Recognition and positioning method and system based on vascular section in ultrasonic image

Abstract

The invention discloses an identification and positioning method and system based on a blood vessel section in an ultrasonic image, and relates to the field of ultrasonic imaging and automatic blood vessel identification. An ultrasonic imaging system is used in the system, and the imaging quality of the short-axis section of the blood vessel is optimal and the contrast ratio is highest by adjusting an imaging mode and imaging parameters. After a high-quality blood vessel section image is obtained, a series of algorithms such as blood vessel image signal enhancement, image noise weakening, blood vessel recognition and segmentation, true and false judgment algorithms of recognized and segmented blood vessels, and position and size calculation of the blood vessels are provided and established. Therefore, the blood vessel is identified and positioned. The blood vessel recognition positioning accuracy gt under various scenes is tested through a large amount of data; the calculation overhead of the recognition algorithm is controllable, and the calculation time overhead is lt; the time is 500ms. Therefore, accurate puncture position planning is provided for precise puncture of an automatic machine.

Description

Technical Field

[0001] The present invention relates to the fields of three-dimensional imaging and precision puncture, and in particular to a method and system for identifying and positioning blood vessel sections in ultrasonic images. Background Art

[0002] The identification and positioning method and system based on blood vessel sections in ultrasound images involve the intersection of medical imaging and image processing technology. Its background technology mainly includes the following aspects: 1. Medical Ultrasound Imaging Technology Medical ultrasound imaging is a non-invasive imaging technique that uses the propagation and reflection of high-frequency sound waves within the human body to form images. Its main advantages are safety, lack of radiation, real-time imaging, and relatively low cost. It is widely used in cardiovascular, abdominal, obstetrics and gynecology, and musculoskeletal fields.

[0003] 2. Vascular imaging and its importance Vascular imaging is a key application of ultrasound imaging, particularly in the diagnosis and treatment of cardiovascular disease. Vascular morphology, inner diameter, wall thickness, and blood flow are all crucial diagnostic indicators. Accurately identifying and locating vascular sections is crucial for early disease detection, surgical navigation, and treatment planning.

[0004] 3. Image Processing and Analysis Ultrasound images present challenges for automated processing due to noise, low contrast, and artifacts. Image processing and analysis techniques, such as edge detection, contour extraction, image segmentation, feature extraction, and pattern recognition, can improve the accuracy and stability of vessel identification to a certain extent.

[0005] 4. Machine Learning and Deep Learning Technologies In recent years, machine learning, particularly deep learning, has made significant progress in image processing. Deep learning models such as convolutional neural networks (CNNs) have demonstrated outstanding performance in image classification, segmentation, and feature extraction. Applying deep learning to vascular identification and localization in ultrasound images can effectively improve the system's automation and recognition accuracy.

[0006] 5. Clinical Application and Demand In clinical practice, doctors need to quickly and accurately obtain information about vascular structures. Traditional manual annotation methods are time-consuming and susceptible to subjective factors. Automated vessel identification and localization systems not only improve efficiency but also provide objective and consistent results, assisting doctors in diagnosis and treatment.

[0007] The current problems faced by manual puncture are: 1. High technical requirements Puncture procedures require extensive experience and a high level of technical skill from doctors or nurses. Even experienced professionals may face difficulties due to the patient's deep or thin blood vessels or other special circumstances.

[0008] 2. Success rate and accuracy The success rate and accuracy of punctures are crucial, especially in emergency and critically ill patients. The time and number of successful punctures directly impact treatment outcomes. Failed punctures not only increase patient pain but can also lead to complications such as local hematoma and infection.

[0009] 3. Individual differences among patients The difficulty of puncture can be affected by factors such as the patient's age, gender, body type, and vascular condition. For example, it is more difficult to locate and successfully puncture blood vessels in children, obese patients, and the elderly.

[0010] 4. Pain and discomfort During the puncture process, patients experience a certain amount of pain and discomfort. If the puncture fails multiple times, the patient's anxiety and discomfort will intensify, which not only affects the patient's cooperation but also increases the pressure on medical staff.

[0011] 5. Complications of puncture Improper puncture technique may lead to several complications, including: Hematoma: The needle punctures a blood vessel, causing local bleeding and hematoma.

[0012] Infection: Incomplete disinfection of the puncture site or improper operation may lead to local or systemic infection.

[0013] Nerve damage: Accidental puncture of a nerve can cause severe pain to the patient and even cause long-term nerve damage.

[0014] Arterial puncture: Accidentally inserting the needle into an artery instead of a vein can have serious consequences.

[0015] 6. Device Dependency Some punctures require ultrasound guidance and other equipment, but not all medical institutions are fully equipped, especially in primary care settings with limited resources. Even if the equipment is available, medical staff still need to receive appropriate training to use it proficiently.

[0016] 7. Workload In a busy clinical environment, frequent puncture operations will increase the workload and professional burnout of medical staff, affecting their work efficiency and service quality.

[0017] 8. Challenges in special circumstances In some special circumstances (such as emergency, ambulance, field hospital, etc.), environmental and conditional limitations will further increase the difficulty of the puncture operation. Summary of the Invention

[0018] To solve the above technical problems, the present invention adopts the following technical solution: a method and system for identifying and locating vascular sections in ultrasound images to achieve mechanical automatic and precise puncture.

[0019] Step 1: The ultrasound device automatically and accurately fits the skin surface for ultrasound imaging. Before this, infrared binocular structured light is used to perform three-dimensional reconstruction of the fitting area. Infrared three-dimensional image recognition locates the initial position of the blood vessel, and the robotic arm carries the ultrasound probe to fit the position. This will obtain an ultrasound image under the skin. Step 2: Optimize the imaging of vascular sections in ultrasound images. Adjust the imaging mode and imaging parameters to optimize the quality of vascular images. The parameters to be adjusted include: gain, depth, dynamic range, frequency, image enhancement, grayscale map, frame correlation, etc. Step 3: Signal enhancement and noise reduction methods for vascular images. The goal is to enhance the signal in the vascular region and reduce other noise signals, thereby enhancing the vascular signal. This method introduces salt-and-pepper noise into large black areas (such as the coupling agent and areas with imperfect adhesion) to distinguish them from the blood vessels. Then, a poetic and filtering operation is performed to preserve the low-frequency signal and the large circular signals in the vascular cross-section.

[0020] Step 4: Vessel recognition and true / false positive detection algorithms. By calculating the gradient of the entire image, the Aixam circle detection operator is proposed, based on the fact that the gradient inside a circle is smaller and the gradient at the edge is larger. A series of circle-like features are identified in the image. A clustering algorithm is used to determine true / false positives based on these features, and the cluster with the strongest signal is selected as the vessel slice identification result.

[0021] Step 5: Calculate the location of the blood vessel section. Using the circle-like clustering results, calculate the circle's pixel coordinates and radius. Then, using the ultrasound imaging device's pixel parameters and the probe's position, calculate the 3D coordinates and radius of the blood vessel's center.

[0022] Step 6: Precise mechanical puncture and blood sampling are achieved through positional planning guidance based on vessel identification. The relative positional relationship between the puncture structure and the ultrasound imaging device is fixed and precise. Mechanical assembly is precise. A spatial pre-calibration algorithm is used to establish the positional mapping relationship and method between the puncture needle tip and each point in the three-dimensional ultrasound image. The three-dimensional puncture point coordinates obtained in Step 5 are used to guide the puncture structure for precise puncture, ensuring that the puncture needle tip reaches the center of the vessel.

[0023] Compared with the prior art, the present invention has the following beneficial effects: 1. Improve the success rate of puncture By accurately identifying and locating vascular sections, the system can provide accurate puncture locations, greatly improving the success rate of puncture, which is especially important in patients with poor vascular conditions (such as children, the elderly or obese patients).

[0024] 2. Reduce patient suffering The automated system can reduce the pain and discomfort caused by multiple puncture failures and improve patients' treatment experience and cooperation.

[0025] 3. Reduce the risk of complications Accurate vascular positioning reduces the chances of mispuncture and repeated puncture, lowers the risk of complications such as hematoma, infection, and nerve damage, and improves the safety of the operation.

[0026] 4. Improve operational efficiency The automated system can quickly identify and locate blood vessels, shorten puncture operation time, and improve the efficiency of medical staff, enabling them to treat more patients more effectively.

[0027] 5. Reduce the pressure on medical staff System assistance can reduce the work pressure of medical staff in complex and high-pressure environments, reduce occupational burnout, and improve their job satisfaction.

[0028] 6. Improve the accuracy of diagnosis and treatment Accurate blood vessel identification and positioning not only facilitates puncture operations, but also provides more reliable data support for disease diagnosis and treatment, such as the detection of lesions such as vascular stenosis and aneurysms.

[0029] 7. Adaptable to various clinical environments The system can be used in a variety of clinical environments, including emergency rooms, operating rooms, ICUs, etc., and is particularly suitable for situations that require rapid response and high-precision operations.

[0030] 8. Supplementary training and education The system can also serve as a training tool to help novice medical staff master vessel identification and puncture techniques, accelerate their learning curve, and improve the technical level of the overall medical team.

[0031] 9. Improve the efficiency of medical resource utilization Reducing puncture failures and related complications can reduce the waste of medical resources, improve the utilization efficiency of medical resources, and reduce medical costs.

[0032] 10. Data Recording and Analysis The system can automatically record the data of each puncture operation, providing valuable information for subsequent clinical research and quality control, and promoting the continuous advancement of medical technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0034] Figure 1 The figure is a flow chart of the steps for implementing the method of the present invention.

[0035] Figure 2 Results of blood vessel identification and positioning for non-contact three-dimensional imaging.

[0036] Figure 3 Schematic diagram of the operation of the ultrasound probe and puncture structure Figure 4 The result of ultrasound image vascular section recognition Figure 5 Schematic diagram of the conversion calculation method between ultrasound image coordinates and physical three-dimensional coordinates Implementation Method

[0037] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0038] See also Figure 1 As shown, the present invention provides a method and system for identifying and locating blood vessel sections in ultrasound images, comprising: Step 1: The ultrasound device automatically and accurately fits the skin surface for ultrasound imaging. Before this, infrared binocular structured light is used to perform three-dimensional reconstruction of the fitting area. Infrared three-dimensional image recognition locates the initial position of the blood vessel, and the robotic arm carries the ultrasound probe to fit the position. This will obtain an ultrasound image under the skin.

[0039] A1. Infrared binocular structured light 3D reconstruction Step description: Use infrared binocular structured light to scan the skin surface and perform three-dimensional reconstruction of the fitting area.

[0040] Technical details: Infrared binocular structured light can obtain high-precision three-dimensional images and generate detailed models of the skin surface and its underlying structures.

[0041] Purpose: To provide accurate positioning of the ultrasound probe and to accurately model the skin surface features.

[0042] A2. Preliminary vascular localization Step description: In the three-dimensional reconstructed image, the preliminary position of the blood vessel is located using image recognition technology.

[0043] Technical details: Utilizes the depth information and texture features of infrared images, combined with image processing algorithms and deep learning models, to identify and locate the potential locations of blood vessels.

[0044] Purpose: To provide a reference for subsequent ultrasound probe positioning.

[0045] A3. The robotic arm carries the ultrasound probe for positioning Step description: The robotic arm carries the ultrasound probe and moves it to the predetermined position based on the preliminary blood vessel positioning information in the infrared three-dimensional image, and then fits it to the skin surface.

[0046] Technical details: The robotic arm has high-precision motion control capabilities and can ensure that the ultrasound probe fits firmly and accurately on the skin surface according to a pre-set path and force feedback mechanism.

[0047] Purpose: To achieve precise fitting of the ultrasound probe and ensure the quality of subsequent ultrasound imaging.

[0048] A4. Ultrasound imaging to obtain images beneath the skin Step description: The ultrasound probe is placed in the fitted position to perform ultrasound imaging and obtain images of blood vessels under the skin.

[0049] Technical details: High-frequency ultrasound is used to penetrate the skin and tissues. The reflected sound wave signals are received by the probe and converted into images.

[0050] Objective: To obtain clear and accurate images of subcutaneous blood vessels to provide a basis for further diagnosis and treatment.

[0051] Step 2: Optimize the imaging of vascular sections in ultrasound images. Adjust the imaging mode and imaging parameters to optimize the quality of vascular images. The parameters to be adjusted include: gain, depth, dynamic range, frequency, image enhancement, grayscale map, frame correlation, etc. B1. Gain Gain refers to the intensity of the amplified ultrasound signal.

[0052] Optimization strategy: Moderately increasing the gain can enhance image brightness and make vascular structures clearer. However, excessive gain can increase noise and degrade image quality. The gain should be adjusted to a level that clearly displays blood vessels without increasing excessive noise.

[0053] B2. Depth The depth setting determines the maximum depth of ultrasonic detection.

[0054] Optimization strategy: Adjust the ultrasound imaging depth parameter according to the depth of the blood vessels so that the blood vessels are located in the center of the image. Too large a depth will result in reduced resolution, while too small a depth will result in incomplete target structures in the image.

[0055] B3. Dynamic Range Dynamic range controls the range of gray levels displayed in an image.

[0056] Optimization strategy: A higher dynamic range can provide more gray levels, resulting in higher contrast between blood vessels and surrounding tissues, thereby improving image details and clarity.

[0057] B4. Frequency The frequency of the ultrasound probe affects the resolution and penetration of the imaging.

[0058] Optimization strategy: High-frequency probes offer higher resolution but lower penetration and are suitable for imaging superficial tissues. Low-frequency probes offer greater penetration but lower resolution and are suitable for imaging deeper tissues. The appropriate frequency should be selected based on the depth of the blood vessels.

[0059] B5. Image Enhancement Image enhancement technology is used to improve the contrast and edge definition of images.

[0060] Optimization strategy: Using image enhancement functions (such as edge enhancement, smoothing filtering, etc.) can highlight blood vessel edges, reduce artifacts and noise, and enhance blood vessel visibility.

[0061] B6. Gray Scale Mapping Grayscale maps are used to adjust the contrast and brightness of images.

[0062] Optimization strategy: Selecting an appropriate grayscale map can enhance image contrast and make the distinction between blood vessels and surrounding tissue more distinct. Avoid using grayscale maps that are too flat or too steep.

[0063] B7. Frame Correlation Frame correlation technology is used to reduce noise and improve image stability.

[0064] Optimization strategy: Through frame correlation processing, the image signals of multiple frames can be averaged to reduce the impact of random noise and improve image stability and clarity. However, excessive frame correlation may cause image blur.

[0065] Step 3: Signal enhancement and noise reduction methods for vascular images. The goal is to enhance the signal in the vascular region and reduce other noise signals, thereby enhancing the vascular signal. This method introduces salt-and-pepper noise into large black areas (such as the coupling agent and areas with imperfect adhesion) to distinguish them from the blood vessels. Then, a poetic and filtering operation is performed to preserve the low-frequency signal and the large circular signals in the vascular cross-section.

[0066]

[0067] C1. Introducing salt and pepper noise First, large black areas (such as coupling agent and areas that are not perfectly fitted) are processed into salt and pepper noise. The purpose of this step is to distinguish these areas and make subsequent processing more efficient.

[0068] C2. Preprocessing Operations Preprocess the image to improve the effectiveness of subsequent filtering and enhancement steps.

[0069] Grayscale conversion: Convert ultrasound images to grayscale images to simplify processing.

[0070] Histogram Equalization: Perform histogram equalization on the image to enhance contrast.

[0071] C3. Morphological Operations Morphological operations are used to preliminarily process the image and enhance the features of the vascular area.

[0072] Erosion: Removes noise points and reduces the impact of salt and pepper noise.

[0073] Dilation: Enhances the continuity of blood vessels and makes the blood vessel boundaries clearer.

[0074] C4. Filtering Operation Select appropriate filters to process the image to reduce noise and preserve vascular signals.

[0075] Median Filtering: Median filtering is very effective in removing salt and pepper noise and can reduce noise interference in images.

[0076] Gaussian Filtering: Gaussian filtering can smooth images and reduce high-frequency noise while preserving the main structure of blood vessels.

[0077] C5. Frequency Domain Processing Convert the image to the frequency domain and achieve signal enhancement and noise reduction through frequency domain processing.

[0078] Fourier Transform: Convert the image to the frequency domain and analyze its frequency components.

[0079] Frequency domain filtering: Design a bandpass filter to retain low-frequency signals and large circular signals of vascular sections, and weaken other high-frequency noise.

[0080] C6. Morphological Enhancement Morphological operations are further used to enhance the blood vessel area.

[0081] Top-Hat Transform: Enhances bright blood vessel areas in images.

[0082] Black-Hat Transform: Enhances dark blood vessel areas in images.

[0083] C7. Feature Extraction Feature extraction technology is used to further enhance vascular signals and suppress noise in other areas.

[0084] Edge detection: Use edge detection algorithms such as Sobel and Canny to extract blood vessel edges.

[0085] Region Growing: Perform region growing based on seed points to extract blood vessel regions.

[0086] C8. Post-processing Finally, the processed images are post-processed to ensure the signal enhancement effect in the vascular area.

[0087] Wavelet Transform: Wavelet transform is used to perform multi-scale analysis on images to further enhance vascular signals.

[0088] Image fusion: Fusion of multiple processing results to obtain the final enhanced vascular image.

[0089] Step 4: Vessel recognition and true / false positive detection algorithms. By calculating the gradient of the entire image, the Aixam circle detection operator is proposed, based on the fact that the gradient inside a circle is smaller and the gradient at the edge is larger. A series of circle-like features are identified in the image. A clustering algorithm is used to determine true / false positives based on these features, and the cluster with the strongest signal is selected as the vessel slice identification result.

[0090]

[0091] D1. Full-image gradient calculation First, edge features are identified by calculating the gradient of the entire image. Using a gradient calculation method (such as the Sobel operator), a gradient magnitude image is generated for each pixel in the image. Regions with larger gradient magnitudes typically correspond to edges in the image, while regions with smaller gradient magnitudes represent relatively flat areas.

[0092] D2. Aixam circle detection operator The Aixam circle detection operator is designed based on the gradient characteristics to detect circular structures in images. This detection operator uses the following features: Smaller internal gradient: The pixel gradient value inside the blood vessel is smaller because the internal area is flatter.

[0093] Large edge gradient: The pixel gradient value at the edge of the blood vessel is large because there is a significant intensity change in the edge area.

[0094] Through the Aixam circle detection operator, a series of circular regions can be identified in the gradient magnitude image, which may correspond to the cross-section of blood vessels.

[0095] D3. Circle-like clustering algorithm Cluster analysis is performed on the detected circular regions to distinguish real blood vessels from pseudo-vessels. The main steps of the clustering algorithm are as follows: Feature extraction: Extract the center coordinates and radius of each class circle as a feature vector.

[0096] Clustering: Use a clustering algorithm (such as DBSCAN or K-means) to group circles with similar characteristics into clusters. Clustering algorithms analyze the spatial distribution and geometric characteristics of circles to identify circles with similar properties.

[0097] D4. Determine whether it is true or false In the clustering results, the circle clustering results with the strongest signal are further screened out. The key steps for judging authenticity include: Cluster center feature analysis: Calculate the center features of each cluster, such as the gradient mean of the center point and the edge gradient mean.

[0098] Signal strength evaluation: Evaluate the signal strength of the circle-like elements within each cluster. Signal strength is measured by the difference between the internal gradient and the edge gradient. Circle-like clusters representing true blood vessels should have low internal gradients and high edge gradients.

[0099] Screening the strongest signal cluster: Based on the signal intensity evaluation results, screen out the circle-like cluster with the strongest signal as the recognition result of the vascular slice.

[0100] Step 5: Calculate the location of the blood vessel section. Using the circle-like clustering results, calculate the circle's pixel coordinates and radius. Then, using the ultrasound imaging device's pixel parameters and the probe's position, calculate the 3D coordinates and radius of the blood vessel's center.

[0101] E1. Extracting pixel coordinates and pixel radius of the circle Extract the pixel coordinates (center coordinates) and pixel radius of each circle from the circle-like clustering results. The clustering results contain multiple circular regions, each defined by the pixel coordinates (x, y) of the center point and the radius (r).

[0102] E2. Obtaining the pixel parameters and probe position of the ultrasound imaging device Obtain relevant parameters of the ultrasound imaging device through device calibration: The physical size of each pixel in the x direction (pixel_size_x) The physical size of each pixel in the y direction (pixel_size_y) The position of the probe in the z direction (probe_position_z) These parameters define the transformation relationship between image coordinates and physical space coordinates.

[0103] E3. Calculate the 3D coordinates of the blood vessel center According to the pixel parameters of the ultrasound imaging device and the position of the probe, the pixel coordinates of the circle are converted into three-dimensional space coordinates: Three-dimensional coordinates in the x-direction: Multiply the x-pixel coordinate of the circle by the physical size of each pixel in the x-direction to obtain the corresponding physical space coordinates.

[0104] Three-dimensional coordinates in the y direction: Multiply the y pixel coordinate of the circle by the physical size of each pixel in the y direction to obtain the corresponding physical space coordinates.

[0105] Three-dimensional coordinates in the z direction: Since the position of the probe is fixed, the z coordinate is the position of the probe in the z direction (probe_position_z).

[0106] E4. Calculate the 3D radius of blood vessels Convert the circle's pixel radius to a 3D radius: By calculating the average of the physical size of each pixel in the x and y directions, a comprehensive pixel physical size is obtained.

[0107] Multiply the pixel radius of the circle by this comprehensive pixel physical size to get the corresponding three-dimensional space radius.

[0108] Step 6: Precise mechanical puncture and blood sampling are achieved through positional planning guidance based on vessel identification. The relative positional relationship between the puncture structure and the ultrasound imaging device is fixed and precise. Mechanical assembly is precise. A spatial pre-calibration algorithm is used to establish the positional mapping relationship and method between the puncture needle tip and each point in the three-dimensional ultrasound image. The three-dimensional puncture point coordinates obtained in Step 5 are used to guide the puncture structure for precise puncture, ensuring that the puncture needle tip reaches the center of the vessel.

[0109] F1. Spatial calibration and mapping relationship establishment Use a spatial calibration algorithm to establish a positional mapping relationship between the puncture needle tip and each point in the three-dimensional ultrasound image. The calibration process includes: Calibrate the positional relationship between the ultrasound imaging device and the puncture structure to ensure that the relative position of the two is fixed and accurate.

[0110] Measure and record the three-dimensional coordinates of multiple known points in the ultrasound image and their pixel positions in the image.

[0111] The transformation model between the image coordinate system and the mechanical coordinate system is established through multi-point calibration data.

[0112] F2. 3D puncture point calculation Using the spatial mapping relationship, the three-dimensional spatial coordinates of the identified blood vessel center are converted into the target position of the puncture needle tip.

[0113] The puncture path is calculated to ensure that the needle tip reaches the center of the blood vessel at the optimal angle and depth to avoid damaging surrounding tissues.

[0114] F3. Mechanical assembly and puncture guidance The ultrasound imaging device is precisely assembled with the puncture structure to ensure the validity of the spatial mapping relationship.

[0115] Control the robotic arm or puncture device to perform precise puncture based on the calculated three-dimensional puncture point and path.

[0116] The puncture process is monitored through a real-time feedback system, and the puncture path is fine-tuned to ensure that the needle tip accurately reaches the center of the blood vessel.

Claims

1. A method and system for identifying and locating blood vessel sections in ultrasound images, characterized in that: include: 1) The ultrasonic device automatically and accurately fits the skin surface for ultrasonic imaging. Before that, infrared binocular structured light is used to perform three-dimensional reconstruction of the fitting area. The infrared three-dimensional image recognition locates the initial position of the blood vessel, and the robotic arm carries the ultrasonic probe to fit the position. This will obtain an ultrasonic image under the skin. ； 2) Optimize the imaging of vascular sections in ultrasound images. Adjust the imaging mode and imaging parameters to optimize the quality of vascular images. The parameters to be adjusted include: gain, depth, dynamic range, frequency, image enhancement, grayscale spectrum, frame correlation, etc. 3) Signal enhancement and noise reduction methods for vascular images. The goal is to enhance the signal in the vascular area and reduce other noise signals to achieve vascular signal enhancement. The method is to introduce salt and pepper noise into large black areas (coupling agent, parts that are not ideally fitted) to distinguish them from blood vessels. Then perform poetry and filtering operations to retain low-frequency signals and large circular signals in vascular sections; ； 4) Blood vessel recognition algorithm and true and false judgment algorithm. Through the gradient calculation of the whole image, the aixam circle detection operator is proposed because the internal gradient of the circle is small and the edge gradient is large. The image will identify a series of circles. Through the clustering algorithm of many circles, the true and false judgment is performed, and the clustering result of the circle with the strongest signal is selected as the recognition result of the blood vessel slice; ； 5) Calculate the position of the blood vessel section. Calculate the pixel coordinates and pixel radius of the circle through the clustering results of the circle-like objects. Then calculate the three-dimensional coordinates and three-dimensional radius of the blood vessel center through the pixel parameters of the ultrasound imaging device and the position of the probe. 6) Through the guidance of the position planning of blood vessel recognition, mechanical precise puncture and blood collection are completed. The relative position relationship between the puncture structure and the ultrasound imaging device is fixed and accurate. The mechanical assembly is accurate. Through the pre-calibration algorithm of space, the position mapping relationship and method between the puncture needle tip and each point of the three-dimensional ultrasound image can be established. After the three-dimensional puncture point spatial coordinates are obtained in step 5, the puncture structure can be guided to puncture accurately, and the puncture needle tip reaches the center point of the blood vessel.

2. An ultrasound imaging system as described in claim 1 needs to have a better sound wave frequency, a rich number of channels, a deeper detection depth, and a portable probe design, so as to achieve real-time, accurate, high-resolution, and portable skin-fitted vascular imaging quality.

3. A vascular section signal enhancement algorithm as claimed in claim 1, involving salt and pepper noise filling of invalid areas, appropriate corrosion operations of the overall image, filtering operations, blurring of over-bright areas, curve adjustment and rationalization of overall image brightness.

4. A blood vessel recognition algorithm as claimed in claim 1. The gradient distribution is obtained by calculating the gradient of the whole image. The aixam circle detection algorithm is proposed based on the principle that the gradient inside the circle is small and the gradient at the edge of the circle is large, and various rich circles in the image are detected by a lower threshold. Then the most effective circle clustering is obtained by the clustering algorithm to extract the optimal circle signal, which is the recognition and extraction result of the blood vessel section.

5. The method of calculating the optimal puncture point position and performing precise puncture as claimed in claim 1. The image coordinates of the blood vessel section circle are converted into three-dimensional spatial coordinates, and then the puncture control coordinates of the puncture mechanism are calculated by a pre-calibrated hand-eye method. The puncture needle tip is then accurately punctured by the puncture mechanism to reach the ideal position, so that subsequent operations can be performed.

Images