Quality control method and system for trauma ultrasound key evaluation image and video

Abstract

The invention relates to a quality control method for trauma ultrasound key assessment images and videos, and belongs to the technical field of trauma ultrasound key assessment, and the method comprises the following steps: S1, obtaining a trauma ultrasound key assessment FAST ultrasound image; s2, identifying and extracting a key anatomical structure in the FAST ultrasonic image by using an FAST-Detect model; s3, performing quality control and scoring on the number of key anatomical structures identified in the FAST ultrasonic image; and S4, outputting a high-quality FAST ultrasonic image according to a scoring result. The invention further provides a quality control system for the wound ultrasonic key evaluation image and video. According to the method, the imaging quality of the FAST ultrasonic image can be quickly and accurately determined, so that the accurate and quick control of the imaging quality of the FAST ultrasonic image is realized, the high-quality FAST ultrasonic image is further obtained, and the accurate wound condition of a patient is favorably obtained.

Description

Technical Field

[0001] The present invention belongs to the technical field of focused ultrasound assessment of trauma, and relates to a quality control method and system for focused ultrasound assessment of trauma (FAST) images and videos. Background Art

[0002] Focused assessment with sonography for trauma (FAST) is an efficient and simple ultrasound examination method that can quickly determine whether there is potentially life-threatening internal bleeding or pericardial tamponade, thereby improving the efficiency and safety of emergency treatment. FAST was originally used in the ultrasound assessment of abdominal trauma to check whether there is fluid accumulation in areas such as the liver, spleen, and pelvis.

[0003] Whether the quality of the acquired images meets the clinical diagnosis requirements is a necessary condition for the accuracy of FAST examinations. In order to obtain high-quality and clear FAST examination images, existing technical solutions mainly focus on the improvement of the imaging system and the optimization of the examination process, quantifying the image quality by clarity, which does not meet the strict clinical definition of standard sections (i.e. the existence of key structures). It also fails to solve the problems caused by the operational dependence and harsh environment of FAST ultrasound examinations in trauma diagnosis and treatment scenarios.

[0004] The FAST examination is a medical examination for emergency situations, which is greatly affected by factors such as the operator's skills, on-site environment, and instrument conditions. Due to 1) the limited level of medical treatment personnel; 2) the urgency of treatment time; 3) the impact and shaking of treatment and transportation vehicles; 4) the harsh environment, etc., the diagnosis and treatment environment and the level of treatment received by the injured and sick are unstable. If quality assessment and control are not carried out, the lesion cross-section obtained will be incorrect, which will directly affect the treatment results. Summary of the invention

[0005] In view of this, the purpose of the present invention is to provide a quality control method and system for focused ultrasound assessment of trauma (FAST) images and videos in order to control and evaluate the image quality of ultrasound examinations in trauma scenarios, so as to improve the quality of FAST imaging obtained in trauma diagnosis and treatment scenarios and enhance the credibility of trauma disease analysis and diagnosis results.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] In one aspect, the present invention provides a method for quality control of trauma ultrasound focused assessment images and videos, comprising the following steps:

[0008] S1: Obtain FAST ultrasound images for focused assessment of trauma ultrasound;

[0009] S2: Use the FAST-Detect model to identify and extract key anatomical structures in FAST ultrasound images;

[0010] S3: Quality control and scoring of the number of key anatomical structures identified in FAST ultrasound images;

[0011] S4: Output high-quality FAST ultrasound images according to the scoring results.

[0012] Furthermore, the parameters of FAST ultrasonic imaging include section parameters and characteristic parameters of the FAST ultrasonic imaging; the section parameters of the FAST ultrasonic imaging include the section score of the FAST ultrasonic imaging; the characteristic parameters of the FAST ultrasonic imaging include the site characteristic parameters or structure characteristic parameters of the FAST ultrasonic imaging.

[0013] Furthermore, the FAST-Detect model in step S2 uses a segmentation-based deep learning method or a target detection-based deep learning method to identify key anatomical structures in the ultrasound image; first, the target detection network needs to be trained using the pre-annotated recognition structure;

[0014] If the acquired FAST ultrasound image is used as the image signal input, the FAST-Detect model analyzes the single image to obtain the identified objects, quantity, and confidence level;

[0015] If the acquired FAST ultrasound image is a sequence video signal input, the FAST-Detect model first cuts the sequence video signal into frames, analyzes each frame of the image, and obtains the identified objects, quantity, and confidence.

[0016] Further, in step S3, for a single input image signal, the scoring rule is that after the FAST-Detect model analysis, 1 point is deducted for each missing structure, the starting score is 10 points, and the final score is the quality score;

[0017] For the input video sequence signal, score each frame after frame cutting. The scoring rule is to deduct 1 point for each missing structure, with a starting score of 10 points. The highest score in each frame is taken as the quality score of the video sequence. If there are multiple frames with the highest score, the one with the highest cumulative confidence of each structure is taken as the quality scoring object.

[0018] The target score calculation formula for each FAST ultrasound image is as follows:

[0019]

[0020] In the formula, Score i is the score of the slice corresponding to each feature category, D iis the structural feature score of the i-th structural feature corresponding to the feature category in the target section, and M is the total number of structural features corresponding to the feature category in the section.

[0021] Further, in step S4, the quality control rule is that if the quality score is less than 10 points, it is judged as unqualified;

[0022] For the sequence video signal, the quality score value and the quality judgment structure calculated and obtained in step S3 are output, and the image frame with the highest quality score value and the highest cumulative confidence of each structure in each frame is output;

[0023] The cumulative confidence calculation formula for each FAST ultrasound image is as follows:

[0024]

[0025] In the formula, Confidence i is the cumulative confidence of the slice corresponding to each feature category, C i is the recognition confidence of the structural feature of the i-th structural feature corresponding to the feature category in the target section, and M is the total number of structural features corresponding to the feature category in the section.

[0026] In another aspect, the present invention provides a quality control system for trauma ultrasound focused assessment images and videos, comprising:

[0027] FAST imaging acquisition module: used to obtain FAST ultrasound images for focused trauma ultrasound assessment;

[0028] FAST-Detect structure recognition module: used to identify and extract key anatomical structures in FAST ultrasound images using the FAST-Detect model;

[0029] Structural information scoring module: used to perform quality control and scoring on the number of key anatomical structures identified in FAST ultrasound images;

[0030] FAST imaging quality control module: used to determine whether FAST imaging is qualified, realize quality control of FAST imaging, and output the final quality control results.

[0031] Furthermore, the FAST imaging acquisition module is also used to adjust the gain, frequency, depth and focus position during the imaging acquisition process, optimize the imaging effect, and perform real-time processing, including denoising, smoothing and enhancement.

[0032] Furthermore, the FAST-Detect structure recognition module firstly trains the target detection network using the pre-annotated recognition structure; and then uses the target detection network to recognize the key anatomical structures in the ultrasound image;

[0033] If the FAST ultrasound image acquired by the FAST imaging acquisition module is input as the image signal, the FAST-Detect structure recognition module analyzes the single image to obtain the identified objects, quantity, and confidence level;

[0034] If the FAST ultrasound image acquired by the FAST imaging acquisition module is a sequence video signal input, the FAST-Detect structure recognition module first cuts the sequence video signal into frames, analyzes each frame of the image, and obtains the identified objects, quantity, and confidence.

[0035] Furthermore, for a single input image signal, the scoring rule of the structural information scoring module is that after the analysis by the FAST-Detect structural recognition module, 1 point is deducted for each missing structure, the starting score is 10 points, and the final score is the quality score;

[0036] For the input video sequence signal, the structural information scoring module scores each frame after frame cutting. The scoring rule is to deduct 1 point for each missing structure, with a starting score of 10 points. The highest score in each frame is taken as the quality score of the video sequence. If there are multiple frames with the highest score, the one with the highest cumulative confidence of each structure is taken as the quality scoring object.

[0037] The target score calculation formula for each FAST ultrasound image is as follows:

[0038]

[0039] In the formula, Score i is the score of the slice corresponding to each feature category, D i is the structural feature score of the i-th structural feature corresponding to the feature category in the target section, and M is the total number of structural features corresponding to the feature category in the section.

[0040] Furthermore, the quality control rule of the FAST imaging quality control module is that if the quality score is less than 10 points, it is judged as unqualified;

[0041] For the sequence video signal, the output structure information scoring module calculates the obtained quality score value and quality judgment structure, and outputs the image frame with the highest quality score value and the highest cumulative confidence of each structure in each frame;

[0042] The cumulative confidence calculation formula for each FAST ultrasound image is as follows:

[0043]

[0044] In the formula, Confidence i is the cumulative confidence of the slice corresponding to each feature category, C iis the recognition confidence of the structural feature of the i-th structural feature corresponding to the feature category in the target section, and M is the total number of structural features corresponding to the feature category in the section.

[0045] The beneficial effects of the present invention are as follows: the present invention can quickly and accurately determine the imaging quality of the FAST ultrasound image by automatically determining the imaging quality of the FAST ultrasound image according to the determined imaging score of the FAST ultrasound image, thereby achieving precise and rapid control of the imaging quality of the FAST ultrasound image, and then obtaining high-quality FAST ultrasound images, which is conducive to obtaining accurate patient trauma conditions; through the determined imaging quality and parameters of the FAST ultrasound image, it is also possible to know the standardization of the staff in the process of detecting the FAST ultrasound image and whether the objects required to be detected by FAST are completely covered.

[0046] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below in conjunction with the accompanying drawings, wherein:

[0048] Figure 1 This is a flow chart of the quality control method for focused assessment with sonography for trauma (FAST);

[0049] Figure 2 A block diagram of the quality control system for Focused Assessment with Sonography for Trauma (FAST);

[0050] Figure 3 This is an example image identified by the FAST-Detect structure recognition module. DETAILED DESCRIPTION

[0051] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0052] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and thus the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0053] In the following description, numerous details are discussed to provide a more thorough explanation of the embodiments of the present invention. However, it is obvious to those skilled in the art that the embodiments of the present invention can be implemented without these specific details. In other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present invention difficult to understand.

[0054] like Figure 1 As shown, the present invention provides a quality control method for FAST images and videos for focused assessment of trauma ultrasound, comprising the following steps:

[0055] S01: Obtain FAST ultrasound images;

[0056] S02: Use the FAST-Detect model to identify and extract key anatomical structures specified in the emergency ultrasound guidelines from FAST ultrasound images;

[0057] S03: Quality control and scoring of the number of key anatomical structures identified in FAST ultrasound images;

[0058] S04: Output high-quality FAST ultrasound images according to the scoring results.

[0059] In this method, the parameter determination model determined by the FAST imaging input is analyzed, and the analysis results are obtained as the parameters of the FAST imaging. Wherein, the parameters of the FAST imaging include the section parameters of the FAST imaging, and the section parameters of the FAST imaging include the section scores of the FAST imaging; and / or, the parameters for FAST imaging sent by the determined terminal device and / or input by the authorized personnel are received as the parameters of the FAST imaging, the parameters of the FAST imaging include the characteristic parameters and / or section parameters of the FAST imaging, the characteristic parameters of the FAST imaging include the part characteristic parameters and / or structure characteristic parameters of the FAST imaging, and the section parameters of the FAST imaging include the section scores of the FAST imaging.

[0060] like Figure 2As shown, the present invention provides a quality control system for FAST images and videos for focused assessment of trauma ultrasound, comprising four modules: a FAST imaging acquisition module 1, a FAST-Detect structure recognition module 2, a structure information scoring module 3, and a FAST imaging quality control module 4, which respectively perform the tasks of acquiring FAST imaging information, identifying and extracting key anatomical structures specified in the emergency ultrasound guidelines in FAST ultrasound images, performing quality control and scoring on the number of key anatomical structures identified in FAST ultrasound images, and outputting high-quality FAST ultrasound images according to the scores.

[0061] The FAST imaging acquisition module 1 is used to acquire FAST ultrasound images, which include two modes: sequence video and image signal, which are ultrasound images of one of the four-chamber heart section, left upper abdominal section, right upper abdominal section, and pelvic section under the xiphoid process in the FAST examination. In this embodiment, the FAST imaging acquisition module can meet different FAST ultrasound imaging analysis requirements by adopting a variety of imaging methods, such as two-dimensional (B-ultrasound), three-dimensional and four-dimensional imaging technologies. The module is compatible with data in different forms, including image formats (JPG, PNG, etc.), DICOM formats, and video modes. During the imaging acquisition process, the gain, frequency, depth, and focus position can be adjusted to optimize the imaging effect, and real-time processing functions such as denoising, smoothing, and enhancement are provided to improve image clarity and ensure stable imaging.

[0062] The FAST-Detect structure recognition module 2 is used to identify and extract the key structures specified in the emergency ultrasound guidelines in FAST imaging. The structure extraction method adopts a deep learning method, one is a deep learning method based on segmentation, and the other is a deep learning method based on target detection. The deep learning method based on segmentation can use U-Net, U-Net++, FCN, R-CNN, etc., and the deep learning method based on detection can use R-CNN, YOLO, SSD, etc. Both of these deep learning-based methods require the annotation of the structure in advance and then training. If the FAST imaging acquisition module obtains the image signal input, the FAST-Detect structure recognition module analyzes the single image to obtain the identified object, quantity, and confidence. If the FAST imaging acquisition module obtains the sequence video signal input, the FAST-Detect structure recognition module first cuts the sequence video signal, analyzes each frame of the image, and obtains the identified object, quantity, and confidence. In this embodiment, the structure can be manually annotated in advance, and the target detection network can be used to identify the key anatomical structures in the ultrasound image, such as YOLO, RCNN, etc. The model structure is improved and adjusted based on the preliminary recognition effect, and then the final training is performed to obtain the final detection model, which can be used as the recognition model in the FAST-Detect structure recognition module. For example, for the left upper abdomen and right upper abdomen sections in FAST, four anatomical structures related to image quality and diagnosis, liver, kidney, diaphragm, and Morrison's pouch, are annotated in advance. Then the YOLOv8 model is used as the benchmark model for adjustment and improvement. To address the problem of low accuracy in small target detection, the study introduced the XIOU loss function and CFNet module to optimize the YOLOv8 model, significantly improving the recognition accuracy. Through the above anatomical structure determination and annotation, model improvement and training, the anatomical structures of the left upper abdomen and right upper abdomen sections, liver, kidney, diaphragm, and Morrison's pouch, are automatically obtained in the FAST-Detect module. Figure 3 shown.

[0063] The structural information scoring module 3 is used to summarize the identified structural information and implement the quality scoring of FAST imaging. For a single input image signal, the scoring rule of the structural information scoring module is that after the analysis by the FAST-Detect structural recognition module, 1 point is deducted for each missing structure, the starting score is 10 points, and the final score is the quality score; for the input video sequence signal, the structural information scoring module scores each frame of the image after frame cutting, and the scoring rule is that 1 point is deducted for each missing structure, the starting score is 10 points, and the highest score in each frame is taken as the quality score value of the video sequence. If there are multiple frames of images with the highest score, the one with the highest cumulative confidence of each structure is taken as the quality scoring object. In this method, the target score calculation formula corresponding to each FAST ultrasound image is as follows:

[0064]

[0065] In the formula, Score i is the score of the slice corresponding to each feature category, D i is the structural feature score of the i-th structural feature corresponding to the feature category in the target section, and M is the total number of structural features corresponding to the feature category in the section.

[0066] In this embodiment, after analysis by the FAST-Detect structure recognition module, 1 point will be deducted for each missing structure, with an initial score of 10 points, and the final score will be used as the quality score. For the input video sequence signal, the structural information scoring module will score each frame of the image, and the scoring rule is also to deduct 1 point for each missing structure, with an initial score of 10 points. The quality score of the video sequence will take the highest score in each frame. If the scores of multiple frames are the same, the structure with the highest cumulative confidence is selected as the final quality scoring object. For example, for the left upper abdominal and right upper abdominal sections in FAST, four anatomical structures related to image quality and diagnosis, liver, kidney, diaphragm, and Morrison's space, are scored. A single score value is obtained for a single image, and for video images, the score value and confidence cumulative value corresponding to the time series are obtained.

[0067] The FAST imaging quality control module 4 is used to determine whether the FAST imaging is qualified, to achieve quality control of the FAST imaging, and to output the final quality control results. The quality control rule of the FAST imaging quality control module is that if the quality score is less than 10 points, it is judged as unqualified. For a single input image signal, the image information, quality score and whether it is qualified are output; for a sequence video signal, the output structure information scoring module calculates the obtained quality score and quality judgment structure, and outputs the quality score in each frame and the image frame with the highest cumulative confidence of each structure. In this method, the cumulative confidence calculation formula corresponding to each FAST ultrasound image is as follows:

[0068]

[0069] In the formula, Confidence i is the cumulative confidence of the slice corresponding to each feature category, C i is the recognition confidence of the structural feature of the i-th structural feature corresponding to the feature category in the target section, and M is the total number of structural features corresponding to the feature category in the section.

[0070] In this embodiment, the FAST imaging quality control module is used to determine the eligibility of FAST imaging, monitor its quality, and output the final quality control results. The quality control rules of this module stipulate that when the quality score value is less than 10 points, the imaging will be judged as unqualified. For a single input image signal, the module will output image information, quality score value and qualified or unqualified judgment result; for a video sequence signal, the quality score value and quality judgment structure calculated by the structural information scoring module are output, and the quality score value of each frame and the image frame with the highest cumulative confidence of each structure are provided. For example, for the left upper abdomen and right upper abdomen sections in FAST. If the input is a single image, the score input from the structural information scoring module is less than 10 points, then the left upper abdomen and right upper abdomen sections are judged to be unqualified, and the quality control image output by the FAST imaging quality control module is empty. If the score input by the structural information scoring module is 10 points, then the left upper abdomen and right upper abdomen sections are judged to be qualified, and the quality control image output by the FAST imaging quality control module is this image. If the input is a time series video, and the score of each time point section input from the structural information scoring module is less than 10 points, the video section is judged to be unqualified, and the quality control image output by the FAST imaging quality control module is empty. If the score of each time point section input from the structural information scoring module is 10 points, the video section is judged to be qualified, and the quality control image output by the FAST imaging quality control module is the section corresponding to the time with a score of 10 points. In addition, if there are multiple time series with a score of 10 points, the FAST imaging quality control module outputs the time series section corresponding to the highest cumulative confidence value in the structural information scoring module.

[0071] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.

Claims

1. A quality control method for trauma ultrasound focused assessment images and videos, characterized by: The following steps are involved: S1: Obtain FAST ultrasound images for focused assessment of trauma ultrasound; S2: Use the FAST-Detect model to identify and extract key anatomical structures in FAST ultrasound images; S3: Quality control and scoring of the number of key anatomical structures identified in FAST ultrasound images; S4: Output high-quality FAST ultrasound images according to the scoring results.

2. The quality control method for trauma ultrasound focused assessment images and videos according to claim 1, characterized in that: The parameters of FAST ultrasound imaging include the section parameters and characteristic parameters of the FAST ultrasound imaging; the section parameters of the FAST ultrasound imaging include the section score of the FAST ultrasound imaging; the characteristic parameters of the FAST ultrasound imaging include the site characteristic parameters or structure characteristic parameters of the FAST ultrasound imaging.

3. The quality control method for trauma ultrasound focused assessment images and videos according to claim 1, characterized in that: The FAST-Detect model in step S2 uses a segmentation-based deep learning method or a target detection-based deep learning method to identify key anatomical structures in ultrasound images; first, the target detection network needs to be trained using the pre-annotated recognition structure; If the acquired FAST ultrasound image is used as the image signal input, the FAST-Detect model analyzes the single image to obtain the identified objects, quantity, and confidence level; If the acquired FAST ultrasound image is a sequence video signal input, the FAST-Detect model first cuts the sequence video signal into frames, analyzes each frame of the image, and obtains the identified objects, quantity, and confidence.

4. The quality control method for trauma ultrasound focused assessment images and videos according to claim 1, characterized in that: In step S3, for a single input image signal, the scoring rule is that after the FAST-Detect model analysis, 1 point is deducted for each missing structure, the starting score is 10 points, and the final score is the quality score; For the input video sequence signal, score each frame after frame cutting. The scoring rule is to deduct 1 point for each missing structure, with a starting score of 10 points. The highest score in each frame is taken as the quality score of the video sequence. If there are multiple frames with the highest score, the one with the highest cumulative confidence of each structure is taken as the quality scoring object. The target score calculation formula for each FAST ultrasound image is as follows: In the formula, Score i is the score of the slice corresponding to each feature category, D i is the structural feature score of the i-th structural feature corresponding to the feature category in the target section, and M is the total number of structural features corresponding to the feature category in the section.

5. The quality control method for trauma ultrasound focused assessment images and videos according to claim 1, characterized in that: In step S4, the quality control rule is that if the quality score is less than 10 points, it is judged as unqualified; For the sequence video signal, the quality score value and the quality judgment structure calculated and obtained in step S3 are output, and the image frame with the highest quality score value and the highest cumulative confidence of each structure in each frame is output; The cumulative confidence calculation formula for each FAST ultrasound image is as follows: In the formula, Confidence i is the cumulative confidence of the slice corresponding to each feature category, C i is the recognition confidence of the structural feature of the i-th structural feature corresponding to the feature category in the target section, and M is the total number of structural features corresponding to the feature category in the section.

6. A quality control system for trauma ultrasound focused assessment images and videos, characterized by: include: FAST imaging acquisition module: used to obtain FAST ultrasound images for focused trauma ultrasound assessment; FAST-Detect structure recognition module: used to identify and extract key anatomical structures in FAST ultrasound images using the FAST-Detect model; Structural information scoring module: used to perform quality control and scoring on the number of key anatomical structures identified in FAST ultrasound images; FAST imaging quality control module: used to determine whether FAST imaging is qualified, realize quality control of FAST imaging, and output the final quality control results.

7. The quality control system for trauma ultrasound focused assessment images and videos according to claim 6, characterized in that: The FAST imaging acquisition module is also used to adjust the gain, frequency, depth and focus position during the imaging acquisition process, optimize the imaging effect, and perform real-time processing, including denoising, smoothing and enhancement.

8. The quality control system for trauma ultrasound focused assessment images and videos according to claim 6, characterized in that: The FAST-Detect structure recognition module first uses the pre-annotated recognition structure to train the target detection network; then uses the target detection network to identify key anatomical structures in the ultrasound image; If the FAST ultrasound image acquired by the FAST imaging acquisition module is input as the image signal, the FAST-Detect structure recognition module analyzes the single image to obtain the identified objects, quantity, and confidence level; If the FAST ultrasound image acquired by the FAST imaging acquisition module is a sequence video signal input, the FAST-Detect structure recognition module first cuts the sequence video signal into frames, analyzes each frame of the image, and obtains the identified objects, quantity, and confidence.

9. The quality control system for trauma ultrasound focused assessment images and videos according to claim 6, characterized in that: For a single input image signal, the scoring rule of the structural information scoring module is that after the FAST-Detect structural recognition module analyzes, 1 point is deducted for each missing structure, the starting score is 10 points, and the final score is the quality score; For the input video sequence signal, the structural information scoring module scores each frame after frame cutting. The scoring rule is to deduct 1 point for each missing structure, with a starting score of 10 points. The highest score in each frame is taken as the quality score of the video sequence. If there are multiple frames with the highest score, the one with the highest cumulative confidence of each structure is taken as the quality scoring object. The target score calculation formula for each FAST ultrasound image is as follows: In the formula, Score i is the score of the slice corresponding to each feature category, D i is the structural feature score of the i-th structural feature corresponding to the feature category in the target section, and M is the total number of structural features corresponding to the feature category in the section.

10. The quality control system for trauma ultrasound focused assessment images and videos according to claim 6, characterized in that: The quality control rule of the FAST imaging quality control module is that if the quality score is less than 10 points, it is judged as unqualified; For the sequence video signal, the output structure information scoring module calculates the obtained quality score value and quality judgment structure, and outputs the image frame with the highest quality score value and the highest cumulative confidence of each structure in each frame; The cumulative confidence calculation formula for each FAST ultrasound image is as follows: In the formula, Confidence i is the cumulative confidence of the slice corresponding to each feature category, C i is the recognition confidence of the structural feature of the i-th structural feature corresponding to the feature category in the target section, and M is the total number of structural features corresponding to the feature category in the section.

Images