data into frequency domain signals to obtain converted data. the calculation of sample feature variation according to the high-frequency components and the low-frequency components by using the mathematical model, and the inference of abnormal conditions or pathogen types existing in the sample to obtain an inference result, comprising: 4.The 5G and microfluidic technology-based intelligent POCT detection method of claim 1, wherein, calculating sample fluctuation according to frequency data corresponding to the high-frequency components and the low-frequency components; calculating comprehensive sample irregularity according to the sample fluctuation and weight coefficients corresponding to the high-frequency components and the low-frequency components to obtain sample irregularity; identifying abnormal conditions or pathogen types encountered by the sample according to the sample fluctuation and the sample irregularity, and classifying specific features into specific pathological states to obtain an inference result. ​ 5.The 5G and microfluidic technology-based smart POCT detection method of claim 4, wherein, The abnormal situation or pathogen type encountered by the sample is identified according to the sample fluctuation and the irregularity of the sample, and specific features are classified into specific pathological states to obtain an inference result, including: Obtaining standard sample data under the same conditions as the current sample but without the influence of the tested substance as a reference; Calculating the specific change value of the pre-processed data and the standard sample data to obtain a change vector; The change vector, the sample fluctuation, and the irregularity of the sample are quantitatively input into an analysis model to determine the abnormal situation or pathogen type encountered by the sample, and specific features are classified into specific pathological states. 6.The 5G and microfluidic technology-based intelligent POCT detection method of claim 1, wherein, The mathematical model is where C(t) represents the concentration of the target substance, S(t) is the transport speed of the sample, P(t) is the device pressure, V(t) is the liquid flow speed, b and c are influence coefficients, respectively, for quantifying the influence of the flow speed and its rate of change on the concentration change; f(S(t), P(t)) represents in the mathematical model the influence of the transport speed of the sample and the device pressure on the concentration of the target substance.

7. An intelligent POCT detection system based on 5G and microfluidic technology, characterized in that, The system uses the intelligent POCT detection method based on 5G and microfluidic technology according to any one of claims 1 to 6, including: An initial data acquisition unit for acquiring initial data of a sample collected in real time using microfluidic technology combined with an integrated sensor; A preprocessing unit for preprocessing the initial data and converting the preprocessed data to obtain converted data; A mathematical model construction unit for constructing a mathematical model between sample feature changes and target analyte concentration according to the converted data; An analysis unit for analyzing the frequency domain corresponding to the time series signal of the initial data and determining the high-frequency component and the low-frequency component in the frequency domain in combination with a set threshold; An inference unit for calculating the change of sample features according to the high-frequency component and the low-frequency component using the mathematical model, and inferring the abnormal situation or pathogen type present in the sample to obtain an inference result; A sending unit for sending the inference result using 5G technology.

8. A computer device, comprising: The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the method according to any one of claims 1 to 6 when executing the computer program.

9. A storage medium, characterized by The storage medium stores a computer program, and the processor implements the method according to any one of claims 1 to 6 when executing the computer program.

Images