Unlock instant, AI-driven research and patent intelligence for your innovation.
A Vein Segmentation Method Based on Deep Learning
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A venous blood vessel and deep learning technology, which is applied in the field of biometrics and medical imaging, can solve the problems of poor quality of venous blood vessel imaging, and achieve the effects of improving quality, reducing noise interference, improving accuracy and application generalization performance
Active Publication Date: 2022-04-22
HARBIN INST OF TECH
View PDF6 Cites 0 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
[0003] The purpose of the present invention is to solve the problem of poor quality of venous vessel imaging by traditional methods, and propose a deep learning-based venous vessel segmentation method
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more
Image
Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
specific Embodiment approach 1
[0012] Specific implementation mode one: combine figure 1 This embodiment will be described. A method for segmenting veins based on deep learning described in this embodiment, the method includes the following steps:
[0013] Step S1, collecting venous blood vessel images, and performing image enhancement processing on the collected images;
[0014] The image collected in step S1 is under the near-infrared light source with a wavelength of 850nm, through the FLIR near-infrared camera (GS3-U3-41C6NIR-C) to collect the arm vein pictures of 40 adults aged between 22 and 50, and 140 images were randomly selected from it as the whole dataset.
[0015] Step S2, labeling the enhanced image, adjusting the labeled image to a uniform size, obtaining a size-adjusted image, and forming a training data set with the size-adjusted image;
[0016] Step S3, using the training data set to train the UNet model of the fusion attention mechanism to obtain the trained UNet model of the fusion at...
specific Embodiment approach 2
[0021] Specific embodiment 2: The difference between this embodiment and specific embodiment 1 is that the image enhancement process in step S1 is performed by Contrast-Limited Adaptive Histogram Equalization (CLAHE), and before the image enhancement process, The collected images need to be converted from RGB format to grayscale images.
specific Embodiment approach 3
[0022] Specific embodiment three: the difference between this embodiment and specific embodiment one is: in the step S2, the enhanced image is marked, and the marked image is adjusted to a uniform size, and the specific process is as follows:
[0023] Experts annotate the enhanced image to obtain the label image of the arm veins, and adjust the annotated image to a 512×512 PNG image.
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More
PUM
Login to View More
Abstract
A vein segmentation method based on deep learning, which belongs to the field of biometrics and medical imaging technology. The invention solves the problem of poor quality of vein imaging by traditional methods. The specific implementation process of the method of the present invention is as follows: step S1, collecting venous blood vessel images, and performing image enhancement processing on the collected images; step S2, marking the enhanced images, adjusting the marked images to a uniform size, and obtaining The image after the size adjustment, the image after the size adjustment forms the training data set; Step S3, utilizes the training data set to train the UNet model of the fusion attention mechanism, obtains the UNet model of the fusion attention mechanism well trained; Step S4, will The collected new image is input into the trained UNet model of the fusion attention mechanism, and the vein vessel segmentation result of the new image is obtained through the trained UNet model of the fusion attention mechanism. The present invention can be applied to vein vessel segmentation.
Description
technical field [0001] The invention belongs to the technical field of biometric identification and medical imaging, and in particular relates to a vein segmentation method based on deep learning. Background technique [0002] Venipuncture is a necessary means for human health assessment, disease diagnosis and treatment. At present, almost all of them are done manually by medical staff, which not only has high work intensity, but also has a high failure rate. Due to the robot's high motion precision, strong perception performance, and good stability, fully automatic venipuncture robots have attracted extensive attention from researchers. Among them, venous blood vessel recognition is one of the important technologies of fully automatic venipuncture robots, because venous blood vessel recognition Accuracy is the key factor affecting the success of robot puncture, and it is also a major difficulty in the research of the robot. Vein recognition mainly utilizes the characterist...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More
Application Information
Patent Timeline
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to View More 
Login to View More