60 results about "Chest ct" patented technology

The invention relates to the technical field of intelligent medical treatment, and discloses a connective tissue disease related lung disease early screening method and system based on deep learning, and the method comprises the steps: obtaining clinical data, a high-resolution chest CT image and biomarker data of a to-be-screened target; the clinical data, the high-resolution chest CT image and the biomarker data are input into a deep learning feature extraction and fusion module, high-dimensional clinical features, image feature vectors and biomarker time sequence feature vectors are extracted and fused, and fused feature vectors are obtained; and inputting the fusion feature vector into an early screening and typing module, and outputting a lung abnormal risk detection result. According to the application, the effects of early diagnosis of connective tissue disease related lung diseases and timely intervention can be achieved.

The invention provides a pneumonia CT image diagnosis model training method, diagnosis method and equipment, and the training method comprises the steps: inputting a 3D chest CT image into a multi-task deep learning model, enabling a shared encoder in the model to extract multi-scale feature data, and enabling a connection module and a decoder to obtain pneumonia focus region prediction result data according to the multi-scale feature data, the classification head obtains pneumonia diagnosis type prediction result data according to the multi-scale feature data; determining the joint loss of the model in the current iteration round and updating model parameters; and if the current multi-task deep learning model satisfies a training termination condition, outputting the current model as a pneumonia CT image diagnosis model. According to the method, the problems of low model feature utilization rate and low pneumonia diagnosis process efficiency caused by incapability of simultaneously completing focus segmentation and type classification due to task simplification of an existing pneumonia diagnosis model can be solved.

The invention discloses a pulmonary interstitial image analysis method and system based on clinical prior guidance feature fusion, and the method comprises the steps: obtaining chest CT images of a user at a plurality of time points and corresponding clinical data, and carrying out the data preprocessing and region segmentation; performing feature extraction on the automatically segmented chest CT image and the corresponding clinical data by using specific indexes; specific time points are coded into time embedding vectors, total image feature vectors are projected to the same dimension as the time embedding vectors through a linear layer, and time coding fusion is generated; the most relevant CT image follow-up visit time points are actively'inquired 'and'weighted' by using clinical risk factors, time sequence image feature fusion is carried out, and fusion features after weighted fusion are output; and carrying out progress probability calculation by using the fusion features, predicting the progress risk in the next one year according to the calculation result, and realizing dynamic time sequence feature selection driven by clinical priori knowledge.

The invention provides a construction method of an IPA recognition model, an IPA recognition method, terminal equipment and a storage medium. The construction method comprises the following steps: constructing a training sample set by collected clinical metadata and chest CT image samples; inputting to a pre-constructed initial neural network model for iterative training; wherein the feature extraction layer outputs a focus area prediction result of a chest CT image sample and corresponding image features based on a preset segmentation network, and extracts clinical features; the feature fusion layer fuses the received features and outputs primary fusion features; and the diagnostic reasoning layer maps the weighted primary fusion features into an IPA disease prediction probability based on a full-connection layer network, the IPA disease prediction probability is used as the output of an initial neural network model, training is ended when the segmentation loss function and the preset constraint term both reach preset values, and an IPA recognition model is obtained. According to the method, the timeliness and the accuracy of identifying the invasive pulmonary aspergillus disease can be improved.

The invention provides a chest CT image automatic segmentation and interstitial pneumonia prediction method and system, and the method and system achieve the automatic segmentation of a CT image through deep learning, greatly improve the image analysis efficiency, reduce the error of manual intervention, and reduce the judgment deviation caused by subjective factors. Quantitative features are combined with a statistical model, small changes of lung images are effectively captured, and the accuracy of IP risk prediction is remarkably improved. According to test data, in a pneumonia patient sample, the accuracy rate of model prediction IP is improved to 85% or above, the method provides timely IP risk prompts for clinicians based on an early warning function of CT image segmentation and feature analysis, and is beneficial to optimizing a treatment scheme, delaying or reducing the probability of occurrence of serious complications, and improving the accuracy rate of the model prediction IP in the pneumonia patient sample. According to the automatic image analysis and prediction system, the working pressure of image doctors can be relieved, the film reading burden is greatly reduced, and the response efficiency of a medical system is improved.

The invention relates to the technical field of medical image processing, and discloses a chronic obstructive pulmonary emphysema distribution quantitative method and system based on multi-scale segmentation, and the method comprises the steps: carrying out the anisotropic diffusion filtering noise reduction of a chest CT image; segmenting a lung field and removing a blood vessel bronchial structure by adopting a region growing algorithm; multi-scale image representation is constructed based on a Gaussian pyramid, an emphysema candidate area is identified in a coarse scale layer, and a boundary is accurately drawn by adopting a self-adaptive threshold value in a fine scale layer; extracting local texture features to distinguish the lobular central emphysema and the total lobular emphysema; dividing severity levels according to spatial aggregation characteristics and density gradient distribution, and calculating an air swelling volume ratio and a distribution heterogeneity index; the three-dimensional pseudo-color volume is used for drawing visualization, a structured quantitative report is generated, accurate segmentation and subtype classification of the emphysema area are achieved, and comprehensive quantitative analysis indexes are provided.

The invention provides an airway geometric parameter measurement method and system, and relates to the technical field of medical image processing, and the method comprises the steps: obtaining airway foreground voxel data corresponding to a chest CT, and converting a voxel index coordinate into a world physical coordinate according to image space meta-information; obtaining a center line point set through three-dimensional skeletonization, establishing candidate connecting edges in a preset neighborhood, generating a topological tree based on a strategy with branch penalty terms, and completing branch segmentation and generation coding; the method comprises the steps that firstly, a tail end branch meeting a preset condition serves as a starting point for backtracking, the section area and the section equivalent diameter of each node slice are calculated, branch node direction vector score cross angle parameters are fitted, finally, a structured file is summarized and output according to algebraic labels and other dimensions, and stable measurement and standardized output of airway multi-scale geometric parameters can be achieved. And the comparability and repeatability of group scale analysis are improved.

The present application relates to the technical field of medical image processing, in particular to a CT image intelligent analysis system for pneumonia auxiliary screening. The present application firstly pre-processes chest CT images and detects candidate lesion regions; then extracts topological features, deep convolution features and texture statistical features based on persistent homology theory for each candidate lesion, performs feature fusion through a multi-head self-attention mechanism to generate a unified lesion representation vector; then maps the lesion representation vector to a pre-constructed radiology knowledge graph, performs reasoning using a graph neural network, and outputs the pneumonia suspected probability and lesion classification for each lesion; finally, the evidence theory is used to fuse and quantify the uncertainty of the analysis results of multiple lesions to generate a comprehensive screening report. The present application effectively identifies complex morphological lesions through topological features, accurately identifies lesion types through knowledge graph reasoning, and provides quantitative diagnosis uncertainty through the evidence theory.

The invention relates to a pulmonary nodule auxiliary detection method based on multichannel nonlinear density adaptive mapping, and belongs to the field of image processing, and the method comprises the steps: processing original chest CT data to generate pulmonary nodule candidate ROI image blocks; simultaneously generating a first reference channel and an enhanced mapping channel for each pulmonary nodule candidate ROI image block; for an enhanced mapping channel, according to physical density distribution of the pulmonary nodules, performing functional segment division on a physical density space, and adopting a differentiated continuous mapping mechanism to enable a target density interval where the pulmonary nodules are located to occupy a larger characterization bandwidth; and inputting into the deep neural network by adopting a joint input mode of a first reference channel and an enhanced mapping channel, and outputting one or more of a pulmonary nodule candidate probability graph, a candidate score and a candidate region. According to the invention, through joint input of the first reference channel and one or more enhanced mapping channels, the discrimination capability of a downstream detection network is improved.

The application discloses a lung bullae recognition and interventional navigation planning method based on medical topology analysis, and comprises the following steps: obtaining an affine omics feature set and a clinical variable set based on chest CT images and clinical text data; performing rule-preferred feature selection and neural symbol reasoning to generate a knowledge-enhanced feature vector; inputting the three kinds of features into a classifier to obtain lung bullae recognition probability and recognition results; inputting the chest CT images into a double-flow cascaded segmentation network to obtain a lung bullae lesion mask and an airway tree mask; performing three-dimensional skeleton extraction on the airway tree mask to obtain a center line, and constructing an airway topology graph, constructing a lung bullae three-dimensional surface based on the lung bullae lesion mask, and determining a drainage entrance node; and performing reverse backtracking search based on the drainage entrance node in the airway topology graph to obtain a target drainage bronchus path. The application breaks through the end-to-end technical chain from transparent diagnostic stratification to high-fidelity three-dimensional intelligent interventional path planning of lung bullae, and significantly reduces the risk of surgical misoperation.

This invention discloses an automated multi-parameter assessment method, device, and medium for aortic imaging based on three-dimensional images. The method includes acquiring three-dimensional chest CT images and preprocessing them; extracting and optimizing an initial centerline based on the obtained three-dimensional vascular mask; dividing the initial centerline into ascending aorta, aortic arch, and descending aorta segments using the optimized initial centerline and the opening positions of branches on the aortic arch; extracting the average CT value of the adipose tissue surrounding the descending aorta based on the divided descending aorta segments; constructing orthogonal cross sections, performing dimensionality reduction and projection processing on each cross-sectional mask, and calculating the major and minor axes based on the farthest point pairs, outputting the major axis diameter sequence of each cross section; calculating the aortic calcification load, assessing aortic dilation and morphological variations, and evaluating the inflammatory status of the perivascular adipose tissue; and generating a comprehensive report of indicators. This invention fully utilizes existing imaging resources, improving the utilization efficiency and clinical value of medical imaging data.

This invention relates to a multi-organ biological age and disease risk assessment system based on chest CT radiomics, belonging to the field of medical image computer-aided analysis technology. The invention aims to address the problem that a single global image age cannot characterize organ-specific aging and lacks a closed-loop clinical risk quantification mechanism. The technical solution includes: receiving chest CT radiomics features and demographic parameters of the subject through an input layer; performing spatial mapping using a feature layer; outputting a 10-dimensional biological age through a three-stage cascaded machine learning model in the age prediction layer; calculating the age acceleration rate through a bias correction layer; quantifying the disease risk multiple by coupling a Cox proportional hazards model through a risk assessment layer; and finally, generating clinical interpretation through a large language model driven by an intelligent reporting layer. This invention achieves low-cost and non-invasive organ-level aging assessment, accurately capturing allometric aging, and improving the efficiency of clinical risk stratification and the standardization of interpretation.

The application relates to an interactive bronchoscope navigation system and method, the navigation system comprising: a CT data acquisition module for acquiring chest CT data of a patient; a navigation data acquisition device for simulating the insertion operation of a camera at the head end of an insertion tube of a bronchoscope based on the chest CT data and obtaining insertion position data and posture data of the camera; a digital bronchial tree generation module for generating a digital bronchial tree based on the chest CT data; a navigation data mapping module for mapping the insertion position data and the posture data of the camera into the digital bronchial tree, performing motion calculation of a virtual camera under the constraint of a digital airway lumen of the digital bronchial tree, obtaining the pose of the virtual camera in the digital airway lumen, and realizing simulation operation. The simulation operation of the bronchoscope can be realized, the digital airway of the patient can be preoperatively preoperated, and the operation efficiency is improved, the operation risk is reduced, and the medical cost is saved.

This invention discloses a method and system for lung interstitial imaging analysis based on clinically prior-guided feature fusion. The method includes: acquiring chest CT images and corresponding clinical data at multiple time points from the user, performing data preprocessing and region segmentation; extracting features from the automatically segmented chest CT images and corresponding clinical data using specific indicators; encoding specific time points into temporal embedding vectors, projecting the total image feature vectors through a linear layer to the same dimension as the temporal embedding vectors to generate temporal encoding fusion; actively "querying" and "weighting" the most relevant CT image follow-up time points using clinical risk factors, performing temporal image feature fusion, and outputting the weighted fused features; calculating the progression probability using the fused features, and predicting the risk of progression in the next year based on the calculation results. This invention achieves dynamic temporal feature selection driven by clinical prior knowledge.

This invention belongs to the field of medical image processing technology, specifically relating to a method and apparatus for quantitative analysis of microvessels in pulmonary nodules in chest CT images. The analysis method of this invention includes the following steps: Step 1, preprocessing the chest CT image; Step 2, segmenting the chest CT image using a convolutional neural network model to obtain images of the pulmonary airways, pulmonary vessels, and lung tissue; Step 3, detecting pulmonary nodules in the chest CT image using the convolutional neural network model, extracting lesions, and extracting microvessels around the lesions and within the pulmonary nodules; Step 4, calculating the microvessel density value based on the extraction results of Step 3. This invention also provides an apparatus suitable for the above analysis method. This invention can non-invasively, rapidly, and accurately display the microvessel density value in pulmonary nodules and around lesions in three dimensions, and provide precise quantitative data. It has significant clinical and socioeconomic value.

This invention discloses a method for constructing an interpretable predictive model of pathogenic bacteria in chest CT scans, belonging to the field of artificial intelligence technology. The method includes: Step S1: acquiring a chest CT image dataset and pathogenic bacteria type labels, and constructing a pathogenic bacteria imaging knowledge graph using a medical knowledge base; Step S2: constructing an initial model, which includes: a noise-invariant encoder, a counterfactual causal intervention module, a knowledge-guided feature compensation module, an interpretability analysis module, and a classifier; Step S3: feeding the spatial attention weight map back to the noise-invariant encoder to update and iterate the anatomical pathological features. Based on the final features obtained after internal iterative optimization, the initial model is trained according to preset training rules until overall convergence. This invention effectively distinguishes between real lesions and noise artifacts in low-dose CT images through causal intervention and attention map difference-driven feature updates, improving the model's recognition accuracy and robustness.

The application relates to the technical field of image analysis, in particular to a respiratory system lung nodule tumor cell detection method and system based on image feature fusion and a storage medium, which comprises the following steps: constructing a two-dimensional pixel gray array of a chest CT image and performing edge positioning, constructing a closed image boundary through gray difference analysis and gradient direction continuity, extracting a lung nodule candidate region image in the closed region, establishing pixel adjacency relations in horizontal, vertical and diagonal directions, constituting a direction difference image set and extracting a continuous texture region with a closed structure, generating a lung nodule edge contour connection image through spatial relationship mapping and pixel connection extension, and finally performing boundary fusion and pixel structure recombination to form a target detection image. In the application, the direction gray difference construction and boundary closure analysis are combined, the connected path judgment and pixel aggregation processing are fused, the texture structure concentration and image recognition definition are effectively enhanced, and the accuracy and readability of lung nodule detection are improved.

This invention discloses a method for constructing a lung weight index based on chest CT images and its application. The method includes: Step 1, low-dose CT image acquisition and preprocessing; Step 2, automatic segmentation of the lung and lung lobes; Step 3, calculation of the average density of lung lobes and conversion to physical density; Step 4, calculation of lung lobe volume; Step 5, calculation of lung lobe weight and whole lung weight; Step 6, construction of the lung weight index; Step 7, risk assessment based on the lung weight index. This method can be applied to practical scenarios such as lung cancer screening, health checkups, and image-assisted diagnosis. This invention comprehensively reflects the state of lung tissue, blood vessels, and interstitium through the lung weight index, providing a more holistic quantitative description. Furthermore, it does not require nodule detection as a necessary prerequisite, making it suitable for individuals with few nodules, small volume, or those who have not yet formed a definite mass, thus better meeting the practical needs of early lung cancer screening.

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.

The application discloses a kind of multi-modal fusion model construction method and system for malnutrition assessment, comprising: obtaining the chest CT image data of evaluation object, clinical index and nutritional status classification label, constructs sample data set;Sample data set is divided into training set, test set and verification set according to preset proportion;Extraction skeletal muscle imageomics features of chest CT image data in training set, target imageomics features are output by multistage dimension reduction screening;From the training set, extract the clinical index matched with the target imageomics features, and splice with the target imageomics features, construct the fusion feature set with label;Based on the fusion feature set, the nutritional status classification label is used as the supervision signal, and the machine learning classification model is trained and optimized to construct a multi-modal fusion model;The T12 level skeletal muscle features and clinical indicators are fused by the application, which significantly improves the prediction performance of malnutrition assessment, and has important clinical guiding value.

The invention relates to a two-stage pulmonary nodule detection method and system based on cross-layer attention fusion, and belongs to the technical field of medical image processing, and the method comprises the steps: carrying out the preprocessing of a three-dimensional chest CT image, segmenting a pulmonary parenchyma region, and carrying out the standardization of the pulmonary parenchyma region, and obtaining a to-be-detected image; inputting the to-be-detected image into the candidate nodule detection network, and generating a position frame of a candidate nodule; extracting a three-dimensional image region corresponding to the candidate nodule position frame, inputting the three-dimensional image region into a false positive suppression network for classification, and filtering out false positive candidates; and outputting a final pulmonary nodule detection result confirmed by the false positive suppression network. Through the design of dual-stage task decoupling, an innovative space and channel residual module and a multi-scale progressive sensing network module and a self-adaptive training strategy, the core problems of sensitivity and specificity imbalance, weak feature discrimination ability, poor multi-scale adaptability and unstable training in intelligent detection of pulmonary nodules are systematically solved.

This invention belongs to the field of medical auxiliary prediction model construction technology. Current prediction models for refractory Mycoplasma pneumoniae pneumonia are all multimodal and rely on time-consuming experimental results, making them unsuitable for the clinical need to quickly determine the probability of having refractory Mycoplasma pneumoniae pneumonia. This invention proposes a method, device, medium, and program product for constructing a prediction model for refractory Mycoplasma pneumoniae pneumonia. By developing a Transformer-based deep learning framework, it achieves automated and efficient risk identification of RMPP using baseline non-contrast chest CT without additional examinations. Furthermore, a first prediction model is constructed based on the risk probability values ​​output by the Transformer-based deep learning framework, and a second prediction model is constructed based on the risk probability values ​​and six clinical data points. The predicted probabilities of each model show good consistency with the observed results.

The invention relates to the technical field of image analysis, in particular to a coronary artery calcification integral prediction method based on a chest CT image, and the method comprises the steps: obtaining chest CT images, corresponding to different preset layers, collected by a to-be-detected patient at the current moment, and carrying out the connected domain extraction of each chest CT image; determining a target similarity between different connected domains in the chest CT image corresponding to each adjacent preset layer; screening out connected domains representing the same tissue structure; performing blood vessel change rule analysis processing on each connected domain sequence; screening out a blood vessel connected domain; and based on the gray value, screening out a target calcification region from a region formed by all the blood vessel connected domains in each chest CT image, and performing blood vessel calcification integral prediction based on the target calcification region. According to the method, the blood vessel calcification region is identified by analyzing the chest CT image, and the efficiency of identifying the blood vessel calcification region is improved, so that the efficiency of predicting the blood vessel calcification integral is improved.

A method for segmenting and 3D visualization of chest CT images, belonging to the field of medical image processing and computer-aided diagnosis technology, comprises the following steps: S1, importing and standardizing preprocessing of medical image data; S2, segmenting of multiple tissue images of the thoracic cavity based on a hybrid strategy; S3, dual-mode 3D geometric reconstruction; and S4, interactive 3D visualization and rendering. Through these steps, this invention directly achieves accurate segmentation of multiple tissues and dual-mode 3D reconstruction of surfaces and volumes without relying on high-end hardware or complex manual operations, thus improving image analysis efficiency.

The application relates to the technical field of image processing, and discloses a CT image reconstruction and bronchoscope image fusion system based on trachea running, which first extracts the three-dimensional structure of a trachea and bronchi from a chest CT original image, generates a cavity perspective CT reconstruction image consistent with the trachea running based on a trachea center line, further acquires a bronchoscope real-time video stream, realizes spatial registration of the cavity perspective CT image and the real-time video through three-dimensional coordinate mapping, and fuses and renders the registered CT image and the bronchoscope video in a gray base and color overlay mode to obtain a fusion image with mucosa details and structural background; the system can also plan a bronchoscope insertion path based on the cavity CT image, and supports interactive and associated display among the original CT image, a three-dimensional trachea tree model and the cavity perspective CT image. The application can form a planning path before bronchoscope examination, provide real-time structural guidance in the process, and form a multi-modal fusion image after completion.

The invention discloses a method for evaluating sepsis lung injury by detecting the crosslinking degree of type VI collagen, and relates to the technical field of medical diagnosis and monitoring. The method at least comprises the following steps: S1, firstly, carrying out sample collection and multi-index detection; s2, performing imaging feature extraction, and automatically identifying fine structure changes in the chest CT image by using a deep learning algorithm; and S3, based on the feature extraction in the S2, establishing a corresponding quality control algorithm, a basic damage scoring model, a dynamic progress prediction model and a comprehensive risk assessment model. The accurate, early-stage and dynamic sepsis lung injury assessment method is provided through multi-dimensional biomarker detection, iconography feature extraction and a dynamic algorithm model, the early diagnosis rate of sepsis lung injury can be remarkably improved, the disease progress can be effectively monitored, the treatment effect can be effectively assessed, and the application prospect is wide. The method has remarkable clinical application prospects and economic benefits, and particularly has important value in intensive care and formulation of personalized treatment schemes.

The application discloses a diffuse lung disease CT classification method based on MambaVision and density-frequency domain double-prior enhancement, acquires a chest CT sequence of a patient to be classified, pre-processes each CT slice, and extracts HU statistics of a lung field region; the pre-processed CT slice and the HU statistics are jointly input into a classification network model, the model takes MambaVision as a backbone network, and fuses a HU-SE module and an MFA module, wherein the HU-SE module generates channel attention weights by using the HU statistics, adaptively recalibrates feature channels, the MFA module decomposes feature maps into low, medium and high frequency bands and fuses the feature maps by using learnable weights, respectively enhances the inter-class discrimination ability of image features from two dimensions of density prior and frequency domain features, and outputs a slice-level classification probability vector; finally, a patient-level aggregation strategy is used to fuse multi-slice prediction results, and a patient-level diffuse lung disease classification diagnosis result is output. The application effectively improves the separability between diffuse lung disease types with similar morphologies.

The utility model discloses a kind of absorbable rib fixation nails, made of high molecular absorbable material integrated, including shank part and the clamping portion of both ends of shank part, clamping portion is equipped with bifurcation structure with opening towards non-connection shank part one end, at least one branch outside of bifurcation structure is equipped with inverted tooth structure, inverted tooth structure end is conical or triangular, the width of bifurcation structure is greater than the width of shank part.The utility model can be made into nail body by absorbable material, clamping portion end bifurcation, inverted tooth structure can be fixed after being expanded in marrow cavity and increasing friction, wide conical end is easily placed into marrow cavity and can play anti-rotation, anti-removal effect, fixed firm and not easy to displace, without suture reinforcement to prevent affecting periosteal blood supply and nerve compression, after fracture healing, fixation nail can be degraded and absorbed, without secondary operation removal, after absorption, chest CT, MRI examination has no influence, more promotional, can better improve the user's use experience.

The invention discloses a system and method for diagnosing pulmonary tuberculosis-pulmonary arterial hypertension through artificial intelligence chest CT imaging, and the system comprises a data acquisition module which is used for collecting a chest CT image of a patient in the supine position full inspiration period, and exporting the CT image in a DICOM format; the data processing module is used for carrying out full-lung automatic segmentation and lung lobe segmentation VTDLT calculation by using an AI algorithm through AI image analysis software; the quantitative analysis module is used for outputting the damage volume and the number and distribution of the affected lung lobes; the risk prediction module is used for combining the VTDLT with clinical parameters to construct a PH risk prediction model; according to the application, the pulmonary arterial hypertension is screened by combining CT imaging with an artificial intelligence algorithm, an original mode of screening through right heart catheter examination is not changed, a noninvasive, early-stage and accurate PH screening tool is provided for TDL patients, the use requirements of RHC are reduced, and the pain and medical burden of the patients are relieved.

The application provides a lung EGFR mutation recognition model based on CT sequence images, and belongs to the field of medical image analysis; the model comprises: a three-window preprocessing module configured to generate three-window CT sequence images according to original chest CT images, wherein the three windows comprise a lung window, a mediastinal window and a bone window; a sequence generation module configured to obtain three-window sequence images with consistent structures according to the three-window CT sequence images; a time sequence modeling module configured to perform frame-by-frame feature coding on the three-window sequence images, model the cross-slice spatio-temporal relationship of the coded slice features, and output sequence-level time sequence features fused with global spatio-temporal information; a sequence aggregation module configured to aggregate the sequence-level time sequence feature vectors, evaluate and weight the contribution of each slice to the final diagnosis, and generate patient-level fusion features; and a nonlinear classification module configured to output a prediction result of EGFR gene mutation according to the patient-level fusion features.