Waist plexus nerve block ultrasonic image recognition and puncture guide method and device

Abstract

The invention discloses an artificial intelligence lumbar plexus block ultrasound image recognition and puncture guiding method and device, and the method comprises the steps: carrying out the image segmentation of a lumbar plexus block ultrasound image when the site type information of the lumbar plexus block ultrasound image is a preset site type, and obtaining a preset site type of the lumbar plexus block ultrasound image; obtaining a plurality of tissue segmentation regions, a puncture needle segmentation region and region categories of the plurality of tissue segmentation regions on the lumbar plexus block ultrasound image; generating a puncture direction line of the puncture needle on the lumbar plexus block ultrasound image based on the puncture needle segmentation region; judging whether the puncture direction line passes through a puncturable type tissue segmentation region or not; if the puncture direction line passes through the puncturable type tissue segmentation area, judging whether the puncture direction line passes through the evasion type tissue segmentation area or not; if the puncture direction line does not pass through the avoiding type of tissue segmentation area, an instruction of continuing puncture according to the current puncture direction is sent out. According to the invention, the accuracy of lumbar plexus block ultrasonic image recognition and puncture guide can be improved.

Description

Technical Field

[0001] The present application relates to the field of image processing technology, and in particular to a method and device for ultrasonic image recognition and puncture guidance of lumbar plexus nerve block. Background Art

[0002] Traditional lumbar plexus block technology mainly relies on surface landmarks and the clinical experience of doctors. This blind exploration operation has a high failure rate and may cause damage to internal organs and large blood vessels. This has led many inexperienced anesthesiologists to not dare to try it. In recent years, ultrasound-guided lumbar plexus block technology has gradually been fully utilized due to its advantages such as visualization, real-time performance and high success rate. Although ultrasound guidance has made progress in lumbar plexus block technology, there are still great challenges in practical application. First, ultrasound images are two-dimensional imaging images based on human body structure, and their correct interpretation requires rich clinical experience and spatial imagination. It is very difficult for inexperienced doctors to accurately identify the lumbar plexus and its surrounding tissue structures. Secondly, anatomical differences and physiological changes between individuals can lead to ever-changing corresponding ultrasound images. For example, fat infiltration in the muscles of obese and elderly patients seriously affects the recognition of ultrasound images, making it more difficult to identify ultrasound images of lumbar plexus block and lower puncture accuracy, which in turn affects the effect of nerve block. In addition, ordinary ultrasound can only display images of local tissue structures, but cannot identify the names of various areas and indicate whether puncture operations can be performed; it cannot identify the puncture needle; and it cannot guide the puncture needle accurately to the area where the lumbar plexus nerves are located. Summary of the invention

[0003] The embodiments of the present application provide a method and device for ultrasonic image recognition and puncture guidance of lumbar plexus nerve block, which can improve the accuracy of ultrasonic image recognition and puncture guidance of lumbar plexus nerve block.

[0004] In a first aspect, the present application provides a method for ultrasonic image recognition and puncture guidance of lumbar plexus nerve block, comprising: Obtaining lumbar plexus block ultrasound images acquired by ultrasound equipment; Inputting the lumbar plexus nerve block ultrasound image into a target image segmentation model to perform image segmentation on the lumbar plexus nerve block ultrasound image, thereby obtaining a plurality of tissue segmentation regions, a puncture needle segmentation region, and a plurality of region categories of the tissue segmentation regions on the lumbar plexus nerve block ultrasound image, wherein the region categories of the tissue segmentation regions include a punctureable category and an avoidance category; generating a puncture direction line of the puncture needle on the lumbar plexus nerve block ultrasound image based on the puncture needle segmentation area; Determining whether the puncture direction line passes through the puncturable tissue segmentation area; If the puncture direction line passes through the tissue segmentation area of ​​the puncturable category, determining whether the puncture direction line passes through the tissue segmentation area of ​​the avoidance category; If the puncture direction line does not pass through the tissue segmentation area of ​​the avoidance category, an instruction to continue puncturing in the current puncture direction is issued.

[0005] In a second aspect, the present application provides a lumbar plexus nerve block ultrasound image recognition and puncture guidance device, comprising: An acquisition module, used for acquiring a lumbar plexus nerve block ultrasound image acquired by an ultrasound device; The classification module is used to input the lumbar plexus block ultrasound image into the site classification model to obtain the site type information of the lumbar plexus block ultrasound image.

[0006] a segmentation module, for inputting the lumbar plexus block ultrasound image into a target image segmentation model to perform image segmentation on the lumbar plexus block ultrasound image when the site type information of the lumbar plexus block ultrasound image is a preset site type, so as to obtain a plurality of tissue segmentation regions, a puncture needle segmentation region, and a plurality of region categories of the tissue segmentation regions on the lumbar plexus block ultrasound image, wherein the region categories of the tissue segmentation regions include a punctureable category and an avoidance category; A generating module, configured to generate a puncture direction line of the puncture needle on the lumbar plexus nerve block ultrasound image based on the puncture needle segmentation area; A first judgment module is used to judge whether the puncture direction line passes through the puncturable tissue segmentation area; A second judgment module is used to judge whether the puncture direction line passes through the tissue segmentation area of ​​the avoidance category if the puncture direction line passes through the tissue segmentation area of ​​the puncturable category; The instruction module is used to issue an instruction to continue puncturing in the current puncture direction if the puncture direction line does not pass through the tissue segmentation area of ​​the avoidance category.

[0007] In a third aspect, the electronic device provided in the present application includes a memory and a processor, wherein the memory stores a computer program, and the processor is used to run the computer program in the memory to implement the steps in the lumbar plexus nerve block ultrasound image recognition and puncture guidance method provided in the present application.

[0008] In a fourth aspect, the computer-readable storage medium provided in the present application stores a plurality of instructions, which are suitable for loading by a processor to implement the steps in the lumbar plexus nerve block ultrasound image recognition and puncture guidance method provided in the present application.

[0009] In a fifth aspect, the computer program product provided in the present application includes a computer program or instructions, which, when executed by a processor, implement the steps in the lumbar plexus nerve block ultrasound image recognition and puncture guidance method provided in the present application.

[0010] In the present application, compared with the related art, a lumbar plexus block ultrasound image collected by an ultrasound device is obtained; the lumbar plexus block ultrasound image is input into a site classification model to obtain site type information of the lumbar plexus block ultrasound image; when the site type information of the lumbar plexus block ultrasound image is a preset site type, the lumbar plexus block ultrasound image is input into a target image segmentation model to perform image segmentation on the lumbar plexus block ultrasound image to obtain multiple tissue segmentation areas, puncture needle segmentation areas and area categories of multiple tissue segmentation areas on the lumbar plexus block ultrasound image, and the area categories of the tissue segmentation areas include punctureable categories and avoidance categories; based on the puncture needle segmentation area, a puncture direction line of the puncture needle is generated on the lumbar plexus block ultrasound image; it is determined whether the puncture direction line passes through the tissue segmentation area of ​​the punctureable category; if the puncture direction line passes through the tissue segmentation area of ​​the punctureable category, it is determined whether the puncture direction line passes through the tissue segmentation area of ​​the avoidance category; if the puncture direction line does not pass through the tissue segmentation area of ​​the avoidance category, an instruction to continue puncturing according to the current puncture direction is issued. This application can improve the accuracy of artificial intelligence recognition and puncture guidance of lumbar plexus nerve block ultrasound images. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0012] Figure 1 This is a scene schematic diagram of a lumbar plexus nerve block ultrasound image recognition and puncture guidance system provided in an embodiment of the present application; Figure 2 It is a flow chart of an embodiment of a method for ultrasonic image recognition and puncture guidance of lumbar plexus nerve block provided in an embodiment of the present application; Figure 3 is a schematic diagram of a lumbar plexus block ultrasound image of a paraspinal transverse process sagittal plane, the image category of which is provided in an embodiment of the present application; Figure 4 This is a schematic diagram of the puncture needle after the direction is adjusted in the lumbar plexus nerve block ultrasound image with the image category of the paraspinal transverse process sagittal plane provided in the embodiment of the present application; Figure 5 is a schematic diagram of a lumbar plexus nerve block ultrasound image of a paraspinal transverse process cross section, the image category of which is provided in an embodiment of the present application; Figure 6 is a schematic diagram of a lumbar plexus nerve block ultrasound image of a cross section of the ventrolateral transverse process, the image category of which is provided in an embodiment of the present application; Figure 7 It is a structural schematic diagram of a lumbar plexus nerve block ultrasound image recognition and puncture guidance device provided in an embodiment of the present application; Figure 8 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0013] It should be noted that the principles of the present application are illustrated by implementing them in an appropriate computing environment. The following description is based on the illustrated specific embodiments of the present application and should not be considered as limiting other specific embodiments of the present application that are not described in detail herein.

[0014] In the following description of the present application, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it can be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0015] In the following description of the present application, the terms "first\second\third" involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\third" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present application described here can be implemented in an order other than that illustrated or described here.

[0016] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0017] In order to improve the effect of lumbar plexus block ultrasound image recognition and puncture guidance, the embodiments of the present application provide a lumbar plexus block ultrasound image recognition and puncture guidance method, a lumbar plexus block ultrasound image recognition and puncture guidance device, an electronic device, a computer-readable storage medium, and a computer program product. Among them, the lumbar plexus block ultrasound image recognition and puncture guidance method can be performed by the lumbar plexus block ultrasound image recognition and puncture guidance device, or by an electronic device integrated with the lumbar plexus block ultrasound image recognition and puncture guidance device.

[0018] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0019] Please refer to Figure 1 The present application also provides a lumbar plexus block ultrasound image recognition and puncture guidance system, such as Figure 1 As shown, the electronic device 100 of the lumbar plexus nerve block ultrasound image recognition and puncture guidance system, the electronic device 100 integrates the lumbar plexus nerve block ultrasound image recognition and puncture guidance device provided by the present application.

[0020] Among them, the electronic device 100 can be any device equipped with a processor and has processing capabilities, such as mobile electronic devices with processors such as smart phones, tablet computers, PDAs, laptops, smart speakers, or fixed electronic devices with processors such as desktop computers, televisions, servers, industrial equipment, etc.

[0021] In addition, if Figure 1 As shown, the lumbar plexus nerve block ultrasound image recognition and puncture guidance system may also include a memory 200 for storing original data, intermediate data and result data.

[0022] In the embodiment of the present application, the memory 200 may be a cloud storage. Cloud storage is a new concept extended and developed from the concept of cloud computing. A distributed cloud storage system (hereinafter referred to as the storage system) refers to a storage system that uses cluster applications, grid technology, and distributed storage file systems and other functions to bring together a large number of storage devices of various types (storage devices are also called storage nodes) in the network through application software or application interfaces to work together and jointly provide data storage and business access functions to the outside world.

[0023] At present, the storage method of the storage system is: create a logical volume, and when creating a logical volume, allocate physical storage space for each logical volume, and the physical storage space may be composed of disks of a storage device or several storage devices. The client stores data on a logical volume, that is, stores the data on the file system. The file system divides the data into many parts, each of which is an object. The object contains not only data but also additional information such as data identification (ID, ID entity). The file system writes each object into the physical storage space of the logical volume, and the file system records the storage location information of each object, so that when the client requests to access the data, the file system can allow the client to access the data according to the storage location information of each object.

[0024] The process of the storage system allocating physical storage space to a logical volume is as follows: based on the estimated capacity of the objects stored in the logical volume (this estimate often has a large margin relative to the actual capacity of the objects to be stored) and the grouping of independent redundant disk arrays (RAID, Redundant Array of Independent Disks), the physical storage space is divided into stripes in advance. A logical volume can be understood as a stripe, thereby allocating physical storage space to the logical volume.

[0025] It should be noted that Figure 1 The scenario diagram of the lumbar plexus block ultrasound image recognition and puncture guidance system shown is merely an example. The lumbar plexus block ultrasound image recognition and puncture guidance system and scenario described in the embodiment of the present application are intended to more clearly illustrate the technical solution of the embodiment of the present application, and do not constitute a limitation on the technical solution provided in the embodiment of the present application. A person of ordinary skill in the art can appreciate that with the evolution of the lumbar plexus block ultrasound image recognition and puncture guidance system and the emergence of new business scenarios, the technical solution provided in the embodiment of the present application is equally applicable to similar technical problems.

[0026] It should be noted that the serial numbers of the following embodiments are not intended to limit the preferred order of the embodiments.

[0027] Please refer to Figure 2 , Figure 2 FIG. 1 is a flow chart of an embodiment of a method for ultrasonic image recognition and puncture guidance of lumbar plexus nerve block provided in an embodiment of the present application. Figure 2 As shown, the process of the lumbar plexus nerve block ultrasound image recognition and puncture guidance method provided by this application is as follows: 201. Obtain lumbar plexus block ultrasound images collected by ultrasound equipment.

[0028] In the embodiment of the present application, the ultrasound device can be operated and moved by a robot or a medical worker, and the ultrasound device collects lumbar plexus block ultrasound images during the movement.

[0029] 202. Input the lumbar plexus nerve block ultrasound image into a site classification model to obtain site type information of the lumbar plexus nerve block ultrasound image.

[0030] Among them, the site classification model is the YOLOv8 classification model.

[0031] Build the YOLOv8 classification model. The network structure of YOLOv8 is mainly composed of the following three parts: Backbone: It uses a series of convolution and deconvolution layers to extract features, and also uses residual connections and bottleneck structures to reduce the size of the network and improve performance.

[0032] Neck: The Neck part adopts the Feature Pyramid Network (FPN) structure and multi-scale feature fusion technology to fuse feature maps from different stages of Backbone to better capture information of targets of different scales.

[0033] Head: Responsible for the final target detection and classification tasks, including a detection head and a classification head. The detection head contains a series of convolutional layers and deconvolutional layers to generate classification results; the classification head uses global average pooling to classify each feature map. The labeled data is input into the classification network for training, and the trained model is used for site recognition tasks. The image of the current image site is output to guide the doctor to quickly and accurately reach the optimal puncture position. Among them, the loss function used by the classification model is the cross entropy loss:

[0034] in, The number of categories represented as classification; Representation sample The real category of Representation sample The predicted probability of .

[0035] 203. When the site type information of the lumbar plexus block ultrasound image is a preset site type, the lumbar plexus block ultrasound image is input into the target image segmentation model to perform image segmentation on the lumbar plexus block ultrasound image to obtain multiple tissue segmentation areas, puncture needle segmentation areas and area categories of multiple tissue segmentation areas on the lumbar plexus block ultrasound image.

[0036] When the image site information prompts that it does not belong to the preset site, a probe movement prompt message is given to guide the operator to reach the preset site.

[0037] The region categories of the tissue segmentation region include a punctureable category and an avoidance category.

[0038] In the embodiment of the present application, the target image segmentation model can be a trained UNet++ segmentation model. Specifically, a UNet++ segmentation model is built. The UNet++ model is an encoder-decoder architecture. The encoder uses a context-aware feature encoder with convolution blocks and residual blocks to extract multi-scale feature maps, that is, shallow image features; the decoder uses a feature decoder with convolution and deconvolution to adjust the size of the multi-scale feature map, that is, deep image features, to achieve end-to-end segmentation; the labeled data is input into the segmentation network for training, the lumbar plexus nerve area and its surrounding important tissue structures are identified, and the contours of the boundaries of each tissue are output.

[0039] In a specific embodiment, an ultrasound scanning video of an ultrasound device is obtained; the ultrasound scanning video is frame-processed to obtain a plurality of continuous single-frame ultrasound images; an annotated ultrasound image and an image category of the annotated ultrasound image obtained by annotating the single-frame ultrasound image are obtained, wherein the single-frame ultrasound images obtained from the ultrasound scanning video of different sections belong to different image categories, and the annotated ultrasound image includes an annotated area of ​​a punctureable category and an annotated area of ​​an avoidance category; a preset segmentation model is trained based on the plurality of annotated ultrasound images to obtain a target image segmentation model, and the lumbar plexus block ultrasound image is input into the target image segmentation model to perform image segmentation on the lumbar plexus block ultrasound image to obtain a plurality of tissue segmentation areas, a puncture needle segmentation area, and area categories of a plurality of tissue segmentation areas on the lumbar plexus block ultrasound image, and the area categories of the tissue segmentation areas include a punctureable category and an avoidance category.

[0040] In another specific embodiment, ultrasound scanning videos of multiple different sections of an ultrasound device are obtained; the ultrasound scanning videos are frame-processed to obtain multiple continuous single-frame ultrasound images; annotated ultrasound images and image categories of the annotated ultrasound images obtained by annotating the single-frame ultrasound images are obtained, wherein the single-frame ultrasound images obtained from the ultrasound scanning videos of different sections belong to different image categories, and the annotated ultrasound images include annotated areas of a punctureable category and annotated areas of an avoidance category; a preset segmentation model is trained based on the multiple annotated ultrasound images to obtain a target image segmentation model; the lumbar plexus block ultrasound image is input into an image classification model to classify the lumbar plexus block ultrasound image to obtain the image category of the lumbar plexus block ultrasound image, and a set of region categories in the lumbar plexus block ultrasound image is determined based on the image category of the lumbar plexus block ultrasound image; the lumbar plexus block ultrasound image is input into a target image segmentation model to perform image segmentation on the lumbar plexus block ultrasound image to obtain multiple tissue segmentation areas, puncture needle segmentation areas, and region categories of multiple tissue segmentation areas on the lumbar plexus block ultrasound image, and the region categories of the tissue segmentation areas include a punctureable category and an avoidance category, wherein the region categories of each tissue segmentation area belong to the region category set. Different image categories correspond to different sets of region categories.

[0041] Specifically, data collection is performed first. Data collection can be scanning videos of three different sections: the sagittal plane of the paraspinal transverse process, the cross section of the paraspinal transverse process, and the cross section of the ventrolateral transverse process. The video is then framed to obtain continuous single-frame ultrasound images. Then, data labeling. The single-frame ultrasound image is classified and processed by a nerve block expert based on the ultrasound video information and the scanning position to mark the site information that needs to be identified. The boundaries of the anatomical structures that need to be identified are manually marked on the ultrasound image of the ideal puncture interface to display the area of ​​the corresponding tissue, informing the target area while also highlighting the dangerous area.

[0042] Specifically, the lumbar plexus block ultrasound image can be images of three different image categories, namely, the paraspinal transverse process sagittal plane, the paraspinal transverse process cross section, and the ventrolateral transverse process cross section, which are obtained by cutting the scanning video of the three different sections, namely, the paraspinal transverse process sagittal plane, the paraspinal transverse process cross section, and the ventrolateral transverse process cross section, into single-frame images. Single-frame ultrasound images obtained from scanning videos of different sections belong to different image categories.

[0043] Further, the image category may be a paravertebral transverse process sagittal plane, a paravertebral transverse process cross section or a ventrolateral transverse process cross section. Different image categories correspond to different region category sets, and the region category set includes multiple region categories.

[0044] In an embodiment of the present application, when the image category of the lumbar plexus block ultrasound image is the sagittal plane of the paraspinal transverse processes, the tissue segmentation areas belonging to the punctureable category on the lumbar plexus block ultrasound image include the erector spinae muscle area and the psoas major muscle area, and the tissue segmentation areas belonging to the avoidance category on the lumbar plexus block ultrasound image include multiple transverse process areas and a dangerous area containing abdominal visceral blood vessels, the multiple transverse process areas are arranged at intervals between the erector spinae muscle area and the psoas major muscle area, and the dangerous area is located on the side of the psoas major muscle area away from the erector spinae muscle area.

[0045] like Figure 3 and Figure 4 As shown in the figure, when the image category of the lumbar plexus block ultrasound image is the paraspinal transverse process sagittal plane, the tissue segmentation areas belonging to the puncture category on the lumbar plexus block ultrasound image include the erector spinae muscle area (ESM) and the psoas muscle area (PM), and the tissue segmentation areas belonging to the avoidance category on the lumbar plexus block ultrasound image include multiple transverse process areas (TP) and risk areas (Risk Area). Multiple transverse process areas (TP) are arranged alternately between the erector spinae muscle area (ESM) and the psoas muscle area (PM), and the risk area (Risk Area) is located on the side of the psoas muscle area (PM) away from the erector spinae muscle area (ESM). TP is a bony structure and is an area that needs to be avoided during puncture. Risk Area is an area that the needle tip cannot enter during lumbar plexus block, which is also a risk area for complications.

[0046] That is, when the image category of the lumbar plexus block ultrasound image is the paraspinal transverse process sagittal plane, the regional category set includes the punctureable category and the avoidance category. Further, when the image category of the lumbar plexus block ultrasound image is the paraspinal transverse process sagittal plane, the regional category set includes the erector spinae category, the psoas category, the transverse process category and the danger category, which can be set according to the specific situation.

[0047] In an embodiment of the present application, when the image category of the lumbar plexus block ultrasound image is a cross-section of the paraspinal transverse processes, the tissue segmentation areas on the lumbar plexus block ultrasound image that belong to the punctureable category include the erector spinae muscle area, the psoas major muscle area, and the quadratus lumborum muscle area, and the tissue segmentation areas on the lumbar plexus block ultrasound image that belong to the avoidance category include the vertebrae and dangerous areas containing abdominal organs and blood vessels.

[0048] like Figure 5 As shown in the figure, when the image category of the lumbar plexus block ultrasound image is the paraspinal transverse process cross section, the tissue segmentation areas belonging to the punctureable category on the lumbar plexus block ultrasound image include the erector spinae region (ESM), the psoas region (PM) and the quadratus lumborum region (QLM), and the tissue segmentation areas belonging to the avoidance category on the lumbar plexus block ultrasound image include the vertebra (Vertebra) and the risk area (Risk Area).

[0049] In an embodiment of the present application, when the image category of the lumbar plexus block ultrasound image is the cross-section of the ventrolateral transverse process, the tissue segmentation areas belonging to the punctureable category on the lumbar plexus block ultrasound image include the erector spinae muscle area, the psoas major muscle area, and the quadratus lumborum muscle area, and the tissue segmentation areas belonging to the avoidance category on the lumbar plexus block ultrasound image include the vertebrae and dangerous areas containing abdominal organs and blood vessels.

[0050] like Figure 6 As shown in the figure, when the image category of the lumbar plexus block ultrasound image is the cross section of the ventrolateral transverse process, the tissue segmentation areas belonging to the punctureable category on the lumbar plexus block ultrasound image include the erector spinae region (ESM), the psoas region (PM) and the quadratus lumborum region (QLM), and the tissue segmentation areas belonging to the avoidance category on the lumbar plexus block ultrasound image include the vertebra and the risk area, and the vertebra, the psoas region (PM) and the quadratus lumborum region (QLM) are arranged in sequence between the erector spinae region (ESM) and the risk area.

[0051] In the embodiment of the present application, a preset segmentation model is trained based on multiple annotated ultrasound images to obtain a target image segmentation model, including: (1) The labeled ultrasound image is input into a preset segmentation model to obtain a predicted pixel category for each pixel point on the labeled ultrasound image, where the predicted pixel category is a punctureable category and an avoidance category.

[0052] In the embodiment of the present application, the predicted pixel categories are punctureable category and avoidance category. Further, when the image category of the lumbar plexus block ultrasound image is the paraspinal transverse process sagittal plane, the predicted pixel categories are erector spinae category, psoas category, transverse process category and danger category, which can be set according to the specific situation.

[0053] (2) Determine a prediction region corresponding to the annotated region on the annotated ultrasound image based on the predicted pixel category of each pixel point on the annotated ultrasound image, wherein the predicted pixel categories within the same prediction region belong to the same category.

[0054] For example, all pixels in the erector spinae muscle region (ESM) are of the erector spinae muscle class.

[0055] (3) Obtain the intersection area and the sum of the area between the marked area and the corresponding predicted area.

[0056] (4) Determine the first loss based on the ratio of the intersection area to the sum of the area of ​​the regions.

[0057] In the embodiment of the present application, the first loss The calculation formula is as follows,

[0058] in, The prediction area is composed of a set of pixels representing the prediction result. A label region representing a collection of pixels that represent labels.

[0059] (5) Determine a second loss based on the annotated pixel category and the predicted pixel category of each pixel point on the annotated ultrasound image using a cross entropy loss function, wherein the annotated pixel categories in the annotated area of ​​the same category belong to the same category.

[0060] The annotated pixel category of each pixel point is determined according to the category of each annotated area.

[0061] Specifically, the second loss The calculation formula is as follows,

[0062] in, Represents the total number of pixels. represents the true label of the pixel, Represents the prediction result of the pixel.

[0063] (6) Determine the total loss based on the first loss and the second loss.

[0064] In a specific embodiment, the first loss and the second loss are added to obtain a total loss.

[0065] In another specific embodiment, the total loss is determined based on the first loss and the second loss; and the preset segmentation model is iteratively updated until the total loss is less than the preset loss value, thereby obtaining the target image segmentation model.

[0066] (7) Iteratively update the preset segmentation model until the total loss is less than the preset loss value, and obtain the target image segmentation model.

[0067] Among them, the preset loss value can be set according to the specific situation.

[0068] Furthermore, in order to improve the segmentation accuracy, in another specific embodiment, ultrasound scanning videos of multiple different sections of an ultrasound device are obtained; the ultrasound scanning videos are framed to obtain multiple continuous single-frame ultrasound images; annotated ultrasound images and image categories of the annotated ultrasound images obtained by annotating the single-frame ultrasound images are obtained, wherein the single-frame ultrasound images obtained from the ultrasound scanning videos of different sections belong to different image categories, and the annotated ultrasound images include annotated areas of the punctureable category and annotated areas of the avoidance category; a preset classification model is trained based on the multiple annotated ultrasound images and the image categories of the annotated ultrasound images to obtain an image classification model; a preset segmentation model is trained based on the multiple annotated ultrasound images to obtain the target The invention discloses a target image segmentation model; inputs the lumbar plexus block ultrasound image into the image classification model to classify the lumbar plexus block ultrasound image and obtain the image category of the lumbar plexus block ultrasound image; determines the region category set in the lumbar plexus block ultrasound image based on the image category of the lumbar plexus block ultrasound image; inputs the lumbar plexus block ultrasound image into the target image segmentation model to perform image segmentation on the lumbar plexus block ultrasound image and obtains multiple tissue segmentation regions, puncture needle segmentation regions and region categories of multiple tissue segmentation regions on the lumbar plexus block ultrasound image, wherein the region categories of the tissue segmentation regions include punctureable categories and avoidance categories, wherein the region categories of each tissue segmentation region belong to the region category set.

[0069] In the embodiment of the present application, different image categories correspond to region category sets.

[0070] For example, when the image category of the lumbar plexus block ultrasound image is the paraspinal transverse process sagittal plane, the tissue segmentation areas belonging to the punctureable category on the lumbar plexus block ultrasound image include the erector spinae muscle area (ESM) and the psoas muscle area (PM), and the tissue segmentation areas belonging to the avoidance category on the lumbar plexus block ultrasound image include multiple transverse process areas (TP) and risk areas. When the image category of the lumbar plexus block ultrasound image is the paraspinal transverse process sagittal plane, the region category set includes the erector spinae muscle category, the psoas muscle category, the transverse process category, and the risk category.

[0071] Specifically, the lumbar plexus block ultrasound image is input into the target image segmentation model to perform image segmentation on the lumbar plexus block ultrasound image, and multiple tissue segmentation regions, puncture needle segmentation regions, and region categories of multiple tissue segmentation regions on the lumbar plexus block ultrasound image are obtained, including: when the region categories of multiple tissue segmentation regions output by the target image segmentation model are not completely the same as the region categories of the region category set, a prompt message that cannot be fully observed is issued, prompting the user to move the ultrasound device to re-shoot the lumbar plexus block ultrasound image until the region categories of multiple tissue segmentation regions output by the target image segmentation model are completely the same as the region categories of the region category set. When the region categories of multiple tissue segmentation regions output by the target image segmentation model are completely the same as the region categories of the region category set, multiple tissue segmentation regions, puncture needle segmentation regions, and region categories of multiple tissue segmentation regions on the lumbar plexus block ultrasound image are output.

[0072] Specifically, the image classification model is a YOLOv8 model. A preset classification model is trained based on a plurality of annotated ultrasound images and image categories of the annotated ultrasound images using a cross entropy loss function to obtain an image classification model.

[0073] 204. Generate a puncture direction line of the puncture needle on the lumbar plexus nerve block ultrasound image based on the puncture needle segmentation area.

[0074] In the embodiment of the present application, it is determined whether the puncture needle segmentation area includes the needle tip area. When the puncture needle segmentation area includes the needle tip area, a puncture direction line of the puncture needle is generated on the lumbar plexus nerve block ultrasound image based on the puncture needle segmentation area.

[0075] In a specific embodiment, the puncture needle segmentation region is input into the puncture needle segmentation model to obtain a segmentation result, and when the segmentation result contains the needle tip region, it is determined that the puncture needle segmentation region includes the needle tip region.

[0076] In a specific embodiment, determining whether the puncture needle segmentation area includes the needle tip area includes: (1) Use the Zhang-Suen thinning algorithm to refine the puncture needle segmentation area and obtain the needle thinning line segment.

[0077] (2) Perform straight line fitting on the needle thinning line segment to obtain the first fitting straight line segment.

[0078] Specifically, the least square method may be used to perform straight line fitting on the needle thinning line segment to obtain a first fitting straight line segment.

[0079] (3) Obtain multiple first needle points with equal spacing on the needle thinning line segment.

[0080] The plurality of first needle points include two end points of the needle thinning line segment.

[0081] (4) Calculate the vertical distances between the first needle points and the first fitting straight line segment respectively and sum them up to obtain a first total distance.

[0082] (5) If the first total distance is greater than the first preset distance value, it is determined that the puncture needle segmentation area includes the needle tip area.

[0083] The first preset distance value is set according to specific circumstances.

[0084] If the first total distance is greater than the first preset distance value, it indicates that the needle thinning line segment is significantly different from the straight line, indicating that the puncture needle segmentation region includes the needle tip region.

[0085] In the embodiment of the present application, before determining whether the puncture needle segmented area includes the needle tip area, the following steps are included: (1) Obtaining the segmentation confidence of each segmented region, wherein each segmented region includes a puncture needle segmented region and a plurality of tissue segmented regions.

[0086] Specifically, the segmentation confidence of the segmented region is the sum of the probabilities of correct pixel category prediction of each pixel point in the segmented region.

[0087] (2) Determine the overall confidence based on the segmentation confidence of each segmented region.

[0088] Specifically, the weight coefficient of each segmented region is determined based on the area size of each segmented region. The segmentation confidence of each segmented region is weighted and summed based on the weight coefficient of each segmented region to obtain the overall confidence. The larger the area of ​​the segmented region, the smaller the weight coefficient of the segmented region.

[0089] (3) If the overall confidence is higher than the preset confidence, it is determined whether the puncture needle segmentation area includes the needle tip area; if the overall confidence is not higher than the preset confidence, an instruction to adjust the posture of the ultrasound device is issued, and the collected lumbar plexus nerve block ultrasound image is reacquired.

[0090] In a specific embodiment, the first fitting straight line segment is determined as the puncture direction line of the puncture needle.

[0091] In another specific embodiment, generating a puncture direction line of the puncture needle on a lumbar plexus nerve block ultrasound image based on the puncture needle segmentation region includes: (1) A normal line is drawn through the midpoint of the first fitted straight line segment to divide the puncture needle segmentation area into two sub-segmentation areas.

[0092] (2) The segment where the needle thinning segment is located in the sub-segmentation area with a larger area is taken as the first segment.

[0093] (3) Perform straight line fitting on the first line segment to obtain a second fitted straight line segment.

[0094] (4) The straight line where the second fitted straight line segment is located is determined as the puncture direction line.

[0095] Since the needle tip is generally triangular, the lines in the smaller sub-segmented area where the needle tip is located are removed, and only the first line segment is retained for fitting, which can improve the accuracy of the puncture direction line.

[0096] 205. Determine whether the puncture direction line passes through a puncturable tissue segmentation area.

[0097] Specifically, if the extension line of the puncture direction line passes through the puncturable tissue segmentation area, the puncture direction line passes through the puncturable tissue segmentation area.

[0098] 206. If the puncture direction line passes through a tissue segmentation area of ​​the punctureable category, determine whether the puncture direction line passes through a tissue segmentation area of ​​the avoidance category.

[0099] Specifically, if the extension line of the puncture direction line passes through the tissue segmentation area of ​​the avoidance category, the puncture direction line passes through the tissue segmentation area of ​​the avoidance category.

[0100] If the puncture direction line passes through the tissue segmentation area of ​​the punctureable category, it is determined whether the puncture direction line passes through the tissue segmentation area of ​​the avoidance category.

[0101] 207. If the puncture direction line does not pass through the tissue segmentation area of ​​the avoidance category, an instruction to continue puncturing in the current puncture direction is issued.

[0102] If the puncture direction line does not pass through the tissue segmentation area of ​​the avoidance category, it means that the puncture can continue, and an instruction to continue the puncture in the current puncture direction is issued.

[0103] If the puncture direction line passes through the tissue segmentation area of ​​the avoidance category, it is determined whether the tissue segmentation area of ​​the avoidance category passed by the puncture direction line is located between the puncture needle segmentation area and the target puncture area. If the tissue segmentation area of ​​the avoidance category passed by the puncture direction line is located between the puncture needle segmentation area and the target puncture area, a prompt signal is issued to prompt the operator to adjust the puncture direction and avoid the avoidance area. If the tissue segmentation area of ​​the avoidance category passed by the puncture direction line is not located between the puncture needle segmentation area and the target puncture area, the real-time distance between the needle tip of the puncture needle segmentation area along the puncture needle direction line and the tissue segmentation area of ​​the avoidance category is calculated; if the real-time distance is not less than the second preset distance value, an instruction to continue puncturing according to the current puncture direction is issued. If the real-time distance is less than the second preset distance value, an alarm signal is issued.

[0104] Furthermore, if the tissue segmentation area of ​​the avoidance category through which the puncture direction line passes is located between the puncture needle segmentation area and the target puncture area, multiple candidate straight line paths are obtained, and the candidate straight line path with the smallest angle with the puncture direction line among the multiple candidate straight line paths is determined as the new puncture direction, prompting the user to perform the puncture.

[0105] like Figure 3 and Figure 4 As shown in the figure, when the image category of the lumbar plexus block ultrasound image is the paraspinal transverse process sagittal plane, the tissue segmentation areas belonging to the punctureable category on the lumbar plexus block ultrasound image include the erector spinae muscle area (ESM) and the psoas muscle area (PM), and the tissue segmentation areas belonging to the avoidance category on the lumbar plexus block ultrasound image include multiple transverse process areas (TP) and the risk area.

[0106] When the image category of the lumbar plexus block ultrasound image is the paraspinal transverse process sagittal plane, if the tissue segmentation area of ​​the avoidance category passed by the puncture direction line is the transverse process area (TP), the transverse process area (TP) is located between the puncture needle segmentation area and the target puncture area. The transverse process area (TP) passed by the puncture direction line is determined as the first transverse process area, and the transverse process areas (TP) closest to the puncture direction line are obtained on both sides of the puncture direction line as the second transverse process area and the third transverse process area, respectively. The edge intersection between the first line segment and the edge of the erector spinae muscle area (ESM) is determined as the edge intersection, and the first candidate straight line and the second candidate straight line are determined, wherein the first candidate straight line and the second candidate straight line both pass through the edge intersection, the first transverse process area and the second transverse process area are located on both sides of the first candidate straight line, and the minimum distance between the first transverse process area and the first candidate straight line is equal to the minimum distance between the second transverse process area and the first candidate straight line; the third transverse process area and the second transverse process area are located on both sides of the second candidate straight line, and the minimum distance between the third transverse process area and the second candidate straight line is equal to the minimum distance between the second transverse process area and the second candidate straight line. If the angle between the first candidate straight line and the puncture direction line is greater than the angle between the second candidate straight line and the puncture direction line, the second candidate straight line is determined as the new puncture direction, and a prompt message is issued to prompt the user to perform puncture according to the second candidate straight line.

[0107] In an embodiment of the present application, the real-time distance between the needle tip of the puncture needle segmentation area along the puncture needle direction line and the tissue segmentation area of ​​the avoidance category is calculated, including: determining the endpoint of the needle thinning line segment located in the sub-segmentation area of ​​the smaller area as the needle tip point, and determining the minimum distance between the needle tip point along the puncture needle direction and the tissue segmentation area of ​​the puncturable category as the real-time distance.

[0108] Furthermore, a lumbar plexus block ultrasound image recognition and puncture guidance method includes: determining a target puncture area on the psoas major muscle area (PM) based on the image category of the lumbar plexus block ultrasound image and the minimum circumscribed rectangle of the psoas major muscle segmentation area; and displaying the target puncture area on the lumbar plexus block ultrasound image.

[0109] Specifically, the target puncture area on the psoas major muscle region (PM) is determined based on the image category of the lumbar plexus block ultrasound image and the position of the puncture needle segmentation area, including: The lumbar plexus nerves are distributed in the posterior and medial 1 / 4-1 / 3 area of ​​the psoas major muscle, such as Figure 4 As shown in the figure, when the image category of the lumbar plexus block ultrasound image is the paraspinal transverse process sagittal plane, and the puncture needle segmentation area is located on the side of the psoas major muscle area (PM) away from the risk area, one-third of the area within the minimum circumscribed rectangle of the psoas major muscle area close to the erector spinae segmentation area is determined as the target puncture area.

[0110] like Figure 5 As shown, when the image category of the lumbar plexus block ultrasound image is the paraspinal transverse process cross section, and the puncture needle segmentation area is located on the side of the psoas major muscle region (PM) away from the vertebra (Vertebra), a quarter of the segmentation area close to the vertebra and erector spinae within the minimum circumscribed rectangle of the psoas major muscle region is determined as the target puncture area.

[0111] Specifically, the minimum circumscribed rectangle of the psoas major region (PM) is equally divided into four quarter-rectangular regions, and a quarter region of the minimum circumscribed rectangle of the psoas major region close to the segmentation region of the vertebrae and the erector spinae muscles is determined as the target puncture region.

[0112] like Figure 6 As shown in the figure, when the image category of the lumbar plexus nerve block ultrasound image is the cross section of the ventral lateral transverse process, and the puncture needle segmentation area is located on the side of the psoas major muscle area (PM) away from the risk area, a quarter of the area close to the vertebral area and the erector spinae muscle area within the minimum circumscribed rectangle of the psoas major muscle area is determined as the target puncture area.

[0113] In order to facilitate better implementation of the lumbar plexus block ultrasound image recognition and puncture guidance method provided in the embodiment of the present application, the embodiment of the present application also provides a lumbar plexus block ultrasound image recognition and puncture guidance device based on the above lumbar plexus block ultrasound image recognition and puncture guidance method. The meanings of the terms are the same as those in the above lumbar plexus block ultrasound image recognition and puncture guidance method. For specific implementation details, please refer to the description in the above method embodiment.

[0114] Please refer to Figure 7 , Figure 7A schematic diagram of the structure of a lumbar plexus nerve block ultrasound image recognition and puncture guidance device provided in an embodiment of the present application, the lumbar plexus nerve block ultrasound image recognition and puncture guidance device may include: An acquisition module 701 is used to acquire a lumbar plexus nerve block ultrasound image acquired by an ultrasound device; The classification module 702 is used to input the lumbar plexus nerve block ultrasound image into the site classification model to obtain the site type information of the lumbar plexus nerve block ultrasound image; The segmentation module 703 is used for inputting the lumbar plexus block ultrasound image into the target image segmentation model to perform image segmentation on the lumbar plexus block ultrasound image when the site type information of the lumbar plexus block ultrasound image is a preset site type, so as to obtain a plurality of tissue segmentation regions, a puncture needle segmentation region, and region categories of the plurality of tissue segmentation regions on the lumbar plexus block ultrasound image, wherein the region categories of the tissue segmentation regions include a punctureable category and an avoidance category; A generating module 704, for generating a puncture direction line of the puncture needle on the lumbar plexus nerve block ultrasound image based on the puncture needle segmentation area; The first judgment module 705 is used to judge whether the puncture direction line passes through the puncturable tissue segmentation area; The second judgment module 706 is used to judge whether the puncture direction line passes through the tissue segmentation area of ​​the avoidance category if the puncture direction line passes through the tissue segmentation area of ​​the punctureable category; The instruction module 707 is used to issue an instruction to continue puncturing in the current puncture direction if the puncture direction line does not pass through the tissue segmentation area of ​​the avoidance category.

[0115] The specific implementation of each of the above modules can be found in the previous embodiments and will not be described in detail here.

[0116] An embodiment of the present application also provides an electronic device, including a memory and a processor, wherein the processor is used to execute the steps of the lumbar plexus nerve block ultrasound image recognition and puncture guidance method provided in this embodiment by calling a computer program stored in the memory.

[0117] Please refer to Figure 8 , Figure 8 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0118] The electronic device may include components such as a processor 101 with one or more processing cores, a memory 102 with one or more computer-readable storage media, a power supply 103, and an input unit 104. Those skilled in the art will appreciate that the electronic device structure shown in the figure does not constitute a limitation on the electronic device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently. Among them: The processor 101 is the control center of the electronic device, which uses various interfaces and lines to connect various parts of the entire electronic device, and executes various functions of the electronic device and processes data by running or executing software programs and / or modules stored in the memory 102, and calling data stored in the memory 102. Optionally, the processor 101 may include one or more processing cores; optionally, the processor 101 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the above-mentioned modem processor may not be integrated into the processor 101.

[0119] The memory 102 can be used to store software programs and modules. The processor 101 executes various functional applications and data processing by running the software programs and modules stored in the memory 102. The memory 102 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device, etc. In addition, the memory 102 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices. Accordingly, the memory 102 may also include a memory controller to provide the processor 101 with access to the memory 102.

[0120] The electronic device also includes a power supply 103 for supplying power to each component. Optionally, the power supply 103 can be logically connected to the processor 101 through a power management system, so as to manage charging, discharging, and power consumption through the power management system. The power supply 103 can also include one or more DC or AC power supplies, recharging systems, power failure detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0121] The electronic device may further include an input unit 104, which may be used to receive input digital or character information and generate keyboard, mouse, joystick, optical or trackball signal input related to user settings and function control.

[0122] Although not shown, the electronic device may also include a display unit 105, an image acquisition component, etc., which will not be described in detail here. Specifically in this embodiment, the processor 101 in the electronic device will load the executable code corresponding to one or more computer programs into the memory 102 according to the following instructions, and the processor 101 will execute the steps in the lumbar plexus nerve block ultrasound image recognition and puncture guidance method provided in this application, such as: Acquire a lumbar plexus block ultrasound image acquired by an ultrasound device; input the lumbar plexus block ultrasound image into a site classification model to obtain site type information of the lumbar plexus block ultrasound image; when the site type information of the lumbar plexus block ultrasound image is a preset site type, input the lumbar plexus block ultrasound image into a target image segmentation model to perform image segmentation on the lumbar plexus block ultrasound image to obtain multiple tissue segmentation areas, puncture needle segmentation areas, and area categories of multiple tissue segmentation areas on the lumbar plexus block ultrasound image, wherein the area categories of the tissue segmentation areas include punctureable categories and avoidance categories; generate a puncture direction line of the puncture needle on the lumbar plexus block ultrasound image based on the puncture needle segmentation area; determine whether the puncture direction line passes through the tissue segmentation area of ​​the punctureable category; if the puncture direction line passes through the tissue segmentation area of ​​the punctureable category, determine whether the puncture direction line passes through the tissue segmentation area of ​​the avoidance category; if the puncture direction line does not pass through the tissue segmentation area of ​​the avoidance category, issue an instruction to continue puncturing in the current puncture direction.

[0123] It should be noted that the electronic device provided in the embodiment of the present application belongs to the same concept as the lumbar plexus nerve block ultrasound image recognition and puncture guidance method in the above embodiment. The specific implementation process is detailed in the above related embodiments and will not be repeated here.

[0124] The present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program stored therein is executed on the processor of the electronic device provided in the embodiment of the present application, the processor of the electronic device executes the steps in the lumbar plexus nerve block ultrasound image recognition and puncture guidance method provided in the present application. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0125] The present application also provides a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes various optional implementations of the above-mentioned lumbar plexus nerve block ultrasound image recognition and puncture guidance method.

[0126] The above is a detailed introduction to the lumbar plexus nerve block ultrasound image recognition and puncture guidance method and device provided by the present application. This article uses specific examples to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

[0127] It should be noted that when the above embodiments of the present application are applied to specific products or technologies, the relevant data of the user is involved, and the user's permission or consent is required, and the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

Claims

1. A method for ultrasonic image recognition and puncture guidance of lumbar plexus nerve block, characterized in that: include: Obtaining lumbar plexus block ultrasound images acquired by ultrasound equipment; Inputting the lumbar plexus nerve block ultrasound image into a site classification model to obtain site type information of the lumbar plexus nerve block ultrasound image; When the site type information of the lumbar plexus block ultrasound image is a preset site type, the lumbar plexus block ultrasound image is input into a target image segmentation model to perform image segmentation on the lumbar plexus block ultrasound image, so as to obtain a plurality of tissue segmentation regions, a puncture needle segmentation region, and a plurality of region categories of the tissue segmentation regions on the lumbar plexus block ultrasound image, wherein the region categories of the tissue segmentation regions include a punctureable category and an avoidance category; generating a puncture direction line of the puncture needle on the lumbar plexus nerve block ultrasound image based on the puncture needle segmentation area; Determining whether the puncture direction line passes through the puncturable tissue segmentation area; If the puncture direction line passes through the tissue segmentation area of ​​the puncturable category, determining whether the puncture direction line passes through the tissue segmentation area of ​​the avoidance category; If the puncture direction line does not pass through the tissue segmentation area of ​​the avoidance category, an instruction to continue puncturing in the current puncture direction is issued.

2. The method for ultrasonic image recognition and puncture guidance of lumbar plexus nerve block according to claim 1, characterized in that: The lumbar plexus nerve block ultrasound image recognition and puncture guidance method comprises: If the puncture direction line passes through the tissue segmentation area of ​​the avoidance category, determining whether the tissue segmentation area of ​​the avoidance category through which the puncture direction line passes is located between the puncture needle segmentation area and the target puncture area; If the tissue segmentation area of ​​the avoidance category that the puncture direction line passes through is not located between the puncture needle segmentation area and the target puncture area, then the real-time distance between the needle tip of the puncture needle segmentation area along the puncture needle direction line and the tissue segmentation area of ​​the avoidance category is calculated; If the real-time distance is not less than the second preset distance value, an instruction to continue puncturing in the current puncture direction is issued.

3. The method for ultrasonic image recognition and puncture guidance of lumbar plexus nerve block according to claim 2, characterized in that: The lumbar plexus nerve block ultrasound image recognition and puncture guidance method comprises: If the tissue segmentation area of ​​the avoidance category through which the puncture direction line passes is located between the puncture needle segmentation area and the target puncture area, a prompt signal is issued to prompt the operator to adjust the puncture direction to avoid the avoidance area.

4. The method for ultrasonic image recognition and puncture guidance of lumbar plexus nerve block according to claim 2, characterized in that: The lumbar plexus nerve block ultrasound image recognition and puncture guidance method comprises: If the real-time distance is less than the second preset distance value, an alarm signal is issued.

5. The method for ultrasonic image recognition and puncture guidance of lumbar plexus nerve block according to claim 4, characterized in that: When the image category of the lumbar plexus block ultrasound image is the paraspinal transverse process sagittal plane, the tissue segmentation area belonging to the punctureable category on the lumbar plexus block ultrasound image includes the erector spinae muscle area and the psoas muscle area, and the tissue segmentation area belonging to the avoidance category on the lumbar plexus block ultrasound image includes multiple transverse process areas and a dangerous area containing abdominal visceral blood vessels, the multiple transverse process areas are arranged at intervals between the erector spinae muscle area and the psoas muscle area, and the dangerous area is located on the side of the psoas muscle area away from the erector spinae muscle area; When the image category of the lumbar plexus block ultrasound image is a paraspinal transverse process cross section, the tissue segmentation areas belonging to the puncture category on the lumbar plexus block ultrasound image include the erector spinae muscle area, the psoas muscle area, and the quadratus lumborum muscle area, and the tissue segmentation areas belonging to the avoidance category on the lumbar plexus block ultrasound image include the vertebrae and dangerous areas including abdominal organs and blood vessels; When the image category of the lumbar plexus block ultrasound image is the cross-section of the ventrolateral transverse process, the tissue segmentation areas belonging to the punctureable category on the lumbar plexus block ultrasound image include the erector spinae muscle area, the psoas major muscle area and the quadratus lumborum muscle area, and the tissue segmentation areas belonging to the avoidance category on the lumbar plexus block ultrasound image include the vertebrae and dangerous areas containing abdominal organs and blood vessels.

6. The method for ultrasonic image recognition and puncture guidance of lumbar plexus nerve block according to claim 5, characterized in that: The lumbar plexus nerve block ultrasound image recognition and puncture guidance method comprises: Determine a target puncture area within the psoas major muscle region based on an image category of the lumbar plexus nerve block ultrasound image and a minimum circumscribed rectangle of the psoas major muscle segmentation area; The target puncture area is displayed on the lumbar plexus nerve block ultrasound image.

7. The method for ultrasonic image recognition and puncture guidance of lumbar plexus nerve block according to claim 6, characterized in that: The method of determining a target puncture area on the psoas major region based on the image category of the lumbar plexus nerve block ultrasound image and the minimum circumscribed rectangle of the psoas major region comprises: When the image category of the lumbar plexus nerve block ultrasound image is the paraspinal transverse process sagittal plane, and the puncture needle segmentation area is located on the side of the psoas major muscle area away from the dangerous area, one-third of the area within the minimum circumscribed rectangle of the psoas major muscle area close to the erector spinae segmentation area is determined as the target puncture area; When the image category of the lumbar plexus nerve block ultrasound image is a paravertebral transverse process cross section, and the puncture needle segmentation area is located on the side of the psoas major muscle area away from the vertebrae, a quarter of the segmentation area close to the vertebrae and the erector spinae muscle within the minimum circumscribed rectangle of the psoas major muscle area is determined as the target puncture area; When the image category of the lumbar plexus nerve block ultrasound image is a cross-section of the ventral lateral transverse process, and the puncture needle segmentation area is located on a side of the psoas major muscle area away from the vertebral area, a quarter area within the minimum circumscribed rectangle of the psoas major muscle area close to the vertebral area and the erector spinae muscle area is determined as the target puncture area.

8. A lumbar plexus nerve block ultrasound image recognition and puncture guidance device, characterized in that: include: An acquisition module, used for acquiring a lumbar plexus nerve block ultrasound image acquired by an ultrasound device; A classification module, used for inputting the lumbar plexus nerve block ultrasound image into a site classification model to obtain site type information of the lumbar plexus nerve block ultrasound image; a segmentation module, for inputting the lumbar plexus block ultrasound image into a target image segmentation model to perform image segmentation on the lumbar plexus block ultrasound image when the site type information of the lumbar plexus block ultrasound image is a preset site type, so as to obtain a plurality of tissue segmentation regions, a puncture needle segmentation region, and a plurality of region categories of the tissue segmentation regions on the lumbar plexus block ultrasound image, wherein the region categories of the tissue segmentation regions include a punctureable category and an avoidance category; A generating module, configured to generate a puncture direction line of the puncture needle on the lumbar plexus nerve block ultrasound image based on the puncture needle segmentation area; A first judgment module is used to judge whether the puncture direction line passes through the puncturable tissue segmentation area; A second judgment module is used to judge whether the puncture direction line passes through the tissue segmentation area of ​​the avoidance category if the puncture direction line passes through the tissue segmentation area of ​​the puncturable category; The instruction module is used to issue an instruction to continue puncturing in the current puncture direction if the puncture direction line does not pass through the tissue segmentation area of ​​the avoidance category.

9. An electronic device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and the processor is used to run the computer program in the memory to execute the steps in the lumbar plexus nerve block ultrasound image recognition and puncture guidance method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the steps in the lumbar plexus nerve block ultrasound image recognition and puncture guidance method according to any one of claims 1 to 7.

Images