35 results about "Retina vessels" patented technology

The invention discloses an automatic retinal vessel segmentation method fusing large-kernel multi-scale cavity convolution and an attention mechanism, which is suitable for accurate extraction of a fine-grained vessel structure in a fundus image. According to the method, an improved deep neural network (LKD-UNet) is constructed, a collaborative receptive field module is embedded in each layer of an encoder, and large kernel convolution and multi-scale cavity convolution are combined to enlarge the receptive field and enhance the texture analysis capability; an efficient Transform module is introduced into a bottleneck layer, a cross-regional semantic relationship is modeled through a multi-head self-attention mechanism, and the topological consistency of microvessels is improved; the decoder adopts a shallow symmetric structure and a kernel scale scheduling strategy, and realizes high-resolution reduction in combination with jump connection. In addition, a four-category pixel-level error visualization mechanism is introduced and is used for assisting segmentation error diagnosis and result interpretation. Experimental results show that the method is superior to an existing mainstream model in multiple public data sets, has remarkable advantages in the aspects of segmentation precision, structural coherence and deployment efficiency, and is suitable for remote fundus screening and blood vessel extraction tasks in embedded medical equipment.

The invention discloses a retinal blood vessel image segmentation method and device, equipment and a medium, and relates to the technical field of image segmentation, and the method comprises the steps: obtaining a retinal blood vessel image, and inputting the retinal blood vessel image into a preset blood vessel image segmentation model; performing convolution operation and axial attention calculation on the retinal blood vessel image to obtain local semantic feature maps of all levels including edge information and texture information of the blood vessel, and a global semantic feature map including a blood vessel communication relationship and a blood vessel geometric morphology association relationship; performing dual-path processing and dual-path fusion on the local semantic feature map to obtain a fused local semantic feature map; step-by-step up-sampling operation is carried out on the global semantic feature map, splicing and convolution fusion are carried out on the up-sampling feature map and the fused local semantic feature map of the corresponding layer until the step-by-step up-sampling operation is finished, and a target feature map is output; a binarization segmentation result of the retinal blood vessel image is generated based on the target feature information, and accurate image segmentation is realized.

The invention provides an OCTA image retinal vessel segmentation method based on a full-resolution network, and the method comprises the steps: carrying out the preprocessing of an input OCTA image through a data preprocessing module, wherein the preprocessing comprises the optical transformation and geometric transformation, and obtaining the preprocessed data; the feature coding module uses five feature extraction units which are arranged in sequence to extract image features of five levels and outputs the image features to the cascade feature enhancement module; a cascade feature enhancement module CFEM combines low-level detail information and high-level semantic information through cross-layer guidance for feature fusion, and then outputs a segmentation result; calculating loss and optimizing the OCTA image retinal vessel segmentation model based on the full-resolution network to obtain an optimized segmentation model; using the optimized segmentation model to obtain a retinal blood vessel image; according to the method, high-precision and high-efficiency segmentation of the retinal blood vessel can be realized, the network complexity is relatively low, and the computing resource consumption is relatively low.

The invention discloses an ophthalmologic tumor image feature extraction method and system based on artificial intelligence, and the method comprises the steps: obtaining ophthalmologic tumor multi-source image data, and carrying out the preprocessing of the data, and generating a standardized image set; constructing a multi-scale feature extraction network, and extracting basic texture features, edge contour features and deep semantic features of the image in a layered manner through an improved convolution module; introducing a double-channel attention mechanism, and performing dynamic weighted fusion on the preliminarily extracted features and the retinal vessel topological features to generate a fusion feature set; and the fusion features are optimized by using a residual enhancement network, key area feature expression is enhanced through multi-branch feature aggregation, and a structured feature map containing multi-dimensional feature information is output. The integrity of irregular tumor edge features is improved, and fusion noise caused by modal dislocation is reduced; in combination with a semi-supervised learning framework and a residual enhancement network, the dependence on annotated data is reduced, and the stability and robustness of feature expression under limited data are improved.

The present invention discloses a fundus image segmentation method based on an uncertainty and shallow feature enhanced attention network, which comprises: selecting fundus retinal image data as a training set and a test set; preprocessing the fundus retinal images in the training set; constructing a fully convolutional neural network model on Pytorch; and segmenting the test set using the trained fully convolutional neural network model to obtain a final segmentation result. The present invention provides a neural network model for automatically segmenting retinal blood vessels from fundus images, which combines an attention mechanism to expand the receptive field so that the model can better understand the local details and global context in the image, and improves the learning ability of the model by fusing deep features in the neural network to guide the expression of shallow features. By quantifying image uncertainty, the adverse effects of the model's overconfidence on the segmentation results are reduced, thereby improving the accuracy of retinal blood vessel segmentation.

This invention provides a method and apparatus for cross-domain segmentation of medical images based on an adaptive frequency domain alignment network. The method includes: acquiring a source domain medical image set and a target domain medical image set; preprocessing the source domain and target domain medical image sets; training an AFDAN model based on the preprocessed source domain and target domain medical image sets; optimizing the trained AFDAN model to obtain a final AFDAN model; and performing segmentation inference on the target domain image based on the final AFDAN model to obtain a high-precision medical image segmentation result. The system, method, and apparatus provided by this invention are suitable for high-precision segmentation of medical images such as vitiligo lesions and retinal vessels in scenarios where labeled data is scarce.

An image processing method comprising acquiring a first fundus image obtained by photographing a fundus using first light having a first wavelength, and a second fundus image obtained by photographing the fundus using second light having a second wavelength that is shorter than the first wavelength; specifying, in the first fundus image, respective positions of retinal blood vessels appearing in the second fundus image; and generating a choroidal blood vessel image by processing the positions identified in the first fundus image.

An enhanced feature consistency preprocessing method for retinal vessel image segmentation provided by the present invention includes the following steps: S1. Obtain a retinal vessel image with labels, and preprocess the retinal vessel image with labels; S2. Construct an image segmentation network based on enhanced feature consistency preprocessing. The image segmentation network includes an enhanced feature consistency preprocessing network and a segmentation network. The output end of the enhanced feature consistency preprocessing network is connected to the input end of the segmentation network, and the segmentation network outputs a segmentation result; S3. Input the preprocessed retinal vessel image with labels into the image segmentation network based on enhanced feature consistency preprocessing to train the image segmentation network based on enhanced feature consistency preprocessing; S4. Obtain a real-time collected retinal vessel image, and input the real-time retinal vessel image into the trained image segmentation network based on enhanced feature consistency preprocessing to obtain a segmentation result after processing.

The present invention provides a boundary attention-assisted dynamic graph convolution retinal vascular image segmentation method, comprising the following steps: S1. preprocessing the retinal vascular image, and flipping and slicing the preprocessed image to obtain an image slice set with a size of 48×48 pixels; S2. inputting the image slice set into a dynamic graph convolution U-type network to extract vascular features and output a vascular feature map; S3. inputting the vascular feature map extracted from the U-type network into a boundary attention network for processing to extract a shallow boundary feature map; S4. fusing the vascular feature map output by the dynamic graph convolution U-type network with the shallow boundary feature map output by the boundary attention network to obtain a retinal vascular image segmentation result.

The invention provides a multi-branch enhanced retinal vessel segmentation network method based on stimulation guidance, and belongs to the technical field of medical image intelligent diagnosis. The technical problems that small blood vessels are difficult to accurately identify and boundaries are not clear in retinal blood vessel segmentation are solved. According to the technical scheme, the method comprises the following steps: S1, collecting fundus color image data to be segmented; s2, constructing a boundary enhancement module; s3, constructing a multi-scale feature aggregation module; s4, constructing a stimulation guide gating fusion module; and S5, after model training is completed, for each to-be-segmented test image. According to the method, boundary gradient enhancement, multi-scale interaction and stimulation guide gating cooperate to improve boundary continuity and detail fidelity, thin blood vessel distinguishability is improved, and robustness under a complex background and weak contrast is enhanced.

The application discloses a kind of based on lightweight retinal real-time segmentation method, system and device of double channel structure, comprising: S1, obtain retinal vascular dataset and divide into training set, verification set and test set;S2, the training set and verification set are preprocessed to obtain preprocessed dataset;S3, the training set and verification set after pre-processing are data augmented;S4, build lightweight real-time segmentation network based on double channel, use augmented training set to train segmentation network;S5, use augmented verification set to evaluate segmentation network, if performance is greater than a certain threshold, then output trained segmentation network, pre-process augmented test set, if performance is lower than a certain threshold, then improve segmentation network and re-execute S5;S6, test set is preprocessed and input to trained segmentation network for segmentation, and evaluation index is calculated after segmentation is completed.The application can realize lightweight retinal real-time segmentation based on double channel structure.

The invention discloses a retinal vessel segmentation method of a multi-scale residual U-Net model based on multi-tense features, and belongs to the field of computer vision and medical image processing. According to the method, double residual convolution (DResConv) is adopted to remarkably enhance the context feature extraction capability and improve the segmentation performance of small blood vessels under a complex background. In addition, a mixed downsampling block (MDB) combining an attention mechanism (SKNet) and a pooling operation is adopted, so that the detail loss is effectively reduced, and abundant blood vessel information is reserved at the same time. And finally, channel and space attention are introduced through a multi-scale multi-time fusion module (MFM), a richer multi-scale feature map is generated, and multi-scale feature expression at different moments is effectively realized. The method is more excellent in color fundus retina blood vessel segmentation tasks.

The application belongs to the field of deep learning image segmentation, and discloses an OCTA image microvessel segmentation method fusing multi-view features, which comprises the following steps: splitting input 3D data and 2D image data as training set, verification set and test set respectively; meanwhile, obtaining the blood vessel label map of the 2D image; pre-processing and data enhancement are performed on the input data; the pre-processed 3D data is input into a 3D feature extraction network to extract multi-scale spatial features of the 3D image; the pre-processed 2D data is input into a 2D feature extraction network to extract multi-scale features of the 2D image; the 3D features and the 2D features are input into a cross-dimension feature fusion network to generate fusion features; the fusion features are input into a decoder to generate a prediction map of retinal blood vessels; the application can provide a more objective and accurate evaluation method for the diagnosis and research of diseases related to retinal blood vessels, and can quickly and accurately segment out retinal microvessels.

This invention relates to a retinal vessel image segmentation method based on a multi-scale attention-gated network, comprising: constructing a retinal vessel image segmentation network model; the retinal vessel image segmentation network model extracting features using multi-scale feature convolutional blocks with convolutional kernels of different sizes, effectively extracting and segmenting the entire vessel and terminal fine vessels using information at different feature scales, and using a non-local attention module to obtain richer global semantic information; then employing a multi-scale attention-gated network in the skip connection part, inputting feature maps of different dimensions, and selectively learning interrelated regions; training the retinal vessel image segmentation network model using a training dataset to obtain a trained retinal vessel image segmentation network model; and performing image segmentation of the retinal vessel image using the trained retinal vessel image segmentation network model. This method is beneficial for more accurate segmentation of retinal vessel images.

The invention relates to the technical field of image matching methods, provides a medical image matching method and device, and solves the problem of insufficient focus area matching accuracy in the prior art. The method comprises the following steps: aligning two fundus images at different time points based on key points of a retinal vascular network; performing multi-scale decomposition on the aligned image to obtain a multi-scale frequency sub-band; forming a propagation path through local extreme points of signal intensity in a cross-scale link sub-band, and respectively determining candidate focus seed points in the two images; taking the seed point as a starting point, performing region growth according to the signal intensity gradient direction of the frequency sub-band, and defining a focus contour; and mapping the lesion contours to a retinal vessel topological space, analyzing and comparing the consistency of the structure and space relationship of topological sub-graphs corresponding to the lesion contours in the two images, and establishing a matching link between lesion areas. According to the invention, accurate matching and association of focus areas in fundus images collected at different time points can be realized.

The invention discloses a needle tube set module, a puncture actuator and a surgical robot, and the needle tube set module is suitable for retinal blood vessel injection and retinal sublayer injection of fundus microsurgery. The inner tube is connected with the needle tip, a bent section is arranged at the joint of the inner tube and the needle tip, the inner tube is made of a shape memory material and is provided with a pre-bending angle, and a medicine injection channel is arranged in the needle tip and the inner tube; at least part of the inner tube is connected into the outer tube in a sliding mode, the needle tip and the inner tube can do telescopic motion in the outer tube, and the needle tip, the inner tube and the outer tube can do rotary motion; and the force sensing element is arranged on the outer pipe. The problem of adjusting the angle of puncturing the retina is solved, the high-precision intraocular flexible arm and the force sensing element can be integrated at the same time, the depth can be accurately sensed through the force sensing element in the retina sublayer injection operation puncturing process, the puncturing path can be prolonged, medicine backflow can be avoided, and therefore the operation success rate is increased.

The application provides a retinal blood vessel intelligent segmentation method, characterized by comprising the following steps: acquiring a given retinal image sample set, preprocessing the retinal image sample set to obtain a pretreatment data set; dividing the pretreatment data into a training set and a test set according to a proportion; obtaining scale feature information corresponding to the retinal image in the training set through a feature extraction network, and further calculating the segmentation difficulty degree corresponding to the scale feature information; taking the retinal image and the segmentation difficulty degree as the network source input, taking the reference blood vessel segmentation image corresponding to the retinal image as the output, combining an adaptive optimization algorithm and a preset loss function, applying the retinal image, training the target neural network, and obtaining a retinal blood vessel segmentation network model; and segmenting the retinal image according to the retinal image and the retinal blood vessel segmentation network model.

The invention provides a diabetes retina vessel segmentation method based on deep learning, and the method comprises the steps: carrying out the processing of a diabetic retinopathy fundus image through a trained optimal UFNet model, and segmenting a fundus vessel image; the UFNet model inherits a Unit + + decoder thought, an SE-CA module is added in each layer of encoder of the Unit + + model, and then the feature is output after passing through a feature fusion module; wherein the SE-CA module is composed of an SE attention module and a CA attention module which are arranged in sequence; the CA attention module generates channel attention features by using two modes of global average pooling and maximum pooling, then performs convolution processing on the two pooling results, and finally obtains a channel weight through a sigmoid function; according to the method, blood vessel segmentation and labeling can be carried out on the diabetic retinopathy image at a higher speed, meanwhile, the accuracy rate is high, the requirement for real-time detection can be met, and the method is suitable for a low-computing-power platform.

Systems, methods, and apparatuses associated with image segmentation, such as tubular structure segmentation, are described herein. An artificial neural network is trained based on annotated images of different types of tubular structures, which can have different contrast and / or appearance than a tubular structure of interest, to segment the tubular structure of interest in medical scan images. The training can be performed in multiple stages during which a segmentation model learned from the annotated images during a first stage can be modified to adapt to the tubular structure of interest in a second stage. In examples, the tubular structure of interest can include coronary arteries, catheters, guide wires, etc., and the annotated images used to train the artificial neural network can include blood vessels, such as retinal blood vessels.

The invention relates to the technical field of cardiovascular disease recognition in cattle breeding, in particular to a deep learning-based recognition method for cardiovascular diseases in cattle breeding. Constructing a cattle cardiovascular disease recognition model, wherein the model comprises a light sensation compensation module, a progressive multi-receptive-field feature amplification module, an image high-definition adaptive module, a retina multistage feature fusion module and a livestock cattle cardiovascular disease recognition module; the light sensation compensation module can strengthen direction recognition of bovine retinal vessel textures through asymmetric convolution and an attention network; the progressive multi-receptive-field feature amplification module can realize cross-level multiplexing of retina image features; and the image high-definition self-adaptive module can realize multi-dimensional image information compensation and dynamic feature calibration so as to restore image details. According to the invention, tiny lesions of cardiovascular diseases in the retina can be better captured, so that the requirements of rapid, low-cost and high-precision screening and early warning on bovine cardiovascular diseases are met.

The application discloses a neural network method for image domain sparse feature segmentation based on soft threshold and CNN module alternating iteration, and the core of the method is that a 1-norm and a learnable network double driving mechanism are adopted to quickly and accurately extract a target and a structure with sparse features. l 1-norm and a learnable network double driving mechanism, which can quickly and accurately extract a target and a structure with sparse features. A variational decomposition model is constructed based on target sparse representation, a deep learning network denoising module is introduced in an alternating minimization method, and an algorithm iteration process is expanded into a network. Then, the extracted sparse features are input into a subsequent connected lightweight U-Net, so that image segmentation is realized. Through retinal blood vessel segmentation and road crack detection verification, the network architecture proposed in the application has a lightweight feature, and compared with a standard U-Net, the number of parameters is smaller, and the performance can be equivalent to or even better than that of the standard U-Net.

The present invention discloses a downsampling method for retinal blood vessel segmentation, comprising: first building a basic segmentation framework U-Net network, then replacing the downsampling layer in the original U-Net network model, changing the original Maxpooling layer to a PF-Pooling module, then preprocessing the retinal fundus image to be processed, obtaining a preprocessing result image to be segmented and sending it to the network model, the model extracts pixels at four positions in the sampling window respectively, and then splices the pixels at corresponding positions to obtain four new images with half the scale, then uses Concat stacking to stack the four images in the channel dimension, and finally uses 1*1 convolution to perform channel compression, ultimately achieving the effect of keeping the number of channels of the original image unchanged and halving the width and height, and using this as the downsampling module of the model. Compared with the Maxpooling method, the PF-Pooling sampling method can retain more original image information and further improve the model segmentation accuracy.

An image processing method comprising acquiring a first fundus image obtained by photographing a fundus using first light having a first wavelength, and a second fundus image obtained by photographing the fundus using second light having a second wavelength that is shorter than the first wavelength; specifying, in the first fundus image, respective positions of retinal blood vessels appearing in the second fundus image; and generating a choroidal blood vessel image by processing the positions identified in the first fundus image.

The application discloses a retinal blood vessel segmentation method and system based on an automatic coding-decoding network of a discrete binary particle swarm optimization, and the segmentation method comprises the following steps: constructing a retinal blood vessel dataset; acquiring a lightweight U-shaped neural network model; acquiring an optimal U-shaped neural network model; and acquiring a segmentation result of a retinal blood vessel image through the optimal U-shaped neural network model. The discrete binary particle swarm optimization algorithm is used to automatically search and optimize the neural network structure, so that the network architecture can be automatically adjusted according to the task requirements, and meanwhile, the FPN Attention Block is introduced, the attention mechanism of the ECA-Net and the CBAM is combined, the attention mechanism is flexibly configured according to a selection factor, the encoder output is weighted, and then the decoder is fused, important local features can be effectively focused, and the segmentation precision of small blood vessels and complex structures is improved.

Described herein is a fundus image analysis system including a pre-segmentation image quality assessment module for receiving a fundus input image, and performing overall retinal image quality assessment and measurement quality assessment on the fundus input image; segmentation module for segmenting retinal vessel, artery, vein and optic disc to produce segmentation maps from the fundus input image; and a measurement module for computing region specific measurements within a standard zone within the fundus input image, and global physical or geometric measures of the whole fundus input image.

The present application relates to the technical field of computer image processing, in particular to a kind of retinal vessel density full-automatic partition measurement method;The present application uses the retinal color photo that ultra-wide-angle fundus camera is shot in multiple imaging modes, uses specific image processing method to combine it into three-dimensional matrix, uses improved convolutional neural network to identify three-dimensional matrix, realizes the full-automatic positioning of each area of retinal periphery, blood vessel identification, blood vessel density calculation;Meanwhile, it also makes up the vacancy of retinal vessel density full-automatic partition measurement based on ultra-wide-angle multi-imaging mode fundus camera, proposes the scheme of fine partitioning of retinal periphery according to field angle and an improved method of convolutional neural network fusing the features of multi-imaging mode image, to realize the full-automatic measurement of blood vessel density of each area of retinal periphery.

The present invention discloses a lightweight retinal vessel segmentation method based on an attention mechanism. First, the original retinal vessel image is preprocessed, and a retinal vessel segmentation model is constructed based on U-Net and a fully convolutional network. The retinal vessel segmentation model includes a front-stage network, a rear-stage network, an encoding-decoding block, and a spatial grouping enhancement module. The enhanced retinal vessel image is input into the front-stage network to obtain an initial feature map. The initial feature map is sampled and output to obtain a complete feature map, and the complete feature map is subjected to convolution and channel splicing operations and then input into the rear-stage network to obtain a retinal vessel segmentation map. By optimizing the U-Net network structure and introducing the attention mechanism, the network complexity is reduced while the segmentation performance is improved, and the problem that the network complexity and the segmentation accuracy cannot be effectively balanced in the field of retinal vessel segmentation is mainly solved.

This invention proposes a retinal vessel segmentation network, segmentation method, and storage medium based on collaborative fusion and adaptive guidance. The dual-branch encoder includes a CNN encoder and a Transformer encoder, used to extract local detail features and global semantic features of retinal vessels, respectively. A collaborative fusion integrator connects the feature outputs of the CNN encoder and the Transformer encoder, used to achieve semantic alignment and fusion of local detail features and global semantic features through a convolution-based cross-attention mechanism. An adaptive feature guidance module connects the collaborative fusion integrator and the decoder, using feature enhancement and fusion based on spatial attention and cross-learning to preserve vascular detail information. The decoder restores the resolution of the features processed by the adaptive feature guidance module and outputs the retinal vessel segmentation result through classification. This invention achieves effective complementarity of dual-branch features, enhances multi-scale detail perception capabilities, and improves segmentation accuracy.

The present invention provides an image and segmentation label generation model for tree-structured data and its application. The model includes a tree-structured image emulator model designed based on a small amount of expert knowledge and a generation network model based on a morphological loss function. The present invention can generate labeled simulated tree-structured images that are extremely similar in style to real target images without any manually labeled information. This technology can generate images and segmentation labels for tree-shaped targets in various medical images (such as: brain neurons, retinal blood vessels, pulmonary trachea, etc.), and the data quality can reach the level of manually labeled data. The present invention is the first model with the ability to automatically generate segmentation-level data, and has the advantages of simple and flexible model configuration, high-quality generated data, and wide application range. It can be used to replace the high-cost and time-consuming process of manually labeled data, and is expected to completely solve the bottleneck of the lack of labeled data in the processing of medical tree-shaped image segmentation by deep learning.

The invention relates to the technical field of medical image processing, in particular to a retinal blood vessel dynamic response evaluation system and method based on a graph neural network, and the method comprises the steps: a video obtaining system collects and preprocesses a blood vessel video of a to-be-detected target; the target identification and blood vessel tracking module receives the preprocessed video, completes feature extraction, target identification and detection, and determines a tracking target; the blood vessel diameter space-time dynamic estimation module receives the tracking target and estimates the blood vessel diameter in real time in combination with the blood vessel video; the blood vessel bifurcation point displacement spatio-temporal dynamic estimation module receives a tracking target and calculates a blood vessel displacement matrix by using a blood vessel pulsation segmentation algorithm and a blood pressure change estimation algorithm. The result display module receives the diameter of the blood vessel and the displacement matrix, displays an evaluation result, a spatial frequency domain blood vessel optical flow enhancement technology and a multi-scale structure tensor algorithm, reduces the measurement error of the diameter of the blood vessel from + / -5%-10% of a traditional method to + / -1%-3%, and can capture the dynamic change of the diameter of the micron-sized blood vessel.