Lung interstitial fibrosis early screening system based on big data analysis

Abstract

The invention provides an early screening system for pulmonary interstitial fibrosis based on big data analysis. The early screening system comprises a data acquisition module, a model inspection module, a data preprocessing module, a data storage module, a data analysis module and a result output module. According to the system, a pulmonary interstitial fibrosis early screening model is constructed based on historical data, big data analysis is performed on features extracted from medical image data of a patient according to the model, so that wrong features are eliminated, a data analysis module is used for calculating pulmonary interstitial fibrosis feature coefficients for the correct features, and the accuracy of the pulmonary interstitial fibrosis early screening is improved. The accuracy of the calculation result is improved, the evaluation based on the calculation result can assist a doctor in more accurately identifying the early characteristics of the pulmonary interstitial fibrosis, the accuracy of early diagnosis is improved, and missed diagnosis and misdiagnosis conditions are reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical diagnosis systems, and particularly relates to a pulmonary interstitial fibrosis early screening system based on big data analysis. BACKGROUND

[0002] Pulmonary interstitial fibrosis is a serious lung disease, which has insidious onset and atypical early symptoms, making early diagnosis a great challenge. Traditional diagnostic methods mainly rely on the clinical experience of doctors, single examination methods and limited case data, and have the problems of high misdiagnosis and missed diagnosis rates.

[0003] With the exponential growth of medical data, including massive clinical cases, medical images, genetic test results, etc., how to effectively utilize these data has become a difficult problem to be solved. At the same time, the wide application of big data analysis technology in various fields has brought new opportunities to the medical field. Through big data analysis, multiple source heterogeneous medical data can be integrated to mine potential disease characteristics and rules. However, the current big data analysis system for pulmonary interstitial fibrosis early screening is not perfect, and cannot fully meet the needs of clinical precise diagnosis and early prevention and control. SUMMARY

[0004] In view of the above, it is necessary to provide a pulmonary interstitial fibrosis early screening system based on big data analysis. The system constructs a pulmonary interstitial fibrosis early screening model based on historical data, performs big data analysis on the features extracted from the patient's medical image data according to the model, eliminates the wrong features, and calculates the pulmonary interstitial fibrosis feature coefficient for the correct features using the data analysis module, thereby improving the accuracy of the calculation results. Based on such calculation results, the evaluation can assist doctors to more accurately identify the early features of pulmonary interstitial fibrosis, improve the accuracy of early diagnosis, and reduce the missed diagnosis and misdiagnosis.

[0005] To achieve the above purpose, the technical solution adopted by the present application is:

[0006] A pulmonary interstitial fibrosis early screening system based on big data analysis, comprising,

[0007] A data acquisition module is used to collect data related to pulmonary interstitial fibrosis, specifically medical image data. Through feature extraction of the medical image data, feature data is obtained and sent to a model verification module.

[0008] The model verification module is connected with the data collection module through data transmission technology, is used for correlating the features of the medical image data extracted from the data collection module to obtain a plurality of correlation images, inputting each correlation image into the constructed early screening model for pulmonary fibrosis, obtaining an evaluation value of the pulmonary fibrosis feature coefficient from the model, pairing the evaluation value with the true value of the training set in the model to obtain true cases, false positives and false negatives, and calculating the error scores of the true cases, false positives and false negatives respectively, processing the plurality of correlation images according to the calculated error scores, and sending the processed correlation images to the data preprocessing module.

[0009] The data preprocessing module is connected with the model verification module through data transmission technology, is used for de-correlating the processed correlation images, calculating feature values based on the de-correlated features, and performing cleaning, denoising and standardization processing on the calculated feature values to obtain standardized feature values.

[0010] The data storage module is connected with the data preprocessing module through data transmission technology, and is used for storing the preprocessed data.

[0011] The data analysis module is connected with the data storage module through data transmission technology, and is used for analyzing and judging the preprocessed data. The specific process is that the pulmonary fibrosis feature coefficient L(W, b) is calculated based on an analysis calculation formula, and the lung condition of the patient is judged according to the calculation result. The analysis calculation formula is:

[0012]

[0013] In the formula, W is a weight vector, b is a bias vector, m is the number of samples, y i is the standardized feature value, is a prediction function, and The calculation formula of is In the formula, T is a transpose, and X is a feature vector of the standardized feature value.

[0014] The result output module is connected with the data analysis module through data transmission technology, is used for generating an early screening result of pulmonary fibrosis according to the pulmonary fibrosis feature coefficient calculated by the data analysis module, and visually displaying the generated result.

[0015] Preferably, in the data collection module, the medical image data includes chest X-ray images and CT images of the patient. When the features of the medical image data are extracted, the features of each image data are extracted and compared with the features of the corresponding image in the normal state to obtain an abnormal feature region map of each image data. The obtained abnormal feature region map is the feature data.

[0016] Preferably, in the model verification module, the method for constructing the early screening model for pulmonary interstitial fibrosis is:

[0017] S1, obtain historical chest X-ray images and historical CT images of a plurality of pulmonary interstitial fibrosis patients, and perform feature extraction on the historical chest X-ray images and the historical CT images respectively to obtain X-ray abnormal feature maps and CT abnormal feature maps, and integrate the X-ray abnormal feature maps and the CT abnormal feature maps of the same patient into a correlation image as a training sample;

[0018] S2, constructing an early screening model for pulmonary interstitial fibrosis based on a deep learning network, and dividing the generated training samples into a training set and a test set according to a ratio of 8:2;

[0019] S3, downloading a weight file and loading it onto a BP neural network for initializing migration network parameters, and determining the number of hidden layer nodes of the BP neural network model according to the number of correlation images in the training set;

[0020] S4, modifying the last fully connected layer of the BP neural network, keeping the input unchanged, setting the output as a pulmonary interstitial fibrosis feature coefficient, performing weight initialization on the last layer, using a gradient descent algorithm for learning, and using fixed step attenuation to optimize the training parameters, retraining the entire network to obtain an early screening model for pulmonary interstitial fibrosis;

[0021] S5, randomly and repeatedly extracting small batches of correlation images from the training set during the training process, and extracting all correlation images in the training set as a training cycle, iterating to a certain cycle to complete the training, and then using the test set to evaluate the effect of the early screening model for pulmonary interstitial fibrosis.

[0022] Preferably, in the model verification module, the error score is calculated by the following calculation method:

[0023]

[0024] In the formula, the error rate represents the proportion of all error classification samples to the total number of samples, which is used to measure the overall error rate of the classifier;

[0025] TP is the true positive, representing the number of samples that are actually positive and correctly classified as positive by the classifier;

[0026] FP is the false positive, representing the number of samples that are actually negative but incorrectly classified as positive by the classifier;

[0027] FN is the false negative, representing the number of samples that are actually positive but incorrectly classified as negative by the classifier;

[0028] TN is the true negative, representing the number of samples that are actually negative and correctly classified as negative by the classifier.

[0029] Preferably, in the model verification module, when processing a plurality of associated images according to the calculated error scores, the processing mode includes image rejection and image re-verification.

[0030] Preferably, in the data preprocessing module, when calculating the feature value, the area of the abnormal feature region in each feature is calculated, and the obtained area data is the feature value.

[0031] Preferably, in the data preprocessing module, the received data is standardized by using the following calculation formula:

[0032]

[0033] In the formula, y is the feature value, y i is the standardized feature value, μ is the mean of the feature, and σ is the standard deviation of the feature.

[0034] Preferably, the data storage module adopts a distributed storage architecture to store different types of data on different nodes in the blockchain.

[0035] Preferably, in the result output module, the generation of the early screening result of pulmonary fibrosis is based on the comparison between the pulmonary fibrosis feature coefficient calculated by the data analysis module and the preset feature judgment range, the evaluation of the risk of suffering from pulmonary fibrosis, and the obtained risk evaluation result, which includes risk categories and risk levels.

[0036] Preferably, in the result output module, the visualization display of the generated result includes the following steps:

[0037] S1, data formatting, the original result output by the data analysis module is arranged to meet the display requirements;

[0038] S2, visualization processing, corresponding charts and reports are generated according to the result data, the charts include but are not limited to histograms, pie charts and line charts, and the reports include but are not limited to risk categories, risk levels, data sources, information explanations and diagnostic results;

[0039] S3, the formatted and visualized result is transmitted to the display interface, which includes but is not limited to the in-hospital system interface and the application program interface of the patient mobile terminal;

[0040] S4, associated feedback, a user feedback entry is provided for filling in questions about the screening result, self-symptom supplement and diagnostic supplement.

[0041] Compared with the prior art, the present application has the following beneficial effects:

[0042] 1、The application constructs an early screening model of pulmonary interstitial fibrosis based on historical data. The model can integrate and analyze the preprocessed data information of patients by big data analysis technology. The reasoning result can more accurately identify the early features of pulmonary interstitial fibrosis from the patient medical image data. Based on such features, the analysis and calculation of the pulmonary interstitial fibrosis feature coefficient are more accurate. Therefore, through the system provided by the application, doctors can more accurately identify whether the patient has the early features of pulmonary interstitial fibrosis, greatly improve the accuracy of early diagnosis, and reduce missed diagnosis and misdiagnosis.

[0043] 2、During the operation of the system, the early screening model of pulmonary interstitial fibrosis can continuously accumulate a large amount of pulmonary interstitial fibrosis related data in its database. These data are analyzed and mined, which helps medical researchers to deeply understand the pathogenesis, risk factors and development process of the disease, and provides strong data support for the development of new treatment methods and drugs, and promotes the development of pulmonary interstitial fibrosis disease research.

[0044] 3、The patient can obtain early and accurate screening results, reduce unnecessary repeated examinations and anxiety about long waiting for diagnosis results. At the same time, user feedback and result interpretation method make the patient better understand the screening results and enhance the cognition of their own health status, improve the medical experience and satisfaction. BRIEF DESCRIPTION OF DRAWINGS

[0045] Figure 1 is a system structure block diagram of the application.

[0046] The following specific embodiments will further illustrate the application in conjunction with the above drawings. DETAILED DESCRIPTION

[0047] Please refer to Figure 1 In a preferred embodiment of the application, an early screening system of pulmonary interstitial fibrosis based on big data analysis includes a data acquisition module, a model verification module, a data preprocessing module, a data storage module, a data analysis module and a result output module. Specifically,

[0048] The data collection module is configured to collect data related to pulmonary fibrosis, specifically medical image data, and to extract features from the medical image data to obtain feature data, which is then sent to the model verification module. The medical image data includes chest X-ray images and CT images of patients. During feature extraction, each image data is extracted and compared with the corresponding image in the normal state to obtain an abnormal feature region map for each image data. The abnormal feature region map is the feature data.

[0049] For example, for a patient's chest X-ray image, features are extracted from the chest X-ray image and compared with the chest X-ray image in the normal state to obtain chest abnormal feature regions. Similarly, for a patient's CT image, features are extracted from the CT image and compared with the CT image in the normal state to obtain CT abnormal feature regions. The same processing method is used for other images. It should be noted that during feature extraction, there may be more than one abnormal feature region map for each type of image data, and multiple abnormal feature regions need to be processed.

[0050] The model verification module is connected to the data collection module through data transmission technology and is configured to associate the feature data of the medical image data extracted by the data collection module to obtain a plurality of association images. Each association image is input into the constructed pulmonary fibrosis early screening model to obtain an evaluation value of the pulmonary fibrosis feature coefficient. The evaluation value is paired with the true value of the training set in the model to obtain true positives, false positives, and false negatives. The error scores of the true positives, false positives, and false negatives are calculated, and the multiple association images are processed according to the calculated error scores. The processing methods include image elimination and image revalidation. The processed association images are sent to the data preprocessing module.

[0051] In the present application, the pulmonary fibrosis early screening model construction method is as follows:

[0052] S1, obtain historical chest X-ray images and historical CT images of a plurality of pulmonary fibrosis patients, and extract features from the historical chest X-ray images and historical CT images to obtain X-ray abnormal feature maps and CT abnormal feature maps. The X-ray abnormal feature map and the CT abnormal feature map of the same patient are integrated into an association image as a training sample.

[0053] S2, construct a pulmonary fibrosis early screening model based on a deep learning network, and divide the generated training sample into a training set and a test set according to a ratio of 8:2.

[0054] S3, download the weight file and load it to the BP neural network, which is used to initialize the migration network parameters, and determine the number of hidden layer nodes of the BP neural network model according to the number of associated images in the training set;

[0055] S4, modify the last fully connected layer of the BP neural network, keep the input unchanged, set the output as the pulmonary fibrosis feature coefficient, initialize the weight of the last layer, use gradient descent algorithm for learning, and adopt fixed step attenuation to optimize the training parameters, retrain the whole network, and obtain the pulmonary fibrosis early screening model;

[0056] S5, during the training process, a small batch of associated images are randomly and repeatedly extracted from the training set, and all the associated images in the training set are extracted as a training cycle, and the training is completed after a certain cycle is iterated, and then the test set is used to evaluate the effect of the pulmonary fibrosis early screening model.

[0057] In the model verification module, the error score is calculated by the following calculation method:

[0058]

[0059] In the formula, the error rate represents the proportion of all error classification samples in the total sample number, which is used to measure the overall error rate of the classifier;

[0060] TP is the true positive, which represents the number of samples that are actually positive and are correctly classified as positive by the classifier;

[0061] FP is the false positive, which represents the number of samples that are actually negative but are incorrectly classified as positive by the classifier;

[0062] FN is the false negative, which represents the number of samples that are actually positive but are incorrectly classified as negative by the classifier;

[0063] TN is the true negative, which represents the number of samples that are actually negative and are correctly classified as negative by the classifier.

[0064] The above TP, FP, FN and TN constitute a confusion matrix, that is, in the classification problem, the confusion matrix is used to calculate the true positive, false positive, false negative and other indicators, and then different error scores are calculated.

[0065] In the present application, the construction of the early screening model of pulmonary interstitial fibrosis is based on historical big data, and since the diagnosis result in the historical big data is a definite result, the data of the patient to be tested is applied to the early screening model of pulmonary interstitial fibrosis, and based on the big data analysis of the historical data, the pulmonary interstitial fibrosis characteristic coefficient obtained by analysis will be a relatively accurate judgment result, therefore, the reasoning result obtained from the model judges the correctness of the characteristic data extracted by the data acquisition module, which can effectively process the defective characteristic data and retain the accurate characteristic data, so that through subsequent data analysis and processing of these accurate characteristic data, the early characteristics of pulmonary interstitial fibrosis can be accurately identified, thereby assisting doctors in diagnosing the early stage of pulmonary interstitial fibrosis, greatly improving the accuracy of the diagnosis and reducing the missed diagnosis and misdiagnosis.

[0066] The data preprocessing module is connected with the model verification module through data transmission technology, and is used for disassociating the processed associated images, calculating characteristic values based on the disassociated features, cleaning, denoising and standardizing the calculated characteristic values to obtain standardized characteristic values. When calculating the characteristic values, the area of the abnormal feature region in each feature is calculated, and the obtained area data is the characteristic value.

[0067] In the present application, the associated image is image information obtained by associating various medical image data, each associated image is composed of multiple feature data, and each feature data is extracted from one kind of medical image data. In the data preprocessing module, each feature data needs to be processed. For example, for the chest X-ray image of a patient, the chest abnormal feature region is obtained by feature extraction based on the chest X-ray image, and the area of the chest abnormal feature region is calculated to obtain a characteristic value one. Similarly, for the CT image of a patient, the CT abnormal feature region is obtained by feature extraction based on the CT image, and the area of the CT abnormal feature region is calculated to obtain a characteristic value two. Since the number of feature data corresponding to each kind of medical image data is not one, the number of characteristic value one and characteristic value two obtained is not one.

[0068] In the present application, after the characteristic values are calculated, the characteristic values need to be preprocessed to obtain standardized characteristic values, wherein the following calculation formula is used to standardize the received data:

[0069]

[0070] In the formula, y is the characteristic value, y i is the standardized characteristic value, μ is the mean of the feature, and σ is the standard deviation of the feature.

[0071] The data storage module is connected with the data preprocessing module through data transmission technology and is used for storing the preprocessed data. In the embodiment, the data storage module adopts a distributed storage architecture and stores different types of data on different nodes in the blockchain.

[0072] The data analysis module is connected with the data storage module through data transmission technology and is used for analyzing and judging the preprocessed data. Specifically, the lung interstitial fibrosis characteristic coefficient L(W,b) is calculated based on an analysis calculation formula, and the lung condition of the patient is judged according to the calculation result. The analysis calculation formula is as follows:

[0073]

[0074] In the formula, W is a weight vector, b is a bias vector, m is the number of samples, y i is a standardized characteristic value, is a prediction function, and The calculation formula of L(W,b) is as follows: In the formula, T is a transpose, and X is a characteristic vector of the standardized characteristic value.

[0075] In the data analysis module, the lung interstitial fibrosis characteristic coefficient needs to be calculated for all the characteristic values obtained above, and the calculated lung interstitial fibrosis characteristic coefficients are divided into multiple categories according to the sources of the characteristic values. Since there is actually a positive correlation between each category of characteristic values, when judging the lung abnormality of the patient, the lung interstitial fibrosis characteristic coefficients of each category need to be considered, rather than only considering the lung interstitial fibrosis characteristic coefficients of a single category. For example, two categories of characteristic values obtained from chest X-ray images and CT images, both of the two categories of lung interstitial fibrosis characteristic coefficients need to be referred to when judging the lung abnormality.

[0076] In the present application, the lung interstitial fibrosis characteristic coefficient is used to reflect the lung abnormality degree of the patient. The greater the value of L(W,b) is, the more serious the lung abnormality degree of the patient is, and vice versa. The smaller the value of L(W,b) is, the more slight the lung abnormality degree of the patient is.

[0077] The result output module is connected with the data analysis module through data transmission technology and is used for generating the lung interstitial fibrosis early screening result according to the lung interstitial fibrosis characteristic coefficient calculated by the data analysis module and visually displaying the generated result.

[0078] Specifically, the generation of the lung interstitial fibrosis early screening result is to evaluate the risk of lung interstitial fibrosis by comparing the lung interstitial fibrosis characteristic coefficient calculated by the data analysis module with a preset characteristic judgment range, to obtain a risk evaluation result. The risk evaluation result includes a risk category and a risk level.

[0079] Further, in the result output module, the visual display of the generated result comprises the following steps:

[0080] S1, data formatting, arranging the original result output by the data analysis module to meet the display requirements;

[0081] S2, visual processing, generating corresponding charts and reports according to the result data, wherein the charts include but are not limited to histograms, pie charts and line charts, and the reports include but are not limited to risk categories, risk levels, data sources and information explanations and diagnostic results;

[0082] S3, transmitting the formatted and visualized result to the display interface, wherein the interface includes but is not limited to the in-hospital system interface and the application program interface of the patient mobile terminal;

[0083] S4, correlation feedback, providing a user feedback entry for filling in questions about the screening result, self-symptom supplement and diagnostic supplement.

[0084] Through the system provided by the application, early and relatively accurate screening structure can be obtained, which can reduce unnecessary repeated examination of patients and alleviate the anxiety of patients waiting for the diagnosis result for a long time. At the same time, the patient can better understand the screening result and enhance the cognition of the health status, and improve the medical experience and satisfaction.

[0085] The above description is a detailed description of the preferred embodiment of the application, but the embodiment is not used to limit the scope of the patent application of the application, and any equivalent changes or modifications made under the technical spirit of the application should belong to the patent scope covered by the application.

Claims

1. An early screening system for pulmonary interstitial fibrosis based on big data analysis, characterized by: include, The data acquisition module is used to collect data related to pulmonary interstitial fibrosis, specifically medical imaging data, extract features from the medical imaging data, obtain feature data, and send it to the model verification module; a model verification module, connected to the data acquisition module through data transmission technology, for correlating the features of the medical imaging data extracted from the data acquisition module to obtain a plurality of correlated images, inputting each correlated image into the constructed early screening model for pulmonary interstitial fibrosis, obtaining an evaluation value of the characteristic coefficient of pulmonary interstitial fibrosis from the model, pairing the evaluation value with the true value of the training set in the model to obtain true positive examples, false positive examples, and false negative examples, and calculating the error scores of the true positive examples, false positive examples, and false negative examples, processing the plurality of correlated images according to the calculated error scores, and sending the processed correlated images to the data preprocessing module; The data preprocessing module is connected to the model verification module through data transmission technology, and is used to disassociate the processed correlated images, calculate eigenvalues ​​based on the disassociated features, and clean, denoise, and standardize the calculated eigenvalues ​​to obtain standardized eigenvalues; A data storage module, connected to the data preprocessing module through data transmission technology, for storing preprocessed data; The data analysis module is connected to the data storage module through data transmission technology and is used to analyze and judge the pre-processed data; The specific process is to calculate the pulmonary interstitial fibrosis characteristic coefficient L(W,b) based on the analytical calculation formula, and judge the patient's lung condition based on the calculation result; wherein, the analytical calculation formula is: Where W is the weight vector, b is the bias vector, m is the number of samples, and y i is the standardized eigenvalue, is the prediction function, and The calculation formula is Where T is the transpose and X is the eigenvector of the normalized eigenvalue; The result output module is connected to the data analysis module through data transmission technology. It is used to generate early screening results for pulmonary interstitial fibrosis based on the pulmonary interstitial fibrosis characteristic coefficient calculated by the data analysis module, and to visualize the generated results.

2. The early screening system for pulmonary fibrosis based on big data analysis according to claim 1, characterized in that: In the data acquisition module, the medical imaging data includes the patient's chest X-ray images and CT images. When extracting the features of the medical imaging data, feature extraction is performed on each image data, and the extracted features are compared with the corresponding images under normal conditions to obtain an abnormal feature area map of each image data. The obtained abnormal feature area map is the feature data.

3. The early screening system for pulmonary fibrosis based on big data analysis according to claim 1, characterized in that: In the model verification module, the method for constructing an early screening model for pulmonary interstitial fibrosis is as follows: S1. Obtain historical chest X-ray images and historical CT images of multiple patients with pulmonary fibrosis, perform feature extraction on the historical chest X-ray images and historical CT images respectively to obtain X-ray abnormality feature maps and CT abnormality feature maps, and integrate the X-ray abnormality feature maps and CT abnormality feature maps of the same patient into associated images as training samples; S2. Build an early screening model for pulmonary interstitial fibrosis based on a deep learning network. Divide the generated training samples into a training set and a test set in a ratio of 8:

2. S3, download the weight file and load it into the BP neural network to initialize the migration network parameters, and determine the number of hidden layer nodes of the BP neural network model according to the number of associated images in the training set; S4. Modify the last fully connected layer of the BP neural network, keep the input unchanged, set the output to the characteristic coefficient of pulmonary interstitial fibrosis, initialize the weights of the last layer, use the gradient descent algorithm for learning, and use fixed step size decay to optimize the training parameters. Retrain the entire network to obtain an early screening model for pulmonary interstitial fibrosis; S5. During the training process, small batches of related images are randomly and non-repeatedly extracted from the training set. Extracting all the related images in the training set is considered a training cycle. The training is completed after a certain number of iterations, and then the test set is used to evaluate the effectiveness of the early screening model for pulmonary interstitial fibrosis.

4. The early screening system for pulmonary fibrosis based on big data analysis according to claim 1, characterized in that: In the model checking module, the error score is calculated as follows: In the formula, the error rate represents the proportion of all misclassified samples to the total number of samples, which is used to measure the overall error rate of the classifier; TP is the true positive example, which indicates the number of samples that are actually positive examples and are correctly classified as positive examples by the classifier; FP stands for false positive, which means the number of samples that are actually negative but are mistakenly classified as positive by the classifier; FN is a false negative example, which means the number of samples that are actually positive examples but are mistakenly classified as negative examples by the classifier; TN is the true negative example, which indicates the number of samples that are actually negative examples and are correctly classified as negative examples by the classifier.

5. The early screening system for pulmonary fibrosis based on big data analysis according to claim 1, characterized in that: In the model verification module, when multiple associated images are processed according to the calculated error scores, the processing methods include image elimination and image re-verification.

6. The early screening system for pulmonary fibrosis based on big data analysis according to claim 1, characterized in that: In the data preprocessing module, when calculating the eigenvalue, the area of ​​the abnormal feature area in each feature is calculated, and the obtained area data is the eigenvalue.

7. The early screening system for pulmonary fibrosis based on big data analysis according to claim 1, characterized in that: In the data preprocessing module, the following calculation formula is used to standardize the received data: In the formula, y is the eigenvalue, y i is the standardized eigenvalue, μ is the mean of the feature, and σ is the standard deviation of the feature.

8. The system for early screening of pulmonary fibrosis based on big data analysis according to claim 1, characterized in that: The data storage module adopts a distributed storage architecture to store different types of data on different nodes in the blockchain.

9. The system for early screening of pulmonary fibrosis based on big data analysis according to claim 1, characterized in that: In the result output module, the generation of early screening results for pulmonary interstitial fibrosis is based on the comparison of the pulmonary interstitial fibrosis characteristic coefficient calculated by the data analysis module with the preset characteristic judgment range to evaluate the risk of pulmonary interstitial fibrosis and obtain the risk assessment result, which includes the risk type and risk level.

10. The system for early screening of pulmonary fibrosis based on big data analysis according to claim 1, characterized in that: In the result output module, the visual display of the generated results includes the following steps: S1. Data formatting: arranging the original results output by the data analysis module to meet the display requirements; S2. Visualization processing: generating corresponding charts and reports based on the result data. The charts include but are not limited to histograms, pie charts, and line charts. The reports include but are not limited to risk type, risk level, data source, information description, and diagnosis results. S3. Transmitting the formatted and visualized results to a display interface, including but not limited to an in-hospital system interface and an application program interface of a patient's mobile terminal; S4. Related feedback provides a user feedback portal for filling in questions about the screening results, supplementing one's own symptoms and supplementing the diagnosis.

Images