Dynamic blood flow imaging system based on artificial intelligence

Abstract

The invention discloses a dynamic blood flow imaging system based on artificial intelligence, and belongs to the technical field of ultrasonic imaging. Comprising an ultrasonic probe, and a multi-mode imaging module is arranged in the ultrasonic probe and used for collecting blood flow data of target tissue and generating a blood flow image according to the blood flow data; the self-adaptive optimization module is connected with the multi-mode imaging module and adjusts acquisition parameters of the ultrasonic probe according to the blood flow data; the image processing module is connected with the ultrasonic probe and is used for automatically identifying and analyzing dynamic blood flow information in the blood flow image through a deep learning model to obtain a processed image; and the image display module is connected with the image processing module and is used for displaying the processed image. The technical scheme has the beneficial effects that by introducing the deep learning algorithm, the real-time performance, the definition and the automation degree of dynamic blood flow imaging are improved, and a more accurate and high-quality blood flow imaging result is provided.

Description

Technical Field

[0001] The present invention relates to the technical field of ultrasonic imaging, and in particular to a dynamic blood flow imaging system based on artificial intelligence. Background Art

[0002] Ultrasound imaging technology plays an important role in medicine, especially in the fields of cardiovascular and tumors. Dynamic blood flow imaging technology can provide detailed information on the dynamic changes of blood flow. However, traditional dynamic blood flow imaging technology has certain limitations in image clarity, real-time performance and automation, and it is difficult to fully meet clinical needs.

[0003] Existing dynamic blood flow imaging technology mainly relies on traditional ultrasound equipment and manual operation, resulting in low image clarity and contrast; existing technology is difficult to achieve real-time imaging with high frame rate, resulting in untimely capture of dynamic blood flow changes; it needs to rely on manual operation and work experience, making it difficult to achieve automated and intelligent imaging; there is a large amount of noise in ultrasound images, which affects the image quality and the accuracy of blood flow information. Summary of the invention

[0004] The purpose of the present invention is to provide a dynamic blood flow imaging system based on artificial intelligence to solve the above technical problems;

[0005] A dynamic blood flow imaging system based on artificial intelligence, comprising:

[0006] An ultrasonic probe, wherein the ultrasonic probe is provided with:

[0007] A multimodal imaging module, used for collecting blood flow data of a target tissue and generating a blood flow image according to the blood flow data;

[0008] An adaptive optimization module, connected to the multimodal imaging module, and adjusting acquisition parameters of the ultrasound probe according to the blood flow data;

[0009] An image processing module, connected to the ultrasound probe, automatically identifies and analyzes the dynamic blood flow information in the blood flow image through a deep learning model to obtain a processed image;

[0010] An image display module is connected to the image processing module and is used to display the processed image.

[0011] Preferably, the multimodal imaging module comprises:

[0012] A first imaging unit extracts the relationship between the target tissue blood flow and the surrounding tissue based on computer tomography or magnetic resonance imaging technology;

[0013] A second imaging unit, for acquiring dynamic blood flow information of the target tissue based on optical coherence tomography;

[0014] The third imaging unit reflects the dynamic blood flow information and tissue structure of the target tissue by measuring the electrical impedance change parameter of the target tissue.

[0015] Preferably, the adaptive optimization module comprises:

[0016] A real-time feedback unit, which evaluates the blood flow data based on an image quality evaluation algorithm and adjusts the scanning parameters of the ultrasound probe according to the evaluation result;

[0017] An intelligent acquisition path planning unit, which analyzes the blood flow change trend based on a path prediction model and plans the moving path and acquisition angle of the ultrasound probe;

[0018] A multi-frequency and multi-angle acquisition unit, wherein the ultrasound probe acquires dynamic blood flow information at different depths and directions through multi-frequency and multi-angle acquisition;

[0019] The automatic calibration and optimization unit adjusts the calibration parameters of the ultrasonic probe through automatic calibration technology.

[0020] Preferably, the image processing module comprises:

[0021] A preprocessing unit, which preprocesses the blood flow image;

[0022] The artificial intelligence processing unit is connected to the preprocessing unit and extracts dynamic blood flow information in the preprocessed image through the deep learning model.

[0023] Preferably, the pre-processing unit comprises:

[0024] A denoising processing subunit, which automatically identifies and removes noise in the blood flow image through a convolutional neural network;

[0025] An image enhancement subunit, connected to the denoising processing subunit, and enhancing the contrast of the blood flow image through an image enhancement algorithm;

[0026] The standardization processing subunit is connected to the image enhancement subunit to unify the size and brightness of the blood flow image.

[0027] Preferably, the artificial intelligence processing unit comprises:

[0028] A model generation subunit constructs and trains an initialization model to obtain the deep learning model;

[0029] A real-time analysis subunit connected to the model generation subunit, wherein the deep learning model analyzes the blood flow image in real time through parallel computing and graphics processor acceleration technology;

[0030] The information extraction subunit is connected to the real-time analysis subunit and extracts the dynamic blood flow information of the blood flow image through the deep learning model.

[0031] Preferably, the image display module includes a color Doppler display mode and a power Doppler display mode.

[0032] Preferably, the artificial intelligence processing unit also includes an abnormality detection subunit, which is connected to the real-time analysis subunit and detects and marks abnormal areas in the blood flow through the deep learning model.

[0033] Preferably, the blood flow image of the ultrasound probe is transmitted to the image processing module via a unified serial bus or optical fiber.

[0034] Preferably, the dynamic blood flow information in the blood flow image includes blood flow velocity, blood flow direction and blood flow volume.

[0035] The beneficial effect of the present invention is that by introducing a deep learning algorithm, the real-time, clarity and automation of dynamic blood flow imaging are improved, and more accurate and high-quality blood flow imaging results are provided. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a connection block diagram of the dynamic blood flow imaging system based on artificial intelligence of the present invention;

[0037] Figure 2 is a schematic diagram of a multimodal imaging module of the present invention;

[0038] Figure 3 is a schematic diagram of an adaptive optimization module of the present invention;

[0039] Figure 4 is a schematic diagram of an image processing module of the present invention;

[0040] Figure 5 is a schematic diagram of a pre-processing unit of the present invention;

[0041] Figure 6 Schematic diagram of an artificial intelligence processing unit of the present invention.

[0042] In the accompanying drawings: 1. Ultrasonic probe; 11. Multimodal imaging module; 111. First imaging unit; 112. Second imaging unit; 113. Third imaging unit; 12. Adaptive optimization module; 121. Real-time feedback unit; 122. Intelligent acquisition path planning unit; 123. Multi-frequency and multi-angle acquisition unit; 124. Automatic calibration and optimization unit; 2. Image processing module; 21. Preprocessing unit; 211. De-noising processing subunit; 212. Image enhancement subunit; 213. Standardization processing subunit; 22. Artificial intelligence processing unit; 221. Model generation subunit; 222. Real-time analysis subunit; 223. Information extraction subunit; 224. Abnormality detection subunit; 3. Image display module. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0044] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0045] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but they are not intended to limit the present invention.

[0046] A dynamic blood flow imaging system based on artificial intelligence, such as Figure 1 As shown, including,

[0047] Ultrasonic probe 1, the ultrasonic probe 1 is provided with:

[0048] The multimodal imaging module 11 is used to collect blood flow data of the target tissue and generate a blood flow image according to the blood flow data;

[0049] The adaptive optimization module 12 is connected to the multimodal imaging module 11 and adjusts the acquisition parameters of the ultrasound probe 1 according to the blood flow data;

[0050] The image processing module 2 is connected to the ultrasound probe 1, and automatically identifies and analyzes the dynamic blood flow information in the blood flow image through a deep learning model to obtain a processed image;

[0051] The image display module 3 is connected to the image processing module 2 and is used to display the processed image.

[0052] Specifically, the present invention provides a dynamic blood flow imaging system based on artificial intelligence, which uses a high-frequency ultrasonic probe 1 to provide high-resolution blood flow images. The probe generates blood flow images by emitting ultrasonic waves and receiving reflected signals. The multimodal imaging module 11 is intended to provide more comprehensive blood flow information. By combining a variety of imaging methods, it is possible to obtain richer and multi-dimensional blood flow data in different imaging modes. The adaptive optimization module 12 is intended to dynamically adjust the acquisition parameters of the probe according to the real-time analysis results to improve the image quality and acquisition efficiency. Combined with the deep learning algorithm, the image clarity and automation of dynamic blood flow imaging are improved, and more accurate, timely and intelligent blood flow imaging results are provided.

[0053] In a preferred embodiment, referring to Figure 2 , the multimodal imaging module 11 includes,

[0054] A first imaging unit 111 extracts the relationship between the target tissue blood flow and the surrounding tissue based on computer tomography or magnetic resonance imaging technology;

[0055] A second imaging unit 112 acquires dynamic blood flow information of a target tissue based on optical coherence tomography;

[0056] The third imaging unit 113 reflects the dynamic blood flow information and tissue structure of the target tissue by measuring the electrical impedance change parameter of the target tissue.

[0057] Specifically, the first imaging unit 111 integrates ultrasound and CT (computed tomography) / MRI (magnetic resonance imaging), and synchronously collects and fuses image data of different modalities through a hardware interface and a data fusion algorithm. Ultrasound provides real-time dynamic blood flow information, and CT / MRI provides high-resolution anatomical structure information. The combination can more comprehensively reflect the relationship between blood flow and surrounding tissues.

[0058] The second imaging unit 112 obtains blood flow microstructure information through optical coherence tomography (OCT). The OCT module is integrated in the ultrasound probe 1, and a higher resolution blood flow image is obtained through the combined imaging technology of optics and ultrasound. It provides more detailed blood flow microstructure information, which is particularly suitable for the study of small blood vessels and microcirculation.

[0059] The third imaging unit 113 is electrical impedance tomography (EIT), which reflects blood flow and tissue structure by measuring changes in tissue electrical impedance. The electrical impedance imaging module is integrated in the ultrasound probe 1, and the electrical impedance and ultrasound imaging technology are combined to provide the electrophysiological characteristics of blood flow. Combined with the mechanical information of ultrasound, the blood flow state can be analyzed more comprehensively.

[0060] In a preferred embodiment, referring to Figure 3 , the adaptive optimization module 12 includes,

[0061] A real-time feedback unit 121 evaluates blood flow data based on an image quality evaluation algorithm and adjusts scanning parameters of the ultrasound probe 1 according to the evaluation results;

[0062] An intelligent acquisition path planning unit 122 analyzes the blood flow change trend based on the path prediction model and plans the moving path and acquisition angle of the ultrasound probe 1;

[0063] The multi-frequency and multi-angle acquisition unit 123, the ultrasonic probe 1 acquires dynamic blood flow information at different depths and directions through multi-frequency and multi-angle acquisition;

[0064] The automatic calibration and optimization unit 124 adjusts the calibration parameters of the ultrasound probe 1 through the automatic calibration technology.

[0065] Specifically, the introduction of adaptive acquisition technology aims to dynamically adjust acquisition parameters according to real-time analysis results to improve image quality and acquisition efficiency. The real-time feedback unit 121 analyzes the acquired image data in real time and provides feedback on the current image quality and the adaptability of acquisition parameters. A real-time image quality evaluation algorithm is integrated in the acquisition module, and the frequency, gain, focal length and other parameters of the ultrasound probe 1 are dynamically adjusted according to the evaluation results. This ensures that the best image quality can be obtained under different imaging conditions.

[0066] The intelligent acquisition path planning unit 122 intelligently plans the acquisition path of the ultrasound probe 1 according to the dynamic changes of the blood flow image. The AI ​​algorithm analyzes the dynamic information of the blood flow image, predicts the change trend of the blood flow, and plans the moving path and acquisition angle of the ultrasound probe 1. This improves the acquisition efficiency, reduces unnecessary repeated acquisition processes, and ensures high-quality imaging of key areas.

[0067] The multi-frequency and multi-angle acquisition unit 123 acquires blood flow information at different depths and directions through a combination of multi-frequency and multi-angle acquisition. The multi-frequency transmitting and multi-angle receiving modules are integrated in the ultrasonic probe 1, and the frequency and angle are dynamically adjusted according to the real-time analysis results. This provides more comprehensive blood flow information and improves the depth resolution and direction resolution of the image.

[0068] The automatic calibration and optimization unit 124 uses automatic calibration technology to adjust and optimize acquisition parameters in real time to ensure stability and consistency during long-term acquisition. An automatic calibration algorithm is integrated in the acquisition module to automatically adjust the calibration parameters of the probe according to real-time acquisition data. This improves the stability and consistency of the acquisition process and reduces errors caused by human operation.

[0069] In a preferred embodiment, referring to Figure 4 , Figure 5 , the image processing module 2 includes,

[0070] A preprocessing unit 21, which preprocesses the blood flow image;

[0071] The artificial intelligence processing unit 22 is connected to the preprocessing unit 21 and extracts dynamic blood flow information from the preprocessed image through a deep learning model;

[0072] The pre-processing unit 21 includes:

[0073] A denoising processing subunit 211 automatically identifies and removes noise in the blood flow image through a convolutional neural network;

[0074] The image enhancement subunit 212 is connected to the denoising processing subunit 211 and enhances the contrast of the blood flow image through an image enhancement algorithm;

[0075] The standardization processing subunit 213 is connected to the image enhancement subunit 212 to unify the size and brightness of the blood flow image.

[0076] Specifically, deep learning algorithms such as convolutional neural networks are used to automatically identify and remove noise in images to improve image clarity. Image enhancement algorithms (such as histogram equalization and contrast enhancement) are used to enhance image details and contrast. Images are standardized in terms of size and brightness to ensure consistency and comparability of image data.

[0077] In a preferred embodiment, referring to Figure 6 , the artificial intelligence processing unit 22 includes,

[0078] The model generation subunit 221 constructs and trains an initialization model to obtain a deep learning model;

[0079] The real-time analysis subunit 222 is connected to the model generation subunit 221, and the deep learning model analyzes the blood flow image in real time through parallel computing and graphics processor acceleration technology;

[0080] The information extraction subunit 223 is connected to the real-time analysis subunit 222, and extracts the dynamic blood flow information of the blood flow image through a deep learning model.

[0081] Specifically, deep learning models (such as U-Net and ResNet) can automatically identify and analyze dynamic information in blood flow images. The training data of the model includes a large number of annotated blood flow images to ensure the high accuracy and robustness of the model. The AI ​​(artificial intelligence) processing module processes and analyzes the collected blood flow images in real time through parallel computing and GPU (graphics processing unit) acceleration technology to ensure the timely capture of dynamic blood flow information.

[0082] In a preferred embodiment, the image display module 3 includes a color Doppler display mode and a power Doppler display mode.

[0083] Specifically, the image display module 3 is a high-definition display, which uses a high-resolution medical display and can accurately display the processed dynamic blood flow image, ensuring that the operator can clearly observe the dynamic changes of the blood flow.

[0084] A user-friendly interface is designed to provide a variety of image display modes (such as color Doppler and power Doppler). The operator can select the appropriate display mode according to needs.

[0085] The image display module 3 can update the processed blood flow image in real time, ensure the timely presentation of dynamic changes, and improve the real-time performance of imaging.

[0086] In a preferred embodiment, the artificial intelligence processing unit 22 also includes an abnormality detection subunit 224, which is connected to the real-time analysis subunit 222 and detects and marks abnormal areas in the blood flow through a deep learning model.

[0087] Specifically, deep learning models can analyze large amounts of data, learn complex patterns and features from them, and reduce errors in human judgment. Reducing human intervention makes the entire analysis process more efficient and reduces the burden on medical workers.

[0088] In a preferred embodiment, the blood flow image of the ultrasound probe 1 is transmitted to the image processing module 2 via a unified serial bus (USB 3.0) or an optical fiber.

[0089] Specifically, the collected ultrasound image data is transmitted to the image processing module 2 via a high-speed data transmission channel (such as USB 3.0 or optical fiber) to ensure the real-time and stability of data transmission.

[0090] In a preferred embodiment, the dynamic blood flow information in the blood flow image includes blood flow velocity, blood flow direction and blood flow volume.

[0091] Specifically, the deep learning model automatically extracts dynamic information such as blood flow speed, direction, and flow rate, and generates corresponding visualization images to facilitate subsequent analysis by the operator.

[0092] The present invention uses a high-frequency ultrasonic probe 1, which can collect high-resolution blood flow images with clear image details, which is convenient for subsequent processing and analysis. The high-speed data transmission channel can ensure the real-time transmission of image data and meet the needs of dynamic blood flow imaging. The system can maintain stable data acquisition performance during long-term operation and reduce the risk of data loss and acquisition interruption.

[0093] The preprocessing unit 21 uses a deep learning algorithm to effectively remove noise from the image and improve the clarity and quality of the image. The image enhancement algorithm significantly improves the contrast and detail of the image, making it easier for operators to observe. The image is standardized to ensure that images collected from different batches and different devices have consistent quality and comparability, thereby improving the versatility and applicability of the system.

[0094] The deep learning model of the artificial intelligence processing unit 22 can accurately identify and analyze dynamic information in blood flow images, extract key parameters such as blood flow speed, direction, and flow rate, and assist operators in making accurate judgments. Through parallel computing and GPU acceleration technology, the AI ​​processing module can achieve real-time analysis to ensure the timely capture and presentation of dynamic blood flow information. The model automatically extracts and analyzes blood flow characteristics, generates visual images and abnormal feature map reports, and reduces the workload of operators.

[0095] The image display module 3 can accurately present the processed dynamic blood flow image with clear details through a high-resolution medical display. A user-friendly interface is designed to provide multiple image display modes. Operators can choose the appropriate display mode according to their needs to improve the user experience.

[0096] Specifically, taking the cardiovascular department of a hospital as the background, the dynamic blood flow imaging system of the present invention is introduced, and the specific implementation steps are as follows:

[0097] High-frequency ultrasound probe 1 and high-resolution medical display are installed in the hospital's cardiovascular department and integrated with existing ultrasound equipment and hospital information system (HIS). In terms of software, the software of image preprocessing and AI processing modules are deployed on the hospital's server to ensure that the hardware and software environment meet the requirements of system operation. At the same time, a high-speed data transmission channel is configured to ensure that the image data collected by ultrasound probe 1 can be transmitted to the processing server in real time.

[0098] The person being scanned undergoes routine ultrasound examination preparation, ensures that the skin of the examination site is clean, and applies an appropriate amount of ultrasound coupling agent. The doctor uses a high-frequency ultrasound probe 1 to scan the cardiovascular system of the person being scanned and collect blood flow images in real time. The collected images are transmitted to the server through a high-speed data transmission channel for processing.

[0099] The preprocessing module automatically removes noise from the collected blood flow images to improve image clarity. The images are enhanced to enhance the contrast and details of the blood flow, allowing doctors to observe the blood flow dynamics more clearly. The AI ​​processing module uses a trained deep learning model to perform real-time analysis on the preprocessed images to extract dynamic information such as blood flow speed, direction, and flow. The model automatically detects abnormal areas in the blood flow, such as thrombus, stenosis, etc., and annotates them to generate a feature map report.

[0100] The processed blood flow images, feature map reports and dynamic information are displayed in real time on a high-definition medical display, allowing doctors to intuitively observe the dynamic changes of blood flow. Doctors can learn more about the dynamic information and abnormal areas of blood flow through the reports generated by the system, which assists them in making judgments. Doctors provide feedback on the performance of the system and make suggestions for improvement during use.

[0101] By introducing multimodal imaging technology and adaptive acquisition technology, the function and performance of the image acquisition module can be significantly improved, providing more comprehensive and high-quality blood flow image data, and providing a solid foundation for subsequent image preprocessing and AI analysis, bringing greater practical value and usage experience to operators and scanned persons.

[0102] In summary, the present application provides a dynamic blood flow imaging system based on artificial intelligence, which integrates CT / MRI, OCT, EIT and other modules in the ultrasound probe 1 to ensure data synchronization and fusion between different imaging devices. Integrate real-time feedback, intelligent path planning, multi-frequency and multi-angle acquisition and automatic calibration modules to ensure dynamic adjustment and optimization of acquisition parameters. In terms of software, develop a multimodal data fusion algorithm to achieve synchronous processing and fusion display of data under different imaging modes. Develop real-time image quality assessment and adaptive acquisition algorithms to ensure dynamic adjustment and optimization of acquisition parameters. The user interface provides synchronous display and interactive operation of multimodal images to help operators comprehensively analyze blood flow information under different imaging modes.

[0103] The above description is only a preferred embodiment of the present invention, and does not limit the implementation mode and protection scope of the present invention. For those skilled in the art, it should be aware that all solutions obtained by equivalent substitutions and obvious changes made using the description and illustrations of the present invention should be included in the protection scope of the present invention.

Claims

1. A dynamic blood flow imaging system based on artificial intelligence, characterized in that: include, An ultrasonic probe (1), wherein the ultrasonic probe (1) is provided with: A multimodal imaging module (11), used for collecting blood flow data of a target tissue and generating a blood flow image based on the blood flow data; An adaptive optimization module (12), connected to the multimodal imaging module (11), and adjusting acquisition parameters of the ultrasound probe (1) according to the blood flow data; An image processing module (2) is connected to the ultrasound probe (1) and automatically identifies and analyzes dynamic blood flow information in the blood flow image through a deep learning model to obtain a processed image; An image display module (3) is connected to the image processing module (2) and is used to display the processed image.

2. The artificial intelligence-based dynamic blood flow imaging system according to claim 1, characterized in that: The multimodal imaging module (11) comprises: A first imaging unit (111) extracts the relationship between the target tissue blood flow and the surrounding tissue based on computer tomography or magnetic resonance imaging technology; A second imaging unit (112) is configured to acquire dynamic blood flow information of the target tissue based on optical coherence tomography; The third imaging unit (113) reflects the dynamic blood flow information and tissue structure of the target tissue by measuring the electrical impedance change parameter of the target tissue.

3. The dynamic blood flow imaging system based on artificial intelligence according to claim 1, characterized in that: The adaptive optimization module (12) comprises: A real-time feedback unit (121) is configured to evaluate the blood flow data based on an image quality evaluation algorithm and to adjust scanning parameters of the ultrasound probe (1) according to the evaluation result; An intelligent acquisition path planning unit (122) analyzes the blood flow change trend based on a path prediction model and plans the movement path and acquisition angle of the ultrasound probe (1); A multi-frequency and multi-angle acquisition unit (123), wherein the ultrasound probe (1) acquires dynamic blood flow information at different depths and in different directions through multi-frequency and multi-angle acquisition; An automatic calibration and optimization unit (124) adjusts the calibration parameters of the ultrasonic probe (1) through an automatic calibration technique.

4. The dynamic blood flow imaging system based on artificial intelligence according to claim 1, characterized in that: The image processing module (2) comprises: A preprocessing unit (21) for preprocessing the blood flow image; An artificial intelligence processing unit (22) is connected to the preprocessing unit (21) and extracts dynamic blood flow information in the preprocessed image through the deep learning model.

5. The artificial intelligence-based dynamic blood flow imaging system according to claim 4, characterized in that: The pre-processing unit (21) comprises: A denoising processing subunit (211) automatically identifies and removes noise in the blood flow image through a convolutional neural network; An image enhancement subunit (212), connected to the denoising processing subunit (211), for enhancing the contrast of the blood flow image through an image enhancement algorithm; The standardization processing subunit (213) is connected to the image enhancement subunit (212) to unify the size and brightness of the blood flow image.

6. The artificial intelligence-based dynamic blood flow imaging system according to claim 4, characterized in that: The artificial intelligence processing unit (22) comprises: A model generation subunit (221) constructs and trains an initialization model to obtain the deep learning model; A real-time analysis subunit (222), connected to the model generation subunit (221), wherein the deep learning model analyzes the blood flow image in real time through parallel computing and graphics processor acceleration technology; The information extraction subunit (223) is connected to the real-time analysis subunit (222) and extracts the dynamic blood flow information of the blood flow image through the deep learning model.

7. The artificial intelligence-based dynamic blood flow imaging system according to claim 1, characterized in that: The image display module (3) comprises a color Doppler display mode and a power Doppler display mode.

8. The artificial intelligence-based dynamic blood flow imaging system according to claim 6, characterized in that: The artificial intelligence processing unit (22) also includes an abnormality detection subunit (224), which is connected to the real-time analysis subunit (222) and detects and marks abnormal areas in the blood flow through the deep learning model.

9. The artificial intelligence-based dynamic blood flow imaging system according to claim 1, characterized in that: The blood flow image of the ultrasonic probe (1) is transmitted to the image processing module (2) via a unified serial bus or optical fiber.

10. The artificial intelligence-based dynamic blood flow imaging system according to claim 1, characterized in that: The dynamic blood flow information in the blood flow image includes blood flow velocity, blood flow direction and blood flow volume.

Images