Pulmonary nodule risk prediction system and method for chronic obstructive pulmonary disease patient

Abstract

The invention relates to the technical field of medical care information processing, and particularly discloses a pulmonary nodule risk prediction system and method for a patient with chronic obstructive pulmonary disease, and the system comprises a historical examination data acquisition module which collects multiple CT examination data, lung function data and lung quantitative data of the patient with chronic obstructive pulmonary disease; the growth inflection point recognition module recognizes a pulmonary nodule growth inflection point of the patient; the risk prediction model building module builds a pulmonary nodule risk prediction model; the current patient risk prediction module inputs the current CT examination data, the current lung function data and the current lung quantitative data of the to-be-detected patient into the pulmonary nodule risk prediction model to obtain a current pulmonary nodule risk prediction result of the to-be-detected patient; the review time recommendation module obtains review recommendation time of the patient to be detected based on the current pulmonary nodule risk prediction result of the patient to be detected and a preset review time recommendation model, effective prediction of the pulmonary nodule risk of the patient with chronic obstructive pulmonary disease is achieved, and an important basis is provided for early diagnosis and treatment.

Description

Technical Field

[0001] The present invention relates to the technical field of medical care information processing, and in particular to a system and method for predicting the risk of pulmonary nodules in patients with chronic obstructive pulmonary disease. Background Art

[0002] Chronic obstructive pulmonary disease (COPD) is a common chronic respiratory disease with a large number of patients, which has brought a heavy burden to the medical system and society. In COPD patients, the appearance of lung nodules is an issue that needs close attention because it may be an early manifestation of lung cancer. Accurate prediction of the risk of lung nodules in COPD patients can help to take timely intervention measures to improve treatment effects and patient survival rates. With the continuous advancement of medical technology, CT examinations, pulmonary function tests and other methods have been widely used in the diagnosis and treatment of COPD patients, and can provide a large amount of valuable data. Through in-depth analysis of these data, we can understand the growth and change patterns of lung nodules, the patient's lung function status and the quantitative characteristics of the lungs.

[0003] However, the use of these data is not yet sufficient and systematic. In clinical practice, doctors usually rely on experience and limited data to assess the risk of lung nodules in COPD patients. This method is highly subjective and lacks unified standards and accurate quantitative analysis. At the same time, due to the lack of effective prediction tools, the prediction results cannot be updated in a timely manner based on the patient's latest examination data, and it is difficult to reflect the real-time changes in the disease. It is difficult to accurately identify high-risk patients in the early stages, causing some patients to miss the best time for treatment.

[0004] Therefore, the present invention proposes a system and method for predicting the risk of pulmonary nodules in patients with COPD. Summary of the invention

[0005] The present invention provides a system and method for predicting the risk of lung nodules in patients with COPD, which is used to collect a large amount of multi-type historical data based on a historical examination data acquisition module, providing a rich and comprehensive data basis for subsequent analysis. The growth inflection point identification module can accurately find the key turning points of lung nodule growth, which helps to have a deeper understanding of the development law of lung nodules. The risk prediction model building module comprehensively utilizes a variety of data to build a prediction model, which improves the accuracy and reliability of the model. The current patient risk prediction module can input the latest data of the patient to be tested into the model to quickly obtain the prediction results of its lung nodule risk. The system realizes the effective prediction of the risk of lung nodules in patients with COPD and the reasonable inference of the re-examination time, which provides an important basis for early diagnosis and treatment, helps to improve the treatment effect and quality of life of patients, and reduce the harm of the disease.

[0006] The present invention provides a pulmonary nodule risk prediction system for patients with chronic obstructive pulmonary disease, comprising: The historical examination data acquisition module is used to collect multiple CT examination data, lung function data, and lung quantitative data of a large number of historical COPD patients at different time points; A growth inflection point identification module is used to identify all lung nodule growth inflection points of each historical COPD patient in all CT examination data of each historical COPD patient; The risk prediction model building module is used to build a lung nodule risk prediction model based on the first CT examination data and the CT examination data at all lung nodule growth inflection points of each historical COPD patient, lung function data, and lung quantitative data; The current patient risk prediction module is used to input the current CT examination data, current lung function data and current lung quantitative data of the patient to be tested into the lung nodule risk prediction model to obtain the current lung nodule risk prediction result of the patient to be tested; The review time recommendation module is used to obtain the recommended review time for the patient to be tested based on the current lung nodule risk prediction result of the patient to be tested and the preset review time recommendation model.

[0007] Preferably, the historical inspection data acquisition module includes: A historical patient screening submodule is used to collect a preset number of historical COPD patients whose number of completed complete lung examinations is not less than a preset number and whose maximum interval between all different complete lung examinations is not less than a preset time interval; The examination data collection submodule is used to collect CT examination data, lung function data and lung quantitative data at the time points of all complete lung examinations completed by all historical COPD patients.

[0008] Preferably, the growth inflection point identification module includes: The CT examination data analysis submodule is used to determine the average CT value of all pixels in the lung nodule area in each CT examination data of each historical COPD patient as the average CT value of each CT examination data, and analyze the lung nodule volume in each CT examination data of each historical COPD patient; A pulmonary nodule mass calculation submodule is used to calculate the pulmonary nodule mass of each CT examination data of each historical COPD patient based on the average CT value and pulmonary nodule volume of each CT examination data of each historical COPD patient; The mass growth ratio calculation submodule is used to calculate the growth ratio of the pulmonary nodule mass of each CT examination data other than the first CT examination data of each historical COPD patient relative to the pulmonary nodule mass of the first CT examination data; The growth inflection point analysis submodule is used to treat the time points corresponding to all CT examination data of each historical COPD patient whose growth ratio of the lung nodule mass relative to the first CT examination data is not less than a preset percentage as the growth inflection points of all lung nodules corresponding to each historical COPD patient.

[0009] Preferably, the pulmonary nodule mass calculation submodule calculates the pulmonary nodule mass of each CT examination data of each historical COPD patient based on the average CT value and pulmonary nodule volume of each CT examination data of each historical COPD patient, comprising: ; Wherein, nodule mass is the mass of the pulmonary nodules in the current CT examination data of the historical COPD patient currently calculated, v is the volume of the pulmonary nodules in the current CT examination data of the historical COPD patient currently calculated, The average CT value of the current CT examination data of historical COPD patients calculated currently.

[0010] Preferably, the risk prediction model building module includes: The historical patient risk prediction submodule is used to obtain the lung nodule risk prediction results of each historical COPD patient at each lung nodule growth inflection point based on the first CT examination data and the CT examination data of each lung nodule growth inflection point, lung function data, and lung quantitative data of each historical COPD patient; The training sample arrangement submodule is used to use the first CT examination data and the CT examination data at each pulmonary nodule growth inflection point of each historical COPD patient, pulmonary function data, and pulmonary quantitative data as the input of the model training samples, and use the pulmonary nodule risk prediction results of the corresponding historical COPD patients at the corresponding pulmonary nodule growth inflection point as the model output to generate multiple training samples; The model training submodule is used to train the model based on all training samples and build a lung nodule risk prediction model.

[0011] Preferably, the historical patient risk prediction submodule includes: a doubling time determination unit, used to regard the time interval between each pulmonary nodule growth inflection point of each historical COPD patient and the corresponding time point of the first CT examination data as the doubling time of each pulmonary nodule growth inflection point of each historical COPD patient; A first risk prediction unit is used to retrieve a preset doubling time-first pulmonary nodule risk prediction result corresponding list based on the doubling time, and determine the first pulmonary nodule risk prediction result of each historical COPD patient at each pulmonary nodule growth inflection point; A second risk prediction unit is used to determine the second pulmonary nodule risk prediction result of each historical COPD patient at each pulmonary nodule growth inflection point based on the pulmonary function data of each historical COPD patient at each pulmonary nodule growth inflection point; A third risk prediction unit is used to determine the third lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point based on the lung quantitative data of each historical COPD patient at each lung nodule growth inflection point; The comprehensive risk prediction unit is used to obtain the lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point based on the first lung nodule risk prediction result, the second lung nodule risk prediction result, and the third lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point.

[0012] Preferably, the second risk prediction unit includes: A subunit for determining a reference value of the same level, for determining a reference value of the forced expiratory volume in the first second of each historical COPD patient based on the basic physiological parameters of each historical COPD patient; A step division threshold determination subunit is used to determine the step division threshold of the second pulmonary nodule risk prediction level for each historical COPD patient based on the forced expiratory volume in the first second reference value of the same level and the preset step percentage of each historical COPD patient; The third risk prediction subunit is used to determine the forced expiratory volume in one second of each historical COPD patient at each lung nodule growth inflection point based on the lung function data of each historical COPD patient at each lung nodule growth inflection point, and to determine the second lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point based on the forced expiratory volume in one second of each historical COPD patient at each lung nodule growth inflection point and the corresponding second lung nodule risk prediction level step division threshold.

[0013] Preferably, the same-level reference value determination subunit includes: A physiological parameter value determination terminal is used to determine the parameter value of each physiological parameter in the basic physiological parameters of each historical COPD patient; The parameter deviation calculation end is used to calculate the deviation between the basic physiological parameters of each healthy patient in the healthy patient database and the basic physiological parameters of each historical COPD patient; The peer patient screening terminal is used to treat all healthy patients in the healthy patient database whose basic physiological parameters do not exceed the deviation threshold with those of a single historical COPD patient as all peer patients of the corresponding historical COPD patient; The same-level reference value calculation terminal is used to calculate the same-level reference value of the forced expiratory volume in the first second of each historical COPD patient based on the basic physiological parameters of each historical COPD patient and the basic physiological parameters of all patients at the same level and the forced expiratory volume in the first second: ; In the formula, is the reference value of the forced expiratory volume in the first second of the historical COPD patient currently calculated, n is the total number of all patients of the same level of the historical COPD patient currently calculated, is the forced expiratory volume in the first second of the i-th patient of the same level of the historical COPD patient currently calculated, m is the total number of physiological parameters included in the basic physiological parameters, is the parameter value of the jth physiological parameter of the i-th patient of the historical COPD patient currently calculated, is the parameter value of the jth physiological parameter of the historical COPD patient currently calculated, is the parameter limit value of the jth physiological parameter, is the negative impact ratio of the parameter attenuation value of the jth physiological parameter on the forced expiratory volume in the first second.

[0014] Preferably, the third risk prediction unit includes: The pulmonary quantitative data interpretation subunit is used to determine the percentage of low attenuation area, percentage of high attenuation area, and average density value of the whole lung at each pulmonary nodule growth inflection point of each historical COPD patient based on the pulmonary quantitative data of each historical COPD patient at each pulmonary nodule growth inflection point; The third risk prediction subunit is used to retrieve a preset list of corresponding lung quantitative data-third lung nodule risk prediction results based on the percentage of low attenuation area, percentage of high attenuation area, and average density value of the whole lung at each lung nodule growth inflection point of each historical COPD patient, and determine the third lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point.

[0015] The present invention provides a method for predicting the risk of pulmonary nodules in patients with chronic obstructive pulmonary disease, which is applied to any of the above-mentioned systems for predicting the risk of pulmonary nodules in patients with chronic obstructive pulmonary disease, comprising: S1: Collect multiple CT examination data, lung function data and lung quantitative data of a large number of historical COPD patients at different time points; S2: Identify the inflection point of lung nodule growth in each historical COPD patient in all CT examination data; S3: A lung nodule risk prediction model was built based on the first CT examination data and the CT examination data of the lung nodule growth inflection point, lung function data, and lung quantitative data of each historical COPD patient; S4: inputting the current CT examination data, current lung function data and current lung quantitative data of the patient to be tested into the lung nodule risk prediction model to obtain the current lung nodule risk prediction result of the patient to be tested; S5: Based on the current lung nodule risk prediction result of the patient to be tested and the preset reexamination time recommendation model, the recommended reexamination time of the patient to be tested is obtained.

[0016] The beneficial effects of the present invention compared to the prior art are as follows: a large amount of multi-type historical data is collected based on the historical examination data acquisition module, providing a rich and comprehensive data basis for subsequent analysis. The growth inflection point identification module can accurately find the key turning points of lung nodule growth, which helps to have a deeper understanding of the development law of lung nodules. The risk prediction model building module comprehensively utilizes a variety of data to build a prediction model, which improves the accuracy and reliability of the model. The current patient risk prediction module can input the latest data of the patient to be tested into the model to quickly obtain the lung nodule risk prediction results. The system realizes the effective prediction of the lung nodule risk of patients with COPD and the reasonable inference of the re-examination time, which provides an important basis for early diagnosis and treatment, helps to improve the treatment effect and quality of life of patients, and reduce the harm of the disease.

[0017] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures specifically pointed out in this application document.

[0018] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 Schematic diagram of a system for predicting lung nodules risk in patients with chronic obstructive pulmonary disease in an embodiment of the present invention; Figure 2 This is a flow chart of a method for predicting the risk of lung nodules in patients with COPD in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0021] Example 1: The present invention provides a system for predicting the risk of pulmonary nodules in patients with chronic obstructive pulmonary disease. Figure 1 ,include: The historical examination data acquisition module is used to collect multiple CT examination data, lung function data, and lung quantitative data of a large number of historical COPD patients at different time points; A growth inflection point identification module is used to identify all lung nodule growth inflection points of each historical COPD patient in all CT examination data of each historical COPD patient; The risk prediction model building module is used to build a lung nodule risk prediction model based on the first CT examination data and the CT examination data at all lung nodule growth inflection points of each historical COPD patient, lung function data, and lung quantitative data; The current patient risk prediction module is used to input the current CT examination data, current lung function data and current lung quantitative data of the patient to be tested into the lung nodule risk prediction model to obtain the current lung nodule risk prediction result of the patient to be tested; The review time recommendation module is used to obtain the recommended review time for the patient to be tested based on the current lung nodule risk prediction result of the patient to be tested and the preset review time recommendation model.

[0022] In this embodiment, historical COPD patients refer to patients who have been diagnosed with COPD in the past and have relevant examination data records.

[0023] In this embodiment, the CT examination data refers to the imaging information about the lungs obtained by computer tomography technology, including the location, size, shape, etc. of lung nodules.

[0024] In this embodiment, the lung function data is data obtained by testing the patient's lung function, such as the GOLD classification based on the FEV1 percentage predicted value, etc., which is used to evaluate the patient's respiratory function status.

[0025] In this embodiment, the lung quantitative data includes whole lung LAA% (to assess the severity of emphysema), whole lung HAA% (to assess lung lesions), whole lung MLD (to assess disease progression), and other data obtained through specific measurement and calculation methods.

[0026] In this embodiment, the inflection point of lung nodule growth refers to the time point when a significant change in the lung nodule mass growth ratio is found during the analysis of multiple CT examination data of historical COPD patients.

[0027] In this embodiment, the lung nodule risk prediction model is a mathematical model built based on relevant examination data and analysis results of historical COPD patients, and is used to predict the risk of lung nodules in new patients.

[0028] In this embodiment, the patient to be tested is a COPD patient for whom the lung nodule risk prediction result has not yet been obtained and whose data needs to be input into the model for prediction.

[0029] In this embodiment, the current CT examination data refers to the lung image data obtained by the current CT examination of the patient to be tested.

[0030] In this embodiment, the current lung function data refers to relevant data reflecting the lung function of the patient to be tested obtained at the current time point.

[0031] In this embodiment, the current lung quantitative data refers to the data of the patient to be tested that is currently measured and calculated, such as whole lung LAA%, whole lung HAA%, whole lung MLD, etc.

[0032] In this embodiment, the current pulmonary nodule risk prediction result is an assessment result of the pulmonary nodule risk of the patient to be tested obtained by inputting the current relevant data of the patient to be tested into the pulmonary nodule risk prediction model.

[0033] In this embodiment, the preset review time recommendation model is a pre-established rule or algorithm system for recommending patient review time based on specific conditions or input information. For example, this model may comprehensively consider multiple factors such as the size, morphology, growth rate, age, and health status of the patient's pulmonary nodules, and give a more reasonable and scientific review time recommendation through complex calculations and judgment logic, such as three months later, half a year later, etc. The training samples of the model include a large number of known patients' pulmonary nodule-related data, such as detailed characteristics of nodules (size, shape, density, edge conditions, etc.), basic information of patients (age, gender, medical history, living habits, etc.), previous diagnosis results and risk assessments, and corresponding actual review time arrangements. These samples should cover various types and risk levels of pulmonary nodules, as well as patients with different physical conditions and backgrounds, to ensure that the model can learn comprehensive and accurate patterns and rules, so as to make reasonable review time recommendations for new patients to be tested.

[0034] In this embodiment, based on the current lung nodule risk prediction result of the patient to be tested and the preset review time recommendation model, the recommended review time for the patient to be tested is obtained, that is, the current lung nodule risk prediction result of the patient to be tested is input into the preset review time recommendation model to obtain the recommended review time for the patient to be tested; according to the predicted current lung nodule risk of the patient to be tested, the pre-set review time recommendation model is used to determine the time recommendation suitable for the review of the patient to be tested.

[0035] For example, if the prediction results show that the patient has a higher risk of lung nodules, the preset re-examination time recommendation model may give a recommendation for re-examination in a shorter period of time (such as one month later); if the risk is lower, the re-examination recommendation time may be longer (such as half a year later).

[0036] The beneficial effects of the above technologies are: based on the historical examination data acquisition module, a large amount of multi-type historical data is collected, providing a rich and comprehensive data basis for subsequent analysis. The growth inflection point identification module can accurately find the key turning points of lung nodule growth, which helps to more deeply understand the development laws of lung nodules. The risk prediction model building module comprehensively utilizes a variety of data to build a prediction model, which improves the accuracy and reliability of the model. The current patient risk prediction module can input the latest data of the patient to be tested into the model to quickly obtain the prediction results of their lung nodule risk. The system realizes the effective prediction of the risk of lung nodules in patients with COPD and the reasonable inference of the re-examination time, which provides an important basis for early diagnosis and treatment, helps to improve the treatment effect and quality of life of patients, and reduce the harm of the disease.

[0037] Embodiment 2: Based on Embodiment 1, the historical inspection data acquisition module includes: A historical patient screening submodule is used to collect a preset number of historical COPD patients whose number of completed complete lung examinations is not less than a preset number and whose maximum interval between all different complete lung examinations is not less than a preset time interval; The examination data collection submodule is used to collect CT examination data, lung function data and lung quantitative data at the time points of all complete lung examinations completed by all historical COPD patients.

[0038] In this embodiment, a complete lung examination refers to a comprehensive examination of the lungs, such as a CT examination, a lung function test, and related quantitative data measurements.

[0039] In this embodiment, the preset number of times is the minimum number of times a historical COPD patient must complete a complete lung examination.

[0040] In this embodiment, the maximum interval time of all different complete lung examinations refers to the maximum value of the time intervals between two different examinations in multiple complete lung examinations.

[0041] In this embodiment, the preset time interval is a preset value that the time interval between at least one set of two complete lung examinations of a patient with historical COPD must reach.

[0042] In this embodiment, the preset number is a predetermined number of historical COPD patients that need to be collected.

[0043] The beneficial effects of the above technologies are as follows: the historical patient screening submodule ensures that the selected historical COPD patient data is sufficiently complete and representative by setting screening conditions. This can exclude patients with incomplete data or unreasonable intervals, and improve the accuracy and reliability of subsequent analysis. The examination data collection submodule can comprehensively collect a variety of examination data of screened patients at different time points. It provides rich and high-quality data support for the subsequent identification of pulmonary nodule growth inflection points and the construction of risk prediction models. It ensures the quality and effectiveness of historical examination data, which helps to improve the performance and practical value of the pulmonary nodule risk prediction system.

[0044] Embodiment 3: Based on Embodiment 1, a growth inflection point identification module includes: The CT examination data analysis submodule is used to determine the average CT value of all pixels in the lung nodule area in each CT examination data of each historical COPD patient as the average CT value of each CT examination data, and analyze the lung nodule volume in each CT examination data of each historical COPD patient; A pulmonary nodule mass calculation submodule is used to calculate the pulmonary nodule mass of each CT examination data of each historical COPD patient based on the average CT value and pulmonary nodule volume of each CT examination data of each historical COPD patient; The mass growth ratio calculation submodule is used to calculate the growth ratio of the pulmonary nodule mass of each CT examination data other than the first CT examination data of each historical COPD patient relative to the pulmonary nodule mass of the first CT examination data; The growth inflection point analysis submodule is used to treat the time points corresponding to all CT examination data of each historical COPD patient whose growth ratio of the lung nodule mass relative to the first CT examination data is not less than a preset percentage as the growth inflection points of all lung nodules corresponding to each historical COPD patient.

[0045] In this embodiment, the average CT value of all pixels in the lung nodule area determined in each CT examination data of each historical COPD patient refers to the average value obtained by summing the CT value of each pixel in the lung nodule area in each CT examination image and dividing it by the number of pixels.

[0046] In this embodiment, the pulmonary nodule region refers to a specific range occupied by the pulmonary nodules displayed in the CT image.

[0047] In this embodiment, the CT value of a pixel is a CT scan measurement value of a tissue represented by each pixel in a CT image, and reflects the density characteristics of the tissue.

[0048] In this embodiment, analyzing the volume of lung nodules in each CT examination data of each historical COPD patient means calculating the size of the space occupied by the lung nodules in each CT examination by a specific image analysis method.

[0049] In this embodiment, the quality of the lung nodule is a quality parameter representing the lung nodule that is calculated based on a specific formula in combination with the volume and average CT value of the lung nodule.

[0050] In this embodiment, the growth ratio of the lung nodule mass of each CT examination data except the first CT examination data of each historical COPD patient relative to the lung nodule mass of the first CT examination data is calculated by comparing the lung nodule mass of each subsequent examination with the lung nodule mass of the first examination to calculate the proportion of mass increase.

[0051] In this embodiment, the preset percentage is a pre-set standard value of the mass growth ratio used to determine whether a lung nodule has reached a growth inflection point.

[0052] The beneficial effects of the above technologies are as follows: The CT examination data analysis submodule determines the average CT value and volume of lung nodules, providing basic data for subsequent calculations. The lung nodule mass calculation submodule calculates the mass of lung nodules by the average CT value and volume, making the description of lung nodules more accurate. The mass growth ratio calculation submodule calculates the mass growth ratio, which can intuitively reflect the degree of development and change of lung nodules. The growth inflection point analysis submodule determines the growth inflection point by setting the growth ratio threshold, and accurately identifies the key nodes of lung nodule growth. This comprehensive and precise analysis method can effectively identify the growth inflection point of lung nodules, provide key information for accurate risk prediction, and help improve the accuracy and reliability of lung nodule risk prediction in patients with COPD.

[0053] Example 4: Based on Example 3, the pulmonary nodule mass calculation submodule calculates the pulmonary nodule mass of each CT examination data of each historical COPD patient based on the average CT value and pulmonary nodule volume of each CT examination data of each historical COPD patient, comprising: ; Wherein, nodule mass is the mass of the pulmonary nodules in the current CT examination data of the historical COPD patient currently calculated, v is the volume of the pulmonary nodules in the current CT examination data of the historical COPD patient currently calculated, The average CT value of the current CT examination data of historical COPD patients calculated currently.

[0054] The beneficial effects of the above technology are: through a clear calculation formula, the lung nodule mass of each historical COPD patient's CT examination data can be accurately and quantitatively calculated. This quantitative calculation method provides accurate data support for the subsequent analysis of the growth changes of lung nodules. It makes the calculation of lung nodule mass scientific and consistent, reducing calculation errors and uncertainties. It helps to judge the development trend of lung nodules more objectively and accurately, thereby improving the accuracy and reliability of lung nodule risk prediction. It provides a strong basis for medical personnel to formulate more accurate treatment plans and decisions.

[0055] Embodiment 5: Based on Embodiment 1, a risk prediction model building module includes: The historical patient risk prediction submodule is used to obtain the lung nodule risk prediction results of each historical COPD patient at each lung nodule growth inflection point based on the first CT examination data and the CT examination data of each lung nodule growth inflection point, lung function data, and lung quantitative data of each historical COPD patient; The training sample arrangement submodule is used to use the first CT examination data and the CT examination data at each pulmonary nodule growth inflection point of each historical COPD patient, pulmonary function data, and pulmonary quantitative data as the input of the model training samples, and use the pulmonary nodule risk prediction results of the corresponding historical COPD patients at the corresponding pulmonary nodule growth inflection point as the model output to generate multiple training samples; The model training submodule is used to train the model based on all training samples and build a lung nodule risk prediction model.

[0056] In this embodiment, the lung nodule risk prediction result of a historical COPD patient at each lung nodule growth inflection point refers to the predicted conclusion of the risk status of the lung nodules for patients who have been diagnosed with COPD when the growth of their lung nodules has an inflection point. For example, suppose we have a historical COPD patient A. In the first CT examination, it was found that the volume of his lung nodules was 5mm³ and the average CT value was 40HU. After a period of time, at the second examination (i.e., a lung nodule growth inflection point), the nodule volume became 8mm³ and the average CT value became 50HU. According to the preset evaluation criteria and algorithms, the analysis shows that the patient's first lung nodule risk prediction result at this growth inflection point is "medium risk". At the same time, the patient's lung function data at this inflection point shows GOLD2 moderate, and according to the corresponding judgment rules, the second lung nodule risk prediction result is "some risk". In addition, the lung quantitative data, such as the whole lung LAA% is 20%, the whole lung HAA% is 10%, and the whole lung MLD is -800HU. By comparing with the standard, the third lung nodule risk prediction result is "need to be closely watched". Combining these three results, the final prediction result of the lung nodule risk of patient A with a history of COPD at this inflection point of lung nodule growth was "high risk, requiring close follow-up and further examination."

[0057] In this embodiment, the first CT examination data and the CT examination data at each lung nodule growth inflection point of each historical COPD patient, lung function data and lung quantitative data are used as input samples of the model training, and the lung nodule risk prediction results of the corresponding historical COPD patients at the corresponding lung nodule growth inflection points are used as model outputs. Generating multiple training samples means taking each patient's initial examination data and the relevant data at each lung nodule growth inflection point (such as CT, lung function, lung quantitative data) as input information, and taking the lung nodule risk prediction results obtained at these inflection points as output information, thereby forming multiple data combinations for training the model.

[0058] In this embodiment, model training is performed based on all training samples to build a lung nodule risk prediction model, which means utilizing the numerous training samples generated previously to optimize and adjust the model through specific algorithms and processes, so that it can accurately predict the risk of lung nodules based on the input patient data, thereby building an effective prediction model.

[0059] The beneficial effects of the above technologies are as follows: the historical patient risk prediction submodule can perform risk prediction for each historical COPD patient at different time points, providing rich reference data for subsequent model training. The training sample organization submodule takes multiple types of data as input and risk prediction results as output, effectively organizing and organizing training samples to make the data more standardized and usable. The model training construction submodule performs model training based on a large number of training samples to build a reliable lung nodule risk prediction model. Through such steps, the data of historical patients can be fully utilized to improve the accuracy and generalization ability of the model. It helps to more accurately predict the risk of lung nodules in patients to be tested, and provide strong support for disease management and treatment of COPD patients.

[0060] Example 6: Based on Example 5, the historical patient risk prediction submodule includes: a doubling time determination unit, used to regard the time interval between each pulmonary nodule growth inflection point of each historical COPD patient and the corresponding time point of the first CT examination data as the doubling time of each pulmonary nodule growth inflection point of each historical COPD patient; A first risk prediction unit is used to retrieve a preset doubling time-first pulmonary nodule risk prediction result corresponding list based on the doubling time, and determine the first pulmonary nodule risk prediction result of each historical COPD patient at each pulmonary nodule growth inflection point; A second risk prediction unit is used to determine the second pulmonary nodule risk prediction result of each historical COPD patient at each pulmonary nodule growth inflection point based on the pulmonary function data of each historical COPD patient at each pulmonary nodule growth inflection point; A third risk prediction unit is used to determine the third lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point based on the lung quantitative data of each historical COPD patient at each lung nodule growth inflection point; The comprehensive risk prediction unit is used to obtain the lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point based on the first lung nodule risk prediction result, the second lung nodule risk prediction result, and the third lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point.

[0061] In this embodiment, the preset doubling time-first pulmonary nodule risk prediction result correspondence list refers to a pre-set list that associates different doubling times with corresponding first pulmonary nodule risk prediction results.

[0062] For example, a doubling time of less than 30 days is an extremely short doubling time, and the corresponding risk prediction result of the first lung nodule is highly suspected of malignancy; a doubling time of 30-100 days is a short doubling time, and the corresponding risk prediction result of the first lung nodule is possibly malignant; a doubling time of 100-400 days is a medium doubling time, and the corresponding risk prediction result of the first lung nodule is that close follow-up is required; a doubling time of more than 400 days is a long doubling time, and the corresponding risk prediction result of the first lung nodule is possibly benign.

[0063] In this embodiment, the first lung nodule risk prediction result is a predicted conclusion about the lung nodule risk based on the doubling time. Assume that the doubling time is 60 days. According to the preset doubling time-first lung nodule risk prediction result corresponding list, a doubling time of 30-100 days is considered "likely to be malignant", so the first lung nodule risk prediction result is "likely to be malignant".

[0064] In this embodiment, the second pulmonary nodule risk prediction result is a pulmonary nodule risk prediction result obtained based on the pulmonary function data of each historical COPD patient at each pulmonary nodule growth inflection point.

[0065] In this embodiment, the third lung nodule risk prediction result is a lung nodule risk prediction result obtained based on the lung quantitative data of each historical COPD patient at each lung nodule growth inflection point.

[0066] In this embodiment, based on the first lung nodule risk prediction result, the second lung nodule risk prediction result, and the third lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point, the lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point is obtained, which means comprehensively considering the three risk prediction results based on doubling time, lung function data, and lung quantitative data, respectively, so as to obtain a more comprehensive and accurate final lung nodule risk prediction result for each historical COPD patient at each lung nodule growth inflection point. For example, suppose there is a historical COPD patient A at a certain lung nodule growth inflection point: The risk prediction result of the first pulmonary nodule was "short doubling time, possibly malignant"; The second lung nodule risk prediction result was "poor lung function, high risk" because his lung function test was GOLD3 severe and FEV1 <50% predicted value; The third lung nodule risk prediction result was "the whole lung LAA% is high, and there is a certain risk of lesions", and the whole lung LAA% was 28%; Combining these three results: "Possible malignancy" indicates that the nodule grows faster and may be malignant; "Higher risk" indicates that poor lung function may increase the risk; "There is a certain risk of lesions" indicates that the overall condition of the lungs is not very good. After comprehensive consideration, the final prediction result of the patient's lung nodule risk at this turning point of lung nodule growth is "Overall high risk, further detailed examination and diagnosis are required as soon as possible, and nodule changes should be closely monitored."

[0067] The beneficial effects of the above technologies are as follows: the doubling time determination unit obtains the doubling time by calculating the time interval, which provides an important time dimension indicator for risk prediction. The first risk prediction unit retrieves the corresponding list based on the doubling time to determine the first risk prediction result, and quickly obtains a preliminary risk assessment. The second risk prediction unit performs risk prediction based on lung function data, supplementing the risk assessment information from a functional perspective. The third risk prediction unit performs predictions based on quantitative lung data, further enriching the basis for risk assessment. The comprehensive risk prediction unit integrates the prediction results from multiple aspects to make the final risk prediction more comprehensive and accurate. Through multi-dimensional risk prediction and comprehensive analysis, the risk of historical COPD patients at the inflection point of lung nodule growth can be more accurately assessed, providing a high-quality data foundation for building an effective risk prediction model, which helps to improve the accuracy and reliability of lung nodule risk prediction for COPD patients.

[0068] Embodiment 7: Based on Embodiment 6, the second risk prediction unit includes: A subunit for determining a reference value of the same level, for determining a reference value of the forced expiratory volume in the first second of each historical COPD patient based on the basic physiological parameters of each historical COPD patient; A step division threshold determination subunit is used to determine the step division threshold of the second pulmonary nodule risk prediction level for each historical COPD patient based on the forced expiratory volume in the first second reference value of the same level and the preset step percentage of each historical COPD patient; The third risk prediction subunit is used to determine the forced expiratory volume in one second of each historical COPD patient at each lung nodule growth inflection point based on the lung function data of each historical COPD patient at each lung nodule growth inflection point, and to determine the second lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point based on the forced expiratory volume in one second of each historical COPD patient at each lung nodule growth inflection point and the corresponding second lung nodule risk prediction level step division threshold.

[0069] In this embodiment, basic physiological parameters refer to some physiological indicators that can reflect the basic physical condition of the patient, such as age, height, weight, gender, etc.

[0070] In this embodiment, the forced expiratory volume in one second peer reference value is a reference value obtained by calculating the forced expiratory volume in one second of healthy patients in a healthy patient database whose basic physiological parameters are similar to those of historical COPD patients.

[0071] In this embodiment, the preset step percentage is a pre-set percentage value for dividing the second pulmonary nodule risk prediction level, including: 30%, 50%, and 80%.

[0072] In this embodiment, the threshold for the step-by-step division of the second lung nodule risk prediction level for patients with a history of COPD is calculated based on the preset step percentage and the reference value of the same level of forced expiratory volume in one second. The boundary values ​​used to divide different risk levels are 30%, 50%, and 80%, respectively, multiplied by the reference value of the same level of forced expiratory volume in one second.

[0073] In this embodiment, based on the lung function data of each historical COPD patient at each lung nodule growth inflection point, the forced expiratory volume in the first second of each historical COPD patient at each lung nodule growth inflection point is determined, that is, by analyzing the patient's lung function test data at a specific time point, the forced expiratory volume in the first second is obtained.

[0074] In this embodiment, based on the forced expiratory volume in the first second of each historical COPD patient at each pulmonary nodule growth inflection point and the corresponding second pulmonary nodule risk prediction level ladder division threshold, the second pulmonary nodule risk prediction result of each historical COPD patient at each pulmonary nodule growth inflection point is determined, which means that the ratio of the patient's forced expiratory volume in the first second at the pulmonary nodule growth inflection point to the reference value of the same level is compared with the ladder division threshold, so as to obtain the prediction result of the second pulmonary nodule risk, such as low risk, medium risk or high risk. For example, according to the GOLD classification (FEV1 percentage predicted value) of the pulmonary function test: GOLD1 (mild): forced expiratory volume in one second ≥ the first-level classification threshold (i.e., the product of 80% and the reference value of the same level of forced expiratory volume in one second) GOLD2 (moderate): The second-step classification threshold (i.e., the product of 50% and the reference value of the same level of forced expiratory volume in one second) ≤ forced expiratory volume in one second < the first-step classification threshold (i.e., the product of 80% and the reference value of the same level of forced expiratory volume in one second) GOLD3 (severe): The third-level classification threshold (i.e., the product of 30% and the reference value of the same level of forced expiratory volume in one second) ≤ forced expiratory volume in one second < the second-level classification threshold (i.e., the product of 50% and the reference value of the same level of forced expiratory volume in one second) GOLD4 (extremely severe): forced expiratory volume in one second < the third-level classification threshold (i.e., the product of 30% and the reference value of the same level of forced expiratory volume in one second) For example, suppose a patient B has a history of COPD, whose basic physiological parameters are: age 50, height 170cm, weight 70kg, and gender male. Through analysis and calculation of the healthy patient database, it is found that his forced expiratory volume in the first second is 2.5L. The preset step percentages are 30%, 50%, and 80%, and the calculated step-division thresholds for the second lung nodule risk prediction level are 0.75L, 1.25L, and 2L. At a certain lung nodule growth inflection point, the patient's lung function data showed that his forced expiratory volume in the first second was 1.8L. Because 1.25L<1.8L<2L, it is determined that the patient's second lung nodule risk prediction result at this lung nodule growth inflection point is "GOLD2 (moderate)", that is, "medium risk".

[0075] The beneficial effects of the above technologies are as follows: the same-level reference value determination subunit determines the same-level reference value of the forced expiratory volume in the first second through the patient's basic physiological parameters, establishing a personalized benchmark for subsequent risk assessment. The step division threshold determination subunit determines the division threshold based on the reference value and the preset step percentage, making the risk assessment more hierarchical and targeted. The third risk prediction subunit determines the risk prediction result based on the ratio of the patient's specific lung function data at the inflection point of lung nodule growth to the reference value and the division threshold, thereby improving the accuracy and refinement of the risk assessment. This risk prediction method based on personalized reference values ​​and refined division thresholds can more accurately assess the impact of lung function on the risk of lung nodules, provide a reliable basis for comprehensive risk prediction, and help improve the scientificity and effectiveness of the entire system in predicting the risk of lung nodules in patients with COPD.

[0076] Embodiment 8: Based on Embodiment 7, the same-level reference value determination subunit includes: A physiological parameter value determination terminal is used to determine the parameter value of each physiological parameter in the basic physiological parameters of each historical COPD patient; The parameter deviation calculation end is used to calculate the deviation between the basic physiological parameters of each healthy patient in the healthy patient database and the basic physiological parameters of each historical COPD patient; The peer patient screening terminal is used to treat all healthy patients in the healthy patient database whose basic physiological parameters do not exceed the deviation threshold with those of a single historical COPD patient as all peer patients of the corresponding historical COPD patient; The same-level reference value calculation terminal is used to calculate the same-level reference value of the forced expiratory volume in the first second of each historical COPD patient based on the basic physiological parameters of each historical COPD patient and the basic physiological parameters of all patients at the same level and the forced expiratory volume in the first second: ; In the formula, is the reference value of the forced expiratory volume in the first second of the historical COPD patient currently calculated, n is the total number of all patients of the same level of the historical COPD patient currently calculated, is the forced expiratory volume in the first second of the i-th patient of the same level of the historical COPD patient currently calculated, m is the total number of physiological parameters included in the basic physiological parameters, is the parameter value of the jth physiological parameter of the i-th patient of the historical COPD patient currently calculated, is the parameter value of the jth physiological parameter of the historical COPD patient currently calculated, is the parameter limit value of the jth physiological parameter, is the negative impact ratio of the parameter attenuation value of the jth physiological parameter on the forced expiratory volume in the first second.

[0077] In this embodiment, the parameter value of each physiological parameter, such as the basic physiological parameters, includes age, height, weight, etc. The age is 50 years old, the height is 170 cm, and the weight is 70 kg. Here, 50 years old, 170 cm, and 70 kg are the parameter values ​​of each physiological parameter.

[0078] In this embodiment, the deviation between the basic physiological parameters of each healthy patient in the healthy patient database and the basic physiological parameters of each historical COPD patient is calculated. Assume that the age of the historical COPD patient is 55 years old, the height is 165 cm, and the weight is 65 kg. The age of a healthy patient in the healthy patient database is 40 years old, the height is 175 cm, and the weight is 70 kg. The differences of the three parameters of age, height, and weight are calculated respectively by a specific formula, and a deviation value is obtained by comprehensive consideration. For example, suppose we use the following formula to calculate the deviation of each parameter: Deviation = |Parameter 1-Parameter 2| / Average value of parameter 1; For age, the average value of parameter 1 is assumed to be 45 years old (which can be the average age of the healthy patient database), then the age deviation = |55-40| / 45≈0.33; For height, the average value of parameter 1 is assumed to be 170 cm (the average value of height in the healthy patient database), then the height deviation = |165-175| / 170≈0.059; For weight, the average value of parameter 1 is assumed to be 60 kg (the average weight in the healthy patient database), then the weight deviation = |65-70| / 60≈0.083; Then, by comprehensively considering these three deviations, the final deviation degree can be obtained by simple averaging or weighted averaging.

[0079] If the simple average is used, the deviation = (0.33 + 0.059 + 0.083) / 3 ≈ 0.157; If age, height, and weight are assigned weights of 0.4, 0.3, and 0.3 respectively, then the deviation of the weighted average = 0.4 0.33+0.3 0.059+0.3 0.083≈0.155.

[0080] In this embodiment, the deviation threshold is set to 20%, for example. If the calculated deviation is less than or equal to 20%, they are considered to be similar and can be treated as patients of the same level.

[0081] In this embodiment, the healthy patient database is a database storing a large amount of information on basic physiological parameters of healthy people, for example, a data set containing 1,000 healthy people with different basic physiological parameters such as age, gender, height, weight, and their forced expiratory volume in the first second.

[0082] In this embodiment, the negative impact ratio of the parameter attenuation value of a single physiological parameter on the forced expiratory volume in one second refers to the quantitative ratio of the adverse effect of a specific physiological parameter (such as weight, height, etc.) on the forced expiratory volume in one second when the value of the parameter attenuates.

[0083] Assuming body weight is a single physiological parameter, for every 10kg reduction in body weight, the forced expiratory volume in one second may decrease by 5%. The 5% here is the negative impact ratio of the attenuation value of the body weight parameter on the forced expiratory volume in one second.

[0084] The beneficial effects of the above technologies are as follows: by determining the parameter values ​​of basic physiological parameters, basic data is provided for subsequent calculations. Calculating the degree of deviation between healthy patients and historical COPD patients can accurately screen out patients of the same level. Taking healthy patients who meet the deviation threshold as patients of the same level improves the reliability and relevance of the reference value. The use of a complex but scientific formula to calculate the same-level reference value of the forced expiratory volume in the first second fully considers the influence of multiple physiological parameters, making the reference value more accurate and scientific. This precise and comprehensive method of determining the same-level reference value that considers multiple factors provides a reliable and accurate basis for lung function-based assessment in lung nodule risk prediction, helps to improve the accuracy and effectiveness of risk prediction, and provides more valuable information for the treatment and management of COPD patients.

[0085] Embodiment 9: Based on Embodiment 6, the third risk prediction unit includes: The pulmonary quantitative data interpretation subunit is used to determine the percentage of low attenuation area, percentage of high attenuation area, and average density value of the whole lung at each pulmonary nodule growth inflection point of each historical COPD patient based on the pulmonary quantitative data of each historical COPD patient at each pulmonary nodule growth inflection point; The third risk prediction subunit is used to retrieve a preset list of corresponding lung quantitative data-third lung nodule risk prediction results based on the percentage of low attenuation area, percentage of high attenuation area, and average density value of the whole lung at each lung nodule growth inflection point of each historical COPD patient, and determine the third lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point.

[0086] In this embodiment, based on the lung quantitative data of each historical COPD patient at each inflection point of lung nodule growth, the percentage of low attenuation area, percentage of high attenuation area, and average density value of the whole lung of each historical COPD patient at each inflection point of lung nodule growth are determined. This means that for each historical COPD patient, the lung quantitative data obtained at those specific time points when the inflection point of lung nodule growth occurs, by analyzing and calculating these data, the proportion of the low attenuation area of ​​the patient's lungs to the whole lung at each inflection point (low attenuation area percentage), the proportion of the high attenuation area to the whole lung (high attenuation area percentage), and the average density value of the whole lung (average density value of the whole lung) are obtained. For example, the lung quantitative data of a historical COPD patient at a certain inflection point of lung nodule growth shows that the low attenuation area accounts for 20% of the whole lung, the high attenuation area accounts for 10%, and the average density value of the whole lung is -850HU.

[0087] In this embodiment, based on the percentage of low attenuation area, percentage of high attenuation area, and average density value of the whole lung for each historical COPD patient at each inflection point of lung nodule growth, the preset lung quantitative data-third lung nodule risk prediction result corresponding list is retrieved to determine the third lung nodule risk prediction result of each historical COPD patient at each inflection point of lung nodule growth. That is, for each historical COPD patient at an inflection point in lung nodule growth, the percentage of low attenuation area, percentage of high attenuation area, and average density value of the whole lung are obtained. These specific data are compared and searched with the pre-set "lung quantitative data-third lung nodule risk prediction result corresponding list" to obtain the third lung nodule risk prediction result of each historical COPD patient at each inflection point of lung nodule growth.

[0088] For example, assuming that the preset lung quantitative data-third lung nodule risk prediction results correspond to the following list: If the percentage of low attenuation area is 15%-25%, the percentage of high attenuation area is 5%-15%, and the average density value of the whole lung is -800HU to -900HU, the corresponding third lung nodule risk prediction result is medium risk; If the percentage of low attenuation area is more than 25%, the percentage of high attenuation area is more than 15%, and the average density value of the whole lung is less than -800HU, the corresponding third lung nodule risk prediction result is high risk; If a patient has a low attenuation area percentage of 22% and a high attenuation area percentage of 12% at the inflection point of lung nodule growth, and an average lung density value of -880HU, by retrieving the above list, it can be determined that the risk prediction result of the patient's third lung nodule is "medium risk".

[0089] The beneficial effects of the above technologies are as follows: the lung quantitative data interpretation subunit can extract key indicators from complex lung quantitative data, such as the percentage of low attenuation areas, the percentage of high attenuation areas, and the average density value of the whole lung, making the data more interpretable and usable. The third risk prediction subunit quickly determines the third lung nodule risk prediction result based on these key indicators by retrieving the corresponding list, thereby improving the efficiency and accuracy of risk assessment. This risk prediction method based on specific quantitative data indicators can more accurately assess the impact of lung structure and density changes on the risk of lung nodules, and provides valuable supplementary information for comprehensive risk prediction. It helps to predict the risk of lung nodules in patients with COPD more comprehensively and accurately, provide a more reliable basis for medical decision-making, and improve patients' treatment effects and quality of life.

[0090] Example 10: The present invention provides a method for predicting the risk of pulmonary nodules in patients with COPD, which is applied to any of the risk prediction systems for pulmonary nodules in patients with COPD in Examples 1 to 9, with reference to Figure 2 ,include: S1: Collect multiple CT examination data, lung function data and lung quantitative data of a large number of historical COPD patients at different time points; S2: Identify the inflection point of lung nodule growth in each historical COPD patient in all CT examination data; S3: A lung nodule risk prediction model was built based on the first CT examination data and the CT examination data of the lung nodule growth inflection point, lung function data, and lung quantitative data of each historical COPD patient; S4: inputting the current CT examination data, current lung function data and current lung quantitative data of the patient to be tested into the lung nodule risk prediction model to obtain the current lung nodule risk prediction result of the patient to be tested; S5: Based on the current lung nodule risk prediction result of the patient to be tested and the preset reexamination time recommendation model, the recommended reexamination time of the patient to be tested is obtained.

[0091] The beneficial effects of the above technology are: based on step S1, a large amount of multi-type historical data is collected, which provides a rich and comprehensive data basis for subsequent analysis. Step S2 can accurately find the key turning points in the growth of lung nodules, which helps to have a deeper understanding of the development laws of lung nodules. Step S3 comprehensively utilizes a variety of data to build a prediction model, which improves the accuracy and reliability of the model. Step S4 can input the latest data of the patient to be tested into the model to quickly obtain the prediction results of his lung nodule risk. The system realizes the effective prediction of the risk of lung nodules in patients with COPD and the reasonable inference of the re-examination time, which provides an important basis for early diagnosis and treatment, helps to improve the treatment effect and quality of life of patients, and reduce the harm of the disease.

[0092] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A pulmonary nodule risk prediction system for patients with chronic obstructive pulmonary disease, characterized in that: include: The historical examination data acquisition module is used to collect multiple CT examination data, lung function data, and lung quantitative data of a large number of historical COPD patients at different time points; A growth inflection point identification module is used to identify all lung nodule growth inflection points of each historical COPD patient in all CT examination data of each historical COPD patient; The risk prediction model building module is used to build a lung nodule risk prediction model based on the first CT examination data and the CT examination data at all lung nodule growth inflection points of each historical COPD patient, lung function data, and lung quantitative data; The current patient risk prediction module is used to input the current CT examination data, current lung function data and current lung quantitative data of the patient to be tested into the lung nodule risk prediction model to obtain the current lung nodule risk prediction result of the patient to be tested; The review time recommendation module is used to obtain the recommended review time for the patient to be tested based on the current lung nodule risk prediction result of the patient to be tested and the preset review time recommendation model.

2. The pulmonary nodule risk prediction system for COPD patients according to claim 1, characterized in that: The historical inspection data acquisition module includes: A historical patient screening submodule is used to collect a preset number of historical COPD patients whose number of completed complete lung examinations is not less than a preset number and whose maximum interval between all different complete lung examinations is not less than a preset time interval; The examination data collection submodule is used to collect CT examination data, lung function data and lung quantitative data at the time points of all complete lung examinations completed by all historical COPD patients.

3. The pulmonary nodule risk prediction system for COPD patients according to claim 1, characterized in that: Growth inflection point identification module, including: The CT examination data analysis submodule is used to determine the average CT value of all pixels in the lung nodule area in each CT examination data of each historical COPD patient as the average CT value of each CT examination data, and analyze the lung nodule volume in each CT examination data of each historical COPD patient; A pulmonary nodule mass calculation submodule is used to calculate the pulmonary nodule mass of each CT examination data of each historical COPD patient based on the average CT value and pulmonary nodule volume of each CT examination data of each historical COPD patient; The mass growth ratio calculation submodule is used to calculate the growth ratio of the pulmonary nodule mass of each CT examination data other than the first CT examination data of each historical COPD patient relative to the pulmonary nodule mass of the first CT examination data; The growth inflection point analysis submodule is used to treat the time points corresponding to all CT examination data of each historical COPD patient whose growth ratio of the lung nodule mass relative to the first CT examination data is not less than a preset percentage as the growth inflection points of all lung nodules corresponding to each historical COPD patient.

4. The pulmonary nodule risk prediction system for COPD patients according to claim 3, characterized in that: The pulmonary nodule mass calculation submodule calculates the pulmonary nodule mass of each CT examination data of each historical COPD patient based on the average CT value and pulmonary nodule volume of each CT examination data of each historical COPD patient, including: ; Wherein, nodule mass is the mass of the pulmonary nodules in the current CT examination data of the historical COPD patient currently calculated, v is the volume of the pulmonary nodules in the current CT examination data of the historical COPD patient currently calculated, The average CT value of the current CT examination data of historical COPD patients calculated currently.

5. The pulmonary nodule risk prediction system for COPD patients according to claim 1, characterized in that: Risk prediction model building module, including: The historical patient risk prediction submodule is used to obtain the lung nodule risk prediction results of each historical COPD patient at each lung nodule growth inflection point based on the first CT examination data and the CT examination data of each lung nodule growth inflection point, lung function data, and lung quantitative data of each historical COPD patient; The training sample arrangement submodule is used to use the first CT examination data and the CT examination data at each pulmonary nodule growth inflection point of each historical COPD patient, pulmonary function data, and pulmonary quantitative data as the input of the model training samples, and use the pulmonary nodule risk prediction results of the corresponding historical COPD patients at the corresponding pulmonary nodule growth inflection point as the model output to generate multiple training samples; The model training submodule is used to train the model based on all training samples and build a lung nodule risk prediction model.

6. The pulmonary nodule risk prediction system for COPD patients according to claim 5, characterized in that: Historical patient risk prediction submodule, including: a doubling time determination unit, used to regard the time interval between each pulmonary nodule growth inflection point of each historical COPD patient and the corresponding time point of the first CT examination data as the doubling time of each pulmonary nodule growth inflection point of each historical COPD patient; A first risk prediction unit is used to retrieve a preset doubling time-first pulmonary nodule risk prediction result corresponding list based on the doubling time, and determine the first pulmonary nodule risk prediction result of each historical COPD patient at each pulmonary nodule growth inflection point; A second risk prediction unit is used to determine the second pulmonary nodule risk prediction result of each historical COPD patient at each pulmonary nodule growth inflection point based on the pulmonary function data of each historical COPD patient at each pulmonary nodule growth inflection point; A third risk prediction unit is used to determine the third lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point based on the lung quantitative data of each historical COPD patient at each lung nodule growth inflection point; The comprehensive risk prediction unit is used to obtain the lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point based on the first lung nodule risk prediction result, the second lung nodule risk prediction result, and the third lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point.

7. The pulmonary nodule risk prediction system for COPD patients according to claim 6, characterized in that: The second risk prediction unit includes: A subunit for determining a reference value of the same level, for determining a reference value of the forced expiratory volume in the first second of each historical COPD patient based on the basic physiological parameters of each historical COPD patient; A step division threshold determination subunit is used to determine the step division threshold of the second pulmonary nodule risk prediction level for each historical COPD patient based on the forced expiratory volume in the first second reference value of the same level and the preset step percentage of each historical COPD patient; The third risk prediction subunit is used to determine the forced expiratory volume in one second of each historical COPD patient at each lung nodule growth inflection point based on the lung function data of each historical COPD patient at each lung nodule growth inflection point, and to determine the second lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point based on the forced expiratory volume in one second of each historical COPD patient at each lung nodule growth inflection point and the corresponding second lung nodule risk prediction level step division threshold.

8. The pulmonary nodule risk prediction system for COPD patients according to claim 7, characterized in that: The subunit for determining the reference value of the same level includes: A physiological parameter value determination terminal is used to determine the parameter value of each physiological parameter in the basic physiological parameters of each historical COPD patient; The parameter deviation calculation end is used to calculate the deviation between the basic physiological parameters of each healthy patient in the healthy patient database and the basic physiological parameters of each historical COPD patient; The peer patient screening terminal is used to treat all healthy patients in the healthy patient database whose basic physiological parameters do not exceed the deviation threshold with those of a single historical COPD patient as all peer patients of the corresponding historical COPD patient; The same-level reference value calculation terminal is used to calculate the same-level reference value of the forced expiratory volume in the first second of each historical COPD patient based on the basic physiological parameters of each historical COPD patient and the basic physiological parameters of all patients at the same level and the forced expiratory volume in the first second: ; In the formula, is the reference value of the forced expiratory volume in the first second of the historical COPD patient currently calculated, n is the total number of all patients of the same level of the historical COPD patient currently calculated, is the forced expiratory volume in the first second of the i-th patient of the same level of the historical COPD patient currently calculated, m is the total number of physiological parameters included in the basic physiological parameters, is the parameter value of the jth physiological parameter of the i-th patient of the historical COPD patient currently calculated, is the parameter value of the jth physiological parameter of the historical COPD patient currently calculated, is the parameter limit value of the jth physiological parameter, is the negative impact ratio of the parameter attenuation value of the jth physiological parameter on the forced expiratory volume in the first second.

9. The pulmonary nodule risk prediction system for COPD patients according to claim 6, characterized in that: The third risk prediction unit includes: The pulmonary quantitative data interpretation subunit is used to determine the percentage of low attenuation area, percentage of high attenuation area, and average density value of the whole lung at each pulmonary nodule growth inflection point of each historical COPD patient based on the pulmonary quantitative data of each historical COPD patient at each pulmonary nodule growth inflection point; The third risk prediction subunit is used to retrieve a preset list of corresponding lung quantitative data-third lung nodule risk prediction results based on the percentage of low attenuation area, percentage of high attenuation area, and average density value of the whole lung at each lung nodule growth inflection point of each historical COPD patient, and determine the third lung nodule risk prediction result of each historical COPD patient at each lung nodule growth inflection point.

10. A method for predicting the risk of pulmonary nodules in patients with chronic obstructive pulmonary disease, characterized in that: The pulmonary nodule risk prediction system for COPD patients as described in any one of claims 1 to 9 comprises: S1: Collect multiple CT examination data, lung function data and lung quantitative data of a large number of historical COPD patients at different time points; S2: Identify the inflection point of lung nodule growth in each historical COPD patient in all CT examination data; S3: A lung nodule risk prediction model was built based on the first CT examination data and the CT examination data of the lung nodule growth inflection point, lung function data, and lung quantitative data of each historical COPD patient; S4: inputting the current CT examination data, current lung function data and current lung quantitative data of the patient to be tested into the lung nodule risk prediction model to obtain the current lung nodule risk prediction result of the patient to be tested; S5: Based on the current lung nodule risk prediction result of the patient to be tested and the preset reexamination time recommendation model, the recommended reexamination time of the patient to be tested is obtained.

Images