Early lung cancer screening method and system based on biomarker in-vitro detection

Abstract

The invention provides a lung cancer early screening method and system based on biomarker in-vitro detection, and relates to the technical field of biomarker research, and the lung cancer early screening method comprises the following steps: collecting a peripheral blood sample of a subject, identifying a target biomarker in the peripheral blood sample, and detecting the concentration information of the target biomarker; the target biomarker comprises a carcino-embryonic antigen, a cytokeratin 19 fragment and a squamous epithelial cell carcinoma antigen; obtaining an optical signal through a surface plasma resonance detection device; noise reduction processing is carried out on the optical signal to obtain core characteristic parameters, and the core characteristic parameters comprise a binding rate, a dissociation rate and a resonance angle deviation amplitude; utilizing an edge computing technology to obtain a corresponding risk score; through a cross comparison mode, obtaining verified risk data; the verified risk data and concentration information are integrated and analyzed to obtain an early screening result of the lung cancer, so that high-sensitivity and high-specificity early screening of the non-invasive lung cancer is realized.

Description

Technical Field

[0001] This application relates to the field of biomarker research technology, and in particular to a method and system for early lung cancer screening based on in vitro detection of biomarkers. Background Technology

[0002] In general health check-up scenarios, early screening for lung cancer urgently requires a high-throughput, non-invasive, accurate in vitro detection method that can be integrated into routine check-up procedures. With the continuous rise in lung cancer incidence, traditional imaging screening, while effective to some extent, suffers from problems such as radiation exposure, high costs, and high false-positive rates, making large-scale deployment difficult.

[0003] To address these needs, a multi-biomarker analysis system based on a microfluidic immunoassay chip combined with fluorescent labeling detection has emerged as a mainstream solution. This system immobilizes capture antibodies for multiple lung cancer biomarkers within a microfluidic channel, introduces fluorescently labeled detection antibodies to form a sandwich immune complex, quantitatively analyzes biomarker concentrations using a high-resolution fluorescence imaging system, and integrates multi-indicator data using a cloud-based machine learning model to output a personalized lung cancer risk assessment. However, existing solutions have certain limitations. For example, the fluorescent labeling process is susceptible to interference from endogenous fluorescent substances in the sample, leading to a decrease in the signal-to-noise ratio, and sensitivity is often insufficient, especially when detecting low-concentration biomarkers. Summary of the Invention

[0004] The purpose of this application is to provide a method and system for early lung cancer screening based on in vitro detection of biomarkers, in order to solve the problems in the existing technology such as susceptibility to interference from endogenous fluorescent substances in samples, insufficient sensitivity when detecting low concentrations of biomarkers, and network latency affecting screening efficiency.

[0005] Firstly, this application provides a method for early lung cancer screening based on in vitro detection of biomarkers, which has low screening efficiency, including:

[0006] Peripheral blood samples were collected from the subjects, target biomarkers in the peripheral blood samples were identified, and the concentration information of the target biomarkers was detected. The target biomarkers included carcinoembryonic antigen, cytokeratin 19 fragment, and squamous cell carcinoma antigen.

[0007] The optical properties of the target biomarker and the corresponding specific antibody during the antigen-antibody binding process are detected using a surface plasmon resonance detection device to obtain an optical signal.

[0008] The optical signal is denoised to obtain a denoised optical signal. Core feature parameters are extracted from the denoised optical signal. The core feature parameters include binding rate, dissociation rate and resonance angle shift amplitude.

[0009] Each core feature parameter is compared with a feature parameter database of a target population including lung cancer early-stage population and healthy population by using edge computing technology to obtain a corresponding risk score;

[0010] The target core feature parameter whose risk score meets a preset threshold is verified by cross comparison to obtain verified risk data;

[0011] The verified risk data and the concentration information are integrated and analyzed to obtain a lung cancer early-stage screening result.

[0012] Optionally, the optical signal is subjected to noise reduction processing to obtain a noise-reduced optical signal, and a core feature parameter is extracted from the noise-reduced optical signal, the core feature parameter including a binding rate, a dissociation rate, and a resonance angle shift amplitude, including:

[0013] According to a biological stage of the target biomarker and a corresponding specific antibody in an antigen-antibody binding process, the optical signal is divided into a plurality of signal segments, the biological stage including an initial binding stage, a rapid binding stage, an equilibrium stage, and a dissociation stage;

[0014] Abnormal fluctuation signals with signal intensity exceeding a preset signal intensity threshold in each signal segment are removed to obtain a plurality of interference-free signal segments;

[0015] All interference-free signal segments are spliced in the order of time intervals of biological stages to obtain spliced interference-free signals, and the spliced interference-free signals are subjected to smoothing processing to obtain a noise-reduced optical signal, and a resonance angle shift amplitude is calculated according to a signal change amount in the noise-reduced optical signal;

[0016] A binding rate is calculated based on signal segments corresponding to the initial binding stage and the rapid binding stage in the noise-reduced optical signal, and a dissociation rate is calculated based on a signal segment corresponding to the dissociation stage in the noise-reduced optical signal;

[0017] The resonance angle shift amplitude, the binding rate, and the dissociation rate are integrated to obtain a core feature parameter.

[0018] In a second aspect, the present application provides a lung cancer early-stage screening system based on biomarker in vitro detection, including:

[0019] The collection module is configured to collect a peripheral blood sample of a subject, identify a target biomarker in the peripheral blood sample, and detect concentration information of the target biomarker, the target biomarker including carcinoembryonic antigen, cytokeratin 19 fragment, and squamous cell carcinoma antigen;

[0020] The detection module is used to detect the changes in optical properties of the target biomarker and the corresponding specific antibody during the antigen-antibody binding process using a surface plasmon resonance detection device, so as to obtain an optical signal;

[0021] The noise reduction module is used to perform noise reduction processing on the optical signal to obtain a noise-reduced optical signal, and to extract core feature parameters from the noise-reduced optical signal. The core feature parameters include binding rate, dissociation rate and resonance angle offset amplitude.

[0022] The comparison module is used to compare each core feature parameter with the feature parameter database of the target population using edge computing technology to obtain the corresponding risk score. The target population includes people with early-stage lung cancer and healthy people.

[0023] The verification module is used to verify the target core feature parameters that meet the preset threshold through cross-comparison, and obtain the verified risk data.

[0024] The analysis module integrates and analyzes the verified risk data and the concentration information to obtain early lung cancer screening results.

[0025] Thirdly, this application provides an electronic device, comprising:

[0026] Memory, used to store computer programs;

[0027] A processor, configured to execute the computer program to implement the steps of a method for early lung cancer screening based on in vitro detection of biomarkers as described in the first aspect above.

[0028] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the steps of the lung cancer early screening method based on biomarker in vitro detection as described in the first aspect above.

[0029] This application provides a method for early lung cancer screening based on in vitro detection of biomarkers. The method involves collecting peripheral blood samples from subjects, identifying target biomarkers in the peripheral blood samples, and detecting the concentration information of the target biomarkers. The target biomarkers include carcinoembryonic antigen (CEA), cytokeratin 19 fragment, and squamous cell carcinoma antigen. A surface plasmon resonance (SPR) detection device is used to detect changes in the optical properties of the target biomarkers and their corresponding specific antibodies during antigen-antibody binding, thereby obtaining an optical signal. The optical signal is then denoised to obtain a denoised optical signal. Core feature parameters are extracted from the denoised optical signal, including binding rate, dissociation rate, and resonance angle shift amplitude.

[0030] Then, using edge computing technology, each core feature parameter is compared with the feature parameter database of the target population to obtain the corresponding risk score. The target population includes individuals with early-stage lung cancer and healthy individuals. Through cross-comparison, the target core feature parameters whose risk scores meet preset thresholds are verified to obtain verified risk data. The verified risk data and the concentration information are integrated and analyzed to obtain the early lung cancer screening results. By collecting peripheral blood samples from subjects and identifying target biomarkers, non-invasive and standardized sample acquisition can be achieved, providing a foundation for subsequent multi-marker joint analysis. The above processing method avoids background interference caused by fluorescent labeling, improves the signal-to-noise ratio of the detection process, and enhances the response capability to low-concentration biomarkers. Early lung cancer screening methods improve the reliability and robustness of screening results and reduce the probability of misdiagnosis; they provide a high-precision decision-making basis for lung cancer risk stratification in general physical examinations; furthermore, based on the biological stage of the target biomarker and corresponding specific antibody binding process, the original optical signal is divided into multiple signal segments; abnormal fluctuation signals with signal intensity exceeding a preset threshold in each segment are removed to obtain multiple interference-free signal segments; these interference-free signal segments are spliced ​​in chronological order and smoothed to form a noise-reduced optical signal; based on this signal, the resonance angle shift amplitude, binding rate, and dissociation rate are calculated and integrated into a complete set of core feature parameters; this solves the technical bottlenecks of existing schemes in terms of sensitivity, specificity, and anti-interference ability. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 A flowchart illustrating an early lung cancer screening method based on in vitro biomarker detection, provided for an embodiment of this application;

[0033] Figure 2 A schematic diagram illustrating a specific implementation of an early lung cancer screening method based on in vitro detection of biomarkers, provided in this application embodiment;

[0034] Figure 3 A case reference data table for an early lung cancer screening method based on in vitro detection of biomarkers provided in the embodiments of this application;

[0035] Figure 4This is a schematic diagram of the structure of an early lung cancer screening system based on in vitro detection of biomarkers, provided as an embodiment of this application. Detailed Implementation

[0036] In response to the urgent need for high-throughput, non-invasive, accurate, and integrable early lung cancer screening methods in general physical examination scenarios, existing multi-marker analysis systems based on microfluidic immunoassay chips and fluorescent labeling detection can achieve a certain degree of automation. However, they are still limited by problems such as low signal-to-noise ratio due to the susceptibility of fluorescent labels to interference from endogenous fluorescence in the sample, insufficient sensitivity for low-concentration marker detection, and response delays caused by reliance on cloud data processing. These issues make it difficult to meet the requirements of efficient, safe, and reliable screening in grassroots physical examinations.

[0037] To this end, this application collects peripheral blood samples and uses surface plasmon resonance technology to capture the changes in optical properties during the binding process of carcinoembryonic antigen, cytokeratin 19 fragment, and squamous cell carcinoma antigen with specific antibodies in real time. It obtains the core kinetic characteristic parameters, including binding rate, dissociation rate, and resonance angle shift amplitude. Locally, edge computing is used to achieve rapid comparison and risk scoring with the characteristic databases of early-stage lung cancer patients and healthy individuals. A cross-validation mechanism is introduced to verify high-risk results. Finally, the kinetic characteristics and concentration information are integrated to output a comprehensive screening conclusion. This approach avoids label interference, improves the sensitivity, specificity, and on-site applicability of the detection, and breaks through the limitations of existing technologies in anti-interference ability and real-time analysis.

[0038] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. 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.

[0039] Example 1

[0040] Embodiment 1 of this application provides a method for early lung cancer screening based on in vitro detection of biomarkers, and a flowchart of a specific implementation is shown below. Figure 1 As shown, the method includes:

[0041] Step 101: Collect peripheral blood samples from the subject, identify target biomarkers in the peripheral blood samples, and detect the concentration information of the target biomarkers. The target biomarkers include carcinoembryonic antigen, cytokeratin 19 fragment, and squamous cell carcinoma antigen.

[0042] In this step, peripheral blood sample refers to a blood sample collected from the subject; target biomarker refers to a substance identified from the peripheral blood sample that is associated with early lung cancer screening; concentration information refers to the content data of the target biomarker in the blood; carcinoembryonic antigen refers to a lung cancer-associated protein among the target biomarkers; cytokeratin 19 fragment refers to a lung cancer-associated cellular structure fragment among the target biomarkers; and squamous cell carcinoma antigen refers to an antigen among the target biomarkers associated with squamous cell-derived lung cancer.

[0043] In this embodiment of the application, a blood sample, i.e. a peripheral blood sample, is obtained from the subject using a collection tool. The peripheral blood sample is purified to remove interfering substances, and three substances, carcinoembryonic antigen, cytokeratin 19 fragment, and squamous cell carcinoma antigen, are identified as target biomarkers. At the same time, a detection tool is used to measure the content of each target biomarker to obtain the corresponding concentration information. This concentration information will be used for subsequent integration and analysis with the validated risk data.

[0044] Step 102: Detect the changes in optical properties of the target biomarker and the corresponding specific antibody during the antigen-antibody binding process using a surface plasmon resonance detection device to obtain an optical signal.

[0045] In this step, the surface plasmon resonance detection device refers to the device used to detect changes in optical properties during antigen-antibody binding; the corresponding specific antibody refers to the antibody that can specifically bind to each target biomarker; the antigen-antibody binding process refers to the process by which the target biomarker and the corresponding specific antibody recognize and bind to each other; the change in optical properties refers to the change in optical properties such as light intensity and resonance angle caused during the antigen-antibody binding process; and the optical signal refers to the signal data reflecting the changes in optical properties captured by the surface plasmon resonance detection device.

[0046] Step 103: Perform noise reduction processing on the optical signal to obtain a noise-reduced optical signal, and extract core feature parameters from the noise-reduced optical signal. The core feature parameters include the binding rate, dissociation rate, and resonance angle offset amplitude.

[0047] In this step, the denoised optical signal refers to the signal obtained after interference removal and smoothing of the optical signal; the core feature parameters refer to the key parameters extracted from the denoised optical signal that can reflect the characteristics of the target biomarker; the binding rate refers to the speed of the reaction during the binding phase of the target biomarker and the corresponding specific antibody; the dissociation rate refers to the speed of the reaction during the dissociation phase of the target biomarker and the corresponding specific antibody; and the resonance angle shift amplitude refers to the maximum difference in the resonance angle shift during the antigen-antibody binding process.

[0048] Step 104: Using edge computing technology, each core feature parameter is compared with the feature parameter database of the target population to obtain the corresponding risk score. The target population includes people with early-stage lung cancer and healthy people.

[0049] In this step, the target population refers to the two groups of people used to establish the feature parameter database; the feature parameter database refers to the data set that stores the reference range of core feature parameters in the target population; the risk score refers to the score reflecting the early risk of lung cancer obtained by comparing the core feature parameters with the feature parameter database; the early lung cancer population refers to people who have been diagnosed with early lung cancer; and the healthy population refers to people who have not had lung cancer and are in good health.

[0050] Step 105: Verify the target core feature parameters that meet the preset threshold through cross-comparison to obtain verified risk data.

[0051] In this step, the preset threshold refers to the score standard used to screen the target core feature parameters; the target core feature parameters refer to the core feature parameters whose risk scores are greater than the preset threshold; and the verified risk data refers to the data set containing risk scores and consistency information obtained after the target core feature parameters have been verified and cross-referenced.

[0052] Step 106: Integrate and analyze the verified risk data and the concentration information to obtain the early lung cancer screening results.

[0053] In this step, the early lung cancer screening result refers to the result of determining the early lung cancer risk level of the examinee by integrating and validating risk data and concentration information.

[0054] The embodiments of this application realize early screening of lung cancer, which solves the problems of excessive signal interference, inaccurate parameter extraction, risk judgment bias and low result reliability in the prior art, and improves the accuracy and effectiveness of screening.

[0055] This application provides a specific embodiment, such as Figure 2 As shown, step 103 involves denoising the optical signal to obtain a denoised optical signal, and extracting core feature parameters from the denoised optical signal. These core feature parameters include the binding rate, dissociation rate, and resonance angle shift amplitude. Specifically, this includes the following steps:

[0056] Step 301: Based on the biological stages of the target biomarker and the corresponding specific antibody in the antigen-antibody binding process, the optical signal is divided into multiple signal segments, the biological stages including the initial binding stage, the rapid binding stage, the equilibrium stage, and the dissociation stage.

[0057] In this step, the biological stage refers to the different stages that the target biomarker and the corresponding specific antibody go through sequentially during the antigen-antibody binding process; the signal fragment refers to the segmented signal obtained after breaking down the optical signal according to the biological stage; the initial binding stage refers to the stage where the target biomarker and the corresponding specific antibody begin to recognize each other and initially bind; the rapid binding stage refers to the stage where the target biomarker and the corresponding specific antibody bind rapidly in large quantities; the equilibrium stage refers to the stage where the amount of binding between the target biomarker and the corresponding specific antibody reaches a dynamic equilibrium; and the dissociation stage refers to the stage where the bound target biomarker and the corresponding specific antibody separate from each other.

[0058] In this embodiment, it is first clarified that the target biomarker and the corresponding specific antibody will sequentially go through four biological stages during the antigen-antibody binding process: the initial binding stage, the rapid binding stage, the equilibrium stage, and the dissociation stage. It is an important feature that the signal segments of the biomarker in different biological stages have important identification functions. Based on this, each biological stage corresponds to a specific time interval. According to the division criteria of these time intervals, the continuous optical signal is divided into four signal segments corresponding to different biological stages. Each signal segment contains only the optical signal data within the corresponding biological stage. These signal segments will be used as the objects for subsequent operations to remove abnormal fluctuation signals.

[0059] Step 302: Remove abnormal fluctuation signals in each signal segment whose signal strength exceeds the preset signal strength threshold to obtain multiple interference-free signal segments.

[0060] In this step, signal strength refers to the quantized value of the optical properties corresponding to each signal point in the signal segment; preset signal strength threshold refers to the numerical standard used to determine whether a signal point is an abnormal fluctuation signal; abnormal fluctuation signal refers to the set of signal points whose signal strength exceeds the preset signal strength threshold; interference-free signal segment refers to the signal segment remaining after removing abnormal fluctuation signals.

[0061] In this embodiment, the signal strength of all signal points in each signal segment is first obtained, and then a preset signal strength threshold is retrieved. The signal strength of each signal point is compared with the preset signal strength threshold one by one. Signal points whose signal strength exceeds the preset signal strength threshold are filtered out and defined as abnormal fluctuation signals. These abnormal fluctuation signals are all removed from the corresponding signal segments. The remaining signal part is the interference-free signal segment. All interference-free signal segments will be used for the next step of splicing.

[0062] Step 303: All interference-free signal segments are spliced ​​together in the order of biological stage time intervals to obtain spliced ​​interference-free signals. The spliced ​​interference-free signals are then smoothed to obtain noise-reduced optical signals. The resonance angle shift amplitude is calculated based on the signal change in the noise-reduced optical signals.

[0063] In this step, the signal change refers to the maximum difference in signal intensity in the denoised optical signal; the preset signal intensity threshold refers to the numerical standard used to determine whether a signal point is an abnormal fluctuation signal; the interference-free signal segment refers to the signal segment remaining after removing abnormal fluctuation signals; the denoised optical signal refers to the optical signal that has been spliced ​​and smoothed to remove interference; the resonance angle offset amplitude refers to the core parameter reflecting the degree of antigen-antibody binding, which is calculated from the signal change in the denoised optical signal.

[0064] In this embodiment, all interference-free signal segments are sequentially connected according to their corresponding biological stage time intervals to form a complete spliced ​​interference-free signal. Then, the spliced ​​interference-free signal is subjected to transition optimization to eliminate traces at the splicing points of different signal segments, resulting in a denoised optical signal. The maximum and minimum signal intensity values ​​in the denoised optical signal are extracted, and the signal change is obtained by calculating the difference between the two. The resonance angle shift amplitude is then calculated based on this signal change amplitude. The resonance angle shift amplitude = maximum signal intensity value of the denoised optical signal - minimum signal intensity value of the denoised optical signal. The obtained resonance angle shift amplitude, together with the subsequently calculated binding rate and dissociation rate, constitutes the core feature parameters.

[0065] Step 304: Calculate the binding rate based on the signal segments corresponding to the initial binding stage and the rapid binding stage in the denoised optical signal, and calculate the dissociation rate based on the signal segments corresponding to the dissociation stage in the denoised optical signal.

[0066] In this embodiment, signal segments corresponding to the initial bonding stage and the rapid bonding stage are extracted from the denoised optical signal. The signal intensity at each time point in these two signal segments is extracted, and the rate of change of signal intensity over time is calculated to obtain the bonding rate: bonding rate = (signal intensity at the end of the rapid bonding stage - signal intensity at the beginning of the initial bonding stage) ÷ (end time of the rapid bonding stage - start time of the initial bonding stage). Simultaneously, signal segments corresponding to the dissociation stage are extracted from the denoised optical signal, and the signal intensity at each time point in these signal segments is extracted. The rate of change of signal intensity over time is calculated to obtain the dissociation rate: dissociation rate = (signal intensity at the beginning of the dissociation stage - signal intensity at the end of the dissociation stage) ÷ (end time of the dissociation stage - start time of the dissociation stage). The obtained bonding rate and dissociation rate are integrated with the resonance angle offset amplitude.

[0067] Step 305: Integrate the resonance angle offset amplitude, the binding rate, and the dissociation rate to obtain the core feature parameters.

[0068] In this embodiment, the previously calculated resonance angle shift amplitude, binding rate, and dissociation rate are summarized and categorized to ensure that the value of each parameter is accurate and corresponds to a unique calculation source. The categorized and categorized three parameters are integrated into a complete set of key data that can reflect the binding characteristics of the target biomarker and the corresponding specific antibody, which is the core feature parameter. This core feature parameter will be used for subsequent comparison with the feature parameter database of the target population.

[0069] The embodiments of this application reduce interference components in optical signals, improve the accuracy and reliability of core feature parameter extraction, and provide solid data support for the accuracy of subsequent risk scoring and early lung cancer screening results.

[0070] This application provides a specific embodiment. Step 202 involves removing abnormal fluctuation signals in each signal segment whose signal strength exceeds a preset signal strength threshold to obtain multiple interference-free signal segments. This specifically includes the following steps:

[0071] Step 211: Integrate the signal points in each signal segment that are in the stable period corresponding to the biological stage to form the baseline signal segment of each signal segment.

[0072] In this step, a signal plateau period refers to a time period in which the signal intensity does not fluctuate significantly within each biological stage; a signal point refers to a single data point in the optical signal arranged in chronological order; and a baseline signal segment refers to a reference signal segment formed by integrating the signal points within the signal plateau period.

[0073] In this embodiment of the application, for each signal segment, the signal stability period in which the signal intensity does not fluctuate significantly within the corresponding biological stage is first determined. All signal points within the signal stability period are extracted from the signal segment, and these signal points are integrated into a continuous signal in chronological order, which is the baseline signal segment of the signal segment. Each signal segment corresponds to a baseline signal segment, and these baseline signal segments will be used for subsequent calculation of the signal intensity threshold.

[0074] Step 212: Calculate the signal strength values ​​of all signal points in each baseline signal segment. Based on the signal strength values, calculate the mean signal strength and standard deviation of each baseline signal segment to determine the signal strength threshold for each signal segment.

[0075] In this step, the signal strength value refers to the quantized signal strength data corresponding to each signal point; the signal strength mean refers to the average value of all signal strength values ​​in the baseline signal segment; the signal strength standard deviation refers to the value reflecting the dispersion of signal strength values ​​in the baseline signal segment; and the signal strength threshold refers to the numerical range determined based on the signal strength mean and signal strength standard deviation for judging whether a signal point is abnormal.

[0076] In this embodiment, for each baseline signal segment, the signal strength values ​​of all signal points are statistically analyzed. These signal strength values ​​are then summed and divided by the total number of signal points to obtain the mean signal strength: Mean signal strength = (Sum of signal strength values ​​of all signal points) ÷ Total number of signal points. Then, the standard deviation of the signal strength is calculated based on the difference between each signal strength value and the mean signal strength. The signal strength threshold is determined based on the mean signal strength and the standard deviation of signal strength. Where k is a preset coefficient used to adjust the threshold range. Each signal segment corresponds to a signal strength threshold, and these thresholds will be used to identify abnormal signal points.

[0077] Step 213: Remove abnormal signal points in each signal segment whose signal strength exceeds the corresponding signal strength threshold to obtain multiple abnormal signal segments. Based on the signal strength of adjacent signal points in each abnormal signal segment, calculate the filling value for each abnormal signal segment and add the filling value to the corresponding abnormal signal segment to obtain multiple interference-free signal segments.

[0078] In this step, an abnormal signal point refers to a signal point whose signal strength exceeds the corresponding signal strength threshold; the signal segment after anomaly removal refers to the remaining signal segment after removing the abnormal signal point; adjacent signal points refer to normal signal points before and after the abnormal signal point; the fill value refers to the calculated value used to fill the missing parts in the signal segment after anomaly removal; and the interference-free signal segment refers to the complete signal segment without abnormal signals after adding the fill value.

[0079] In this embodiment of the application, the signal points in each signal segment are compared with the signal strength threshold corresponding to the signal segment, and abnormal signal points whose signal strength exceeds the threshold are filtered out and removed from the signal segment to obtain an abnormal signal segment containing the missing part.

[0080] For each signal segment after anomaly removal, the signal strength values ​​of adjacent signal points before and after the missing part are extracted. The filling value is obtained by calculating the average signal strength of adjacent signal points. The filling value = (signal strength value of the previous adjacent signal point + signal strength value of the next adjacent signal point) ÷ 2. The filling value is added to the missing part of the signal segment after anomaly removal to restore the continuity of the signal segment and obtain interference-free signal segments. All interference-free signal segments will be used for subsequent splicing processing.

[0081] The embodiments of this application remove abnormal interference in the signal, ensuring the continuity and reliability of the signal, providing a high-quality signal basis for subsequent extraction of core feature parameters, and improving the accuracy of early lung cancer screening.

[0082] This application provides a specific embodiment. Step 104 involves using edge computing technology to compare each core feature parameter with the feature parameter database of the target population to obtain the corresponding risk score. This specifically includes the following steps:

[0083] Step 401: Using edge computing technology, retrieve the feature parameter database of the target population. The feature parameter database of the target population stores the first feature parameter range of the early lung cancer population and the second feature parameter range of the healthy population.

[0084] In this step, the first feature parameter range refers to the numerical range of the core feature parameters of the early-stage lung cancer population stored in the feature parameter database; the second feature parameter range refers to the numerical range of the core feature parameters of the healthy population stored in the feature parameter database.

[0085] In this embodiment of the application, edge computing technology is used to access a feature parameter database that stores core feature parameter data of the target population. Two types of data ranges are retrieved and extracted from the database: a first feature parameter range corresponding to early-stage lung cancer patients and a second feature parameter range corresponding to healthy patients. These two ranges will serve as reference standards for subsequent comparison of core feature parameters.

[0086] Step 402: Calculate the first binding matching degree between the binding rate in the core feature parameters and the binding rate reference range of the first feature parameter range. Based on the preset matching degree scoring conversion rule, convert the first binding matching degree into a first original risk score. Calculate the second binding matching degree between the binding rate in the core feature parameters and the binding rate reference range of the second feature parameter range. Based on the preset matching degree coefficient conversion rule, convert the second binding matching degree into a first correction coefficient. Combined with the first original risk score, calculate the first risk score of the binding rate.

[0087] In this step, the binding rate reference range refers to the numerical ranges of binding rates contained in the first feature parameter range and the second feature parameter range, respectively; the first binding matching degree refers to the degree of fit between the binding rate in the core feature parameter and the binding rate reference range of the first feature parameter range; the preset matching degree score conversion rule refers to the pre-set rule for converting the binding matching degree into a risk score; the first original risk score refers to the risk score obtained by converting the first binding matching degree through the preset matching degree score conversion rule; the second binding matching degree refers to the degree of fit between the binding rate in the core feature parameter and the binding rate reference range of the second feature parameter range; the preset matching degree coefficient conversion rule refers to the pre-set rule for converting the binding matching degree into a correction coefficient; the first correction coefficient refers to the correction coefficient obtained by converting the second binding matching degree through the preset matching degree coefficient conversion rule; and the first risk score refers to the risk score corresponding to the binding rate calculated by combining the first original risk score and the first correction coefficient.

[0088] In this embodiment, binding rate reference ranges are extracted from the first feature parameter range and the second feature parameter range, respectively. The binding rate in the core feature parameter is substituted into the calculation formula to obtain the first binding matching degree with the first binding rate reference range. Based on the preset matching score conversion rules, the first binding matching degree is converted into a first original risk score, where the first original risk score = first binding matching degree × 100; then, the second binding matching degree between the binding rate and the second binding rate reference range is calculated. According to the preset matching degree coefficient conversion rule, the second combination matching degree is converted into the first correction coefficient, the first correction coefficient = 1 - the second combination matching degree; finally, the first risk score of the combination rate is calculated, the first risk score = the first original risk score × the first correction coefficient.

[0089] Step 403: Calculate the first dissociation matching degree between the dissociation rate and the dissociation rate reference range of the first feature parameter range. Based on the preset matching degree scoring conversion rule, convert the first dissociation matching degree into a second original risk score. Calculate the second dissociation matching degree between the dissociation rate and the dissociation rate reference range of the second feature parameter range. Based on the preset matching degree coefficient conversion rule, convert the second dissociation matching degree into a second correction coefficient. Combined with the second original risk score, calculate the second risk score of the dissociation rate.

[0090] In this step, the dissociation rate reference range refers to the numerical ranges of dissociation rates contained in the first feature parameter range and the second feature parameter range, respectively; the first dissociation matching degree refers to the degree of fit between the dissociation rate in the core feature parameter and the dissociation rate reference range of the first feature parameter range; the second original risk score refers to the risk score obtained by converting the first dissociation matching degree through a preset matching degree score conversion rule; the second dissociation matching degree refers to the degree of fit between the dissociation rate in the core feature parameter and the dissociation rate reference range of the second feature parameter range; the second correction coefficient refers to the correction coefficient obtained by converting the second dissociation matching degree through a preset matching degree coefficient conversion rule; and the second risk score refers to the risk score corresponding to the dissociation rate calculated by combining the second original risk score and the second correction coefficient.

[0091] In this embodiment, dissociation rate reference ranges are extracted from the first feature parameter range and the second feature parameter range, respectively. The dissociation rate in the core feature parameter is substituted into the calculation formula to obtain the first dissociation matching degree with the first dissociation rate reference range. According to the preset matching degree score conversion rules, the first dissociation matching degree is converted into a second original risk score, and the second original risk score = first dissociation matching degree × 100; then the second dissociation matching degree between the dissociation rate and the second dissociation rate reference range is calculated, and the second dissociation matching degree = 1 - |dissociation rate - mean of the second dissociation rate reference range| ÷ range of the second dissociation rate reference range; according to the preset matching degree coefficient conversion rules, the second dissociation matching degree is converted into a second correction coefficient, and the second correction coefficient = 1 - second dissociation matching degree; finally, the second risk score of the dissociation rate is calculated, and the second risk score = second original risk score × second correction coefficient.

[0092] Step 404: Calculate the first offset matching degree between the resonance angle offset amplitude and the reference range of the resonance angle offset amplitude of the first feature parameter range. Based on the preset matching degree scoring conversion rule, convert the first offset matching degree into a third original risk score. Calculate the second offset matching degree between the resonance angle offset amplitude and the reference range of the resonance angle offset amplitude of the second feature parameter range. Based on the preset matching degree coefficient conversion rule, convert the second offset matching degree into a third correction coefficient. Combined with the third original risk score, calculate the third risk score of the resonance angle offset amplitude.

[0093] In this step, the resonance angle offset amplitude reference range refers to the numerical range of resonance angle offset amplitude contained in the first feature parameter range and the second feature parameter range, respectively; the first offset matching degree refers to the degree of fit between the resonance angle offset amplitude in the core feature parameter and the resonance angle offset amplitude reference range of the first feature parameter range; the third original risk score refers to the risk score obtained by converting the first offset matching degree through a preset matching degree score conversion rule; the second offset matching degree refers to the degree of fit between the resonance angle offset amplitude in the core feature parameter and the resonance angle offset amplitude reference range of the second feature parameter range; the third correction coefficient refers to the correction coefficient obtained by converting the second offset matching degree through a preset matching degree coefficient conversion rule; and the third risk score refers to the risk score corresponding to the resonance angle offset amplitude calculated by combining the third original risk score and the third correction coefficient.

[0094] In this embodiment, a reference range for the resonance angle offset amplitude is extracted from the first feature parameter range and the second feature parameter range, respectively. The resonance angle offset amplitude in the core feature parameter is substituted into the calculation formula to obtain the first offset matching degree with the first resonance angle offset amplitude reference range. According to the preset matching degree scoring conversion rules, the first offset matching degree is converted into the third original risk score, and the third original risk score = first offset matching degree × 100; then the second offset matching degree between the resonance angle offset amplitude and the second offset amplitude reference range is calculated, and the second offset matching degree = 1 - |resonance angle offset amplitude - mean of the second offset amplitude reference range| ÷ range of the second offset amplitude reference range; according to the preset matching degree coefficient conversion rules, the second offset matching degree is converted into the third correction coefficient, and the third correction coefficient = 1 - second offset matching degree; finally, the third risk score of the resonance angle offset amplitude is calculated, and the third risk score = third original risk score × third correction coefficient.

[0095] The embodiments of this application solve the risk assessment bias problem caused by the single reference system in the prior art, improve the accuracy of risk scoring, and provide a reliable basis for subsequent selection of target core feature parameters.

[0096] This application provides a specific embodiment. Step 105 involves verifying the target core feature parameters that meet the preset threshold for risk scores through cross-comparison to obtain verified risk data. This specifically includes the following steps:

[0097] Step 501: Take the core feature parameter whose risk score is greater than the preset threshold as the target core feature parameter, and verify the denoised optical signal and optical signal corresponding to each target core feature parameter to obtain the target verified feature parameter.

[0098] In this step, the target verification feature parameters refer to the target core feature parameters that have been verified by the noise-reduced optical signal and the optical signal corresponding to the target core feature parameters and that meet the verification standards.

[0099] In this embodiment, a preset threshold is first set, and the risk score corresponding to each core feature parameter is compared with the preset threshold. Core feature parameters with risk scores greater than the preset threshold are selected as target core feature parameters. For each target core feature parameter, its corresponding noise-reduced optical signal and optical signal are retrieved respectively. The number of signal points, signal continuity, and signal strength of these two signals are checked respectively. Target core feature parameters that pass the checks for both signals are selected, which are the target checked feature parameters. These parameters will be used for subsequent cross-comparison.

[0100] Step 502: Cross-compare each target's verified feature parameter with the first feature parameter range of the early-stage lung cancer population and the second feature parameter range of the healthy population in the feature parameter database to obtain the consistency score corresponding to each target's verified feature parameter.

[0101] In this step, the consistency score refers to the quantitative score reflecting the degree of fit between the target verified feature parameters and the two reference ranges after cross-comparison between the target verified feature parameters and the first feature parameter range and the second feature parameter range.

[0102] In this embodiment, for each target verified feature parameter, it is first compared with the first feature parameter range of early-stage lung cancer population in the feature parameter database to calculate the first degree of fit; then it is compared with the second feature parameter range of healthy population to calculate the second degree of fit; a consistency score is calculated based on the first and second degrees of fit. The weights 1 and 2 are preset proportional coefficients used to adjust the proportion of the two types of fit in the score. The resulting consistency score will be used for the integration of risk data after subsequent verification.

[0103] Step 503: Integrate the target verified feature parameters, the risk score of the target verified feature parameters, and the consistency score to obtain the verified risk data.

[0104] In this embodiment of the application, the numerical value of the verified feature parameter of each target, the risk score corresponding to the verified feature parameter of the target, and the calculated consistency score are collected. These three types of data are classified and associated to ensure that each verified feature parameter of the target corresponds to a unique risk score and consistency score. The complete set of associated data is integrated into the verified risk data, which will be used for subsequent integration and analysis with concentration information.

[0105] The embodiments of this application eliminate unreliable feature parameters, improve the authenticity and effectiveness of risk data, and provide strong support for the accuracy of subsequent early lung cancer screening results.

[0106] This application provides a specific embodiment. Step 501 involves verifying the denoised optical signal and the optical signal corresponding to each target core feature parameter to obtain the target verified feature parameters. This specifically includes the following steps:

[0107] Step 511: Compare the total number of first signal points and the total number of second signal points of the denoised optical signal corresponding to each target core feature parameter with the preset signal point number threshold to obtain multiple quantity-verified signals.

[0108] In this step, the first total number of signal points refers to the total number of signal points contained in the denoised optical signal corresponding to the core feature parameters of the target; the second total number of signal points refers to the total number of signal points contained in the optical signal corresponding to the core feature parameters of the target; the preset signal point number threshold refers to a pre-set numerical standard for judging whether the number of signal points meets the standard; the signal after quantity verification refers to the signal after comparing the first total number of signal points, the second total number of signal points, and the preset signal point number threshold.

[0109] In this embodiment, for each target core feature parameter, the total number of first signal points and the total number of second signal points of the corresponding denoised optical signal are counted. A preset threshold for the number of signal points is retrieved, and the total number of first signal points is compared with the preset threshold for the number of signal points. At the same time, the total number of second signal points is compared with the preset threshold for the number of signal points. Signals in which the total number of first signal points and the total number of second signal points are both not lower than the preset threshold for the number of signal points are retained. These signals are the signals after quantity verification, and these signals will be used for subsequent length verification.

[0110] Step 512: Based on the continuity of the timestamps of the signal points in the quantity-verified signals corresponding to each target core feature parameter, calculate the signal gap length of each quantity-verified signal, compare the signal gap length with the preset gap threshold, and obtain multiple length-verified signals.

[0111] In this step, the signal gap length refers to the number of signal points corresponding to the gap portion of the signal with no consecutive signal points after quantitative verification; the preset gap threshold refers to the pre-set numerical standard used to determine whether the signal gap length meets the requirements; the signal after length verification refers to the signal after comparing the signal gap length with the preset gap threshold.

[0112] In this embodiment of the application, for each quantity-verified signal, the timestamps of all signal points in the signal are extracted, the gaps in the signal are identified based on the continuity of the timestamps, and the number of signal points contained in each gap is counted as the signal gap length. A preset gap threshold is retrieved, and each signal gap length is compared with the preset gap threshold. The quantity-verified signals in which all signal gap lengths do not exceed the preset gap threshold are retained as the length-verified signals. These signals will be used for subsequent strength verification.

[0113] Step 513: Extract the extreme values ​​of the signal strength of the length-verified signal corresponding to each target core feature parameter, and compare the extreme values ​​of the signal strength with the preset allowable range of signal strength to obtain multiple strength-verified signals.

[0114] In this step, the extreme values ​​of signal strength refer to the maximum and minimum values ​​of signal strength in the signal after length verification; the preset allowable range of signal strength refers to the pre-set numerical range used to determine whether the signal strength meets the requirements; the signal after strength verification refers to the signal after comparing the extreme values ​​of signal strength with the preset allowable range of signal strength.

[0115] In this embodiment, for each length-verified signal, the signal strength of all signal points in the signal is extracted, the maximum and minimum values ​​of the signal strength are selected as signal strength extrema, a preset allowable range of signal strength is retrieved, and the maximum and minimum values ​​of the signal strength are compared with the preset allowable range of signal strength respectively. The length-verified signals whose maximum and minimum values ​​of signal strength are both within the preset allowable range of signal strength are retained, which are the strength-verified signals. These signals will be used for the subsequent selection of target-verified feature parameters.

[0116] Step 514: Based on the quantity-verified signal, length-verified signal, and intensity-verified signal, the denoised optical signal and the target core feature parameters corresponding to the optical signal that have all passed the quantity, length, and intensity verifications are used as the target verified feature parameters.

[0117] In this embodiment of the application, the quantity-verified signal, length-verified signal, and intensity-verified signal corresponding to each target core feature parameter are associated to determine whether the three types of signals corresponding to the same target core feature parameter are the same signal. That is, if all three types of signals pass the quantity verification, length verification, and intensity verification, the target core feature parameters that are all the same verified signal are selected. These parameters are determined as target verified feature parameters, and these parameters will be used for subsequent cross-comparison with the feature parameter database.

[0118] The embodiments of this application eliminate unreliable feature parameters such as insufficient signal points, poor continuity, and abnormal intensity, thereby improving the accuracy of subsequent cross-comparison and risk assessment, and providing a basic guarantee for the reliability of early lung cancer screening results.

[0119] This application provides a specific embodiment. Step 106 involves integrating and analyzing the verified risk data and the concentration information to obtain early lung cancer screening results, specifically including the following steps:

[0120] Step 601: Standardize the concentration values ​​of carcinoembryonic antigen, cytokeratin 19 fragment, and squamous cell carcinoma antigen in the concentration information to obtain a set of standardized concentration values.

[0121] In this step, the standardized concentration value set refers to the set of concentration values ​​obtained after standardizing the concentration values ​​of carcinoembryonic antigen, cytokeratin 19 fragment, and squamous cell carcinoma antigen.

[0122] In this embodiment, the concentration values ​​of carcinoembryonic antigen (CEA), cytokeratin 19 fragment, and squamous cell carcinoma antigen (SCCA) are retrieved from the concentration information. The upper and lower limits of the reference range for the concentration of each target biomarker are obtained. Normalization is then applied to convert each concentration value to the same numerical range, and a standardized concentration value is calculated. The three standardized concentration values ​​after processing are organized into a set of standardized concentration values, which will be used for the subsequent calculation of the comprehensive index.

[0123] Step 602: Based on the risk score and consistency coefficient of the target verified feature parameters in the verified risk data, calculate the weighted risk value of each target core feature parameter.

[0124] In this step, the consistency coefficient refers to the coefficient that reflects the degree of fit between the target verified feature parameters and the ranges of the first and second feature parameters; the weighted risk value refers to the risk value calculated by combining the risk score of the target verified feature parameters and the consistency coefficient.

[0125] In this embodiment of the application, for each target verified feature parameter, its corresponding risk score and consistency coefficient are extracted from the verified risk data, and the weighted risk value is calculated by multiplying the two. The weighted risk value is calculated as: weighted risk value = risk score × consistency coefficient. Each target verified feature parameter corresponds to a weighted risk value. These values ​​will be used for subsequent multiplication with the standardized concentration value.

[0126] Step 603: Multiply the weighted risk value of each target core feature parameter with the standardized concentration value of the corresponding target biomarker in the standardized concentration value set to obtain the comprehensive index of each target core feature parameter.

[0127] In this step, the standardized concentration value of the target biomarker refers to the standardized concentration data corresponding to the target core feature parameter in the standardized concentration value set; the comprehensive index refers to the value obtained by multiplying the weighted risk value by the standardized concentration value of the corresponding target biomarker.

[0128] In this embodiment, a correspondence is established between target core feature parameters and target biomarkers. That is, each target core feature parameter is associated with a specific target biomarker. The standardized concentration value of the target biomarker corresponding to each target core feature parameter is extracted from the standardized concentration value set. The weighted risk value of each target core feature parameter is multiplied by the standardized concentration value of the corresponding target biomarker to calculate a comprehensive index. The comprehensive index = weighted risk value × standardized concentration value of the target biomarker. Each target core feature parameter corresponds to a comprehensive index, and these indices will be used for subsequent cumulative calculations.

[0129] Step 604: Accumulate all comprehensive indices to obtain the total comprehensive index. Based on the range of the total comprehensive index, determine the early lung cancer screening level to obtain the early lung cancer screening result.

[0130] In this step, the total comprehensive index refers to the sum of all comprehensive indices; the numerical range refers to the preset range of the total comprehensive index used to classify the early screening level of lung cancer; and the early screening level of lung cancer refers to the risk level determined based on the numerical range of the total comprehensive index.

[0131] In this embodiment, a comprehensive index corresponding to all target core feature parameters is collected, and these comprehensive indices are summed sequentially to obtain a total comprehensive index. Multiple preset numerical ranges are retrieved, and each numerical range corresponds to an early lung cancer screening level. The total comprehensive index is compared with these numerical ranges to determine the numerical range in which the total comprehensive index falls, and then the corresponding early lung cancer screening level is determined. This level is the early lung cancer screening result.

[0132] The embodiments of this application achieve deep integration of risk data and concentration information, overcome the limitations of single data assessment, and improve the accuracy and reliability of early lung cancer screening results.

[0133] In a specific case, steps 102 and 103 involve: optical signal processing and core feature parameter extraction, whereby an optical signal of the binding of a patient's "carcinoembryonic antigen (CEA)" and its corresponding antibody is obtained using a surface plasmon resonance (SPR) detection device; see [link to relevant documentation]. Figure 3Before this step, the following preparations are required: receiving target subject information: age: 55 years old; smoking history: yes; detection biomarker: carcinoembryonic antigen (CEA); steps 101 and 102: concentration detection and optical signal acquisition, acquisition result: CEA concentration = 8.5 ng / mL (exceeding the normal reference value < 5 ng / mL); optical signal acquisition: the raw optical signal is acquired through a surface plasmon resonance (SPR) device. After noise reduction, segmentation, and calculation in step 103 above, the following core characteristic parameters are obtained: binding rate, dissociation rate, and resonance angle shift amplitude; where the binding rate refers to the speed of the reaction stage when the target biomarker binds to the corresponding specific antibody; in the above case, the core characteristic parameter is as follows: binding rate (Ka) dissociation rate Resonance angle offset amplitude: 7000RU; Step 305: Integration of core feature parameters, where the core feature parameter set Binding rate: Dissociation rate: Resonance angle offset amplitude: 7000 ;

[0134] Then, step 104 is executed, which uses edge computing technology to compare each core feature parameter with the feature parameter database of the target population to obtain the corresponding risk score; Step 401: Retrieve the database, and then the feature parameter range of the early lung cancer population is: Ka: [X1, X2]; Kd: [Y1, Y2]; offset: [6500, 9000]; where, the feature parameter range of the healthy population is: Ka: [Xj1, Xj2]; Kd: [Yj1, Yj2]; offset: [3000, 6000];

[0135] Steps 402-404: Includes calculating the risk score for each parameter, and the assemblage rate Ka: for example, the first assemblage matching degree. First original risk score Points; second combination matching degree First correction factor First risk score Score; Dissociation rate Kd: Assuming calculation yields: Second risk score Points; Resonance angle offset amplitude: Assuming calculation results: Third risk score point;

[0136] Step 105: Includes performing cross-validation; Step 501: Filtering and validation, including: preset thresholds. The binding rate (9.5 points) was discarded, and the dissociation rate (85 points) and resonance angle shift amplitude (88 points) were retained as the target core feature parameters. Then, step 502 was performed: the consistency score was calculated, including the consistency score of the dissociation rate. Consistency score for resonance angle offset amplitude Then proceed to step 503: Integrate the validated risk data, where the validated risk data... Target parameter: [Dissociation rate:] Offset: 7000], Risk Score: [85, 88], Consistency Coefficient: Then proceed to step 106: Integrate analysis and final results, involving step 601: Standardize concentration values. Step 602: Calculate the weighted risk value, where the weighted risk value of the dissociation rate is... Weighted risk value of resonance angle offset amplitude Step 603: Calculate the composite index, including: calculating the composite index of the dissociation rate. ; Calculate the comprehensive index of resonance angle offset amplitude .

[0137] Step 604: Determine the screening results. Calculate the total comprehensive index: 65.69 + 65.05 = 130.74. Screening result: Total comprehensive index 130.74 > 120 (where 120 is the preset range for high-risk levels), therefore, it is classified as "high-risk". See also Figure 3 , Figure 3 It shows the relevant information of the examinee, experimental data, and final screening results.

[0138] Example 2

[0139] Figure 4 This is a schematic diagram of a specific implementation of an early lung cancer screening system based on in vitro biomarker detection, provided in Embodiment 2 of this application. (Refer to...) Figure 4 The system may include:

[0140] The acquisition module 21 is used to collect peripheral blood samples from the subject, identify target biomarkers in the peripheral blood samples, and detect the concentration information of the target biomarkers. The target biomarkers include carcinoembryonic antigen, cytokeratin 19 fragment, and squamous cell carcinoma antigen.

[0141] Detection module 22 is used to detect the changes in optical properties of the target biomarker and the corresponding specific antibody during the antigen-antibody binding process using a surface plasmon resonance detection device, so as to obtain an optical signal;

[0142] The noise reduction module 23 is used to perform noise reduction processing on the optical signal to obtain a noise-reduced optical signal, and to extract core feature parameters from the noise-reduced optical signal. The core feature parameters include binding rate, dissociation rate and resonance angle offset amplitude.

[0143] The comparison module 24 is used to compare each core feature parameter with the feature parameter database of the target population using edge computing technology to obtain the corresponding risk score. The target population includes people with early-stage lung cancer and healthy people.

[0144] Verification module 25 is used to verify the target core feature parameters that meet the preset threshold through cross-comparison to obtain verified risk data;

[0145] Analysis module 26 integrates and analyzes the verified risk data and the concentration information to obtain early lung cancer screening results.

[0146] This application provides an embodiment of an early lung cancer screening system based on in vitro biomarker detection to implement the aforementioned method for early lung cancer screening based on in vitro biomarker detection. Therefore, the specific implementation of the early lung cancer screening system based on in vitro biomarker detection can be found in the embodiment section of the method for early lung cancer screening based on in vitro biomarker detection described above. The specific implementation can be referred to the description of the corresponding embodiments, which will not be repeated here.

[0147] This application also provides an electronic device, comprising: a memory for storing a computer program; and a processor for executing the computer program to implement the steps of any of the above-described methods for early lung cancer screening based on in vitro detection of biomarkers.

[0148] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of any of the above-described methods for early lung cancer screening based on in vitro detection of biomarkers.

[0149] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory, random access memory, portable hard drives, magnetic disks, or optical disks.

[0150] Embodiments of the present invention also provide a computer program product, which includes a computer program that, when executed by a processor, implements the steps in any of the embodiments of the above-described method for early lung cancer screening based on in vitro detection of biomarkers.

[0151] Those skilled in the art will further 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.

[0152] The foregoing has provided a detailed description of a method and system for early lung cancer screening based on in vitro detection of biomarkers, as provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A method for early lung cancer screening based on in vitro detection of biomarkers, characterized in that, include: Peripheral blood samples were collected from the subjects, target biomarkers in the peripheral blood samples were identified, and the concentration information of the target biomarkers was detected. The target biomarkers included carcinoembryonic antigen, cytokeratin 19 fragment, and squamous cell carcinoma antigen. The optical properties of the target biomarker and the corresponding specific antibody during the antigen-antibody binding process are detected using a surface plasmon resonance detection device to obtain an optical signal. The optical signal is denoised to obtain a denoised optical signal. Core feature parameters are extracted from the denoised optical signal. The core feature parameters include binding rate, dissociation rate and resonance angle shift amplitude. Using edge computing technology, each core feature parameter is compared with the feature parameter database of the target population to obtain the corresponding risk score. The target population includes people with early-stage lung cancer and healthy people. By cross-comparison, the target core feature parameters that meet the preset threshold for risk scores are verified to obtain verified risk data. The validated risk data and concentration information are integrated and analyzed to obtain early lung cancer screening results.

2. The method according to claim 1, characterized in that, The optical signal is denoised to obtain a denoised optical signal. Core feature parameters are extracted from the denoised optical signal, including the binding rate, dissociation rate, and resonance angle shift amplitude. Based on the biological stages of the antigen-antibody binding process between the target biomarker and the corresponding specific antibody, the optical signal is divided into multiple signal segments, including the initial binding stage, the rapid binding stage, the equilibrium stage, and the dissociation stage. Abnormal fluctuation signals with signal strength exceeding a preset signal strength threshold are removed from each signal segment to obtain multiple interference-free signal segments. All interference-free signal segments are spliced ​​together in the order of time intervals of biological stages to obtain spliced ​​interference-free signals. The spliced ​​interference-free signals are then smoothed to obtain noise-reduced optical signals. The resonance angle shift amplitude is calculated based on the signal change in the noise-reduced optical signals. Based on the signal segments corresponding to the initial binding stage and the rapid binding stage in the denoised optical signal, the binding rate is calculated, and based on the signal segments corresponding to the dissociation stage in the denoised optical signal, the dissociation rate is calculated. The core characteristic parameters are obtained by integrating the resonance angle offset amplitude, the binding rate, and the dissociation rate.

3. The method according to claim 2, characterized in that, Abnormal fluctuation signals exceeding a preset signal strength threshold are removed from each signal segment, resulting in multiple interference-free signal segments, including: The signal points in each signal segment that are in the stable period corresponding to the biological stage are integrated to form the baseline signal segment of each signal segment; The signal strength values ​​of all signal points in each baseline signal segment are statistically analyzed. Based on the signal strength values, the mean signal strength and standard deviation of each baseline signal segment are calculated to determine the signal strength threshold of each signal segment. Abnormal signal points whose signal strength exceeds the corresponding signal strength threshold in each signal segment are removed to obtain multiple abnormal signal segments. Based on the signal strength of adjacent signal points in each abnormal signal segment, the filling value of each abnormal signal segment is calculated. The filling value is added to the corresponding abnormal signal segment to obtain multiple interference-free signal segments.

4. The method according to claim 1, characterized in that, By cross-referencing, the target core feature parameters that meet the preset threshold for risk scores are verified to obtain verified risk data, including: The core feature parameters whose risk scores are greater than a preset threshold are used as target core feature parameters. The denoised optical signal and optical signal corresponding to each target core feature parameter are verified to obtain the target verified feature parameters. Each target's verified feature parameter is cross-compared with the first feature parameter range of early-stage lung cancer population and the second feature parameter range of healthy population in the feature parameter database to obtain the consistency score corresponding to each target's verified feature parameter. The post-verification feature parameters, the risk scores of the post-verification feature parameters, and the consistency scores are integrated to obtain the post-verification risk data.

5. The method according to claim 1, characterized in that, The denoised optical signal and the optical signal corresponding to each core feature parameter of the target are verified to obtain the verified feature parameters of the target, including: The total number of first signal points and the total number of second signal points of the denoised optical signal corresponding to each target core feature parameter are compared with the preset signal point number threshold to obtain multiple quantity-verified signals. Based on the continuity of the timestamps of the signal points in the quantity-verified signals corresponding to each target core feature parameter, the signal gap length of each quantity-verified signal is calculated, and the signal gap length is compared with a preset gap threshold to obtain multiple length-verified signals. Extract the extreme values ​​of the signal strength of the length-verified signal corresponding to each target core feature parameter, and compare the extreme values ​​of the signal strength with the preset allowable range of signal strength to obtain multiple strength-verified signals; Based on the signal after quantity verification, the signal after length verification, and the signal after intensity verification, the denoised optical signal and the target core feature parameters corresponding to the optical signal that have all passed quantity verification, length verification, and intensity verification are used as the target verification feature parameters.

6. The method according to claim 1, characterized in that, Using edge computing technology, each core feature parameter is compared with a database of feature parameters of the target population to obtain a corresponding risk score, including: Using edge computing technology, the feature parameter database of the target population is retrieved. The feature parameter database of the target population stores the first feature parameter range of the early lung cancer population and the second feature parameter range of the healthy population. Calculate the first binding match degree between the binding rate in the core feature parameters and the binding rate reference range of the first feature parameter range. Based on a preset matching degree scoring conversion rule, convert the first binding match degree into a first original risk score. Calculate the second binding match degree between the binding rate in the core feature parameters and the binding rate reference range of the second feature parameter range. Based on a preset matching degree coefficient conversion rule, convert the second binding match degree into a first correction coefficient. Combined with the first original risk score, calculate the first risk score of the binding rate. Calculate the first dissociation matching degree between the dissociation rate and the dissociation rate reference range of the first feature parameter range. Based on a preset matching degree scoring conversion rule, convert the first dissociation matching degree into a second original risk score. Calculate the second dissociation matching degree between the dissociation rate and the dissociation rate reference range of the second feature parameter range. Based on a preset matching degree coefficient conversion rule, convert the second dissociation matching degree into a second correction coefficient. Combine the second original risk score to calculate the second risk score of the dissociation rate. The first offset matching degree between the resonance angle offset amplitude and the reference range of the resonance angle offset amplitude of the first feature parameter range is calculated. Based on the preset matching degree scoring conversion rule, the first offset matching degree is converted into a third original risk score. The second offset matching degree between the resonance angle offset amplitude and the reference range of the resonance angle offset amplitude of the second feature parameter range is calculated. Based on the preset matching degree coefficient conversion rule, the second offset matching degree is converted into a third correction coefficient. Combined with the third original risk score, the third risk score of the resonance angle offset amplitude is calculated.

7. The method according to claim 1, characterized in that, The validated risk data and concentration information are integrated and analyzed to obtain early lung cancer screening results, including: The concentration values ​​of carcinoembryonic antigen, cytokeratin 19 fragment, and squamous cell carcinoma antigen in the concentration information are standardized to obtain a set of standardized concentration values. Based on the risk score and consistency coefficient of the target verified feature parameters in the verified risk data, the weighted risk value of each target core feature parameter is calculated. The weighted risk value of each target core feature parameter is multiplied by the standardized concentration value of the corresponding target biomarker in the standardized concentration value set to obtain the comprehensive index of each target core feature parameter; All comprehensive indices are summed to obtain a total comprehensive index. Based on the range of values ​​of the total comprehensive index, the early lung cancer screening level is determined to obtain the early lung cancer screening result.

8. A lung cancer early screening system based on in vitro biomarker detection, characterized in that, include: The acquisition module is used to collect peripheral blood samples from the subject, identify target biomarkers in the peripheral blood samples, and detect the concentration information of the target biomarkers. The target biomarkers include carcinoembryonic antigen, cytokeratin 19 fragment, and squamous cell carcinoma antigen. The detection module is used to detect the changes in optical properties of the target biomarker and the corresponding specific antibody during the antigen-antibody binding process using a surface plasmon resonance detection device, so as to obtain an optical signal; The noise reduction module is used to perform noise reduction processing on the optical signal to obtain a noise-reduced optical signal, and to extract core feature parameters from the noise-reduced optical signal. The core feature parameters include binding rate, dissociation rate and resonance angle offset amplitude. The comparison module is used to compare each core feature parameter with the feature parameter database of the target population using edge computing technology to obtain the corresponding risk score. The target population includes people with early-stage lung cancer and healthy people. The verification module is used to verify the target core feature parameters that meet the preset threshold through cross-comparison, and obtain the verified risk data. The analysis module integrates and analyzes the verified risk data and the concentration information to obtain early lung cancer screening results.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the steps of a method for early lung cancer screening based on in vitro detection of biomarkers as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, enables an early lung cancer screening method based on in vitro detection of biomarkers as described in any one of claims 1 to 7.

Images