Expiration component analysis system for lung cancer screening

Abstract

The invention discloses an expiration component analysis system for lung cancer screening, which belongs to the technical field of lung cancer screening and comprises a data acquisition module, a data processing module, a model training module, a test optimization module, an analysis prediction module and an early warning guidance module. The problems that in the prior art, expiration components cannot be effectively analyzed, the lung cancer screening condition cannot be accurately known, so that a user cannot be timely examined and treated, and the cure rate and the survival rate of the lung cancer are reduced are solved. According to the method, the expiration component analysis prediction result is determined, early warning examination and treatment guidance is performed on the user based on the lung cancer examination and treatment scheme, the expiration components can be effectively analyzed, and the lung cancer screening condition can be accurately known, so that the user can be timely examined and treated, and early discovery and early diagnosis of the lung cancer can be realized; the cure rate and the survival rate of the lung cancer can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of lung cancer screening, and in particular to a breath component analysis system for lung cancer screening. Background Art

[0002] Lung cancer, also known as primary bronchial lung cancer, is the most common lung malignancy originating from the trachea, bronchial mucosa or glands. According to the histopathological characteristics, it is mainly divided into non-small cell carcinoma (including adenocarcinoma and squamous cell carcinoma) and small cell carcinoma. Lung cancer is not contagious, but it has obvious familial aggregation and genetic susceptibility. Smoking, occupational exposure, air pollution, etc. are all common causes. Globally, the incidence and mortality of lung cancer are extremely high and are on the rise. For the treatment of lung cancer, commonly used methods include surgery, radiotherapy and chemotherapy. The prognosis of the disease varies depending on the patient's condition, such as tumor type, tumor stage, age, gender, etc., but overall, the five-year survival rate of lung cancer is not high. In terms of prevention and treatment, prevention is the key, including reducing smoking, avoiding occupational and environmental exposure, and regular screening.

[0003] A Chinese patent with publication number CN210931411U discloses a graphene sensor-based breath detector for lung cancer screening, including: a gas collection device, including a connector, an air intake pipe, an air pump and a helium source; the air pump is connected to the first connection area, for forming a negative pressure in the first connection area; a helium source, for providing helium to the second connection area; the breath collection device includes a graphene sensor, and the graphene sensor detects the gas in the second connection area according to the resistance change information of the gas transmitted from the second connection area to screen for tuberculosis or lung cancer. Among them, the gas to be detected is inhaled from the air intake pipe into the first connection area, and the first connecting valve is opened so that the gas in the first connection area can be mixed with the helium in the second connection area, so that the gas to be detected is detected in a stable air pressure state under a certain dilution, thereby improving the reliability of the detection; however, the patent has the following defects:

[0004] The existing methods cannot effectively analyze the components of exhaled breath and cannot accurately understand the lung cancer screening situation, so that users cannot receive timely examinations and treatments, cannot achieve early detection and early diagnosis of lung cancer, and reduce the cure rate and survival rate of lung cancer. Summary of the invention

[0005] The purpose of the present invention is to provide an exhaled breath component analysis system for lung cancer screening, which can effectively analyze the exhaled breath components and accurately understand the lung cancer screening situation, so that users can receive timely examinations and treatments, achieve early detection and early diagnosis of lung cancer, improve the cure rate and survival rate of lung cancer, and solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] Breath composition analysis system for lung cancer screening, including:

[0008] A data collection module is used to collect and monitor the user's exhaled gas and determine the real-time data of exhaled breath components based on lung cancer screening;

[0009] A data processing module, used for searching, sorting and feature extraction of real-time data of breath components based on lung cancer screening, and determining feature data of breath components based on lung cancer screening;

[0010] A model training module is used to train the selected appropriate model architecture and determine a breath component analysis model suitable for lung cancer screening;

[0011] A test optimization module is used to test and optimize the breath component analysis model suitable for lung cancer screening, and determine the best breath component analysis model after optimization and adjustment;

[0012] An analysis and prediction module is used to analyze and predict the characteristic data of breath components based on lung cancer screening, and determine the breath component analysis prediction results;

[0013] The early warning guidance module is used to formulate lung cancer examination and treatment plans, and provide early warning examination and treatment guidance to users based on the lung cancer examination and treatment plans.

[0014] Preferably, the data acquisition module includes:

[0015] A gas collection unit, used to collect the user's exhaled gas and determine the exhaled gas component sample;

[0016] A concentration monitoring unit is used to monitor the concentration of breath component samples in real time and determine breath component concentration data based on lung cancer screening;

[0017] A type monitoring unit, used to monitor the types of breath component samples in real time and determine breath component type data based on lung cancer screening;

[0018] Among them, based on the breath component concentration data and the breath component type data, the real-time data of the breath components based on lung cancer screening is determined.

[0019] Preferably, the data processing module includes:

[0020] A data retrieval unit, used for retrieving real-time data of breath components based on lung cancer screening;

[0021] Obtain real-time data on breath components for lung cancer screening;

[0022] Based on the sequential retrieval method, the real-time data of exhaled breath components are retrieved;

[0023] Conduct consistency checks on real-time data of exhaled breath components;

[0024] Check whether the real-time data of breath components meet the requirements according to the reasonable value range and mutual relationship of each variable in the real-time data of breath components;

[0025] Remove data that is beyond the normal range, logically unreasonable or contradictory in the real-time data of exhaled breath components;

[0026] Determine the real-time data of breath composition with data consistency;

[0027] Process invalid and missing values ​​for real-time data of breath components with data consistency;

[0028] Remove invalid data and missing data in the real-time data of breath components that are of no value to breath component analysis;

[0029] Determine the real-time data of breath components that are valuable for breath component analysis;

[0030] A data sorting unit, used for sorting the retrieved real-time data of breath components;

[0031] Obtain real-time data of breath components that are valuable for breath component analysis;

[0032] Sorting valuable real-time data of breath components for breath component analysis;

[0033] Determine the real-time data of breath components with arrangement order;

[0034] A feature extraction unit, used for extracting features from the sorted real-time data of breath components;

[0035] Acquire real-time data of breath components with ranking order;

[0036] Extract features from real-time data of breath components with arrangement order;

[0037] Determine breath composition characteristics for lung cancer screening.

[0038] Preferably, the model training module includes:

[0039] A data acquisition unit, used for acquiring historical data of breath components;

[0040] According to the breath component analysis requirements for lung cancer screening, breath component samples are obtained from different users in different regions to determine the breath component historical data;

[0041] A data division unit, used for dividing the historical data of breath components;

[0042] According to the needs of breath component analysis for lung cancer screening, the historical data of breath components are divided to determine the training set and test set;

[0043] The architecture selection unit is used to select the appropriate model architecture;

[0044] According to the needs of breath component analysis for lung cancer screening, select a model architecture suitable for lung cancer screening and determine the appropriate model architecture;

[0045] Model training unit, used to train the model architecture;

[0046] Get the appropriate model architecture and training set for your selection;

[0047] Based on the training set, train the selected appropriate model architecture;

[0048] A breath component analysis model suitable for lung cancer screening was determined.

[0049] Preferably, the data partitioning unit comprises:

[0050] A sample quantity acquisition module is used to retrieve the sample quantity corresponding to the training set after determining the training set and the test set;

[0051] The sample requirement quantity extraction module is used to extract the sample requirement quantity corresponding to model training;

[0052] A comparison module, used to compare the number of samples corresponding to the training set with the number of samples required for model training;

[0053] A sample difference quantity acquisition module is used to acquire the sample difference quantity according to the sample number corresponding to the training set and the sample requirement corresponding to the model training when the sample number corresponding to the training set is lower than the sample requirement corresponding to the model training;

[0054] A reference sample data acquisition module, used for randomly extracting sample data corresponding to the number of sample differences from the sample data of the training set as reference sample data;

[0055] A sample data compensation coefficient acquisition module, used to acquire a sample data compensation coefficient corresponding to each reference sample data according to the gas concentration data corresponding to the reference sample data and the sample quantity of the reference sample data;

[0056] The supplementary sample data acquisition module is used to adjust the sample data of each reference sample data using the sample data compensation coefficient to obtain the adjusted supplementary sample data; wherein the adjusted supplementary sample data is obtained by the following formula:

[0057]

[0058] Where D represents the adjusted gas concentration corresponding to each gas component of each reference sample data; D 0 represents the gas concentration before adjustment corresponding to each gas component of each reference sample data; D min and D max They represent the minimum gas concentration value and the maximum gas concentration value of each gas component in the reference sample data respectively; S represents the sample data compensation coefficient;

[0059] The supplement execution module is used to add the adjusted supplementary sample data to the training set.

[0060] Preferably, the sample data compensation coefficient acquisition module includes:

[0061] A gas concentration data extraction module, used to extract gas concentration data corresponding to the reference sample data;

[0062] The first sample data difference compensation coefficient acquisition module is used to acquire the first sample data difference compensation coefficient corresponding to each gas component of the reference sample data according to the gas concentration data corresponding to the reference sample data, wherein the first sample data difference compensation coefficient is acquired by the following formula:

[0063]

[0064] Among them, S 01 represents the first sample data difference compensation coefficient corresponding to the reference sample data; D maxj and D minj They represent the maximum and minimum gas concentrations of the j-th gas component contained in the reference sample data; m represents the type of gas component contained in the reference sample data; n represents the number of times the j-th gas component appears in the reference sample data; D i represents the gas concentration corresponding to the i-th occurrence of the j-th gas component; D p represents the average gas concentration of the jth gas component in the reference sample data; D z Indicates the median gas concentration of the jth gas component in the reference sample data;

[0065] A sample quantity information extraction module, used to extract the sample quantity of the reference sample data;

[0066] The second sample data difference compensation coefficient acquisition module is used to acquire the second sample data difference compensation coefficient according to the sample data volume of the reference sample data and the sample data volume of the existing sample data in the current training set, wherein the second sample data difference compensation coefficient is acquired by the following formula:

[0067]

[0068] Among them, S 02 represents the second sample data difference compensation coefficient; M represents the sample data volume of the reference sample data; N represents the sample data volume of the sample data already in the current training set; m represents the type of gas components contained in the reference sample data; f cj represents the frequency of occurrence of the jth gas component in the reference sample data; f xj represents the frequency of occurrence of the jth gas component in the current training set; D cj represents the average concentration of the jth gas component in the reference sample data; D xj represents the average concentration of the jth gas component in the current training set;

[0069] The sample data compensation coefficient acquisition module is used to acquire the sample data compensation coefficient corresponding to each reference sample data by using the first sample data difference compensation coefficient and the second sample data difference compensation coefficient; wherein the sample data compensation coefficient corresponding to each reference sample data is acquired by the following formula:

[0070] S=(1-S 01 )·S 02

[0071] Where S represents the sample data compensation coefficient; S 01 represents the first sample data difference compensation coefficient corresponding to the reference sample data; S 02 Represents the second sample data difference compensation coefficient.

[0072] Preferably, the test optimization module includes:

[0073] A performance test unit for testing a breath component analysis model suitable for lung cancer screening;

[0074] Obtain a breath component analysis model suitable for lung cancer screening;

[0075] Based on the test set, the performance of the breath component analysis model suitable for lung cancer screening was tested;

[0076] Determine the performance test results based on the breath composition analysis model;

[0077] An optimization and adjustment unit, used to optimize the breath component analysis model suitable for lung cancer screening;

[0078] Obtain performance test results based on breath component analysis model;

[0079] Conduct in-depth mining and correlation analysis on the performance test results based on the breath component analysis model;

[0080] Determine the optimal adjustment plan based on the breath component analysis model;

[0081] Based on the optimization and adjustment plan, the breath component analysis model suitable for lung cancer screening is optimized and adjusted;

[0082] Determine the best exhaled breath component analysis model after optimization and adjustment.

[0083] Preferably, the analysis and prediction module includes:

[0084] A data extraction unit, used for extracting characteristic data of breath components based on lung cancer screening;

[0085] According to the needs of breath component analysis for lung cancer screening, the breath component characteristic data based on lung cancer screening is extracted;

[0086] An analysis and prediction unit, used for analyzing and predicting characteristic data of breath components based on lung cancer screening;

[0087] Acquire extracted breath component characteristic data;

[0088] Inputting the breath component characteristic data into the best breath component analysis model after optimization and adjustment;

[0089] Analyze and predict the characteristic data of breath components based on the breath component analysis model;

[0090] Determine the predicted results of breath composition analysis.

[0091] Preferably, the early warning guidance module includes:

[0092] a strategy-making unit to develop lung cancer screening and treatment plans;

[0093] Obtain the predicted results of breath component analysis;

[0094] Conduct in-depth mining and correlation analysis on the predicted results of breath component analysis;

[0095] Determine the lung cancer examination and treatment plan;

[0096] An early warning guidance unit, used to provide early warning guidance to users;

[0097] Get lung cancer screening and treatment options;

[0098] Provide early warning examination and treatment guidance to users based on lung cancer examination and treatment plans.

[0099] Preferably, the user is given early warning guidance and the following operations are performed:

[0100] When the breath component analysis predicts that the user has abnormal breath components and is at risk of lung cancer, a warning alarm is automatically issued, and the breath component analysis prediction results are transmitted to the user in a visual form, intelligently guiding the user to proceed to the next step of examination and treatment.

[0101] Compared with the prior art, the present invention has the following beneficial effects:

[0102] The present invention collects and monitors the user's exhaled gas to determine the real-time data of exhaled components, retrieves, sorts and extracts features from the real-time data of exhaled components to determine the characteristic data of the exhaled components, determines the exhaled component analysis model by training a selected appropriate model architecture, analyzes and predicts the exhaled component characteristic data based on the exhaled component analysis model, determines the exhaled component analysis prediction results, and formulates a lung cancer examination and treatment plan, and provides early warning examination and treatment guidance to the user based on the lung cancer examination and treatment plan, can effectively analyze the exhaled components, can accurately understand the lung cancer screening situation, so that the user can receive timely examination and treatment, can achieve early detection and early diagnosis of lung cancer, and can improve the cure rate and survival rate of lung cancer. BRIEF DESCRIPTION OF THE DRAWINGS

[0103] Figure 1 This is a module structure diagram of the breath component analysis system for lung cancer screening of the present invention. DETAILED DESCRIPTION

[0104] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0105] In order to solve the existing problem that the breath components cannot be effectively analyzed, the lung cancer screening situation cannot be accurately understood, so that users cannot receive timely examination and treatment, cannot achieve early detection and early diagnosis of lung cancer, and reduce the cure rate and survival rate of lung cancer, please refer to Figure 1 , this embodiment provides the following technical solutions:

[0106] The breath component analysis system for lung cancer screening includes: a data acquisition module, a data processing module, a model training module, a test optimization module, an analysis and prediction module and an early warning guidance module.

[0107] It should be noted that through the interaction between the data acquisition module, data processing module, model training module, test optimization module, analysis and prediction module and early warning guidance module, the exhaled breath components can be effectively analyzed, and the lung cancer screening situation can be accurately understood, so that users can receive timely examination and treatment, and early detection and early diagnosis of lung cancer can be achieved, which can improve the cure rate and survival rate of lung cancer.

[0108] Among them, the data acquisition module is used to collect and monitor the user's exhaled gas and determine the real-time data of exhaled components based on lung cancer screening;

[0109] In this embodiment, the data acquisition module includes:

[0110] A gas collection unit, used to collect the user's exhaled gas and determine the exhaled gas component sample;

[0111] A concentration monitoring unit is used to monitor the concentration of breath component samples in real time and determine breath component concentration data based on lung cancer screening;

[0112] A type monitoring unit, used to monitor the types of breath component samples in real time and determine breath component type data based on lung cancer screening;

[0113] Among them, based on the breath component concentration data and the breath component type data, the real-time data of the breath components based on lung cancer screening is determined.

[0114] Among them, the data processing module is used to retrieve, sort and extract features of the real-time data of breath components based on lung cancer screening, and determine the characteristic data of breath components based on lung cancer screening;

[0115] In this embodiment, the data processing module includes:

[0116] A data retrieval unit, used for retrieving real-time data of breath components based on lung cancer screening;

[0117] Obtain real-time data on breath components for lung cancer screening;

[0118] Based on the sequential retrieval method, the real-time data of exhaled breath components are retrieved;

[0119] Conduct consistency checks on real-time data of exhaled breath components;

[0120] Check whether the real-time data of breath components meet the requirements according to the reasonable value range and mutual relationship of each variable in the real-time data of breath components;

[0121] Remove data that is beyond the normal range, logically unreasonable or contradictory in the real-time data of exhaled breath components;

[0122] Determine the real-time data of breath composition with data consistency;

[0123] Process invalid and missing values ​​for real-time data of breath components with data consistency;

[0124] Remove invalid data and missing data in the real-time data of breath components that are of no value to breath component analysis;

[0125] Determine the real-time data of breath components that are valuable for breath component analysis;

[0126] It should be noted that by searching the real-time data of exhaled breath components and determining the real-time data of exhaled breath components that are valuable for exhaled breath component analysis, the subsequent processing speed and processing accuracy of the real-time data of exhaled breath components can be improved.

[0127] A data sorting unit, used for sorting the retrieved real-time data of breath components;

[0128] Obtain real-time data of breath components that are valuable for breath component analysis;

[0129] Sorting valuable real-time data of breath components for breath component analysis;

[0130] Determine the real-time data of breath components with arrangement order;

[0131] A feature extraction unit, used for extracting features from the sorted real-time data of breath components;

[0132] Acquire real-time data of breath components with ranking order;

[0133] Extract features from real-time data of breath components with arrangement order;

[0134] Determine breath composition characteristics for lung cancer screening.

[0135] Among them, the model training module is used to train the selected appropriate model architecture and determine the breath component analysis model suitable for lung cancer screening;

[0136] In this embodiment, the model training module includes:

[0137] A data acquisition unit, used for acquiring historical data of breath components;

[0138] According to the breath component analysis requirements for lung cancer screening, breath component samples are obtained from different users in different regions to determine the breath component historical data;

[0139] A data division unit, used for dividing the historical data of breath components;

[0140] According to the needs of breath component analysis for lung cancer screening, the historical data of breath components are divided to determine the training set and test set;

[0141] The architecture selection unit is used to select the appropriate model architecture;

[0142] According to the needs of breath component analysis for lung cancer screening, select a model architecture suitable for lung cancer screening and determine the appropriate model architecture;

[0143] Model training unit, used to train the model architecture;

[0144] Get the appropriate model architecture and training set for your selection;

[0145] Based on the training set, train the selected appropriate model architecture;

[0146] A breath component analysis model suitable for lung cancer screening was determined.

[0147] Specifically, the data partitioning unit includes:

[0148] A sample quantity acquisition module is used to retrieve the sample quantity corresponding to the training set after determining the training set and the test set;

[0149] The sample requirement quantity extraction module is used to extract the sample requirement quantity corresponding to model training;

[0150] A comparison module, used to compare the number of samples corresponding to the training set with the number of samples required for model training;

[0151] A sample difference quantity acquisition module is used to acquire the sample difference quantity according to the sample number corresponding to the training set and the sample requirement corresponding to the model training when the sample number corresponding to the training set is lower than the sample requirement corresponding to the model training;

[0152] A reference sample data acquisition module, used for randomly extracting sample data corresponding to the number of sample differences from the sample data of the training set as reference sample data;

[0153] A sample data compensation coefficient acquisition module, used to acquire a sample data compensation coefficient corresponding to each reference sample data according to the gas concentration data corresponding to the reference sample data and the sample quantity of the reference sample data;

[0154] The supplementary sample data acquisition module is used to adjust the sample data of each reference sample data using the sample data compensation coefficient to obtain the adjusted supplementary sample data; wherein the adjusted supplementary sample data is obtained by the following formula:

[0155]

[0156] Where D represents the adjusted gas concentration corresponding to each gas component of each reference sample data; D 0 represents the gas concentration before adjustment corresponding to each gas component of each reference sample data; D min and D max They represent the minimum gas concentration value and the maximum gas concentration value of each gas component in the reference sample data respectively; S represents the sample data compensation coefficient;

[0157] The supplement execution module is used to add the adjusted supplementary sample data to the training set.

[0158] The technical effect of the above technical solution is: by comparing the number of samples in the training set with the number of samples required for model training, the technical solution can identify whether the training set is sufficient. If the training set samples are insufficient, it will supplement the sample data through a series of operations to ensure that the training set has a sufficient number of samples, which helps to improve the generalization ability and accuracy of the model.

[0159] In the field of machine learning and deep learning, sufficient and diverse data sets are key to training high-quality models. By automatically supplementing sample data, this technical solution reduces the need for manual intervention, speeds up the data preparation process, and thus improves the efficiency of model training.

[0160] By adjusting the reference sample data based on the sample data compensation coefficient of the gas concentration data, this technical solution can generate more diverse supplementary sample data. These supplementary sample data not only increase the size of the training set, but also increase the diversity of the data, which helps the model learn more comprehensive features, thereby enhancing the model's adaptability to unknown data.

[0161] Since the supplementary sample data is based on the intelligent adjustment of the existing training set samples, these newly generated sample data can simulate the complex situations in the real world to a certain extent. Therefore, adding these supplementary sample data to the training set can further improve the performance of the model, including indicators such as accuracy, recall rate, and F1 score.

[0162] The entire technical solution is an automated process, which does not require much manual intervention from obtaining the number of samples to generating supplementary sample data and then adding the supplementary sample data to the training set. In addition, the technical solution is also scalable and can be adjusted and optimized according to different data sets and model requirements.

[0163] In summary, this technical solution optimizes the integrity of the data set, improves the model training efficiency and performance, enhances the adaptability of the model, and realizes the automation and scalability of data processing by intelligently supplementing the training set sample data.

[0164] Specifically, the sample data compensation coefficient acquisition module includes:

[0165] A gas concentration data extraction module, used to extract gas concentration data corresponding to the reference sample data;

[0166] The first sample data difference compensation coefficient acquisition module is used to acquire the first sample data difference compensation coefficient corresponding to each gas component of the reference sample data according to the gas concentration data corresponding to the reference sample data, wherein the first sample data difference compensation coefficient is acquired by the following formula:

[0167]

[0168] Among them, S 01 represents the first sample data difference compensation coefficient corresponding to the reference sample data; D maxj and D minj They represent the maximum and minimum gas concentrations of the j-th gas component contained in the reference sample data; m represents the type of gas component contained in the reference sample data; n represents the number of times the j-th gas component appears in the reference sample data; D i represents the gas concentration corresponding to the i-th occurrence of the j-th gas component; D p represents the average gas concentration of the jth gas component in the reference sample data; D z Indicates the median gas concentration of the jth gas component in the reference sample data;

[0169] A sample quantity information extraction module, used to extract the sample quantity of the reference sample data;

[0170] The second sample data difference compensation coefficient acquisition module is used to acquire the second sample data difference compensation coefficient according to the sample data volume of the reference sample data and the sample data volume of the existing sample data in the current training set, wherein the second sample data difference compensation coefficient is acquired by the following formula:

[0171]

[0172] Among them, S 02 represents the second sample data difference compensation coefficient; M represents the sample data volume of the reference sample data; N represents the sample data volume of the sample data already in the current training set; m represents the type of gas components contained in the reference sample data; f cj represents the frequency of occurrence of the jth gas component in the reference sample data; f xj represents the frequency of occurrence of the jth gas component in the current training set; D cjrepresents the average concentration of the jth gas component in the reference sample data; D xj represents the average concentration of the jth gas component in the current training set;

[0173] The sample data compensation coefficient acquisition module is used to acquire the sample data compensation coefficient corresponding to each reference sample data by using the first sample data difference compensation coefficient and the second sample data difference compensation coefficient; wherein the sample data compensation coefficient corresponding to each reference sample data is acquired by the following formula:

[0174] S=(1-S 01 )·S 02

[0175] Where S represents the sample data compensation coefficient; S 01 represents the first sample data difference compensation coefficient corresponding to the reference sample data; S 02 Represents the second sample data difference compensation coefficient.

[0176] The technical effect of the above technical solution is: by introducing the difference compensation coefficient based on gas concentration data (the first sample data difference compensation coefficient and the second sample data difference compensation coefficient), the technical solution can more accurately evaluate the difference between the reference sample data and the existing training set, and adjust the reference sample data accordingly, thereby generating more diverse and representative supplementary sample data. This diversity helps the model learn a wider range of features and improves the generalization ability of the model.

[0177] The first sample data difference compensation coefficient takes into account the maximum, minimum, average, median, number of occurrences and frequency of gas concentration data, which together reflect the distribution of gas components in the reference sample data. Through the comprehensive consideration of these factors, the difference of each gas component can be more accurately evaluated and compensated accordingly. The second sample data difference compensation coefficient further considers the difference in sample data volume between the reference sample data and the training set, as well as the difference in the frequency of occurrence and average concentration of gas components in the two, thereby ensuring that the supplementary sample data can match the existing training set in both quantity and quality.

[0178] Since the supplementary sample data are precisely calculated and adjusted, they can better reflect the actual data distribution and reduce noise and bias in the model training process. Therefore, adding these supplementary sample data to the training set can accelerate the model convergence speed and improve training efficiency.

[0179] By introducing a difference compensation coefficient to adjust the reference sample data, more challenging sample data can be generated. These challenging sample data help the model learn more complex features and rules during the training process, thereby improving the model's predictive ability and robustness for unknown data.

[0180] The difference compensation coefficient calculation formula in this technical solution can be adjusted and optimized according to the specific data set and model requirements. For example, the parameters or weights in the formula can be adjusted according to different gas component characteristics or model performance indicators to adapt to different application scenarios and requirements. This customizability and flexibility makes this technical solution more practical and effective.

[0181] In summary, this technical solution optimizes the generation process of supplementary sample data by introducing a difference compensation coefficient based on gas concentration data, improves data diversity, data quality and model training efficiency, and enhances the robustness and customizability of the model. These technical effects together improve the overall performance and practicality of the model.

[0182] Among them, the test optimization module is used to test and optimize the breath component analysis model suitable for lung cancer screening, and determine the best breath component analysis model after optimization and adjustment;

[0183] In this embodiment, the test optimization module includes:

[0184] A performance test unit for testing a breath component analysis model suitable for lung cancer screening;

[0185] Obtain a breath component analysis model suitable for lung cancer screening;

[0186] Based on the test set, the performance of the breath component analysis model suitable for lung cancer screening was tested;

[0187] Determine the performance test results based on the breath composition analysis model;

[0188] An optimization and adjustment unit, used to optimize the breath component analysis model suitable for lung cancer screening;

[0189] Obtain performance test results based on breath component analysis model;

[0190] Conduct in-depth mining and correlation analysis on the performance test results based on the breath component analysis model;

[0191] Determine the optimal adjustment plan based on the breath component analysis model;

[0192] Based on the optimization and adjustment plan, the breath component analysis model suitable for lung cancer screening is optimized and adjusted;

[0193] Determine the best exhaled breath component analysis model after optimization and adjustment.

[0194] Among them, the analysis and prediction module is used to analyze and predict the characteristic data of breath components based on lung cancer screening, and determine the breath component analysis prediction results;

[0195] In this embodiment, the analysis and prediction module includes:

[0196] A data extraction unit, used for extracting characteristic data of breath components based on lung cancer screening;

[0197] According to the needs of breath component analysis for lung cancer screening, the breath component characteristic data based on lung cancer screening is extracted;

[0198] An analysis and prediction unit, used for analyzing and predicting characteristic data of breath components based on lung cancer screening;

[0199] Acquire extracted breath component characteristic data;

[0200] Inputting the breath component characteristic data into the best breath component analysis model after optimization and adjustment;

[0201] Analyze and predict the characteristic data of breath components based on the breath component analysis model;

[0202] Determine the predicted results of breath composition analysis.

[0203] Among them, the early warning guidance module is used to formulate lung cancer examination and treatment plans, and provide early warning examination and treatment guidance to users based on the lung cancer examination and treatment plans.

[0204] In this embodiment, the early warning guidance module includes:

[0205] a strategy-making unit to develop lung cancer screening and treatment plans;

[0206] Obtain the predicted results of breath component analysis;

[0207] Conduct in-depth mining and correlation analysis on the predicted results of breath component analysis;

[0208] Determine the lung cancer examination and treatment plan;

[0209] An early warning guidance unit, used to provide early warning guidance to users;

[0210] Get lung cancer screening and treatment options;

[0211] Provide early warning examination and treatment guidance to users based on lung cancer examination and treatment plans.

[0212] In this embodiment, the user is given early warning guidance by performing the following operations:

[0213] When the breath component analysis predicts that the user has abnormal breath components and is at risk of lung cancer, a warning alarm is automatically issued, and the breath component analysis prediction results are transmitted to the user in a visual form, intelligently guiding the user to proceed to the next step of examination and treatment.

[0214] Therefore, by collecting and monitoring the user's exhaled gas, the real-time data of exhaled components is determined, and by retrieving, sorting and extracting features from the real-time data of exhaled components, the characteristic data of exhaled components is determined. By training the selected appropriate model architecture, the exhaled component analysis model is determined, and the exhaled component characteristic data is analyzed and predicted based on the exhaled component analysis model to determine the exhaled component analysis prediction results, and a lung cancer examination and treatment plan is formulated. Based on the lung cancer examination and treatment plan, the user is given early warning examination and treatment guidance, the exhaled components can be effectively analyzed, the lung cancer screening situation can be accurately understood, so that the user can receive timely examination and treatment, and early detection and early diagnosis of lung cancer can be achieved, which can improve the cure rate and survival rate of lung cancer.

[0215] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0216] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A breath component analysis system for lung cancer screening, characterized in that: include: A data collection module is used to collect and monitor the user's exhaled gas and determine the real-time data of exhaled breath components based on lung cancer screening; A data processing module, used for searching, sorting and feature extraction of real-time data of breath components based on lung cancer screening, and determining feature data of breath components based on lung cancer screening; A model training module is used to train the selected appropriate model architecture and determine a breath component analysis model suitable for lung cancer screening; A test optimization module is used to test and optimize the breath component analysis model suitable for lung cancer screening, and determine the best breath component analysis model after optimization and adjustment; An analysis and prediction module is used to analyze and predict the characteristic data of breath components based on lung cancer screening, and determine the breath component analysis prediction results; The early warning guidance module is used to formulate lung cancer examination and treatment plans, and provide early warning examination and treatment guidance to users based on the lung cancer examination and treatment plans.

2. The breath component analysis system for lung cancer screening according to claim 1, characterized in that: The data acquisition module comprises: A gas collection unit, used to collect the user's exhaled gas and determine the exhaled gas component sample; A concentration monitoring unit is used to monitor the concentration of breath component samples in real time and determine breath component concentration data based on lung cancer screening; A type monitoring unit, used to monitor the types of breath component samples in real time and determine breath component type data based on lung cancer screening; Among them, based on the breath component concentration data and the breath component type data, the real-time data of the breath components based on lung cancer screening is determined.

3. The breath component analysis system for lung cancer screening according to claim 2, characterized in that: The data processing module comprises: A data retrieval unit, used for retrieving real-time data of breath components based on lung cancer screening; Obtain real-time data on breath components for lung cancer screening; Based on the sequential retrieval method, the real-time data of exhaled breath components are retrieved; Conduct consistency checks on real-time data of exhaled breath components; Check whether the real-time data of breath components meet the requirements according to the reasonable value range and mutual relationship of each variable in the real-time data of breath components; Remove data that is beyond the normal range, logically unreasonable or contradictory in the real-time data of exhaled breath components; Determine the real-time data of breath composition with data consistency; Process invalid and missing values ​​for real-time data of breath components with data consistency; Remove invalid data and missing data in the real-time data of breath components that are of no value to breath component analysis; Determine the real-time data of breath components that are valuable for breath component analysis; A data sorting unit, used for sorting the retrieved real-time data of breath components; Obtain real-time data of breath components that are valuable for breath component analysis; Sorting valuable real-time data of breath components for breath component analysis; Determine the real-time data of breath components with arrangement order; A feature extraction unit, used for extracting features from the sorted real-time data of breath components; Acquire real-time data of breath components with ranking order; Extract features from real-time data of breath components with arrangement order; Determine breath composition characteristics for lung cancer screening.

4. The breath component analysis system for lung cancer screening according to claim 3, characterized in that: The model training module includes: A data acquisition unit, used for acquiring historical data of breath components; According to the breath component analysis requirements for lung cancer screening, breath component samples are obtained from different users in different regions to determine the breath component historical data; A data division unit, used for dividing the historical data of breath components; According to the needs of breath component analysis for lung cancer screening, the historical data of breath components are divided to determine the training set and test set; The architecture selection unit is used to select the appropriate model architecture; According to the needs of breath component analysis for lung cancer screening, select a model architecture suitable for lung cancer screening and determine the appropriate model architecture; Model training unit, used to train the model architecture; Get the appropriate model architecture and training set for your selection; Based on the training set, train the selected appropriate model architecture; A breath component analysis model suitable for lung cancer screening was determined.

5. The breath component analysis system for lung cancer screening according to claim 4, characterized in that: The data partitioning unit includes: A sample quantity acquisition module is used to retrieve the sample quantity corresponding to the training set after determining the training set and the test set; The sample requirement quantity extraction module is used to extract the sample requirement quantity corresponding to model training; A comparison module, used to compare the number of samples corresponding to the training set with the number of samples required for model training; A sample difference quantity acquisition module is used to acquire the sample difference quantity according to the sample number corresponding to the training set and the sample requirement corresponding to the model training when the sample number corresponding to the training set is lower than the sample requirement corresponding to the model training; A reference sample data acquisition module, used for randomly extracting sample data corresponding to the number of sample differences from the sample data of the training set as reference sample data; A sample data compensation coefficient acquisition module, used to acquire a sample data compensation coefficient corresponding to each reference sample data according to the gas concentration data corresponding to the reference sample data and the sample quantity of the reference sample data; The supplementary sample data acquisition module is used to adjust the sample data of each reference sample data using the sample data compensation coefficient to obtain the adjusted supplementary sample data; wherein the adjusted supplementary sample data is obtained by the following formula: Wherein, D represents the adjusted gas concentration corresponding to each gas component of each reference sample data; D0 represents the gas concentration before adjustment corresponding to each gas component of each reference sample data; D min and D max They represent the minimum gas concentration value and the maximum gas concentration value of each gas component in the reference sample data respectively; S represents the sample data compensation coefficient; The supplement execution module is used to add the adjusted supplementary sample data to the training set.

6. The breath component analysis system for lung cancer screening according to claim 5, characterized in that: The sample data compensation coefficient acquisition module includes: A gas concentration data extraction module, used to extract gas concentration data corresponding to the reference sample data; The first sample data difference compensation coefficient acquisition module is used to acquire the first sample data difference compensation coefficient corresponding to each gas component of the reference sample data according to the gas concentration data corresponding to the reference sample data, wherein the first sample data difference compensation coefficient is acquired by the following formula: Among them, S 01 represents the first sample data difference compensation coefficient corresponding to the reference sample data; D maxj and D minj They represent the maximum and minimum gas concentrations of the j-th gas component contained in the reference sample data; m represents the type of gas component contained in the reference sample data; n represents the number of times the j-th gas component appears in the reference sample data; D i represents the gas concentration corresponding to the i-th occurrence of the j-th gas component; D p represents the average gas concentration of the jth gas component in the reference sample data; D z Indicates the median gas concentration of the jth gas component in the reference sample data; A sample quantity information extraction module, used to extract the sample quantity of the reference sample data; The second sample data difference compensation coefficient acquisition module is used to acquire the second sample data difference compensation coefficient according to the sample data volume of the reference sample data and the sample data volume of the existing sample data in the current training set, wherein the second sample data difference compensation coefficient is acquired by the following formula: Among them, S 02 represents the second sample data difference compensation coefficient; M represents the sample data volume of the reference sample data; N represents the sample data volume of the sample data already in the current training set; m represents the type of gas components contained in the reference sample data; f cj represents the frequency of occurrence of the jth gas component in the reference sample data; f xj represents the frequency of occurrence of the jth gas component in the current training set; D cj represents the average concentration of the jth gas component in the reference sample data; D xj represents the average concentration of the jth gas component in the current training set; The sample data compensation coefficient acquisition module is used to acquire the sample data compensation coefficient corresponding to each reference sample data by using the first sample data difference compensation coefficient and the second sample data difference compensation coefficient; wherein the sample data compensation coefficient corresponding to each reference sample data is acquired by the following formula: S=(1-S 01 )·S 02 Where S represents the sample data compensation coefficient; S 01 represents the first sample data difference compensation coefficient corresponding to the reference sample data; S 02 Represents the second sample data difference compensation coefficient.

7. The breath component analysis system for lung cancer screening according to claim 4, characterized in that: The test optimization module includes: A performance test unit for testing breath component analysis models suitable for lung cancer screening; Obtain a breath component analysis model suitable for lung cancer screening; Based on the test set, the performance of the breath component analysis model suitable for lung cancer screening was tested; Determine the performance test results based on the breath composition analysis model; An optimization and adjustment unit, used to optimize the breath component analysis model suitable for lung cancer screening; Obtain performance test results based on breath component analysis model; Conduct in-depth mining and correlation analysis on the performance test results based on the breath component analysis model; Determine the optimal adjustment plan based on the breath component analysis model; Based on the optimization and adjustment plan, the breath component analysis model suitable for lung cancer screening is optimized and adjusted; Determine the best exhaled breath component analysis model after optimization and adjustment.

8. The breath component analysis system for lung cancer screening according to claim 7, characterized in that: The analysis and prediction module comprises: A data extraction unit, used for extracting breath component characteristic data based on lung cancer screening; According to the needs of breath component analysis for lung cancer screening, the breath component characteristic data based on lung cancer screening is extracted; An analysis and prediction unit, used for analyzing and predicting characteristic data of breath components based on lung cancer screening; Acquire extracted breath component characteristic data; Inputting the breath component characteristic data into the best breath component analysis model after optimization and adjustment; Analyze and predict the characteristic data of breath components based on the breath component analysis model; Determine the predicted results of breath composition analysis.

9. The breath component analysis system for lung cancer screening according to claim 8, characterized in that: The early warning guidance module includes: a strategy-making unit, which is used to develop lung cancer screening and treatment plans; Obtain the predicted results of breath component analysis; Conduct in-depth mining and correlation analysis on the predicted results of breath component analysis; Determine the lung cancer examination and treatment plan; An early warning guidance unit, used to provide early warning guidance to users; Get lung cancer screening and treatment options; Provide early warning examination and treatment guidance to users based on lung cancer examination and treatment plans.

10. The breath component analysis system for lung cancer screening according to claim 9, characterized in that: Provide early warning guidance to users and perform the following operations: When the breath component analysis predicts that the user has abnormal breath components and is at risk of lung cancer, a warning alarm is automatically issued, and the breath component analysis prediction results are transmitted to the user in a visual form, intelligently guiding the user to proceed to the next step of examination and treatment.

Images