92 results about "Blood vessel feature" patented technology

A histoengineered blood vessel features that its internal and external layers are made of the composite superfine fibre membrane and different cells are transplanted in different layers. Said composite superfine fibre membrane is composed of a core made of natural or artificial biodegradable material and the shell layer made of natural biodegradable material. Its external construction method includes such steps as preparing scaffold step by step, transplanting cells layer by layer, and integrally shaping.

The invention discloses a subcutaneous vein three-dimensional reconstruction method based on a hybrid matching strategy and aims to obtain the three-dimensional information of veins. The method includes the steps of firstly, using IUWT and Hessian matrix analysis to respectively obtain the blood vessel segmentation result and the related blood vessel feature image in each view; secondly, extracting and dividing blood vessel central lines through morphology and a blood vessel tracking algorithm to obtain the radius and blood vessel direction of each central line branch; thirdly, using epipolar constraint to calculate the candidate point set, of points in the central line branch of single view, in another view; respectively extracting SURF in the blood vessel similarity image of each view, completing SURF feature point matching, and using a Ransac method to calculate the homograph matrix among views; fifthly, using the layered matching strategy from part to whole to realize point-point matching between the blood vessel central lines of the views of two eyes according to the homograph matrix and the candidate matching point set, and completing the optimization of the homograph matrix during matching; sixthly, completing the three-dimensional reconstruction of the matching central line points according to the principle of triangular measuring, and recovering a three-dimensional blood vessel surface according to two-dimensional blood vessel diameter information.

A method for implanting the endothelial cells on the inner surface of artificial blood vessel features that a biomechanical method is used, that is, an external pulsive fluid shearing force increased step by step acts on the planted endothelial cells to make them be attached well on the inner surface of artificial blood vessel in advance.

A detecting system for ultrasonic blood vessel boundaries comprises an image extraction module, a tunica media and tunica adventitia boundary detecting subsystem, and a lumen boundary detecting subsystem, wherein the image extraction module is used for extracting blood vessel feature images from ultrasonic images, the tunica media and tunica adventitia boundary detecting subsystem is used for detecting the boundaries of blood vessel tunica media and blood vessel tunica adventitia from the blood vessel feature images through a region growing method, and the lumen boundary detecting subsystem is used for further detecting the blood vessel lumen boundaries in the inner regions, detected by the tunica media and tunica adventitia boundary detecting subsystem, of the boundaries of the tunica media and the tunica adventitia through a clustering method. The invention further provides a corresponding detecting method for the ultrasonic blood vessel boundaries. According to the detecting system for the ultrasonic blood vessel boundaries and the detecting method for the ultrasonic blood vessel boundaries, the images inside ultrasonic blood vessels are detected through the region growing method, so that images on two sides of each boundary of the blood vessel tunica media and tunica adventitia are easy to tell and images on two sides of each lumen boundary of the blood vessel images are easy to tell through the clustering method, and therefore detected coordinates of the boundaries of the tunica media and the tunica adventitia and the lumen boundaries are accurate.

The invention discloses an optic disc positioning method for a retina image. The optic disc positioning method for the retina image comprises the following steps that (a) bright plaques in the background are erased, and meanwhile blood vessel features are reserved; (b) candidate positions of the center of the optic disc are determined based on blood vessel convergence tendency features; (c) the finial position of the center of the optic disc is determined from the candidate positions of the center of the optic disc based on blood vessel direction features. The optic disc positioning method for the retina image can solve the problem that the positioning of the optic disc is prone to interference brought by various lesions and other noise in the background, and the accuracy and the speed of the positioning of the optic disc under the complex background can be improved.

An artificial blood vessel features that the spiral slot with arc cross-section is made on its inner surface for changing the flow field distribution of small-diameter artificial blood vessel to generate the flush of blood to inner surface and increase the shear force of blood, so preventing the deposition of harmful substances on the inner surface.

The invention relates to a coronary artery sequence blood vessel segmentation method based on space-time discriminative feature learning, which is used for carrying out blood vessel segmentation processing on a cardiac coronary artery angiography sequence image, and includes processing a current frame of image and several adjacent frames of images based on a pre-trained improved Unet network model, and obtaining blood vessel segmentation result of current frame image, wherein the improved Unet network model comprises a coding part, a jump connection layer and a decoding part, the coding part adopts a 3D convolution layer to perform time-space feature extraction, the decoding part is provided with a channel attention module, and the jump connection layer aggregates features extracted by thecoding part, thus obtaining an aggregation feature map and transmitting the aggregation feature map to the decoding part. Compared with the prior art, the cardiac coronary artery blood vessel segmentation method introduces the spatial-temporal features to perform cardiac coronary artery blood vessel segmentation, reduces the interference of time domain noise, emphasizes the blood vessel features,alleviates the problem of class imbalance in blood vessel segmentation, and has higher blood vessel segmentation accuracy.

The invention relates to a retinal blood vessel classification method and a retinal blood vessel classification device. The retinal blood vessel classification method combines a local binary algorithm with an independent component analysis technique for integrated use in the aspect of fundus blood vessel feature extraction, since the local binary algorithm and the independent component analysis technique performs structural description of blood vessel features from different angles, the method inherites the advantages of each single technology, supplements the blood vessels to certain extent, and filters the influence of more noise; and the retinal blood vessel classification method finally improves the precision of fundus blood vessel arteriovenous classification, reduces the proportion of blood vessels without classification marks, and provides a solid data basis for the follow-up diagnosis of diseases related to fundus blood vessels.

The invention provides an eyeground image processing method and device, a storage medium, and a processor. The method comprises the steps: carrying out the denoising of an eyeground image, and obtaining a preset eyeground image; carrying out the segmentation of a blood vessel in the preset eyeground image; extracting a blood vessel feature set in the preset eyeground image after segmentation; carrying out the matching of the blood vessel feature set with a preset eyeground image set; carrying out the affine transformation of a matching result, and obtaining a blood vessel fusion image of the eyeground image. According to the invention, the method can solve a problem that an eyeground image based on the blood vessel features cannot be effectively registered, and achieves the effective registration of the eyeground image according to the blood vessel features.

The embodiment of the invention provides a method and a device for segmenting a blood vessel wall and a blood flow area in a blood vessel OCT image, and the method comprises the steps: segmenting a contour map of the blood vessel wall by utilizing a first-stage full convolutional neural network in a cascade full convolutional neural network model based on a blood vessel Doppler OCT intensity map;and based on the contour map of the blood vessel wall, carrying out background noise removal processing on the blood vessel Doppler OCT phase map, and based on the processed blood vessel Doppler OCT phase map, segmenting the contour map of the blood flow region by using a second-stage full convolutional neural network in a cascade full convolutional neural network model. According to the embodiment of the invention, the cascade full convolution network is used for Doppler OCT blood vessel feature extraction, the segmentation of the blood flow area is more accurate in combination with the Doppler signal, the segmentation difficulty of the blood vessel wall and the blood flow area in the blood vessel OCT image can be effectively reduced, and the blood vessel wall and the blood flow area in the image can be conveniently and effectively segmented.

The embodiment of the invention relates to an eye fundus image blood vessel segmentation method. Each blood vessel feature processing layer sends an up-sampling image self-obtained to a connected blood vessel feature optimization layer; a blood vessel feature optimization layer at the lowest layer also acquires a feature image of a blood vessel feature extraction layer at the lowest layer; the blood vessel feature processing layer at the highest layer sends the up-sampling image self-obtained to the blood vessel feature optimization layer at the lowest layer by virtue of a backward short connection; each blood vessel feature optimization layer sequentially performs blood vessel feature extraction and non-linear processing on each acquired image and obtains a non-linear image correspondingto each image; and each blood vessel feature optimization layer sends each acquired image to the blood vessel feature optimization layer one layer higher by virtue of a forward short connection. The embodiment of the invention sends high layer information to lower layers by virtue of the backward short connection and sends low layer information to higher layers by virtue of the forward short connection and fully fuses features of all the levels, so that blood vessel segmentation is more accurate.

The invention discloses a method for determining branch points and branch segments of a lumen image, and belongs to the technical fields of medicine, three-dimensional imaging, digital image processing and the like. The method comprises the following steps: firstly, according to the characteristic that branch points are three or four blood vessel joint points, making a circle by taking each single-pixel blood vessel feature point as a circle center; judging whether a circumferential point set meets two conditions at the same time, wherein under the condition of meeting, forming a quasi-branchpoint neighborhood bya plurality of adjacent single-pixel blood vessel feature points, taking the sub-pixel center of the quasi-branch point neighborhood as a branch point, after the branch point neighborhood is finally determined, detecting a branch segment alongadjacent pixels with the branch point neighborhood as the starting point, if another branch point neighborhood is met, judging the branch segment as a complete branch segment, and otherwise, judging the branch segment as a half branch segment. The method provided by the invention not only can avoid the problem that the interruption point is misjudged as the branch point, but also can avoid the problem that a small amount of other micro features in the lumen are misjudged as the branch point.

The invention provides a blood vessel segmentation method and device and electronic equipment, and the method comprises the steps of obtaining a target region which comprises a to-be-segmented region;determining blood vessel candidate pixel points based on a preset blood vessel model and a fuzzy clustering algorithm; extracting a first blood vessel according to the blood vessel candidate pixel points; extracting second blood vessels based on the first blood vessels, and the density of the first blood vessels is larger than that of the second blood vessels; and combining the first blood vesseland the second blood vessel to obtain a blood vessel segmentation result. According to the invention, the influence of other tissues with similar blood vessel characteristics on blood vessel segmentation can be effectively reduced, and the influence of blood vessel characteristics on blood vessel segmentation can be effectively reduced. The invention can improve precision of blood vessel segmentation.

A method and a computer system are disclosed for automatic vectorization of the profile of a vessel tree and at least one of its properties on the basis of tomographic images of an examined patient. In at least one embodiment, using previously established location probabilities of landmarks in the vessel tree, there is an automatic determination of a plurality of distinctive landmarks in the current tomographic image data record of the patient, a registration of the current tomographic image data record to the statistical vessel model, an automatic determination of previously unidentified landmarks in the registered tomographic image data record using characteristic identification features of the previously unidentified landmarks from the statistical vessel model and the statistical location probability thereof, and a determination of at least one current vessel model using the identified landmarks and at least one vessel property at and/or between the identified landmarks.

The invention relates to a neural network model for blood vessel segmentation of fundus oculi images. The blood vessel feature processing layer of the highest layer is connected with the blood vesselfeature optimization layer of the lowest layer through a backward short connection. Each vessel feature optimization layer is connected with a vessel feature optimization layer of a higher layer through a forward short connection. Each vessel feature optimization layer for acquiring an upsampled image of the connected vessel feature processing layer; The lowest vessel feature optimization layer isalso used for acquiring the feature image of the lowest vessel feature extraction layer; Each vessel feature optimization layer is also used for sequentially performing vessel feature extraction andnonlinearization processing on each acquired image to obtain the corresponding nonlinearized image of each image, and sending each acquired image to the vessel feature optimization layer of a higher layer through a forward short connection. The invention transmits the high-level information to the low-level through the backward short connection, and transmits the low-level information to the high-level through the forward short connection, fully fuses the characteristics of each level, and makes the blood vessel segmentation more accurate.

The invention discloses a method for detecting a single-pixel blood vessel, and belongs to the technical fields of medicine, three-dimensional imaging, digital image processing and the like. Accordingto the method, Frangi blood vessel characteristic F characteristic parameters are calculated firstly, then the Frangi blood vessel characteristic quantity F is replaced with a blood vessel characteristic center line, and finally in order to overcome a large number of misjudgments caused by microvessels or background noise, the center line R is calculated in a weighted mode; according to the method for detecting the single-pixel blood vessel, a weighted calculation center line is adopted, so that the influence of microvessel and background noise can be inhibited; the technology is applied to abody surface projection virtual transparent observation inner cavity method in a minimally invasive surgery provided by an applicant. As a core part of the three-dimensional modeling method for the inner cavity in the method, the method lays a solid theoretical foundation for obtaining a more real virtual abdominal cavity transparent effect and meeting the requirement of doctors for assisting inobserving the inner cavity in a common minimally invasive surgery on the premise of not influencing the doctors and the operation environment.

The spectrum blood flow detection method of the color spectrogram based on ultrasonic Doppler comprises the following steps: firstly, obtaining a Doppler blood flow spectrogram by adopting an ultrasonic Doppler imaging technology, and measuring a blood flow speed maximum value and a blood flow speed minimum value in the Doppler blood flow spectrogram; carrying out edge extraction on the ultrasonicspectrogram to obtain a region of interest; performing feature extraction on the region of interest to obtain edge features and region features corresponding to the region of interest; constructing aradial basis function neural network model by taking the edge features and the region features as input layer vectors, and analyzing the input layer vector features in a neural network to obtain output quantities related to the blood vessel features; correcting the blood vessel characteristic related output quantity according to the blood flow speed maximum value and the blood flow speed minimumvalue to obtain blood flow speed output; the blood vessel area can be accurately marked and distinguished, the blood flow speed is corrected, and auxiliary diagnosis of the exact illness state is facilitated.

The embodiment of the invention provides a high-precision fundus blood vessel extraction method, a device, a medium and equipment. The method comprises the steps of obtaining a blood vessel pre-selection area image according to a fundus image; extracting a blood vessel center line of the blood vessel pre-selected region image; performing image segmentation processing on the blood vessel pre-selected region image to obtain an overall blood vessel region image; obtaining a first fundus blood vessel image according to the overall blood vessel area image and the blood vessel center line; extracting a second fundus blood vessel image from the fundus image according to a deep learning blood vessel feature detection model; and obtaining a final blood vessel segmentation image according to the first fundus blood vessel image and the second fundus blood vessel image. According to the method, based on combination of computer vision and artificial intelligence deep learning, the blood vessel segmentation image of the fundus image is extracted, organic combination and balance of the extraction precision and the extraction speed are achieved, and the robustness of the extraction result is improved.

The invention provides a fundus image blood vessel feature extraction method, device and system and a storage medium, and the method comprises the following steps: receiving a fundus image, and carrying out the preprocessing of the image, and obtaining a preprocessed image; extracting fundus texture features and fundus color features in the preprocessed image, and fusing the fundus texture features and the fundus color features to obtain a saliency image; carrying out threshold value cutting on the saliency image to obtain a cut image, and determining the position of a visual disk in the cut image; taking the position of the optic disc as a reference point, respectively calculating the average size of the pipe diameters of the branch blood vessels in the cutting image and outputting. According to the invention, the fundus image can be acquired, and the change of retinal vessel characteristics is automatically analyzed, so that the small vessel lesion condition is discovered in time, auxiliary diagnosis information is provided for doctors, and personalized health services of hypertension and other chronic diseases under precision medicine are realized.

The invention reveals a method, device, equipment and medium for predicting FFR (Fractional Flow Reserve) based on the GAN. The method comprises the steps: calculating an FFR calculation value of blood vessels in a target designated region by using a generated network model in a GAN model and a blood vessel feature vector of the target designated region; obtaining an FFR invasive measurement value, measured in an invasive testing mode, of the blood vessels of the target designated region; taking the FFR calculation value and the FFR invasive measurement value as input vectors of a discriminantnetwork model in the GAN model, and performing the iterative training of the GAN model; and performing the prediction and conversion of the FFR value of the blood vessels in the target designated region according to the trained GAN model. The method can accurately predict the FFR value, has a wide application range, has more freedom, avoids the risk of surgery, does not need to use a vasodilator,and is safer for patients. The method is low in computational complexity and is high in calculation speed.