35 results about "Chest radiograph" patented technology

The invention discloses an X-ray chest radiography report automatic generation method, system and device based on factual description enhancement and a medium. The method comprises the following steps: firstly, constructing a medical entity extraction method based on a RadGraph model, and carrying out identification and structured extraction on clinical keywords to obtain factual description consisting of key medical entities; secondly, establishing a comparative learning method guided by factual description, enhancing semantic consistency between the image and the text from global and local levels, and extracting visual features with diagnostic value; establishing a historical similar case retrieval strategy independent of disease tags, and calculating visual semantic similarity to realize automatic retrieval of historical cases; and finally, proposing an evidence-driven chest radiography report generation method, constructing a cross-modal fusion network, and generating a chest radiography report with clinical accuracy and consistency. The system, the equipment and the medium automatically generate an X-ray chest radiography report based on factual description enhancement based on the method; according to the method, efficient and stable automatic retrieval is realized, the clinical accuracy of the generated chest radiograph report and the reliability of evaluation are improved, and the universality and robustness of the model are remarkably improved.

The invention discloses a chest radiograph report generation and focus positioning method and system based on reinforcement learning, and belongs to the technical field of artificial intelligence and medical image analysis crossing. According to the method, firstly, a basic model is supervised and fine-tuned by using a multi-source chest radiograph image and a query text; and then optimizing the strategy model by adopting a GRPO reinforcement learning framework, introducing an indicator function guided by token entropy into a target function, and finally realizing parallel generation of an end-to-end chest radiograph report text and focus bounding box coordinates by the optimized strategy model. According to the method, visual positioning of the focus is realized while the generation quality of the chest radiograph report is improved, and the integrated output form of the method better fits the clinical actual working process.

The invention discloses a multi-label classification method based on key region selection and cross-layer feature fusion. Comprising the following steps: carrying out standard preprocessing operation on a chest radiograph image, and carrying out multi-label coding processing on an image label; inputting the preprocessed image into an OfficientNet-B0 network to extract features, performing dimensionality reduction to obtain an original feature sequence, and splicing the original feature sequence with learnable category embedding and position codes to form an embedding sequence; inputting the embedded sequence into a Transform encoder, and extracting deep features of the image through a multi-head attention mechanism and a multi-layer perceptron; constructing a key area selection module, selecting a discriminative area by using an attention mechanism of an encoder and filtering irrelevant information; constructing a cross-layer feature fusion module, fusing features of different levels in the encoder, and inputting the features into a classifier to obtain a multi-label classification result; and designing adaptive class balance loss, dynamically adjusting the attention on different classes of positive and negative samples, and optimizing model parameters through back propagation. The method is an optimization method for effectively improving image multi-label classification performance.

The invention discloses a chest radiograph report generation method based on thinking chain reasoning and active learning. The method comprises the following steps: collecting a multi-view chest radiograph image and a corresponding report; performing view angle identification and view angle classification on the multi-view-angle chest radiograph image; a shared visual encoder is adopted to extract sequence features of all visual angles, and a multi-visual-angle visual token is generated through cross-visual-angle feature fusion; constructing a systematic hierarchical diagnosis process, sequentially executing structured five-stage chain reasoning for each anatomical structure in the multi-view chest radiography image, and generating a structured reasoning prompt; inputting the multi-view visual token and the structured reasoning prompt into a large language model, and generating a target sequence through an autoregression mode; constructing a scoring function based on joint scoring of uncertainty and inconsistency; labeling the samples in the unlabeled sample pool according to the sample scores in each round of iteration; and inputting a to-be-tested multi-view chest radiograph image into the trained language model, and generating a thinking chain and a complete report text.

The invention discloses an intelligent assessment method and system for coronary stenosis based on chest radiography, and relates to the technical field of health risk assessment, and the method comprises the steps: obtaining chest radiography image data of a target patient; based on the time sequence fluctuation of the calcification focus area in the chest radiograph image data, calculating to obtain a calcification activity index, determining a cardiac pulsation key point in the chest radiograph image data, and based on the displacement change of the cardiac pulsation key point between each image frame, calculating to obtain a collateral circulation compensation index of the calcification focus area; based on the collateral circulation compensation index and the calcification activity index, calculating a dynamic risk factor corresponding to the coronary artery stenosis; and inputting the dynamic risk factor into a preset neural network model for iterative training, and performing prediction processing on the chest radiograph image data based on the trained preset neural network model to obtain a coronary artery stenosis risk assessment result. The technical effect of reducing the clinical misjudgment risk of coronary artery stenosis is achieved.

The invention provides a chest radiograph image recognition method and device, and relates to the technical field of image processing. The method comprises the following steps: acquiring a chest radiograph image to be recognized; inputting the chest radiograph image into a pre-trained identification model to obtain an identification probability of each preset disease tag corresponding to the chest radiograph image; the recognition model is used for performing recursive multi-scale feature division and spatial reconstruction on the chest radiograph image, and performing interactive updating between a preset disease tag and features in the feature division process of each scale level to obtain tag semantics with global perception; and determining the recognition probability of each preset disease tag corresponding to the chest radiograph image based on the tag semantics with global perception. According to the invention, the recognition precision of complex focuses with variable scales and different shapes in the chest radiograph can be improved.

The invention relates to an X-ray chest radiography spine offset detection method and system based on a prior probability graph, and the method comprises the steps: obtaining a to-be-detected X-ray chest radiography image, and carrying out the normalization of the transverse pixels of the image; obtaining a corresponding prior probability value through linear interpolation in pre-constructed one-dimensional prior probability distribution based on the normalized position of each column of transverse pixels of the image, constructing a one-dimensional array, copying the one-dimensional array in the vertical direction, and constructing a two-dimensional prior probability graph; inputting an X-ray chest radiograph image to be detected into the GPU accelerated segmentation network to obtain a spine binary mask; and based on the two-dimensional prior probability graph, calculating an average prior probability and an offset probability in the mask area of the spine binary mask, comparing the average prior probability and the offset probability with a preset offset threshold value, if the offset probability is greater than the offset threshold value, determining that the spine is offset, otherwise, determining that the spine is basically centered. Compared with the prior art, the method has the advantages that the offset detection result has better interpretability, stability and quantificaiton.

The invention provides a chest radiography multi-label classification method and system based on a cross-modal memory network, and the method comprises the steps: obtaining chest radiography data, and carrying out the preprocessing of the chest radiography data; the chest radiograph data comprises an original chest radiograph image and corresponding label information; escaping the label information to map the label information to a uniquely corresponding text identifier; training a chest radiograph classification model established based on a cross-modal memory network (CMN) by using the preprocessed chest radiograph image and the transferred label information; and realizing a multi-label classification task of the chest radiograph by utilizing the chest radiograph classification model. According to the chest radiograph multi-label classification method and system, key features in chest radiograph images can be effectively captured and fused, effective alignment and fusion between medical terms and labels are achieved at the same time, and therefore the accuracy and interpretability of a classification model are improved.

The application discloses a kind of based on chest radiograph coronary artery stenosis intelligent evaluation method and system, it is related to health risk assessment technical field, the method includes: obtaining the chest radiograph image data of target patient;Based on the time sequence fluctuation of calcification focus area in chest radiograph image data, the calcification activity index is calculated, determine the heart beat key point in chest radiograph image data, based on the displacement change between each image frame of heart beat key point, the collateral circulation compensation index of calcification focus area is calculated;Based on collateral circulation compensation index and calcification activity index, the dynamic risk factor corresponding to coronary stenosis is calculated;Dynamic risk factor is input to the preset neural network model and is iteratively trained, and based on the preset neural network model after training is completed to chest radiograph image data is predicted to process, obtains coronary stenosis risk assessment result.The application reaches the technical effect of reducing the clinical misjudgment risk of coronary artery stenosis.

The invention provides a chest radiograph abnormal region detection method and a program product, and the method comprises the steps: adding an FAHG module on the basis of YOLOv11, and carrying out the frequency domain enhancement and noise suppression of a feature map outputted to a detection head; in the FAHG module, Fourier transform is firstly carried out on the feature map; then low-pass, band-pass and high-pass filtering is executed respectively, and low-frequency, intermediate-frequency and high-frequency sub-band components are obtained through decomposition; enhancement processing is carried out on the three sub-band components, cross-channel information interaction and feature expression are enhanced for the low-frequency sub-band components and the intermediate-frequency sub-band components, and noise is suppressed for the high-frequency sub-band components; and after the enhanced low-frequency sub-band component, the enhanced intermediate-frequency sub-band component and the enhanced high-frequency sub-band component are respectively subjected to inverse Fourier transform, weighted fusion is carried out. Compared with other existing YOLO series models, the method for detecting the abnormal area of the chest radiograph image has high detection precision and generalization ability, the detection precision is remarkably improved under the condition that the calculated amount is not obviously increased, and good balance is achieved between efficiency and precision.

The invention discloses a tuberculosis auxiliary diagnosis system fusing chest radiography and breath sound characteristics. The tuberculosis auxiliary diagnosis system comprises a breath sound acquisition module, a breath sound dynamic analysis module, a chest radiography acquisition module, a chest radiography detection module, an activity discrimination fusion module and a result output module. Breathing sound signals are collected at a plurality of standard point positions on the body surface of a patient, dynamic abnormal indexes are extracted, meanwhile, chest radiography images are obtained, focus attribute information is recognized, an activity judgment fusion module conducts consistency analysis on the chest radiography images and the dynamic abnormal indexes, standardized risk scores are generated, and high-risk, low-risk or uncertain grading results are output. According to the method, the misdiagnosis and missed diagnosis risks can be effectively reduced under the complex conditions of old focus combined activities, early hidden cases, combined chronic obstructive pulmonary disease and the like, and a visual diagnosis report containing a chest radiograph thermodynamic diagram and a breath sound frequency spectrum is provided. Compared with single chest radiography or single breath sound analysis, the sensitivity and specificity of active tuberculosis diagnosis are improved, and the method has high clinical application value.

The application discloses a chest radiograph rib extraction method based on an adversarial generation network, and comprises the following steps: obtaining a chest radiograph image pair through dual-energy subtraction to construct a data set, wherein the chest radiograph image pair comprises one frame of standard chest radiograph I C , one frame of soft tissue image I S and one frame of bone image I B ; performing gray histogram equalization processing on the standard chest radiograph I C , the soft tissue image I S and the bone image I B respectively to obtain I CB and I CS ; using pairs of I C , I CB and I CS from the same individual to construct a rib generation data set, constructing rib boundary supervision data sets according to I C , I B and I CS of each individual in the data set according to a construction rule; inputting the rib boundary supervision data set into a rib supervision network, calculating the similarity loss between three feature maps of the image pair, inputting the constructed gray image into the rib supervision network for training, and completing the training of the rib supervision network; constructing a rib image generation network and training the rib image generation network; and inputting the collected chest radiograph image into the trained rib image generation network to obtain the rib image in the chest radiograph image.

This invention proposes an automated diagnostic report generation method based on pixel-level localization evidence from chest X-rays, belonging to the fields of medical image processing and artificial intelligence. By constructing a neural network comprising multimodal feature extraction, a large language model, image segmentation, and collaborative output modules, this invention achieves pixel-level localization evidence for lesion text, enabling reliable annotation of lesion descriptions in the diagnostic report within the image. During the training phase, visual and text encoders extract image and text features respectively, the large language model generates a structured report, the image segmentation module generates pixel-level masks using language embedding as cues, and the collaborative output module establishes a correspondence between the report and the mask through special markers, achieving localizable joint prediction. This invention requires no manual prompts, supports fully automated analysis, and possesses good interpretability and clinical deployment adaptability, making it suitable for intelligent assisted diagnostic scenarios involving medical images such as chest X-rays.

The application discloses a chest radiograph report generation and lesion positioning method and system based on reinforcement learning, and belongs to the technical field of cross between artificial intelligence and medical image analysis. The method first uses a plurality of source chest radiograph images and query texts to supervise and fine-tune a basic model; then a GRPO reinforcement learning framework is used to optimize a strategy model, and an indicator function guided by token entropy is introduced into a target function, and finally, the optimized strategy model is used to realize end-to-end parallel generation of chest radiograph report texts and lesion boundary box coordinates. The method improves the quality of chest radiograph report generation while realizing the visualization and positioning of lesions, and the integrated output form is more in line with the actual clinical workflow.

The invention discloses a child chest radiograph evaluation method, device and equipment based on deep learning and a medium, and the method comprises the steps: dividing a pre-collected child chest X-ray image into a training set, an internal verification set and an external verification set, and constructing a U-Net segmentation model, the method comprises the steps that a training set is established, region segmentation is conducted on the training set through a U-Net segmentation model, a DenseNetECA model is established, a core architecture of the DenseNetECA model is composed of four ECABlock modules, dense connection and an efficient channel attention mechanism are integrated in the ECABlock modules, feature dimensionality reduction and space down-sampling are conducted between the two ECABlock modules through a Transformation module, and the two ECABlock modules are connected through a Decision module; and inputting the processed training set, the internal verification set and the external verification set into a DenseNetECA model to complete model training, inputting a to-be-detected X-ray image into a U-Net segmentation model to perform region segmentation, and inputting the segmented to-be-detected X-ray image into the DenseNetECA model to perform evaluation to obtain an image evaluation result. The accuracy of the chest image of the child can be improved.

The present application belongs to the technical field of deep learning and medical image processing, and specifically relates to a chest radiograph abnormality recognition domain adaptive method and system combining Wasserstein distance and difference measurement. The method comprises the following steps: S1, chest radiograph data preparation and preprocessing; S2, multi-scale feature extraction based on Swin Transformer; S3, loss minimization based on Wasserstein distance and contrast domain difference; S4, model verification for chest radiograph prediction. The present application has the characteristics that not only the source domain sample closest to the target domain sample can be selected, but also the distance between the same categories of samples in the target domain and the source domain can be shortened in the feature space, and the distance between different categories can be widened, and at the same time, the extracted multi-scale features are used for chest radiograph classification task, which can effectively improve the receptive field and capture more information beneficial to the chest radiograph classification task.

The invention discloses an image processing method and system, computer equipment and a storage medium, and relates to the technical field of image processing, and the method comprises the steps: obtaining a chest radiograph CXR image and a real CT image of a chest and a lung, carrying out the structure registration of the real CT image, and obtaining the real CT image with an aligned structure; the method comprises the following steps: performing preliminary 2D feature extraction on a chest radiograph CXR image, performing down-sampling, secondary feature extraction and feature refining processing on preliminary 2D features in multiple stages in sequence, performing dimension expansion on the refined features, performing 3D decoding, generating a 3D simulation CT image, training a multi-task interactive learning model through a real CT image and the 3D simulation CT image which are structurally aligned, and obtaining a multi-task interactive learning model through the real CT image and the 3D simulation CT image which are structurally aligned. And a multi-task interactive learning model capable of carrying out image segmentation and classification is obtained. According to the invention, the 3D simulation CT image is generated, the dimensions of the simulation CT image and the real CT image are unified, and the dependence on the real CT image data is reduced.

The application discloses a longitudinal chest radiograph progress monitoring method and system based on anatomical anchor semantic difference, relates to the field of medical image processing, and generates a shared anatomical support mask by performing spatial union on candidate boxes of front and rear chest radiographs in a unified coordinate system; the mask and the original image are soft fused through a learnable gate to obtain an image enhanced in anatomical anchoring; longitudinal difference processing is performed on the front and rear enhanced images in an image-report semantic embedding space to obtain report-aware semantic difference, a visual encoder is used to extract multi-scale visual features to calculate difference features, the difference features are refined in nonlinearity, the refined difference information is asymmetrically injected into current visual features, and finally, multi-scale fusion, semantic recalibration and progress classification are completed. The application can stably, accurately and explainably monitor the disease performance changes in the front and rear chest images of the same patient without relying on the input of radiology reports in the reasoning stage.

The application discloses an AI model-based perioperative medical detection image accurate analysis and management system and method, and relates to the technical field of medical image processing. The AI model-based perioperative medical detection image accurate analysis and management system and method comprises the following steps: S1, preprocessing chest radiograph structure feature data and historical statistical data of a surgical patient in a perioperative period; S2, discriminating and segmenting a structure overlapping area by gradient field response and geometric density correction for the overlapping structure area; S3, constructing a time sequence registration criterion by adopting a structure-guided rigid and affine joint registration algorithm; S4, fusing time sequence labels by fusing response and space fluctuation suppression; and S5, constructing, training and reasoning a structure function evaluation model by adopting multi-modal feature enhancement. The technical problem that data labels are confused due to structure overlapping and time sequence changes in the prior art AI analysis of chest radiographs in a perioperative period, and the stability and accuracy of AI model training and reasoning are affected is solved.

A Wasserstein distance and difference metric-combined chest radiograph anomaly identification domain adaptation method and a corresponding system, a Wasserstein distance and difference metric-combined chest radiograph anomaly identification domain adaptation device, and a non-transitory computer-readable storage medium are provided. The method includes the following steps: step 1, data preparation and data pre-processing for chest radiographs; step 2, multi-scale feature extraction based on a swin transformer network; step 3, loss minimization based on a Wasserstein distance and a contrastive domain discrepancy; and step 4, using the model to perform chest radiograph prediction after verifying the model.

The invention discloses a chest radiography image report generation method and device based on retrieval enhancement generation, and belongs to the field of medical image intelligent processing. The method comprises the following steps: firstly, constructing a retrieval database containing atomized clinical discovery; then extracting and fusing multi-scale visual features from the pre-trained visual model by using a multi-stage visual fusion encoder; efficient interaction of query and visual features is realized by using a decoder based on deformable convolution, and multi-view feature differences are automatically decoupled by introducing a hybrid expert system; and finally, introducing a semantic denoising training strategy to accelerate model convergence. According to the method, the generation accuracy of the chest radiography image analysis report under multi-view input is effectively improved, the training efficiency is effectively improved, and an efficient solution is provided for a medical image intelligent analysis system.

The utility model belongs to the field of medical instruments, and particularly relates to a standing position chest radiography shooting or all-lower-limb standing position assisting device which comprises a base, a plurality of supporting columns are fixedly connected to the top of the base, a top frame is fixedly connected to the tops of the supporting columns, a protective pad is fixedly connected to the top of the top frame, a grip is installed on the top of the top frame, and the protective pad is fixedly connected to the top of the base. The grip is located at one end of the protective pad, and an elastic belt is installed on the side wall of the top frame. Through the arrangement of the base, the top frame, the protective pad, the grip and other structures, a patient can be supported in an auxiliary mode and can be assisted to stand for a long time, and the situation that the patient is weak after an operation and cannot stand independently for a long time during chest radiography or lower limb film shooting, and the normal radiography process is affected is avoided. By arranging a bidirectional lead screw, an extrusion seat, a limiting plate, a positioning seat and other structures, the whole device is limited and fixed, the anti-skid capacity of the device is enhanced, and the situation that the device accidentally slides in the supporting and standing process of the patient, and consequently the patient falls down is avoided.

The application discloses a kind of based on layered multi-scale feature fusion network's children pneumonia classification system and method, set up multiple levels to extract different scale feature map, on global branch, extract the global feature map of children chest radiograph image based on window multi-head self-attention mechanism, on local branch, extract the local feature map of children chest radiograph image using convolution operation, in fusion branch, the global feature map, local feature map extracted by each level and the fused feature map of previous level are fused, then classification is carried out again, and the classification result of children pneumonia is obtained.Extract the global feature map, local feature map of different scale, and fuse the global feature map, local feature map of current level with the fused feature map of last level, so that the feature information extracted by each level can interact, and through the feature extraction and interaction of multiple levels, the features of children chest radiograph image can be fully extracted, and the accuracy of classification result is improved.

The invention provides a multi-view chest radiograph report generation method based on disease state retrieval prompt. According to the method, a multi-modal language model is used as a text generation decoder, a medical image encoder is used for encoding an anterior chest radiography image and a lateral chest radiography image, and encoded image features are fused to serve as image input of the multi-modal language model; the disease state information obtained through retrieval is fused to serve as text input of a multi-modal language model, and the two drive a report generation process together. According to the method, complementary information of the anterior chest radiography and the lateral chest radiography is effectively combined, and higher accuracy is achieved in the aspect of fine-grained feature capture. According to the method, the weight is distributed to the retrieved disease state information and the weight fusion is carried out, so that the prompt information not only can reserve diversified clinical semantics, but also can better highlight the part closely related to the current chest radiograph image, and the interference of irrelevant or noise information on report generation is effectively reduced.

The utility model discloses a kind of children's chest fixators, it is related to medical auxiliary instrument field, including the CT seat of taking chest radiograph, the top of the CT seat of taking chest radiograph is fixedly installed with base, the top of the base is fixedly connected with antiskid layer, and recess is opened in the inside symmetry of antiskid layer, the top of the antiskid layer is fixedly installed with connecting plate, the inside height of the both sides of the connecting plate is installed with magic tape hook face. The children's chest fixator, recess, magic tape nap and magic tape hook face are arranged, the feet of children are fixed, and protective pad, fixed ring plate, first magic tape band and second magic tape band are arranged, the body of children is fixed, so that fixator can adapt to the fixing needs of children of different body shapes, and moving plate, clamping rod, strong spring and extension rod are arranged, the height of fixator is adjusted, and it is guaranteed that the arm of children can be effectively supported.

The invention discloses a pneumoconiosis chest radiograph staging method based on deep label relaxation learning. The pneumoconiosis chest radiograph staging method comprises the following steps: step 1, pre-training a backbone network; 2, dividing samples meeting a formula into untrusted subsets, and dividing all the remaining samples into trusted subsets; 3, calculating an average prediction entropy of the three-backbone network on a sample x in the trusted subset; 4, performing descending order arrangement on the samples in the credible subset, and calculating a relaxation value corresponding to the sample x; step 5, calculating loss corresponding to the sample; step 6, calculating mean square errors of the three-backbone network on the samples x and xa; and 7, calculating the total loss of the three-backbone network on the trusted subset and the untrusted subset, and optimizing the three-backbone network. The deep neural network is trained in combination with multi-network collaborative learning, adaptive label relaxation learning and self-supervised learning, and effective generalization of the model on a pneumoconiosis chest radiography data set with noise labels can be realized.

The invention discloses a chest radiography protection screen integrating light following and voice, and relates to the technical field of medical auxiliary equipment. The protective screen comprises a screen body, a light-following sensing system, a voice recognition system, a control system, a power system and a power supply system. The light following control system detects the position change of a light pillar of the X-ray machine through a photosensitive sensor to realize automatic following lifting of the screen; the voice control system receives personnel instructions through the voice recognition module and controls the screen to ascend and descend. Through deep fusion of light following and voice control, intelligent operation of protection equipment in radioexamination is realized, energy distribution optimization is guaranteed through a multi-stage power supply design, control precision is guaranteed through a sensing mechanism and a logical judgment algorithm, and system toughness is enhanced through a dual-mode redundancy design. The control stability and reliability are improved, the operation process is simplified, and the radiation protection efficiency and safety in medical examination are remarkably improved.

The invention discloses a medical diagnosis collaborative decision-making system for multi-modal data dynamic maintenance, and relates to the technical field of medical diagnosis collaborative decision-making, and the system comprises a data obtaining module which is used for obtaining dual-energy chest radiography collection time and synchronous cardiac and respiratory cycles; the phase mismatch index calculation module is used for calculating a motion mismatch quantitative index; the oblique double-line-pitch candidate point detection module is used for detecting a candidate point set in the heart edge and diaphragm muscle areas; the mismatching mask generation module is used for screening mismatching points and generating a mismatching mask; the evidence state marking module is used for projecting the computed tomography diagnosis area to the chest radiograph and marking the evidence state through overlapping judgment; and the collaborative decision option generation module is used for generating decision options containing finalizing and evidence obtaining according to the evidence state. The problem of motion artifact interference of the dual-energy chest radiograph can be solved, and diagnosis accuracy and decision reliability are improved.

The present invention relates to noise label based machine learning for anomaly assessment in medical imaging. For machine learning and application of machine learning models for anomaly assessment in medical imaging, the machine learning uses regularization of loss, such as regularization used to train anomaly classification in chest radiographs. The regularization can be noise and / or correlation regularization, which targets noisy ground truth labels of the training data. As a result of using noise and / or correlation regularization in the training, the resulting machine learning model can better classify anomalies in medical images.

The application discloses a chest radiograph screening classification prediction method and device, electronic equipment and a storage medium, and comprises the following steps: contrast enhancement processing is performed on an acquired chest radiograph image to be screened to obtain a contrast enhancement image; and the contrast enhancement image is input into a pre-trained classification model to determine the category of the chest radiograph image and the corresponding score and the corresponding explainable model activation schematic diagram. In this way, the 'findings' and 'no findings' in the chest radiograph can be quickly separated, the work efficiency of doctors is improved, and misdiagnosis and missed diagnosis are reduced.