A method and system for early warning of copd based on chest ct parameters

Abstract

The present application relates to the field of medical imaging technology, specifically to a COPD early warning method and system based on chest CT parameters, comprising the following steps: based on chest CT image data, extracting pixel data of lung lobe anatomical region, segmenting airway tree structure, identifying bronchial wall boundary, screening gradient change points, and generating airway wall boundary point set. In the present application, by refining the lung lobe anatomical region, segmenting the airway tree structure, accurately positioning the bronchial wall boundary, calculating the airway wall thickness, correcting abnormal fluctuations, improving the stability of the thickness parameter, using the thickness parameter to calculate the airflow passage cross-sectional area, analyzing the lung lobe ventilation volume and ventilation change, quantifying the lung ventilation imbalance degree, screening the ventilation restricted area, analyzing the relationship between the wall thickness and the ventilation change rate, extracting the abnormal growth area, marking the ventilation restriction severity, setting the early warning threshold, calculating the abnormal index mean value, marking the risk area, enhancing the ventilation abnormality recognition hierarchy, and improving the early warning ability of COPD.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of medical imaging technology, and in particular to a COPD early warning method and system based on chest CT parameters. BACKGROUND

[0002] The field of medical imaging technology includes the use of imaging technology for the visualization analysis of the structure and function of various parts of the human body. The core content of this field includes early identification of different diseases, disease monitoring, and lesion evaluation based on medical imaging. Medical imaging technology is widely used in multiple disciplines and mainly involves the use of imaging devices such as CT, MRI, and X-ray to obtain high-resolution images. Through image processing, analysis, and optimization techniques, image data is extracted and analyzed to assess and research health conditions. The core goal of medical imaging technology is to provide accurate visual information to assist research and application in related medical fields.

[0003] Among them, the COPD early warning method based on chest CT parameters refers to processing chest CT image data to extract feature parameters related to COPD (Chronic Obstructive Pulmonary Disease), thereby providing support for COPD early warning. This patent subject involves using image data obtained from CT images to conduct detailed analysis of lung structures using medical image analysis methods, extracting key parameters such as lung ventilation, lung morphology, and airway diameter. Through comprehensive evaluation of the parameters and quantitative data analysis, COPD early warning information is generated to help relevant personnel identify potential risks early. This patent solves the problem of how to effectively evaluate the potential risks of COPD through CT image data and provides early warning through quantitative analysis, avoiding excessive reliance on traditional imaging methods and subjective judgments.

[0004] Existing technologies rely on static analysis methods of medical images for lung structure and airway morphology evaluation, which overly relies on image visual features and fails to establish a stable quantitative analysis system. Airway wall thickness evaluation is affected by subjective judgments, making it difficult to accurately reflect the subtle structural changes in the lesion site. Ventilation restriction cannot effectively correlate with the dynamic changes in airway wall thickness, leading to one-sidedness in risk assessment. Lung ventilation analysis is mainly based on overall lung ventilation function and fails to refine the spatial distribution of regional ventilation status, ignoring the development trend of local ventilation abnormalities. The identification method of ventilation-restricted areas cannot accurately match the development process of airway lesions, and no effective risk index calculation model is provided, resulting in lag in COPD early warning. SUMMARY

[0005] The purpose of the present application is to solve the shortcomings in the prior art and to provide a COPD early warning method and system based on chest CT parameters.

[0006] In order to achieve the above object, the present application adopts the following technical scheme: A COPD early warning method based on chest CT parameters, comprising the following steps:

[0007] S1: based on chest CT image data, extracting pixel data of lung lobe anatomical region, segmenting airway tree structure, identifying bronchial wall boundary, screening gradient change point, and generating airway wall boundary point set;

[0008] S2: based on the airway wall boundary point set, extracting the boundary points of the measuring region in the bronchial branch region, identifying the local curvature radius, adjusting the measuring path direction, analyzing the pixel gradient change of adjacent points on the measuring path, screening the boundary points meeting the measurement conditions, calculating the pixel gradient change rate, correcting the abnormal fluctuation, and obtaining the airway wall thickness parameter;

[0009] S3: calling the airway wall thickness parameter, segmenting the alveolar ventilation region, calculating the airflow channel cross-sectional area data, analyzing the lung lobe ventilation volume and ventilation change amplitude, and obtaining the lung ventilation imbalance coefficient;

[0010] S4: calling the lung ventilation imbalance coefficient, screening the ventilation restricted region, analyzing the thickness and ventilation change rate, extracting the abnormal growth region, and obtaining the airway wall ventilation coupling abnormality index;

[0011] S5: calling the airway wall ventilation coupling abnormality index, screening the abnormal distribution region, setting the early warning threshold, calculating the abnormal index mean, marking the threshold exceeding region and performing early warning, and outputting the COPD risk early warning response.

[0012] As a further scheme of the present application, the airway wall boundary point set includes bronchial wall boundary points, gradient change points and pixel gradient change rates, the airway wall thickness parameter includes measuring region boundary points, local curvature radius and adjacent point pixel gradient change, the lung ventilation imbalance coefficient includes alveolar ventilation region, airflow channel cross-sectional area data and lung lobe ventilation change amplitude, the airway wall ventilation coupling abnormality index includes ventilation restricted region, thickness change rate and ventilation change rate, and the COPD risk early warning response includes abnormal distribution region, abnormal index mean and threshold exceeding region.

[0013] As a further scheme of the present application, the airway wall boundary point set is obtained by the following steps:

[0014] S111: based on chest CT image data, extracting pixel data of lung lobe anatomical region, calculating the pixel gradient change amount of the region, screening the pixel points meeting the threshold, and generating the lung lobe pixel gradient change point set;

[0015] S112: based on the lung lobe pixel gradient change point set, segmenting the airway tree structure, extracting the boundary pixel points, calculating the gradient direction vector, screening the gradient change rate, and generating the airway wall candidate point set;

[0016] S113: based on the airway wall candidate point set, calculating the local gradient mean value of the boundary point, analyzing the gradient distribution, identifying the spatial offset, and according to the gradient and offset relationship, using the formula:

[0017] ;

[0018] The boundary gradient offset coefficient of the airway wall boundary point is calculated, the boundary points that meet the threshold value are screened, and the airway wall boundary point set is generated;

[0019] wherein, the boundary gradient offset coefficient of the airway wall boundary point, the gradient value of the airway wall candidate boundary point , the gradient mean value of the candidate boundary point, and the offset of the boundary point in the horizontal and vertical directions, a small positive number is used to avoid the denominator being zero, the total number of airway wall candidate boundary points.

[0020] As a further scheme of the present application, the airway wall thickness parameter acquisition step is specifically:

[0021] S211: based on the airway wall boundary point set, extracting the measurement boundary points in the bronchial branch region, calculating the local gradient change value of the boundary points, screening the boundary points that meet the set threshold value, and obtaining the measurement region boundary point set;

[0022] S212: based on the measurement region boundary point set, adjusting the measurement path direction, matching the local curvature trend, analyzing the pixel gradient change of adjacent points on the path, and using the formula:

[0023] ;

[0024] The gradient change curvature value of the measurement path is calculated, the boundary points that meet the measurement condition are screened, and the measurement path boundary point set is generated;

[0025] wherein, the gradient change curvature value of the measurement path, the gradient values of the differentiated boundary points on the measurement path, the offset of the corresponding boundary points in the horizontal and vertical directions, a small positive number is used to avoid the denominator being zero;

[0026] S213: Based on the measurement path boundary point set, analyze the pixel gradient change rate, correct the abnormal fluctuation, call the corrected gradient change rate data, calculate the local thickness data of the boundary point, and obtain the airway wall thickness parameter.

[0027] As a further scheme of the present application, the lung ventilation imbalance degree coefficient obtaining step is specifically:

[0028] S311: Based on the airway wall thickness parameter, segment the alveolar ventilation area, calculate the local airflow channel cross-sectional area of each region, screen combined with the regional airflow channel morphological characteristics, remove outliers, and obtain the airflow channel cross-sectional area data;

[0029] S312: Call the airflow channel cross-sectional area data, identify the lung lobe corresponding to the lung lobe ventilation volume, and normalize the lung lobe ventilation volume. The formula is:

[0030] ;

[0031] The ventilation change amplitude value of the lung lobe is calculated;

[0032] Wherein, represents the ventilation change amplitude value of the lung lobe, represents the current lung lobe ventilation volume, represents the initial lung lobe ventilation volume;

[0033] S313: Based on the ventilation change amplitude of the lung lobe, analyze the difference degree between the lung lobes, and combine the lung lobe ventilation volume to obtain the lung ventilation imbalance degree coefficient.

[0034] As a further scheme of the present application, the airway wall ventilation coupling abnormality index obtaining step is specifically:

[0035] S411: Call the lung ventilation imbalance degree coefficient, screen the uneven ventilation distribution area between each lung lobe, and remove the area in the ventilation balanced range to obtain the ventilation restricted area;

[0036] S412: Based on the ventilation restricted area, calculate the airway wall thickness in the area, monitor the ventilation change rate over time, and use the formula:

[0037] ;

[0038] The ventilation change rate value is calculated;

[0039] Wherein, represents the ventilation change rate value, represents the ventilation volume at time , and represents the ventilation volume at time , Represents the total time interval;

[0040] S413: Analyze the changing trends of the ventilation rate and airway wall thickness, extract abnormal growth areas, and calculate the airway wall ventilation coupling anomaly index.

[0041] As a further aspect of the present invention, the step of obtaining the COPD risk warning response specifically includes:

[0042] S511: Based on the airway wall ventilation coupling anomaly index, call the airway wall thickness, ventilation distribution value and local flow rate value, calculate the anomaly distribution value, filter out the anomaly areas that exceed the threshold, and obtain the average value of the anomaly areas;

[0043] S512: Compare the average value of the abnormal area with the warning threshold, extract the area that exceeds the warning range, call the ventilation distribution status and local resistance coefficient, and filter out the data of the area that exceeds the threshold.

[0044] S513: Based on the data exceeding the threshold region, call the environmental variables and sample data, using the formula:

[0045] ;

[0046] Calculate the mean COPD abnormality index and output the COPD risk warning response;

[0047] in, This represents the mean of the COPD abnormality index. Representing the The anomaly index value of each sample. The representative sample's anomaly index mean. Represents the total number of samples. To measure the degree of sample bias.

[0048] A COPD early warning system based on chest CT parameters, wherein the system is used to execute the aforementioned COPD early warning method based on chest CT parameters, and the system includes:

[0049] The anatomical region analysis module is based on chest CT image data. It extracts lung pixel data, filters lung parenchyma boundaries, segments bronchial tree structure, analyzes gray-level gradient, detects bronchial wall boundaries, analyzes gradient changes at boundary points, calculates bronchial branch angles, and obtains a set of bronchial wall boundary points.

[0050] The airway wall thickness calculation module extracts the measurement region boundary points based on the bronchial wall boundary point set, calculates the lumen radial distance, analyzes the gray scale gradient, screens the boundary points meeting the standard, calculates the thickness difference between the boundary points, analyzes the local thickness change, analyzes the bronchial wall continuous profile thickness distribution, adjusts the abnormal fluctuation section, corrects the measurement path, counts the differentiated branch thickness change characteristics, and obtains the airway wall thickness parameters;

[0051] The lung airflow distribution analysis module calculates the airway branch cross-sectional area based on the airway wall thickness parameters, counts the lung lobe ventilation channel change, analyzes the airway ventilation flow rate difference, and obtains the lung ventilation limitation distribution data;

[0052] The ventilation abnormality recognition module screens the abnormal region based on the lung ventilation limitation distribution data, analyzes the corresponding relationship between the wall thickness change rate and the ventilation limitation rate, extracts the ventilation resistance increased region, recognizes the lung lobe ventilation uniformity, and obtains the airway wall ventilation abnormality index;

[0053] The COPD risk grading module sets the early warning threshold based on the airway wall ventilation abnormality index, screens the threshold-exceeding region, counts the spatial distribution, screens the risk region and generates the early warning, and obtains the COPD risk early warning response.

[0054] Compared with the prior art, the advantages and positive effects of the present application are that:

[0055] In the present application, by analyzing the chest CT image data, the lung lobe anatomical region is refined, the airway tree structure is segmented, and the key gradient change points are screened to ensure accurate positioning of the bronchial wall boundary. Based on the boundary point set, the measurement region boundary points are extracted, the airway wall thickness is calculated, and the abnormal fluctuation is corrected to improve the stability and continuity of the thickness parameters. The thickness parameters are used to calculate the airflow channel cross-sectional area of the alveolar ventilation region, analyze the lung lobe ventilation volume and ventilation change, accurately quantify the lung ventilation imbalance degree, further screen the ventilation limitation region, analyze the correlation between the airway wall thickness and the ventilation change rate, extract the abnormal growth region, calibrate the severity of the ventilation limitation, finally set the early warning threshold, calculate the abnormal index mean value, mark the risk region and generate the early warning scheme, realize the quantitative COPD risk assessment based on the CT image, optimize the image data extraction method, enhance the accuracy of the airway wall thickness measurement, establish a systematic evaluation mechanism for lung ventilation characteristics, make the ventilation abnormality recognition more hierarchical, and improve the early warning ability of COPD. BRIEF DESCRIPTION OF DRAWINGS

[0056] Figure 1 The workflow schematic diagram of the present application is shown in the figure;

[0057] Figure 2 The flowchart of the airway wall boundary point set acquisition in the present application is shown in the figure;

[0058] Figure 3Flow chart for obtaining airway wall thickness parameter in the application;

[0059] Figure 4 Flow chart for obtaining lung ventilation imbalance degree coefficient in the application;

[0060] Figure 5 Flow chart for obtaining airway wall ventilation coupling abnormality index in the application;

[0061] Figure 6 Flow chart for obtaining COPD risk early warning response in the application. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0063] In the description of the present application, it should be understood that the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application. In addition, in the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise specifically limited.

[0064] Example one

[0065] Please refer to Figure 1 The present application provides a technical scheme: a COPD early warning method based on chest CT parameters, comprising the following steps:

[0066] S1: based on chest CT image data, extracting pixel data of lung lobe anatomical region, segmenting airway tree structure, identifying bronchial wall boundary, screening gradient change point, and generating airway wall boundary point set;

[0067] S2: based on the airway wall boundary point set, extracting the boundary points of the measurement region in the bronchial branch region, identifying the local curvature radius, adjusting the measurement path direction, analyzing the pixel gradient change of adjacent points on the measurement path, screening the boundary points meeting the measurement conditions, calculating the pixel gradient change rate, correcting the abnormal fluctuation, and obtaining the airway wall thickness parameter;

[0068] S3: calling the airway wall thickness parameter, segmenting the alveolar ventilation region, calculating the airflow channel cross-sectional area data, analyzing the lung lobe ventilation volume and ventilation change amplitude, and obtaining the lung ventilation imbalance degree coefficient;

[0069] S4: calling lung ventilation imbalance coefficient, screening ventilation-restricted area, analyzing thickness and ventilation change rate, extracting abnormal growth area, and obtaining airway wall ventilation coupling abnormality index;

[0070] S5: calling airway wall ventilation coupling abnormality index, screening abnormal distribution area, setting early warning threshold, calculating abnormal index mean, marking threshold-exceeding area and giving early warning, and outputting COPD risk early warning response.

[0071] The airway wall boundary point set includes bronchial wall boundary points, gradient change points, and pixel gradient change rate, the airway wall thickness parameter includes measurement area boundary points, local curvature radius, and adjacent point pixel gradient change, the lung ventilation imbalance coefficient includes alveolar ventilation area, airflow channel cross-sectional area data, and lung lobe ventilation change amplitude, the airway wall ventilation coupling abnormality index includes ventilation-restricted area, thickness change rate, and ventilation change rate, and the COPD risk early warning response includes abnormal distribution area, abnormal index mean, and threshold-exceeding area.

[0072] Please refer to Figure 2 The acquisition step of the airway wall boundary point set is specifically as follows:

[0073] S111: based on chest CT image data, extracting pixel data of lung lobe anatomical area, calculating pixel gradient change amount of the area, screening pixel points meeting the threshold, and generating lung lobe pixel gradient change point set;

[0074] By analyzing the CT image through specific software, different densities of lung tissue are recognized and distinguished. This process is completed by extracting pixels within a specific threshold range in the image. Pixels represent normal or abnormal lung tissue. The pixel gradient change amount of each area is calculated. This value is obtained by comparing the density difference of adjacent pixel points. This helps to depict the detailed outline of the lung structure, and then screen out pixel points with significant changes. For example, in the differentiation of tumors or lesions, significant pixel points indicate the boundaries of abnormal areas. In this way, the lung lobe pixel gradient change point set is generated. The point set is the key data for determining the location and range of lesions.

[0075] S112: based on the lung lobe pixel gradient change point set, segmenting the airway tree structure, extracting boundary pixel points, calculating gradient direction vectors, screening gradient change rate, and generating airway wall candidate point set;

[0076] In practical applications, it is very important to identify airway obstruction and evaluate airway lesions. The boundary pixel points of the airway region are extracted, and the pixel points are determined by calculating the direction and size of the gradient of the adjacent region pixels. The gradient direction vector of the boundary pixel points is calculated, which is used to evaluate the overall direction and morphology of the airway wall. In combination with the distribution characteristics of the gradient direction vector, the region with abnormal gradient change is identified, which indicates the specific position of the disease affecting the airway. The gradient change rate with important clinical diagnostic significance is screened, and the rate data helps doctors judge the severity and type of airway disease. A candidate point set of the airway wall is generated, which can be used for further diagnostic analysis and disease evaluation.

[0077] S113: Based on the candidate point set of the airway wall, the local gradient mean of the boundary point is calculated, the gradient distribution is analyzed, the spatial offset is identified, and the formula is used according to the gradient and offset relationship:

[0078] ;

[0079] The boundary gradient offset coefficient of the airway wall boundary point is calculated, the boundary points meeting the threshold value are screened, and the airway wall boundary point set is generated;

[0080] Wherein, represents the boundary gradient offset coefficient of the airway wall boundary point, represents the gradient value at the candidate boundary point of the airway wall, represents the gradient mean of the candidate boundary point, represents the gradient mean of the candidate boundary point, and represents the offset of the boundary point in the horizontal and vertical directions, represents a small positive number for avoiding zero denominator, represents the total number of candidate boundary points of the airway wall;

[0081] And substitute the specific numerical value for calculation to obtain the boundary gradient offset coefficient of the airway wall boundary point;

[0082] (Gradient value at the candidate boundary point of the airway wall): calculated from CT image data, obtained by calculating the gradient change of adjacent pixels;

[0083] (Gradient mean of the candidate boundary point): calculated by taking the mean value of the gradient values of all candidate points;

[0084] , (offset of the boundary point in the horizontal and vertical directions): the calculation method is: , wherein is the airway centerline coordinate,​ ) coordinates of candidate boundary points;

[0085] (prevent small positive number for denominator): set to ;

[0086] (total number of candidate boundary points): set to 5 boundary points for example calculation;

[0087] Assumption: gradient values of 5 candidate points:

[0088] ;

[0089] Calculate mean of gradient:

[0090] ;

[0091] Coordinates and offset calculation of 5 candidate points: ;

[0092] ;

[0093] Calculate :

[0094] , ;

[0095] , ;

[0096] , ;

[0097] , ;

[0098] , ;

[0099] Calculate :

[0100] ;

[0101] ;

[0102] ;

[0103] ;

[0104] ;

[0105] Calculate :

[0106] ;

[0107] ;

[0108] Calculate each item:

[0109] ,

[0110] ,

[0111] ;

[0112] ;

[0113] The results show that the boundary gradient offset coefficient of the airway wall boundary point is 1.665, and the numerical size reflects the deviation of the boundary point gradient change from the average value. In medical image analysis, the numerical value can be used to judge whether the airway wall has pathological thickening or stenosis. By comparing with the set standard threshold (for example, 1.5), it can be judged whether the current airway has abnormal conditions. By normalizing the gradient offset value to the spatial offset value, the false judgment caused by the local gradient anomaly of the airway wall can be avoided, and the detection ability of the airway wall thickness anomaly can be improved. The formula uses the gradient mean value as the reference value for calculation, which can reduce the interference of individual abnormal points on the overall evaluation, making the detection more stable and reliable.

[0114] Please refer to Figure 3 , the step of obtaining the airway wall thickness parameter is specifically:

[0115] S211: Based on the airway wall boundary point set, extract the measurement boundary points in the bronchial branch region, calculate the local gradient change value of the boundary points, and select the boundary points meeting the set threshold to obtain the measurement region boundary point set;

[0116] First, the coordinate information of all points in the airway wall boundary point set is obtained, and the point set belonging to the bronchial branch region is determined according to the position relationship, so as to generate the preliminary boundary of the measurement region. Then, the local gradient change value of each boundary point is calculated. The value is obtained by calculating the gray value difference of adjacent pixel points. In CT image, the gray value can represent the change of tissue density. For example, if the gray values of adjacent pixel points are set to 180 and 150, the gradient change value is calculated as For the entire boundary point set, each point performs the calculation to build a complete gradient change distribution map, and then the boundary points meeting the set threshold range are screened, which is used to distinguish boundary points from non-boundary points. The reference basis for setting the threshold is the gray difference between normal tissue and lesion area. Assuming that the set threshold is 25, all points with gradient change value greater than 25 are retained, and points below 25 are excluded, and finally the boundary point set of the measurement area is obtained.

[0117] S212: Based on the boundary point set of the measurement area, adjust the measurement path direction to match the local curvature trend, analyze the pixel gradient change of adjacent points on the path, and use the formula:

[0118] ;

[0119] Calculate the gradient change curvature value of the measurement path, screen the boundary points meeting the measurement condition, and generate the boundary point set of the measurement path;

[0120] Wherein, represents the gradient change curvature value of the measurement path, respectively represent the gradient value of the differentiated boundary point on the measurement path, represents the offset of the corresponding boundary point in the horizontal and vertical directions, represents a small positive number to avoid zero denominator;

[0121] Based on the boundary point set of the measurement area, calculate the local curvature radius of the boundary point, obtain the boundary point coordinates and calculate the curvature change of adjacent points. The curvature radius is calculated from the coordinates of three adjacent points. Adjust the measurement path direction to optimize the path along the local curvature change trend. In the path adjustment process, the pixel gradient change of adjacent points on the path needs to be analyzed to calculate the gradient change rate on the measurement path.

[0122] Set an example data, assuming that the gray value of the boundary point is: ;

[0123] Offset ;

[0124] Small positive number To prevent the denominator from being zero, substitute the calculation:

[0125] ;

[0126] ;

[0127] ;

[0128] ;

[0129] The result shows that the gradient change curvature value of the measurement path is 11.56, which can be used to judge the smoothness of the path. If the value is larger, it means that the boundary curvature changes sharply, and the measurement path direction needs to be adjusted.

[0130] S213: Based on the measurement path boundary point set, analyze the pixel gradient change rate, correct abnormal fluctuations, call the corrected gradient change rate data, calculate the local thickness data of the boundary point, and obtain the airway wall thickness parameter;

[0131] Get the pixel gray value of each point on the measurement path, calculate the gradient change ratio between adjacent points, correct abnormal fluctuations, compare the gradient change rate of consecutive points on the measurement path, and detect abnormal jump points. If the gradient change rate of adjacent points exceeds the set range, smooth correction is performed, for example, the allowed range is set to ±10%, and if the change rate of adjacent points is 25%, it is adjusted to within 10%. Call the corrected gradient change rate data to calculate the local thickness parameter of the boundary point, and finally obtain the airway wall thickness parameter.

[0132] Please refer to Figure 4 , the steps for obtaining the lung ventilation imbalance coefficient are as follows:

[0133] S311: Based on the airway wall thickness parameter, segment the alveolar ventilation area, calculate the local airflow channel cross-sectional area of each region, and screen out abnormal values based on the morphological characteristics of the region airflow channel to obtain the airflow channel cross-sectional area data;

[0134] First, obtain the patient's chest high-resolution CT (HRCT) image data, segment the airway wall using image processing software, extract the airway wall thickness information, and apply image segmentation algorithms such as region growing or level set method to separate the alveolar ventilation area from the image, ensuring accurate identification of the alveolar structure. Calculate the local airflow channel cross-sectional area of each segmented region, which can be calculated by measuring the airway diameter and applying the formula , where represents the cross-sectional area, represents the airway diameter, and the morphological characteristics of the region airflow channel, such as branch angle and length, are combined to screen out abnormal values. For example, if a cross-sectional area value deviates significantly from the range of two standard deviations from the mean, it is considered an abnormal value and is removed to obtain the airflow channel cross-sectional area data.

[0135] S312: Call the airflow channel cross-sectional area data, identify the corresponding lung lobe ventilation volume of the lung lobe, and normalize the lung lobe ventilation volume using the formula:

[0136] ;

[0137] Calculate the ventilation change amplitude value of the lung lobe;

[0138] wherein, represents the ventilation change amplitude value of the lung lobe, represents the current lung lobe ventilation, represents the initial lung lobe ventilation;

[0139] First, the cross-sectional areas of each airflow channel are classified and summarized according to the lung lobe to which they belong, and the total ventilation of each lung lobe is calculated. This can be achieved by multiplying the cross-sectional area of each channel by the corresponding airflow velocity and summing all channels within the same lung lobe. Assuming that there are three airflow channels in a lung lobe with cross-sectional areas of 2 cm 2 , 1.5 cm 2 , and 1 cm 2 , and the corresponding airflow velocities are all 5 cm / s, then the total ventilation of the lung lobe is . Next, the ventilation of each lung lobe is normalized, i.e., the ventilation of each lung lobe is divided by the sum of the total ventilation of all lung lobes to obtain the relative ventilation proportion. The ventilation change amplitude of each lung lobe is calculated using the formula:

[0140] ;

[0141] wherein, represents the ventilation change amplitude, represents the current lung lobe ventilation, represents the initial lung lobe ventilation, assuming the initial ventilation is 20 cm 3 / s and the current ventilation is 22.5 cm 3 / s, then the ventilation change amplitude is , and the ventilation change amplitude value of the lung lobe is calculated.

[0142] S313: Based on the ventilation change amplitude of the lung lobe, analyze the difference degree between the lung lobes, and combine the lung lobe ventilation to obtain the lung ventilation imbalance degree coefficient;

[0143] First, the difference degree of the ventilation change amplitude between each lung lobe is calculated, which can be achieved by calculating the standard deviation of the ventilation change amplitude of each lung lobe. The larger the standard deviation, the more significant the ventilation change difference between the lung lobes. Combined with the ventilation of each lung lobe, the lung ventilation imbalance degree coefficient is calculated. The ventilation change amplitude of each lung lobe is multiplied by the proportion of its corresponding ventilation to the total ventilation, and then the results of all lung lobes are summed. Assuming that there are three lung lobes with ventilation change amplitudes of 10%, 15%, and 20%, and the corresponding ventilation proportions are 30%, 40%, and 30%, respectively, then the lung ventilation imbalance degree coefficient is , and the lung ventilation imbalance degree coefficient is obtained.

[0144] Please refer to Figure 5 , the steps for obtaining the airway wall ventilation coupling abnormality index are as follows:

[0145] S411: Call the lung ventilation imbalance coefficient to filter the areas of uneven ventilation distribution between each lung lobe, remove the areas within the ventilation balance range, and obtain the ventilation-restricted areas.

[0146] First, using the patient's lung CT data, ventilation distribution information for each lung lobe is extracted. The average ventilation volume of each lung lobe is calculated and normalized to the total lung ventilation volume to obtain the ventilation proportion of each lung lobe. By comparing the ventilation proportions of each lung lobe, the ventilation imbalance between lung lobes is calculated. If the ventilation proportion of a certain lung lobe is lower than a certain threshold (such as a set lower limit for normal ventilation proportion), it is determined that the lung lobe has ventilation limitation. Assuming the total lung ventilation volume is 5000 mL, of which the ventilation volume of the left upper lobe is 800 mL, the ventilation volume of the right upper lobe is 900 mL, and the ventilation volume of the right middle lobe is... The ventilation volume is 1000 mL for the right lower lobe, 1400 mL for the right lower lobe, and 900 mL for the left lower lobe. The ventilation percentages for each lobe are 16%, 18%, 20%, 28%, and 18%, respectively. If a threshold is set such that a ventilation percentage below 15% is considered a restricted ventilation area, then although the left upper lobe has the lowest ventilation percentage, it is still within the normal range. Otherwise, it is classified as a restricted area. The variation in ventilation volume for each lobe is calculated. If the variation in ventilation volume of a certain lobe at multiple time points exceeds the set variation threshold (e.g., exceeding 20%), then ventilation in that area can also be considered restricted.

[0147] S412: Based on the ventilated restricted area, calculate the airway wall thickness within the area and monitor the rate of change of ventilation over time using the formula:

[0148] ;

[0149] The ventilation rate of change was calculated.

[0150] in, This represents the rate of change in ventilation. Representative moment ventilation volume, Representative moment ventilation volume, Represents the total time interval;

[0151] First, CT image data corresponding to the ventilated restricted area is extracted. The airway wall contours are extracted, and the thickness of each airway wall is calculated. The airway wall thickness distribution in this area is obtained by measuring the difference in contour distance between the outer and inner walls. The rate of change of ventilation over time is monitored, and the ventilation volume at multiple time points is calculated. The rate of change of ventilation is then determined using the formula:

[0152] ;

[0153] in, Indicates the rate of change in ventilation. representative time point representative time point representative time point representative time point representative total time interval, for example, if the ventilation of a certain ventilation-restricted region is 200 mL, 180 mL and 150 mL at 0 s, 10 s and 20 s respectively, the ventilation change rate is calculated as follows:

[0154]

[0155] indicates that the ventilation of the region is decreasing, and the ventilation rate is decreasing. Combined with the airway wall thickness data of the region, it is analyzed whether the airway wall thickening is positively correlated with the decrease of the ventilation rate, and finally the ventilation change rate is calculated.

[0156] S413: analyze the ventilation change rate value and the change trend of the airway wall thickness, extract the abnormal growth region, and calculate the airway wall ventilation coupling abnormality index;

[0157] First, compare the airway wall thickness values at multiple time points, calculate the change rate, and judge the growth speed of the airway wall thickness. If the airway wall thickness of a certain ventilation-restricted region grows more than a set threshold (such as 0.1 mm / month), it is marked as an abnormal growth region. Combined with the ventilation change rate data, it is calculated whether the airway wall thickening is accompanied by a decrease in the ventilation rate. If the airway wall thickness growth speed of a certain region is fast, and the ventilation change rate is decreasing, it is considered that the region has pathological abnormalities. For example, if the airway wall thickness of a certain ventilation-restricted region increases from 1.2 mm to 1.6 mm in three months, the growth rate is mm / month, which exceeds the set threshold, and the ventilation change rate of the region decreases to -3.0 mL / s, which is determined as an abnormal growth region. Finally, the airway wall ventilation coupling abnormality index is calculated.

[0158] Please refer to Figure 6 , the steps for obtaining the COPD risk warning response are as follows:

[0159] S511: based on the airway wall ventilation coupling abnormality index, call the airway wall thickness, ventilation distribution value and local flow value, calculate the abnormal distribution value, screen the abnormal regions exceeding the threshold, and obtain the average value of the abnormal regions;

[0160] ​By collecting airway wall thickness data of multiple patients, extracting ventilation distribution values at each position, and calculating local flow values, the thickness of different branches of the bronchus is measured using CT scan data of different patients, the airway wall thickness information is obtained, the ventilation distribution is quantified based on the physiological characteristics of different individuals, the ventilation indicators are normalized for comparison between different individuals, the local airway resistance is calculated through airflow dynamics, and the airway patency is judged in combination with the ventilation data. After normalizing all individual data, the abnormal distribution values of each sample are calculated, the areas where the ventilation and airway wall thickness distribution do not match are screened out, for example, when the airway wall of a certain area is thickened but the ventilation is reduced, the area is abnormal. The data of all abnormal areas are integrated, and the average abnormality index is calculated, which can be used for abnormality judgment of different individuals, and finally the average of abnormal areas is obtained.

[0161] S512: Compare the average of abnormal areas with the warning threshold, extract the areas that exceed the warning range, call the ventilation distribution state and local resistance coefficient, and select the threshold area data that exceeds the threshold;

[0162] First, set the warning threshold, use statistical methods to extract the normal ventilation distribution range from healthy individual data, calculate the mean and standard deviation in this range to set the upper and lower limits of the normal value, and then compare the mean of the abnormal area with the set threshold. If the average of an area exceeds the warning range, mark the area as a risk area, call the ventilation distribution state, calculate the ventilation change at different time points, analyze the local resistance coefficient, measure the airway resistance by fluid dynamics method, and calculate the correlation between airway resistance and ventilation. All areas are screened to extract areas that meet the risk characteristics, and finally the threshold area data that exceeds the threshold is obtained.

[0163] S513: Based on the threshold area data that exceeds the threshold, call the environmental variables and sample data, and use the formula:

[0164] ;

[0165] Calculate the average of COPD abnormality index, and output the COPD risk warning response;

[0166] Wherein, represents the average of the COPD abnormality index, represents the abnormality index value of the th sample, represents the average of the sample abnormality index, represents the total number of samples, measures the degree of sample deviation;

[0167] Set five sample data , calculate the average of the abnormality index:

[0168] ;

[0169] Calculate the deviation square sum:

[0170] ;

[0171] ;

[0172] Finally calculate the COPD abnormal index mean: ;

[0173] The results show that the deviation degree of the abnormal index is small, if the value exceeds the set risk limit, further screening is needed, and finally the COPD risk warning response is output.

[0174] The COPD warning system based on chest CT parameters is used to execute the COPD warning method based on chest CT parameters described above, and the system comprises:

[0175] The anatomical region analysis module extracts the lung pixel data based on the chest CT image data, filters the lung parenchyma boundary, segments the bronchial tree structure, analyzes the gray scale gradient, detects the bronchial wall boundary, analyzes the gradient change of the boundary points, calculates the bronchial branch angle, and obtains the bronchial wall boundary point set;

[0176] The airway wall thickness calculation module extracts the measurement region boundary points based on the bronchial wall boundary point set, calculates the lumen radial distance, analyzes the gray scale gradient, filters the boundary points that meet the standard, calculates the thickness difference between the boundary points, analyzes the local thickness change, analyzes the continuous cross-sectional thickness distribution of the bronchial wall, adjusts the abnormal fluctuation section, corrects the measurement path, counts the differentiated branch thickness change characteristics, and obtains the airway wall thickness parameter;

[0177] The lung airflow distribution analysis module calculates the airway branch cross-sectional area based on the airway wall thickness parameter, counts the lung lobe ventilation channel change, analyzes the airway ventilation flow rate difference, and obtains the lung ventilation restriction distribution data;

[0178] The ventilation abnormality recognition module filters the abnormal region based on the lung ventilation restriction distribution data, analyzes the corresponding relationship between the wall thickness change rate and the ventilation restriction rate, extracts the region where the ventilation resistance increases, identifies the lung lobe ventilation uniformity, and obtains the airway wall ventilation abnormality index;

[0179] The COPD risk grading module sets the warning threshold based on the airway wall ventilation abnormality index, filters the region exceeding the threshold, counts the spatial distribution, filters the risk region and gives a warning, and obtains the COPD risk warning response.

[0180] The above merely describes the preferred embodiments of the present application, and is not intended to limit the present application in other forms. Any skilled person in the art can modify or change the disclosed technical content into equivalent embodiments with equivalent changes, and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application, without departing from the technical solution content of the present application, still falls within the protection scope of the present application.

Claims

1. A COPD early warning method based on chest CT parameters, characterized in that, Includes the following steps: S1: Based on chest CT image data, extract pixel data of the lung lobe anatomical region, segment the airway tree structure, identify bronchial wall boundaries, filter gradient change points, and generate a set of airway wall boundary points. S2: Based on the airway wall boundary point set, extract the boundary points of the measurement area within the bronchial branch area, identify the local radius of curvature, adjust the measurement path direction, analyze the pixel gradient changes of adjacent points on the measurement path, filter the boundary points that meet the measurement conditions, calculate the pixel gradient change rate, correct abnormal fluctuations, and obtain the airway wall thickness parameters. S3: Call the airway wall thickness parameter, segment the alveolar ventilation area, calculate the cross-sectional area data of the airflow channel, analyze the lobar ventilation and ventilation variation, and obtain the lung ventilation imbalance coefficient. S4: Call the lung ventilation imbalance coefficient to screen ventilation-restricted areas, analyze the thickness and ventilation change rate, extract abnormal growth areas, and obtain the airway wall ventilation coupling abnormality index. S5: Call the airway wall ventilation coupling anomaly index, filter the abnormal distribution area, set the warning threshold, calculate the average value of the anomaly index, mark the area exceeding the threshold and issue a warning, and output the COPD risk warning response. The specific steps for obtaining the airway wall boundary point set are as follows: S111: Based on chest CT image data, extract pixel data of the anatomical region of the lung lobe, calculate the pixel gradient change of the region, filter pixels that meet the threshold, and generate a set of lung lobe pixel gradient change points. S112: Based on the set of lung lobe pixel gradient change points, segment the airway tree structure, extract boundary pixel points, calculate gradient direction vector, filter gradient change rate, and generate a set of airway wall candidate points. S113: Based on the candidate point set of the airway wall, calculate the local gradient mean of the boundary points, analyze the gradient distribution, identify the spatial offset, and use the formula based on the relationship between gradient and offset: ； Calculate the boundary gradient offset coefficients of the airway wall boundary points, filter the boundary points that meet the threshold, and generate a set of airway wall boundary points; in, The boundary gradient offset coefficient representing the boundary points of the airway wall. Representative airway wall candidate boundary points gradient value at, The mean gradient of the candidate boundary points. and Represents boundary points Offset in the horizontal and vertical directions, Small positive numbers are used to avoid the denominator being zero. This represents the total number of candidate boundary points for the airway wall.

2. The COPD early warning method based on chest CT parameters according to claim 1, characterized in that, The airway wall boundary point set includes bronchial wall boundary points, gradient change points, and pixel gradient change rate. The airway wall thickness parameters include measurement area boundary points, local radius of curvature, and pixel gradient change of adjacent points. The lung ventilation imbalance coefficient includes alveolar ventilation area, airflow channel cross-sectional area data, and lobar ventilation change amplitude. The airway wall ventilation coupling abnormality index includes ventilation-restricted area, thickness change rate, and ventilation change rate. The COPD risk warning response includes abnormal distribution area, abnormal index mean, and area exceeding threshold.

3. The COPD early warning method based on chest CT parameters according to claim 1, characterized in that, The steps for obtaining the airway wall thickness parameter are as follows: S211: Based on the airway wall boundary point set, extract the measurement boundary points within the bronchial branch region, calculate the local gradient change value of the boundary points, and filter the boundary points that meet the set threshold to obtain the measurement region boundary point set; S212: Based on the set of boundary points of the measurement area, adjust the direction of the measurement path, match the local curvature trend, analyze the pixel gradient changes of adjacent points on the path, and use the formula: ； Calculate the gradient curvature value of the measurement path, select boundary points that meet the measurement conditions, and generate a set of boundary points for the measurement path; in, The curvature value representing the gradient change of the measurement path. These represent the gradient values ​​at the differential boundary points along the measurement path. This represents the offset of the corresponding boundary point in the horizontal and vertical directions. Small positive numbers are used to avoid the denominator being zero; S213: Based on the set of boundary points of the measurement path, analyze the pixel gradient change rate, correct abnormal fluctuations, call the corrected gradient change rate data, calculate the local thickness data of the boundary points, and obtain the airway wall thickness parameters.

4. The COPD early warning method based on chest CT parameters according to claim 3, characterized in that, The specific steps for obtaining the lung ventilation imbalance coefficient are as follows: S311: Based on the airway wall thickness parameter, the alveolar ventilation area is segmented, the local airflow channel cross-sectional area of ​​each area is calculated, and the airflow channel morphological characteristics of the area are combined to filter out outliers and obtain the airflow channel cross-sectional area data. S312: Call the cross-sectional area data of the airflow channel, identify the lobar ventilation corresponding to the lung lobe, and normalize the lobar ventilation using the following formula: ； The amplitude of ventilation change in the lung lobe was calculated; in, Values ​​representing the amplitude of ventilation changes in lung lobes. Represents the current lobe ventilation. Represents the initial lobar ventilation; S313: Based on the ventilation variation range of the lung lobes, analyze the degree of difference between lung lobes, and combine the lung lobe ventilation volume to obtain the lung ventilation imbalance coefficient.

5. The COPD early warning method based on chest CT parameters according to claim 4, characterized in that, The specific steps for obtaining the airway wall ventilation coupling anomaly index are as follows: S411: Call the lung ventilation imbalance coefficient to filter the areas of uneven ventilation distribution between each lung lobe, remove the areas within the ventilation balance range, and obtain the ventilation-restricted areas. S412: Based on the ventilation-restricted area, calculate the airway wall thickness within the area, monitor the rate of change of ventilation over time, and use the formula: ； The ventilation rate of change was calculated. in, This represents the rate of change in ventilation. Representative moment ventilation volume, Representative moment ventilation volume, Represents the total time interval; S413: Analyze the changing trends of the ventilation rate and airway wall thickness, extract abnormal growth areas, and calculate the airway wall ventilation coupling anomaly index.

6. The COPD early warning method based on chest CT parameters according to claim 5, characterized in that, The specific steps for obtaining the COPD risk warning response are as follows: S511: Based on the airway wall ventilation coupling anomaly index, call the airway wall thickness, ventilation distribution value and local flow rate value, calculate the anomaly distribution value, filter out the anomaly areas that exceed the threshold, and obtain the average value of the anomaly areas; S512: Compare the average value of the abnormal area with the warning threshold, extract the area that exceeds the warning range, call the ventilation distribution status and local resistance coefficient, and filter out the data of the area that exceeds the threshold. S513: Based on the data exceeding the threshold region, call the environmental variables and sample data, using the formula: ； Calculate the mean COPD abnormality index and output the COPD risk warning response; in, This represents the mean of the COPD abnormality index. Representing the The anomaly index value of each sample. The representative sample's anomaly index mean. Represents the total number of samples. To measure the degree of sample bias.

7. A COPD early warning system based on chest CT parameters, characterized in that, The COPD early warning method based on chest CT parameters according to any one of claims 1-6, the system comprising: The anatomical region analysis module is based on chest CT image data. It extracts lung pixel data, filters lung parenchyma boundaries, segments bronchial tree structure, analyzes gray-level gradient, detects bronchial wall boundaries, analyzes gradient changes at boundary points, calculates bronchial branch angles, and obtains a set of bronchial wall boundary points. The airway wall thickness calculation module extracts the boundary points of the measurement area based on the set of bronchial wall boundary points, calculates the radial distance of the lumen, analyzes the gray-scale gradient, filters boundary points that meet the standards, calculates the thickness difference between boundary points, analyzes local thickness changes, analyzes the thickness distribution of the continuous bronchial wall profile, adjusts abnormal fluctuation sections, corrects the measurement path, statistically analyzes the characteristics of differential branch thickness changes, and obtains airway wall thickness parameters. Based on the airway wall thickness parameter, the lung airflow distribution analysis module calculates the cross-sectional area of ​​airway branches, statistically analyzes changes in lung lobe ventilation channels, analyzes differences in airway ventilation velocity, and obtains data on the distribution of lung ventilation limitation. Based on the lung ventilation restriction distribution data, the ventilation abnormality identification module filters abnormal areas, analyzes the relationship between the wall thickness change rate and the ventilation restriction rate, extracts areas with increased ventilation resistance, identifies the lung lobe ventilation balance, and obtains the airway wall ventilation abnormality index. The COPD risk grading module sets an early warning threshold based on the airway wall ventilation anomaly index, filters areas exceeding the threshold, statistically analyzes the spatial distribution, filters risk areas and issues early warnings, and obtains COPD risk early warning responses.

Images