75 results about "Automated segmentation" patented technology

The application discloses a multi-modal image data segmentation method and related products, which comprises the following steps: acquiring multi-modal image data; using a segmentation model to perform segmentation processing on the multi-modal image data to obtain a segmentation result of the multi-modal image data output by the segmentation model; the segmentation model at least comprises a multi-level encoding layer and a multi-level decoding layer; the multi-level encoding layer is used for performing cross-modal and cross-window feature enhancement processing on the multi-modal image data; and the multi-level decoding layer is used for performing semantic complementary fusion processing on the enhanced features of the multi-modal image data. The application realizes high-precision and high-stability multi-modal medical image automatic segmentation without significantly increasing the calculation complexity, and overcomes the defects of the prior art, such as insufficient cross-modal alignment, weak semantic consistency and limited fine-grained structure reconstruction capability.

This invention provides a multi-dimensional user profiling automated segmentation and precise targeting system and method, aiming to solve the problems of single user feature dimensions, lagging updates, and inaccurate matching in existing advertising targeting. The method includes: collecting multimodal data such as user text, behavior, and images; extracting feature vectors from each data source using a model and fusing them into a unified high-dimensional representation; inputting this data into a multi-label neural network to output user interest, behavior, and intent tags; using an improved clustering algorithm to form a strategic audience package; and combining this with ad placement and resource allocation for ad scheduling, while simultaneously updating the user profile and model through a closed-loop user feedback mechanism. This method features innovative aspects such as multi-source heterogeneous feature modeling, minute-level tag updates, and tag-driven clustering and ad delivery, significantly improving audience identification accuracy and ad ROI performance.

The application belongs to the technical field of image processing, and discloses a chest and abdominal cavity bleeding automatic segmentation method and device based on deep learning, which comprises the following steps: obtaining and normalizing chest cavity or abdominal cavity CT volume data, and then performing linear embedding to generate initial features; inputting the initial features into an encoder for multi-stage encoding, in which the encoding features of different levels are respectively enhanced in the frequency domain according to different semantic levels to generate level-aware frequency domain enhanced features; the input features are modeled by a deformable mixed window multi-head self-attention mechanism in at least one stage of the encoder, and the attention calculation results are weighted and fused by using the frequency domain guide weight generated based on the frequency domain enhancement results; the encoding features are input into a decoder for multi-stage decoding and fusion with the corresponding level features of the encoder to gradually restore the spatial resolution; and the bleeding area segmentation results corresponding to the CT volume data are generated according to the final output features of the decoder. The application can efficiently and accurately segment the bleeding area and assist clinical decision-making.

The application provides an aorta morphological feature automatic measurement and clinical auxiliary decision-making method and system based on artificial intelligence, which is suitable for three-dimensional medical image analysis of aorta and its main branches. The method comprises the following steps: image data acquisition and preprocessing, automatic blood vessel structure segmentation, center line extraction and key point positioning, multi-parameter automatic measurement, risk assessment and clinical auxiliary decision-making, data standardization output and system feedback optimization. The corresponding system comprises an image acquisition and processing module, a segmentation module, a center line extraction module, a parameter measurement module, an auxiliary decision-making module, a data output module and a verification and optimization module. Through a deep learning model and morphological calculation means, efficient identification and quantitative analysis of blood vessel structure are realized, and individualized risk suggestions and surgical planning are output based on statistical and machine learning models. The application improves the automation, standardization and intelligence level of blood vessel analysis, and enhances the clinical auxiliary decision-making ability under complex cases.

The application discloses an automatic intracranial hemorrhage area segmentation method based on multi-layer CT images and belongs to the technical field of medical image processing. The method comprises the following steps: inputting a target CT image with a labeled hemorrhage area and upper and lower layer CT images before and after the target CT image sequence into a hemorrhage area segmentation model for training; obtaining fusion features considering multi-layer correlation by fusing multi-layer CT image features; combining a hybrid loss function composed of cross entropy and dice and a gradient back propagation algorithm to obtain a hemorrhage area segmentation model with minimum loss value; inputting a to-be-detected CT image and upper and lower layer CT images before and after the to-be-detected CT image sequence into the model; and finally outputting a hemorrhage area probability graph of the to-be-detected CT image of all layers in the to-be-detected CT sequence. The application fully utilizes the information of upper and lower layers in the CT sequence under the premise of only increasing a small amount of operation amount, and improves the accuracy of the hemorrhage area segmentation and the consistency between adjacent layers.

The application relates to a brain region segmentation method based on multi-sequence magnetic resonance imaging collaborative feature fusion, which comprises the following steps: collecting brain amplitude graph data, phase graph data and T1 structure image data; carrying out pretreatment and reconstruction processing on the amplitude graph data and the phase graph data to obtain a quantitative magnetization rate image; carrying out segmentation labeling after spatial registration of the T1 structure image data and the quantitative magnetization rate image to construct a brain region segmentation label data set; training a CIA-Net deep learning model through a training set in the brain region segmentation label data set to obtain a trained CIA-Net deep learning model; inputting the quantitative magnetization rate image to be segmented and the T1 structure image data into the trained CIA-Net deep learning model to output an automatic segmentation result of the brain region. The application can realize automatic fine segmentation of a cranial magnetic resonance image under the conditions of fine granularity and a large number of brain regions, and improve the accuracy, structural consistency and stability of the segmentation result.

The application belongs to the field of image processing, and discloses a knee joint image accurate automatic segmentation auxiliary method and system. The network structure is constructed, and the network structure is optimized according to the required result, so that the network structure is more matched with the knee joint image accurate automatic segmentation demand. Under the same data condition, more effective feature expression and more consistent segmentation output are obtained, and the dependence on single data distribution is reduced. The application obtains the source domain data set for training by processing the preprocessed image data, and trains the optimized network structure by using the source domain data set, so that the training data organization and the training process are more standardized and controllable, and the model is more likely to learn stable segmentation related features, thereby improving the adaptability and generalization ability to different source data, and solving the problems of insufficient model reliability under the conditions of insufficient adaptation of multi-center heterogeneous data and limited scale of high-quality labeling.

The invention relates to the technical field of computer vision and artificial intelligence, in particular to a YOLO automatic labeling tool and method based on SAM2 segmentation. Comprising an SAM2 automatic segmentation module, a mask image processing module, a morphological corrosion processing and target separation module, a connected domain analysis and bounding box generation module, an interactive label correction module and a YOLO format conversion and data set generation module. The SAM2 automatic segmentation module segments an input video and generates a binary mask image, and supports target tracking; the mask image processing module completes format conversion and preprocessing; the morphological corrosion treatment and target separation module is used for separating an adhered target; the connected domain analysis and bounding box generation module generates an initial bounding box; the interactive label correction module supports visual correction and batch adjustment; and the YOLO format conversion and data set generation module outputs a YOLO label and completes normalization, category coding and training / verification set division.

The invention relates to a method for automatic segmentation of a dental arch that comprises acquiring a three-dimensional surface of the dental arch, in order to obtain a three-dimensional representation comprising a set of vertices, generating virtual views from the three-dimensional representation, projecting the three-dimensional representation onto each two-dimensional virtual view, in order to obtain an image representing each vertex on the virtual view, processing each image by means of a deep learning network, carrying out inverse projection of each image in order to assign, to each vertex of the three-dimensional representation, one or more pixels of the images in which the vertex appears and to which it corresponds, and assigning one or more probability vectors to each vertex, determining the class of dental tissue to which each vertex most probably belongs based on the probability vector or vectors.

This invention relates to the field of medical image processing technology, specifically to a method for constructing an intelligent segmentation model for magnetic resonance imaging (MRI) images. The method includes: acquiring LGE-CMR images; constructing a cardiac region localization network, which is used to locate the cardiac region in the LGE-CMR image; and constructing a region-of-interest (ROI) intelligent enhancement module, which enhances the RIO based on the output of the cardiac region localization network. This invention achieves automatic cardiac region localization by constructing a cardiac region localization network, and combines this with the RIO intelligent enhancement module to selectively adjust contrast and sharpness to highlight lesion areas. Furthermore, it incorporates a feature fusion module from a multi-scale, multi-class segmentation network to enhance the feature expression of small regions. This allows for automated multi-class segmentation of myocardial scars and microvascular occlusions, solving the problems of time-consuming manual segmentation and inter-observer errors. It also avoids the difficulty of existing automatic segmentation models in simultaneously and accurately identifying three types of regions and the possibility of missed or false detections of small regions.

This invention discloses a breast cancer image segmentation method and system based on parameter sharing and prior guidance, belonging to the field of medical image processing technology. The method includes the following steps: acquiring the image to be segmented; processing the image to be segmented using an image segmentation model to obtain a tumor segmentation map; wherein, the image segmentation model training process involves: extracting features from known noisy segmentation labels and corresponding medical images; inputting denoised features into a denoising channel for noise processing, inputting pathological features into a conditional channel for semantic information processing, and using a cross-attention mechanism to achieve parameter sharing between the denoising channel and the conditional channel; using a denoised stream to predict the added noise, and using a prior guidance strategy to identify and guide the tumor location during the prediction process, ultimately obtaining the segmented image. This invention utilizes the advantages of diffusion denoising technology combined with a prior knowledge guidance strategy to achieve more accurate and automated segmentation, improving the efficiency of breast cancer image detection and segmentation.

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.

The application discloses a model training method and a segmentation method for automatic segmentation of a choroid plexus image. The method comprises the following steps: acquiring a preprocessed data set; constructing a to-be-trained Swin-UNETR network architecture; training the to-be-trained Swin-UNETR network architecture through the preprocessed data set, so as to obtain a trained model for automatic segmentation of a medical image. Compared with a traditional UNETR method, the model training method for automatic segmentation of the choroid plexus image adopts a Swin Transformer architecture, can more effectively capture multi-scale features of an image, and the Dice coefficient is improved by 5-10%.

The disclosure provides a kind of cerebral hemorrhage analysis method and system based on image registration, comprising: automatically segmenting cerebral hemorrhage area in brain CT image;The cerebral CT image and the segmented cerebral hemorrhage image are pretreated;Cerebral CT image is registered according to MRI template, and the deformation field is obtained;The cerebral hemorrhage image is transformed using the deformation field;Analysis patient bleeding level and the brain area involved.Use normal brain MRI image as registration template, register patient brain CT with MRI template, extract patient hematoma feature, can quickly and effectively analyze bleeding level and the brain area involved by hematoma, provide guidance for doctors, and can be used for subsequent visualization modeling, help doctors to formulate more comprehensive treatment plan.

The application discloses a screen surface bubble detection method and system, comprising the following steps: performing frequency domain processing based on fast Fourier transform and Gaussian high-pass filtering on the effective display area of a first image and a second image respectively to obtain a first spatial domain image and a second spatial domain image; performing multi-peak automatic segmentation on the first spatial domain image and the second spatial domain image respectively to obtain bubble candidate regions, and performing morphological dilation and intersection operation on the bubble candidate regions to connect adjacent regions while keeping the original area of each bubble candidate region unchanged, thereby forming complete bubble regions; screening a first candidate bubble region set from the complete bubble regions corresponding to the first spatial domain image, and screening a second candidate bubble region set from the complete bubble regions corresponding to the second spatial domain image; and determining that any coincident region in the first candidate bubble region set and the second candidate bubble region set is a bubble defect when the coincidence rate of the coincident region is greater than a preset ratio threshold.

This invention discloses a lightweight automatic MRI segmentation method for breast masses based on MobileSAM. The method inputs the breast MRI image to be detected into a breast mass MRI image segmentation model, outputting the localization and prediction results of the breast mass. The model uses the MobileSAM lightweight image coding network as its backbone, and introduces an adapter to achieve adaptation and efficient fine-tuning of the breast image domain. A separable hollow pyramid fusion module is designed in the decoding branch, employing multi-branch deep separable hollow convolution to aggregate multi-scale context, enhancing the feature discrimination of small masses and blurred boundaries. Furthermore, a coordinate-channel-space joint attention gating module is introduced, using decoded semantics as a guide to filter skip features, strengthening the spatial location and detail reconstruction of the mass boundary. This invention achieves end-to-end automatic segmentation without interactive prompts, significantly improving the accuracy and robustness of breast MRI mass segmentation while maintaining lightweight design and low training costs.

This invention discloses an automatic segmentation system and method for brain tumor MRI images, belonging to the field of medical image processing technology. It includes an encoder, a decoder, and the processing pathways connecting them. A feature calibration module is integrated into the encoder's downsampling stage. Through three-dimensional wavelet transform, spatial features are decomposed into sub-bands of different frequencies and differentially calibrated and fused to explicitly preserve and enhance detail information during feature compression. A cross-scale context gating unit is embedded in the skip connections. Attention weights are generated by aggregating multi-scale context information, intelligently filtering the transmitted encoder features before fusing them with decoder features. This method effectively solves the problems of detail loss due to downsampling and coarse feature fusion in traditional segmentation networks, significantly improving the accuracy and boundary fit of multi-sub-region segmentation of brain tumors.

This invention belongs to the field of image segmentation technology, specifically relating to an MRI segmentation method for small structures of the knee joint based on a 3D residual multi-scale dynamic attention network. The method includes: acquiring an MRI image of the knee joint to be segmented; constructing a 3D residual multi-scale dynamic attention network model; inputting the MRI image of the knee joint to be segmented into the 3D residual multi-scale dynamic attention network model to achieve automatic segmentation of small structures of the knee joint in the MRI image. This invention enhances multi-scale feature extraction capabilities and reduces computational overhead by introducing a residual multi-scale expanded depthwise separable convolution module; simultaneously, by combining a dynamic gated attention mechanism, adaptive fusion between encoder and decoder features is achieved, thereby improving boundary localization capabilities. Furthermore, by introducing a multi-scale auxiliary output structure incorporating boundary-aware loss, the segmentation results are further constrained, improving the ability to characterize complex structural contours.

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 application discloses a kind of glass wart measurement and subtype classification method based on physician mark mask, belong to medical image processing and ophthalmic artificial intelligence auxiliary diagnosis technical field.This method takes the binary mask marked by physician as core input, in turn through image reading and abnormal compatibility processing, mask pretreatment and lesion contour extraction, morphological parameter calculation and physical unit conversion, automatic classification based on AREDS standard Subtype, finally complete visual annotation, structured report generation and batch processing.The application realizes Chinese path compatibility by binary stream decoding, extracts lesion contour using OpenCV related interface, calculates area, maximum diameter, height, roundness and other parameters, completes the conversion of pixels and actual physical units combined with calibratable scale, strictly follows AREDS standard to divide subtype according to maximum diameter of lesion, and supports batch processing and abnormal protection.The application avoids the error of automatic segmentation algorithm, improves the measurement accuracy and result consistency, realizes full-process automation, adapts to clinical diagnosis and treatment and scientific research demand, and provides reliable tool for early quantitative evaluation of age-related macular degeneration.

The application discloses a return scattering ionization map F area automatic segmentation method based on Haar wavelet downsampling, to solve the problems of fuzzy F area boundary, low contrast and high frequency information loss in the prior art. The method comprises the following steps: step 1, data acquisition, labeling and pretreatment; step 2, constructing a SegNext segmentation network based on Haar wavelet downsampling; step 3, decoding fusion and segmentation prediction output; step 4, training optimization based on edge weighted focus joint loss. The method retains the high frequency characteristics of the ionization map by using the wavelet downsampling module, and combines the focus loss and the edge weighting mechanism, so that the segmentation result is more accurate and stable in the boundary area.

The invention discloses a lumbar disc herniation interpretable diagnosis system based on cross-style consistency semi-supervised segmentation. The method comprises the following steps: firstly, carrying out style diversification processing on a lumbar axial MRI image through Bessel transformation to generate an image pair for consistency learning; the aggressive student model processes the stylized image and carries out cooperative training through cross pseudo supervision, and the robust student model provides a reliable prediction target through smooth parameter updating; the trained model can realize automatic segmentation of an intervertebral disc and a posterior spinous process area. Based on the segmentation result, the system adopts a geometric rule algorithm to quantify the prominence degree, and automatically generates a grading result and a visual report according to a clinical MSU grading standard. According to the method, the dependence on labeled data is reduced, the interpretability of the model is improved, and efficient and accurate lumbar disc herniation auxiliary diagnosis with clinical guiding significance can be realized.

This invention discloses an intelligent segmentation method, system, device, and medium for laparoscopic colorectal surgery images. The method includes: acquiring images during laparoscopic colorectal surgery; performing frame-by-frame processing on the images; and annotating keyframes in the frame-by-frame images and presetting Regions of Interest (ROIs) to achieve segmentation of the laparoscopic colorectal surgery images. This invention, through real-time laparoscopic visualization, provides guidance for locating the inferior mesenteric artery, automatic segmentation of the inferior mesenteric artery, automatic segmentation of the Toldts space, and automatic segmentation of instrument forceps during the inferior mesenteric artery dissection stage, effectively achieving intelligent segmentation of laparoscopic colorectal surgery images. This invention achieves automated keyframe extraction with high processing efficiency and can quickly construct large-scale, standardized keyframe datasets.

This invention discloses an intelligent monitoring and early warning method for small urban water bodies based on visible light drones, belonging to the field of water quality prediction and management technology. The method acquires water body images using a drone equipped with a visible light sensor, automatically stitches them together to generate orthophotos, automatically segments water body areas and outputs boundary vector layers, extracts band DN values ​​and constructs a water quality parameter feature dataset through normalization, band combination, and chromaticity angle calculation, uses a machine learning model to invert chlorophyll a concentration, turbidity, and phycocyanin concentration, and optimizes the model accuracy by combining measured data, performs statistical analysis and future trend prediction on the inversion results, and generates thematic maps, completes water quality evaluation according to standards, and automatically outputs reports and pushes them to relevant units when early warning conditions are met. This invention uses low-cost visible light equipment to achieve fully automated intelligent monitoring, effectively improving inversion accuracy and regulatory efficiency, and adapting to the needs of large-scale, routine monitoring and early warning of small urban water bodies.

This application relates to the field of medical image processing technology, and discloses a fully automated lung nodule segmentation system and method based on ultra-high resolution target scanning CT. The system includes a data preprocessing module, a 3D target detection module, an adaptive resampling module, a 3D fine segmentation module, and a post-processing and feature extraction module. The detection module outputs a 3D bounding box to provide nodule in-plane size information for adaptive resampling; the resampling module maintains the Z-axis unchanged while adaptively downsampling the XY axes; the segmentation module adopts a dual-cascaded feature pyramid network architecture to achieve two-stage optimization of coarse and fine segmentation; the post-processing module generates a binarized segmentation mask and calculates quantitative features. This application effectively mitigates the impact of some volume effects on the segmentation accuracy of subsolid nodule boundaries.

This application discloses a non-invasive method and system for predicting the ALK rearrangement status of lung adenocarcinoma based on a large-scale medical model. The method includes: acquiring chest medical imaging data and corresponding clinical information, wherein the chest medical imaging data includes lesion regions; inputting the chest medical images into a fine-tuned large-scale automatic segmentation model for lung adenocarcinoma lesions to obtain lesion segmentation mask images, wherein the segmentation model is optimized through a fine-tuning strategy using lung adenocarcinoma-specific cue tokens and a low-rank adapter module; based on the lesion segmentation mask images, cropping standardized lesion region images from the chest medical images; inputting the lesion region images and clinical information into a trained large-scale ALK rearrangement status prediction model for lung adenocarcinoma, and outputting the ALK rearrangement status prediction result for the lesion. The technical solution of this application effectively improves the segmentation accuracy of lung adenocarcinoma lesions and the accuracy and efficiency of ALK rearrangement status prediction.

This invention discloses an automatic segmentation method, system, and storage medium for breast cancer ultrasound images. The method includes: acquiring a breast ultrasound image dataset, performing standardized preprocessing, multi-expert annotation integration, and data augmentation to obtain high-quality data; constructing a basic Attention U-Net model and two improved models, CSWin-Unet and MRCST-Net, respectively, to achieve gradient optimization of segmentation performance through attention gating mechanism, cross-shaped window attention, and residual convolution-Transformer parallel module; using a unified training framework and four indicators—mean intersection-over-union (MIoU), accuracy (Acc), Kappa coefficient, and Dice coefficient—and selecting the optimal model by averaging the results after validation; and performing end-to-end tumor region segmentation on the input breast ultrasound image based on the optimal model. This invention can help improve the efficiency and accuracy of early breast cancer diagnosis.

This invention addresses the limitations of existing technologies by proposing an adaptive dual-threshold segmentation method for ultrasonic phased array C-scan images of battery busbars. It primarily targets the technical problem of blurred edges and difficulty in distinguishing foreground and background in ultrasonic phased array C-scan images of multi-station battery busbars. By enhancing blurred edges, redundant information in the ultrasonic phased array C-scan images is effectively eliminated, enhancing the detectability of useful information and simplifying the data volume to the greatest extent. This improves the reliability of image segmentation, feature detection, and feature recognition, avoiding the shortcomings of existing segmentation methods. Using this method to segment ultrasonic phased array C-scan images of power battery busbars yields significant results, enabling high-speed and automated segmentation. This facilitates subsequent image detection algorithms, improving the reliability and accuracy of detection, and is particularly suitable for segmenting ultrasonic phased array C-scan images with severe noise pollution.

The application discloses a method for automatically constructing a digital twin model of a bridge prefabricated component under sparse point cloud data, and steps include: S1, acquiring a sparse point cloud data set of the bridge prefabricated component; S2, improving a region growing algorithm and combining principal component analysis to dynamically search for an optimal growing threshold, thereby realizing automatic segmentation of sparse point cloud data of a component end face; S3, determining a main shaft direction of the component, arranging a slice plane along the main shaft direction, adaptively determining a slice thickness in combination with an end face geometric thickness, and extracting a measured slice contour through boundary point detection; S4, introducing a design section contour as a geometric prior, registering the measured slice contour extracted in S3, constructing a parameterized section contour, and optimizing and updating geometric parameters of the contour; and S5, constructing a digital twin model of the bridge prefabricated component. The method realizes automatic processing of sparse point cloud and robust extraction of key geometric elements, and provides method support for geometric review and construction geometric control of the prefabricated component.