Carotid artery plaque identification method and device

Abstract

The present invention provides a method and device for identifying carotid artery plaques. The method comprises the following steps: generating an ultrasonic screening queue according to an ultrasonic plaque screening image sent by a labeling request terminal, sending an ultrasonic screening set in the ultrasonic screening queue to a labeling node in a dynamic ultrasonic labeling tree for processing in sequence, and obtaining a plaque labeling image; fusing the labeled plaque contours in the plaque labeling image based on a plaque fusion strategy, and obtaining a fused plaque image, which is sent to the labeling request terminal; determining the basic number of ultrasonic plaque screening images received by a dynamic labeling combination in the dynamic ultrasonic labeling tree within a preset rotation period, and retrieving the fused number of fused plaque images corresponding to the corresponding dynamic labeling combination within the preset rotation period; and calculating according to the fused number and the basic number to obtain a score value corresponding to each labeling terminal.

Description

Technical Field

[0001] The present invention relates to data processing technology, and in particular to a method and device for identifying carotid artery plaque. Background Art

[0002] Carotid artery lumen stenosis caused by carotid artery sclerosis plaque is an important factor in the occurrence of ischemic stroke. Therefore, early identification of vulnerable carotid artery plaques through imaging methods is of great significance for stroke prevention and early clinical intervention.

[0003] In the existing technology, high-resolution magnetic resonance imaging can be used for examination. However, due to its high cost and complicated operation, rural clinics have not yet popularized the corresponding examination method. In addition, since CT angiography has problems such as radiation and side effects of iodine contrast agents, rural clinics do not have corresponding detection equipment for accuracy inspection. Therefore, rural clinics can use conventional ultrasound examination to screen carotid artery disease. This method has the advantages of being non-invasive, radiation-free, real-time, economical and convenient, but the sensitivity and specificity for identifying components within plaques are limited. Therefore, doctors need to have sufficient experience to diagnose the disease through ultrasound examination.

[0004] Therefore, this plan proposes that rural clinics can conduct preliminary examinations through ultrasound, send the ultrasound examination results to senior doctors for judgment, and integrate the diagnostic comments of multiple doctors to prevent misdiagnosis and wrong diagnosis, thereby improving the convenience of identifying and detecting carotid artery plaques in rural areas. Summary of the invention

[0005] The embodiments of the present invention provide a method and device for identifying carotid artery plaques, which can fuse the diagnostic annotations of multiple doctors to prevent misdiagnosis and wrong diagnosis, thereby improving the convenience of identifying and detecting carotid artery plaques in rural areas.

[0006] A first aspect of an embodiment of the present invention provides a method for identifying carotid artery plaques, comprising:

[0007] Generate an ultrasound screening queue according to the ultrasound plaque screening image sent by the annotation request end, and send the ultrasound screening set in the ultrasound screening queue to the annotation end of the annotation node in the dynamic ultrasound annotation tree for processing in sequence to obtain a plaque annotation image;

[0008] Based on the plaque fusion strategy, the annotated plaque contours in the plaque annotated image are fused to obtain a fused plaque image and send it to the annotation request end;

[0009] Determine the basic number of ultrasound plaque screening images received by the dynamic annotation combination in the dynamic ultrasound annotation tree within a preset rotation period, and retrieve the fused number of fused plaque images corresponding to the dynamic annotation combination within the preset rotation period;

[0010] The score value corresponding to each annotation end is obtained by calculating according to the fusion quantity and the basic quantity.

[0011] Optionally, in a possible implementation of the first aspect, the dynamic ultrasound annotation tree is constructed by the following steps, including:

[0012] Counting all preliminary review annotation ends to obtain a preliminary review annotation set, selecting any two preliminary review annotation ends in the preliminary review annotation set based on a preset rotation period to combine them to obtain multiple preliminary review annotation combinations, and constructing preliminary review association nodes based on the preliminary review annotation ends in the preliminary review annotation combinations;

[0013] Select any review annotation terminal according to the preset rotation period to construct a review annotation node, and connect the initial review association node corresponding to the initial review annotation combination with the review annotation node;

[0014] Constructing a final review annotation node based on the final review annotation end, connecting the review annotation node with the final review annotation node to generate a dynamic ultrasonic annotation tree;

[0015] The method generates an ultrasound screening queue according to the ultrasound plaque screening image sent by the annotation request end, and sequentially sends the ultrasound screening set in the ultrasound screening queue to the annotation end of the annotation node in the dynamic ultrasound annotation tree for processing to obtain a plaque annotation image, including:

[0016] Counting the ultrasonic plaque screening images sent by each annotation request end, and obtaining the ultrasonic screening set corresponding to each annotation request end;

[0017] Determine the sending time of the ultrasonic screening set, and sort the ultrasonic screening set in ascending order based on the sending time to obtain an ultrasonic screening queue;

[0018] Acquire an idle primary review associated node in the dynamic ultrasonic annotation tree as the current primary review node, and take the review annotation node connected to the current primary review node as the current review node, and the final review annotation node as the current final review node;

[0019] Sending the ultrasound screening set in the ultrasound screening queue to the preliminary review annotation end corresponding to the current preliminary review node in sequence;

[0020] Receiving the annotation processing of the ultrasound plaque screening image in the ultrasound screening set by the preliminary review annotation end of the current preliminary review node, and obtaining the initial annotation image corresponding to each of the preliminary review annotation ends;

[0021] Using the initial annotation image corresponding to any one of the preliminary review annotation ends in the preliminary review annotation combination as the first annotation image, and using the initial annotation images corresponding to the remaining preliminary review annotation ends in the preliminary review annotation combination as the second annotation image;

[0022] A plaque annotated image is obtained according to the first annotated image and the second annotated image.

[0023] Optionally, in a possible implementation manner of the first aspect, obtaining the plaque annotated image according to the first annotated image and the second annotated image includes:

[0024] Updating the second annotation contour in the second annotation image to the first annotation image to obtain a preliminary annotation image;

[0025] Obtaining a preliminary horizontal difference value and a preliminary vertical difference value corresponding to the preliminary annotated image according to comparison information of the first contour coordinate set and the second contour coordinate set in the preliminary annotated image;

[0026] When it is determined that both the preliminary review transverse difference and the preliminary review longitudinal difference are less than or equal to the preset difference, the preliminary review annotated image is used as a plaque annotated image;

[0027] When it is determined that the preliminary review horizontal difference or the preliminary review vertical difference is greater than a preset difference, the preliminary review annotated image is sent to the review annotated end corresponding to the current review node, and the review annotated end of the current review node receives the annotation processing of the preliminary review annotated image, so as to obtain the review annotated image corresponding to the review annotated end;

[0028] Acquire the contour of the marked patch marked at the review mark end in the review mark image as a third marked contour, determine a third contour coordinate set of the third marked contour, and obtain a first review transverse difference and a first review longitudinal difference according to comparison information of the third contour coordinate set and the first contour coordinate set;

[0029] Based on the comparison information of the third contour coordinate set and the second contour coordinate set, obtaining a second review transverse difference and a second review longitudinal difference;

[0030] When it is determined that the first review transverse difference and the first review longitudinal difference are less than or equal to a preset difference, or the second review transverse difference and the second review longitudinal difference are less than or equal to a preset difference, the review annotated image is used as a plaque annotated image.

[0031] Optionally, in a possible implementation of the first aspect, the method further includes:

[0032] When it is determined that the first review horizontal difference or the first review vertical difference is greater than a preset difference, and the second review horizontal difference or the second review vertical difference is greater than the preset difference, sending the review annotated image to the final review annotated end corresponding to the current final review node;

[0033] Receiving the annotation processing of the review annotation image by the final review annotation end of the current final review node, and obtaining the final review annotation image corresponding to the final review annotation end;

[0034] Acquire the contour of the marked patch marked at the final marking end in the final marking image as a fourth marked contour, determine a fourth contour coordinate set of the fourth marked contour, and obtain a first final horizontal difference and a first final vertical difference according to comparison information of the fourth contour coordinate set and the first contour coordinate set;

[0035] Obtaining a second final horizontal difference and a second final vertical difference according to comparison information of the fourth contour coordinate set and the second contour coordinate set;

[0036] Obtaining a third final horizontal difference and a third final vertical difference according to comparison information of the fourth contour coordinate set and the third contour coordinate set;

[0037] When it is determined that the first final horizontal difference and the first final vertical difference, the second final horizontal difference and the second final vertical difference, or the third final horizontal difference and the third final vertical difference are less than or equal to the preset difference, the final annotated image is used as the plaque annotated image.

[0038] Optionally, in a possible implementation manner of the first aspect, updating the second annotation contour in the second annotation image to the first annotation image to obtain a preliminary annotation image includes:

[0039] Acquire a first center point of the first annotated image and a second center point of the second annotated image;

[0040] Constructing a first updated coordinate system based on the first center point, and constructing a second updated coordinate system based on the second center point;

[0041] Performing coordinate processing on the first annotated image according to the first updated coordinate system, taking the annotated patch contour in the first annotated image as a first annotated contour, and counting the coordinates of the first annotated contour to obtain a first contour coordinate set;

[0042] Performing coordinate processing on the second annotated image based on the second updated coordinate system, taking the annotated patch contour in the second annotated image as a second annotated contour, and obtaining a second contour coordinate set by counting the coordinates of the second annotated contour;

[0043] The second annotated contour is updated into the first annotated image according to the second contour coordinate set to obtain a preliminary annotated image.

[0044] Optionally, in a possible implementation manner of the first aspect, obtaining a preliminary horizontal difference value and a preliminary vertical difference value corresponding to the preliminary reviewed annotated image according to comparison information of a first contour coordinate set and a second contour coordinate set in the preliminary reviewed annotated image includes:

[0045] Acquire a first minimum abscissa, a first maximum abscissa, a first minimum ordinate and a first maximum ordinate of a first contour coordinate set in the preliminary annotated image, and a second minimum abscissa, a second maximum abscissa, a second minimum ordinate and a second maximum ordinate of a second contour coordinate set;

[0046] Obtain a left difference value based on an absolute value of a difference between the first minimum horizontal coordinate and the second minimum horizontal coordinate, and obtain a right difference value based on an absolute value of a difference between the first maximum horizontal coordinate and the second maximum horizontal coordinate;

[0047] Based on the absolute value of the difference between the first minimum ordinate and the second minimum ordinate, a lower difference is obtained; based on the absolute value of the difference between the first maximum ordinate and the second maximum ordinate, an upper difference is obtained;

[0048] A preliminary lateral difference is obtained based on the sum of the left side difference and the right side difference, and a preliminary longitudinal difference is obtained based on the lower side difference and the upper side difference.

[0049] Optionally, in a possible implementation manner of the first aspect, the fusing the annotated plaque contours in the plaque annotated image based on the plaque fusion strategy to obtain a fused plaque image and sending it to the annotation request end includes:

[0050] When it is determined that the plaque annotated image is a preliminary annotated image, two annotated plaque contours in the corresponding plaque annotated image are used as contours to be fused;

[0051] When the plaque annotated image is determined to be a review annotated image, a first review difference value is obtained according to the sum of the first review transverse difference value and the first review longitudinal difference value in the corresponding plaque annotated image, and a second review difference value is obtained according to the sum of the second review transverse difference value and the second review longitudinal difference value;

[0052] Selecting the minimum value of the first review difference value and the second review difference value as the screening review difference value, and determining two annotated plaque contours in the plaque annotated image as contours to be fused based on the screening review difference value;

[0053] When the plaque-annotated image is determined to be a final-reviewed annotated image, a first final-reviewed difference value is obtained according to the sum of the first final-reviewed transverse difference value and the first final-reviewed longitudinal difference value in the corresponding plaque-annotated image, a second final-reviewed difference value is obtained according to the sum of the second final-reviewed transverse difference value and the second final-reviewed longitudinal difference value, and a third final-reviewed difference value is obtained according to the sum of the third final-reviewed transverse difference value and the third final-reviewed longitudinal difference value;

[0054] Selecting the minimum value among the first final difference value, the second final difference value and the third final difference value as the screening final difference value, and determining two annotated plaque contours in the plaque annotated image as contours to be fused based on the screening final difference value;

[0055] Acquire the positional relationship of the contours to be fused, where the positional relationship includes a containment relationship and a non-containment relationship;

[0056] When it is determined that the positional relationship is a containment relationship, the contours to be fused in the patch annotation image are fused according to a nested fusion strategy to obtain a fused patch image and send it to the annotation request end;

[0057] When it is determined that the positional relationship is a non-inclusion relationship, the contours to be fused in the patch annotation image are fused according to an overlapping fusion strategy to obtain a fused patch image which is sent to the annotation request end.

[0058] Optionally, in a possible implementation manner of the first aspect, when determining that the positional relationship is a containment relationship, performing fusion processing on the contour to be fused in the patch annotated image according to a nested fusion strategy to obtain a fused patch image and sending it to the annotation request end, including:

[0059] When it is determined that the positional relationship is a containing relationship, the contour to be fused located on the outside is obtained as the containing contour, and the contour to be fused located on the inside is obtained as the contained contour;

[0060] Acquire the center point of the included contour as the positioning center point, construct a positioning straight line based on a preset angle and the positioning center point, and determine the outer intersection point of the positioning straight line with the included contour, and the inner intersection point of the positioning straight line with the included contour;

[0061] Calculate the contour fusion coordinates based on the outer coordinates of the adjacent outer intersection points and the inner coordinates of the inner intersection points on the same side of the positioning straight line;

[0062] The adjacent contour fusion coordinates are sequentially connected to each other to obtain a fused intermediate contour, and the annotated plaque contour in the plaque annotated image is deleted to obtain a fused plaque image and send it to the annotation request end;

[0063] When it is determined that the positional relationship is a non-inclusion relationship, the contour to be fused in the patch annotated image is fused according to an overlapping fusion strategy to obtain a fused patch image and send it to the annotation request end, including:

[0064] When it is determined that the positional relationship is a non-inclusion relationship, the intersection points of the two contours to be fused are obtained, the intersection points are connected to obtain an intersection connection line, and the midpoint of the intersection connection line is obtained as the positioning midpoint;

[0065] Take any one intersection point as the first contour intersection point, and take the other intersection point as the second contour intersection point;

[0066] Taking any one of the contours to be fused as the first intersecting contour, and taking another contour to be fused as the second intersecting contour;

[0067] Constructing an overlapping positioning line perpendicular to the intersection connection line based on the positioning midpoint, determining a first positioning intersection point between the overlapping positioning line and the first intersection contour, and a second positioning intersection point between the overlapping positioning line and the second intersection contour;

[0068] The center positioning coordinates corresponding to the center positioning point are obtained by performing mean calculation based on the first coordinates of the first positioning intersection point and the second coordinates of the second positioning intersection point adjacent to each other on the same side of the overlapping positioning line;

[0069] Connecting the center positioning points corresponding to the two center positioning coordinates to obtain a fused long axis;

[0070] Performing mean calculation on the center positioning coordinates to obtain fusion midpoint coordinates, and performing calculation based on the fusion midpoint coordinates and the coordinates of the first contour intersection point to obtain a first fusion semi-minor axis distance;

[0071] Calculating based on the coordinates of the fusion midpoint and the coordinates of the second contour intersection point to obtain a second fusion semi-minor axis distance;

[0072] A special-shaped fused ellipse contour is generated according to the fused midpoint coordinates, the first fused semi-minor axis distance, the second fused semi-minor axis distance and the fused major axis, and the annotated plaque contour in the plaque annotated image is deleted to obtain a fused plaque image and send it to the annotation request end.

[0073] Optionally, in a possible implementation manner of the first aspect, before the step of determining the basic number of ultrasound plaque screening images received by the dynamic annotation combination in the dynamic ultrasound annotation tree within the preset rotation period, and retrieving the fused number of fused plaque images corresponding to the dynamic annotation combination within the preset rotation period, the step includes:

[0074] Obtaining a preliminary review annotation combination corresponding to a preliminary review associated node in the dynamic ultrasonic annotation tree within the preset rotation period;

[0075] Determine a review annotation end corresponding to a review annotation node connected to the corresponding preliminary review associated node in the dynamic ultrasonic annotation tree, and a final review annotation end corresponding to a final review annotation node connected to the corresponding review annotation node;

[0076] Adding the review annotation end to the preliminary review annotation combination to obtain a review annotation combination, and adding the final review annotation end to the review annotation combination to obtain a final review annotation combination;

[0077] Obtaining a dynamic annotation combination according to the initial review annotation combination, the review annotation combination and the final review annotation combination within the preset rotation period;

[0078] The calculation is performed according to the fusion quantity and the basic quantity to obtain the score value corresponding to each annotation end, including:

[0079] The score corresponding to each annotation end is obtained by the following formula:

[0080] in, is the score value corresponding to each annotation end, is the number of fusions of the initial review annotation combination corresponding to the annotation end, is the basic number of initial review annotation combinations corresponding to the annotation end. is the initial review weight, is the number of fusions of the corresponding review annotation combination on the annotation side, is the basic number of review annotation combinations corresponding to the annotation end, For review weight, is the number of fusions of the final annotation combination corresponding to the annotation end, is the basic number of final review annotation combinations corresponding to the annotation end. The final weight.

[0081] A second aspect of the present invention provides a carotid artery plaque identification device, comprising:

[0082] A sending module is used to generate an ultrasound screening queue according to the ultrasound plaque screening image sent by the annotation request end, and sequentially send the ultrasound screening set in the ultrasound screening queue to the annotation end of the annotation node in the dynamic ultrasound annotation tree for processing to obtain a plaque annotation image;

[0083] A fusion module is used to fuse the annotated plaque contours in the plaque annotated image based on a plaque fusion strategy, and obtain a fused plaque image to send to the annotation request end;

[0084] A determination module, used to determine the basic number of ultrasound plaque screening images received by the dynamic annotation combination in the dynamic ultrasound annotation tree within a preset rotation period, and to retrieve the fused number of fused plaque images corresponding to the dynamic annotation combination within the preset rotation period;

[0085] The calculation module is used to calculate according to the fusion quantity and the basic quantity to obtain the score value corresponding to each annotation end.

[0086] According to a third aspect of the present invention, there is provided an electronic device, comprising: a memory, a processor and a computer program, wherein the computer program is stored in the memory, and the processor runs the computer program to execute the first aspect of the present invention and various methods that may be involved in the first aspect.

[0087] According to a fourth aspect of the present invention, a storage medium is provided, in which a computer program is stored. When the computer program is executed by a processor, it is used to implement the first aspect of the present invention and various methods that may be involved in the first aspect.

[0088] The beneficial effects of the present invention are as follows:

[0089] 1. The present invention can compare and fuse the diagnostic annotations of multiple doctors, reduce misdiagnosis and wrong diagnosis, and thus improve the convenience of identifying and detecting carotid plaques in rural areas. First, the present invention can sequentially send the ultrasound screening set in the ultrasound screening queue to the annotation end of the corresponding annotation node, so that the annotation end can perform annotation processing on the ultrasound plaque screening image, thereby performing fusion processing on the annotated plaque contour in the plaque annotation image, and then obtain a fused plaque image, so that the medical staff corresponding to the annotation request end can view and diagnose, and the present invention can determine the score value of each annotation end by obtaining the basic number of ultrasound plaque screening images received by each annotation end within a preset rotation period and the fusion number of fused plaque images obtained, thereby determining the combination with the largest score value of each annotation end, and then the corresponding annotation ends can be combined to improve the annotation efficiency of the ultrasound screening plaque image, and the corresponding fused plaque image can be quickly obtained, so that it can be quickly sent to the corresponding annotation request end, so that the corresponding medical staff can view it in time and improve the diagnostic efficiency.

[0090] 2. The present invention can adopt corresponding strategies to fuse contours according to the positional relationship of the contours to be fused, thereby improving the efficiency of contour fusion and ensuring the accuracy of the displayed plaque image contour. Among them, the present invention can adopt corresponding nested fusion strategies according to the inclusion relationship of the contours to be fused, so as to obtain the corresponding fused intermediate contours, and then obtain the fused plaque image, so as to send it to the annotation request end so that the corresponding medical staff can view it in time. When the contours to be fused are non-inclusion relationships, the corresponding overlapping fusion strategy can be adopted, so as to obtain the corresponding fusion midpoint coordinates, the first fusion semi-minor axis distance, the second fusion semi-minor axis distance and the fusion major axis according to the intersection of the corresponding contours to be fused, so as to generate a special-shaped fused elliptical contour, and then obtain the corresponding fused plaque image and send it to the annotation request end so that the corresponding medical staff can view the plaque image.

[0091] 3. The present invention can calculate the score value of each annotation end to determine the accuracy of plate identification and annotation by different annotation ends. In addition, the present invention can also select another annotation end that maximizes the score value of each annotation end according to the corresponding score values ​​of each annotation end in multiple preset rotation time periods to combine them, so as to improve the efficiency of plaque annotation. BRIEF DESCRIPTION OF THE DRAWINGS

[0092] Figure 1 A flow chart of a method for identifying carotid artery plaque provided by the present invention;

[0093] Figure 2 A schematic diagram of a dynamic ultrasonic labeling tree provided by the present invention;

[0094] Figure 3 A schematic diagram of a fused intermediate contour provided by the present invention;

[0095] Figure 4 This is a schematic structural diagram of a carotid artery plaque identification device provided by the present invention;

[0096] Figure 5 A schematic diagram of the hardware structure of an electronic device provided by the present invention. DETAILED DESCRIPTION

[0097] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0098] like Figure 1 As shown, the present invention provides a method for identifying carotid artery plaques, the method comprising:

[0099] S1, generating an ultrasound screening queue according to the ultrasound plaque screening image sent by the annotation request end, and sending the ultrasound screening set in the ultrasound screening queue to the annotation end of the annotation node in the dynamic ultrasound annotation tree for processing in sequence to obtain a plaque annotation image.

[0100] It should be noted that after completing the ultrasonic examination of the carotid artery in the rural clinic, the corresponding rural medical staff can send the corresponding ultrasonic plaque screening image to the superior physician with diagnostic experience for plaque annotation, so that the subsequent rural medical staff can identify the carotid artery plaque for diagnosis and treatment, saving the patient's time for round-trip examinations. In addition, when annotating plaques, a plaque annotation image will only be generated when the similarity of the annotation information of multiple physicians is high. If the annotation information does not meet certain judgment standards after being reviewed and annotated by multiple physicians, it means that the corresponding ultrasonic plaque screening image cannot be accurately diagnosed, and it is recommended that personnel go to a well-equipped hospital for high-precision examinations.

[0101] It can be understood that the annotation requesting end is the information terminal of the medical staff who requests plaque annotation, such as a mobile phone, a computer, etc., and the ultrasonic plaque screening image is an ultrasonic plaque examination image.

[0102] Since there are a large number of rural clinics that send plaque annotations to superior doctors, an ultrasound screening queue can be generated according to the time when different annotation request terminals are sent, so that the corresponding ultrasound screening set can be assigned to the annotation terminal for physician annotation in sequence later, wherein the ultrasound screening queue is an image sequence of ultrasound plaque screening images sent by different annotation request terminals. Since a large number of ultrasound images can be obtained during the ultrasound detection process, rural medical staff can select a part of the images to send requests for annotations. The ultrasound screening set is a collection of multiple ultrasound plaque screening images sent by the same annotation request terminal. The dynamic ultrasound annotation tree is a structure tree that displays the connection relationship between different annotation terminals, which can be dynamically changed. The annotation node is the node corresponding to the annotation terminal. The annotation terminal is the information terminal of the superior physician who can annotate the ultrasound screening image, such as the mobile phone and computer of the doctor in the municipal hospital. The plaque annotation image is an image with multiple annotation information that meets the judgment criteria.

[0103] Through the above implementation, the present invention can obtain a plaque annotation image, so as to facilitate the subsequent fusion of corresponding annotations, thereby determining a more accurate plaque annotation.

[0104] In some embodiments, a dynamic ultrasound annotation tree may be constructed by the following steps, including:

[0105] A1, counting all preliminary review annotation ends to obtain a preliminary review annotation set, selecting any two preliminary review annotation ends in the preliminary review annotation set based on a preset rotation period to combine, obtaining multiple preliminary review annotation combinations, and constructing preliminary review association nodes based on the preliminary review annotation ends in the preliminary review annotation combinations.

[0106] It should be noted that in order to ensure the accuracy of the annotation of the ultrasound plaque screening image and reduce the possibility of annotation errors by a single physician, multiple physicians can annotate the same ultrasound plaque screening image, and then compare the different annotations to improve the accuracy of the annotation information. The annotation ends corresponding to the physicians can be assigned in a hierarchical manner to obtain the corresponding dynamic ultrasound annotation tree, which is convenient for the subsequent annotation of the ultrasound plaque screening image.

[0107] It can be understood that all preliminary review annotation ends are counted to obtain a preliminary review annotation set, and then the preliminary review annotation ends in the preliminary review annotation set are combined in pairs to obtain multiple preliminary review annotation combinations, so that the same ultrasound screening set can be subsequently sent to the preliminary review annotation end corresponding to the same preliminary review annotation combination, so that medical staff can perform plaque annotation.

[0108] Among them, the preliminary review annotation end is the information terminal of the medical staff who conducts the review and annotation for the first time, such as the mobile phone and computer corresponding to staff A and staff B. The preliminary review annotation set is the set of annotation ends that perform the preliminary review annotation. The preset rotation period is the time period for exchanging annotation partners, which is pre-set, such as 1 week. The preliminary review annotation combination is a combination of any two preliminary review annotation ends, that is, two preliminary review annotation ends that annotate the same ultrasound plaque screening image, such as staff A and staff B are a combination. The preliminary review associated node is the node corresponding to the preliminary review annotation end.

[0109] It is not difficult to understand that in order to reduce the large differences in annotations caused by differences in technical levels between two doctors corresponding to the same preliminary review annotation combination, which affects the diagnosis efficiency, the preliminary review annotation ends in the preliminary review annotation combination can be rotated according to the preset rotation period, so as to subsequently determine the preliminary review annotation end combination with the highest diagnostic efficiency.

[0110] A2, selecting any review annotation end according to a preset rotation period to construct a review annotation node, and connecting the initial review association node corresponding to the initial review annotation combination with the review annotation node.

[0111] It can be understood that, within the preset rotation period, any one of the review annotation terminals is selected to construct a corresponding review annotation node, so that Figure 2 As shown, the review annotation node is connected to the initial review associated node corresponding to the initial review annotation combination, so that the annotation results of the corresponding initial review annotation end can be sent to the corresponding review annotation end for secondary annotation.

[0112] Among them, the preset rotation period is a pre-set rotation time period, the review annotation end is the information terminal of the person who performs secondary annotation on images that are not similar to the initial review annotations, and the review annotation node is the node corresponding to the review annotation end, so as to connect the initial review association node corresponding to the initial review annotation combination with the corresponding review annotation node, so as to obtain the corresponding dynamic ultrasonic annotation tree later.

[0113] A3, constructing a final review annotation node based on the final review annotation end, and connecting the review annotation node with the final review annotation node to generate a dynamic ultrasonic annotation tree.

[0114] It can be understood that the final review annotation end is the information terminal of the physician who makes the final review and annotation of the ultrasound plaque screening image, and the final review annotation node is the node corresponding to the final review annotation end, so that Figure 2 As shown, by connecting each review annotation node with the final review annotation node, a dynamic ultrasound annotation tree can be obtained.

[0115] Through the above implementation, the present invention can obtain a dynamic annotation tree, so that the corresponding ultrasound plaque screening image can be quickly viewed and annotated according to the dynamic annotation tree in the future, so as to improve the diagnosis efficiency.

[0116] In some embodiments, the specific implementation of step S1 (generating an ultrasound screening queue according to the ultrasound plaque screening image sent by the annotation request end, and sequentially sending the ultrasound screening set in the ultrasound screening queue to the annotation end of the annotation node in the dynamic ultrasound annotation tree for processing to obtain a plaque annotation image) includes:

[0117] S11, counting the ultrasonic plaque screening images sent by each annotation request end, and obtaining an ultrasonic screening set corresponding to each annotation request end.

[0118] It can be understood that the ultrasound plaque screening images sent by each annotation request end are counted to obtain the corresponding ultrasound screening set, so that the ultrasound screening set can be arranged in chronological order to obtain the corresponding ultrasound screening queue, which is convenient for subsequent allocation to the corresponding preliminary review personnel corresponding to the preliminary review associated node for review and annotation.

[0119] S12, determining the sending time of the ultrasonic screening set, and sorting the ultrasonic screening set in ascending order based on the sending time to obtain an ultrasonic screening queue.

[0120] It can be understood that the sending time is the time when the annotation request end sends the ultrasound plaque screening image in the ultrasound screening set to the superior request for annotation, so that the ultrasound screening set can be sorted in ascending order according to the sending time to obtain the ultrasound screening queue, and then the corresponding ultrasound screening set is allocated in turn according to the sending time to the preliminary review annotation end corresponding to each idle preliminary review associated node.

[0121] S13, obtaining an idle primary review associated node in the dynamic ultrasonic annotation tree as the current primary review node, and taking the review annotation node connected to the current primary review node as the current review node, and the final review annotation node as the current final review node.

[0122] It can be understood that in order to assign the ultrasound screening set in the ultrasound screening queue to the corresponding preliminary review annotation end for initial annotation, it is necessary to obtain the preliminary review associated node that currently has no annotation task, so as to facilitate the subsequent allocation of the corresponding ultrasound screening queue, so that the corresponding annotation request end can obtain the annotated feedback image in time for the medical staff to make a judgment.

[0123] Among them, when in the dynamic ultrasonic annotation tree, the current preliminary review node is an idle preliminary review associated node in the dynamic ultrasonic annotation tree, the current review node is a review annotation node connected to the current preliminary review node, and the current final review node is a corresponding final review annotation node.

[0124] S14, sending the ultrasound screening set in the ultrasound screening queue to the preliminary review annotation end corresponding to the current preliminary review node in sequence.

[0125] It can be understood that the ultrasound screening set in the ultrasound screening queue is sent to the preliminary review and annotation end corresponding to the current preliminary review node in sequence, so that the subsequent preliminary review and annotation end can review and annotate the corresponding ultrasound plaque annotation image in time, so as to obtain the corresponding plaque annotation image to assist medical staff in diagnosis.

[0126] S15, receiving the annotation processing of the ultrasound plaque screening image in the ultrasound screening set by the preliminary review annotation end of the current preliminary review node, and obtaining the initial annotation image corresponding to each of the preliminary review annotation ends.

[0127] It can be understood that when the preliminary review annotation end completes the annotation processing of the ultrasound plaque screening image in the ultrasound screening set, the server will obtain the initial annotation image corresponding to each preliminary review annotation end, so that the initial annotation image corresponding to the preliminary review annotation combination can be processed and compared subsequently to determine the plaque annotation image.

[0128] The initial annotated image is an image obtained after the initial annotating end annotates the ultrasound plaque screening image.

[0129] S16, taking the initial annotation image corresponding to any one of the preliminary examination annotation ends in the preliminary examination annotation combination as the first annotation image, and taking the initial annotation images corresponding to the remaining preliminary examination annotation ends in the preliminary examination annotation combination as the second annotation image.

[0130] It is understandable that the initial annotation image corresponding to any one of the preliminary review annotation ends in the preliminary review annotation combination can be used as the first annotation image, and the second annotation image can be the initial annotation images corresponding to the remaining preliminary review annotation ends in the preliminary review annotation combination.

[0131] Through the above implementation, the present invention can obtain the first annotated image and the second annotated image, so as to facilitate subsequent processing of the first annotated image and the second annotated image, thereby determining the plaque annotated image.

[0132] S17: Obtain a plaque annotated image according to the first annotated image and the second annotated image.

[0133] It can be understood that the plaque annotated image is an image in which the similarity of the annotation information in the first annotated image and the second annotated image exceeds a certain degree, and the plaque annotated image is obtained to facilitate the diagnosis of rural medical personnel.

[0134] In some embodiments, the specific implementation of step S17 (updating the second annotated contour in the second annotated image to the first annotated image to obtain a preliminary annotated image) includes:

[0135] S171: Update the second annotation contour in the second annotation image to the first annotation image to obtain a preliminary annotation image.

[0136] It is understandable that in order to ensure the accuracy of the physician's annotation of the ultrasound plaque screening image, the preliminary annotation end of the same group is asked to annotate the same ultrasound plaque screening image. Therefore, in order to facilitate the determination of the similarity between the annotated contour of the carotid artery plaque in the first annotated image and the annotated contour of the carotid artery in the second annotated image, the second annotated contour in the second annotated image can be added to the first annotated image, thereby obtaining a preliminary annotation image with two standard contours, so as to facilitate the processing and identification of the preliminary annotation image.

[0137] The second annotated contour is the contour marked by the initial review annotator in the second annotated image, and the initial review annotated image is an image with contours annotated by two initial review annotators, that is, an image with the second annotated contour updated in the first annotated image.

[0138] In some embodiments, the specific implementation of step S171 (obtaining the plaque annotated image according to the first annotated image and the second annotated image) includes:

[0139] S1711: Acquire a first center point of the first annotated image and a second center point of the second annotated image.

[0140] It should be noted that in order to update the second annotation contour in the second annotation image at the corresponding position in the first annotation image, the first center point in the first annotation image and the second center point corresponding to the second annotation image can be obtained respectively, so as to facilitate the subsequent establishment of corresponding coordinate systems, thereby updating the second annotation contour in the corresponding first annotation image, thereby facilitating contour comparison processing.

[0141] It can be understood that the first center point is the center point of the first annotated image, and the second center point is the center point of the second annotated image.

[0142] S1712: construct a first updated coordinate system based on the first center point, and construct a second updated coordinate system based on the second center point.

[0143] It can be understood that the first updated coordinate system is a coordinate system established in the first annotated image with the first center point as the origin, and the second updated coordinate system is a coordinate system established in the second annotated image with the second center point as the origin, so as to facilitate the subsequent updating of the second annotated contour at the corresponding position in the first annotated image.

[0144] S1713: coordinate the first annotated image according to the first updated coordinate system, use the annotated patch contour in the first annotated image as a first annotated contour, and count the coordinates of the first annotated contour to obtain a first contour coordinate set.

[0145] It can be understood that the first annotated image is coordinate-processed by the first updated coordinate system so as to obtain corresponding contour coordinates and to facilitate subsequent updating of the corresponding second annotated contour into the first annotated image.

[0146] The annotated plaque contour is the contour used by a person to annotate the plaque in the ultrasound plaque screening image, the first annotated contour is the annotated plaque contour in the first annotated image, and the first contour coordinate set is the coordinate set of the first annotated contour.

[0147] S1714: coordinate the second annotated image based on the second updated coordinate system, use the annotated patch contour in the second annotated image as a second annotated contour, and count the coordinates of the second annotated contour to obtain a second contour coordinate set.

[0148] It is understandable that the second annotated image is coordinate processed by the second updated coordinate system to obtain the coordinates corresponding to the second annotated contour, thereby facilitating updating the corresponding second annotated contour in the first annotated image according to the contour coordinates to obtain the preliminary annotated image.

[0149] The second contour coordinate set is a coordinate set of the second labeled contour.

[0150] S1715: Update the second annotated contour into the first annotated image according to the second contour coordinate set to obtain a preliminary annotated image.

[0151] It can be understood that the second annotated contour is updated to the first annotated image according to the second contour coordinate set to obtain a preliminary annotated image, so as to facilitate the subsequent comparison of the first contour coordinate set and the second contour coordinate set in the preliminary annotated image, so as to subsequently determine the plaque marking image.

[0152] S172, obtaining a preliminary horizontal difference value and a preliminary vertical difference value corresponding to the preliminary reviewed annotated image according to the comparison information of the first contour coordinate set and the second contour coordinate set in the preliminary reviewed annotated image.

[0153] It can be understood that the first contour coordinate set and the second contour coordinate set in the preliminary annotated image can be compared to obtain the difference between the first annotated contour and the second annotated contour in the horizontal direction, as well as the difference between the first annotated contour and the second annotated contour in the vertical direction, so as to facilitate the subsequent determination of a more accurate plaque contour and improve the diagnostic accuracy of personnel.

[0154] The preliminary horizontal difference is the horizontal distance difference between the first and second annotated contours in the preliminary annotated image, and the preliminary vertical difference is the vertical distance difference between the first and second annotated contours in the preliminary annotated image.

[0155] In some embodiments, the specific implementation of step S172 (obtaining the plaque annotated image according to the first annotated image and the second annotated image) includes:

[0156] S1721, obtaining the first minimum horizontal coordinate, the first maximum horizontal coordinate, the first minimum vertical coordinate and the first maximum vertical coordinate of the first contour coordinate set in the preliminary annotated image, and the second minimum horizontal coordinate, the second maximum horizontal coordinate, the second minimum vertical coordinate and the second maximum vertical coordinate of the second contour coordinate set.

[0157] It can be understood that the first minimum horizontal coordinate is the minimum horizontal coordinate in the first contour coordinate set, the first maximum horizontal coordinate is the maximum horizontal coordinate in the first contour coordinate set, the first minimum vertical coordinate is the minimum vertical coordinate in the first contour coordinate set, the first maximum vertical coordinate is the maximum vertical coordinate in the first contour coordinate set, the second minimum horizontal coordinate is the minimum horizontal coordinate in the second contour coordinate set, the second maximum horizontal coordinate is the maximum horizontal coordinate in the second contour coordinate set, the second minimum vertical coordinate is the minimum vertical coordinate in the second contour coordinate set, and the second maximum vertical coordinate is the maximum vertical coordinate in the second contour coordinate set.

[0158] Through the above implementation, the present invention obtains the minimum value and the maximum value in each direction in the corresponding contour coordinate set, so as to subsequently calculate the corresponding preliminary horizontal difference and preliminary vertical difference.

[0159] S1722: Obtain a left difference value based on the absolute value of the difference between the first minimum horizontal coordinate and the second minimum horizontal coordinate, and obtain a right difference value based on the absolute value of the difference between the first maximum horizontal coordinate and the second maximum horizontal coordinate.

[0160] It can be understood that the difference between the first minimum horizontal coordinate and the second minimum horizontal coordinate is calculated, so as to obtain the left difference, wherein the left difference is the absolute value of the difference between the first minimum horizontal coordinate and the second minimum horizontal coordinate, that is, the distance difference between the first marked contour and the second marked contour in the left direction, and the difference between the first maximum horizontal coordinate and the second maximum horizontal coordinate is calculated, so as to obtain the right difference, wherein the right difference is the absolute value of the difference between the first maximum horizontal coordinate and the second maximum horizontal coordinate, that is, the distance difference between the first marked contour and the second marked contour in the right direction.

[0161] S1723: Obtain a lower difference based on the absolute value of the difference between the first minimum ordinate and the second minimum ordinate, and obtain an upper difference based on the absolute value of the difference between the first maximum ordinate and the second maximum ordinate.

[0162] It can be understood that the lower side difference is the absolute value of the difference between the first minimum ordinate and the second minimum ordinate, that is, the distance difference between the first marked contour and the second marked contour in the lower side direction.

[0163] The upper difference is the absolute value of the difference between the first maximum ordinate and the second maximum ordinate, that is, the distance difference between the first marked contour and the second marked contour in the upper direction.

[0164] S1724, obtaining a preliminary horizontal difference based on the sum of the left difference and the right difference, and obtaining a preliminary vertical difference based on the lower difference and the upper difference.

[0165] It can be understood that the preliminary horizontal difference is the sum of the left difference and the right difference, and the preliminary vertical difference is the sum of the lower difference and the upper difference.

[0166] S173, when it is determined that the preliminary review transverse difference and the preliminary review longitudinal difference are both less than or equal to a preset difference, the preliminary review annotated image is used as a plaque annotated image.

[0167] It should be noted that when the preliminary horizontal difference and the preliminary vertical difference corresponding to the first annotated contour and the second annotated contour are both less than or equal to the preset difference, it can be said that the first annotated contour and the second annotated contour are extremely close, which can indicate that the contour of the annotated plaque image in the corresponding image has a high accuracy. Therefore, the preliminary annotated image can be used as the plaque annotated image.

[0168] It can be understood that the preset difference is a preset reference value for determining the difference in the horizontal direction and the vertical direction, and can be preset manually.

[0169] Through the above implementation, the present invention can obtain the corresponding plaque annotation image, so as to facilitate the subsequent fusion processing of the contours in the plaque annotation image, thereby obtaining a more accurate plaque image contour.

[0170] S174, when it is determined that the preliminary review horizontal difference or the preliminary review vertical difference is greater than the preset difference, the preliminary review annotation image is sent to the review annotation end corresponding to the current review node, and the annotation processing of the preliminary review annotation image by the review annotation end of the current review node is received to obtain the review annotation image corresponding to the review annotation end.

[0171] It can be understood that when there is a preliminary review horizontal difference or a preliminary review vertical difference greater than the preset difference, it can be said that there are large differences in the plaque contours marked by the two preliminary review labelers in the corresponding preliminary review labeling combination. Therefore, in order to ensure the accuracy of the labeling information, the corresponding preliminary review labeling image needs to be sent to the review standard end corresponding to the current review node of the superior for secondary labeling.

[0172] It is not difficult to understand that after the review annotation end of the current review node annotates the initial review annotation image, a review annotation image with review annotations can be obtained.

[0173] The reviewed annotated image is an image obtained by performing secondary plaque annotation on the initially reviewed annotated image by the review annotation end.

[0174] S175, obtaining the annotation patch contour annotated at the review annotation end in the review annotation image as the third annotation contour, determining a third contour coordinate set of the third annotation contour, and obtaining a first review lateral difference and a first review longitudinal difference based on comparison information between the third contour coordinate set and the first contour coordinate set.

[0175] It can be understood that the third annotation contour is the annotation patch contour annotated by the review annotation end in the review annotation image, and the third contour coordinate set is the coordinate set corresponding to the third annotation contour.

[0176] It is not difficult to understand that after obtaining the third contour coordinate set, it can be compared with the first contour coordinate set to determine the difference between the third annotation contour marked at the review annotation end and the first annotation contour marked at the preliminary review annotation end, thereby obtaining the first review horizontal difference and the first review vertical difference.

[0177] Among them, the first review horizontal difference is the horizontal difference between the first marked contour and the third marked contour, and the first review vertical difference is the vertical difference between the first marked contour and the third marked contour. The specific implementation method of obtaining it is consistent with the above-mentioned steps of obtaining the preliminary review horizontal difference and the preliminary review vertical difference, and will not be repeated here.

[0178] S176, obtaining a second review transverse difference and a second review longitudinal difference based on the comparison information of the third contour coordinate set and the second contour coordinate set.

[0179] Similarly, by comparing the third contour coordinate set and the second contour coordinate set, a second review horizontal difference and a second review vertical difference are obtained to determine the difference between the third annotation contour marked at the review annotation end and the second annotation contour marked at the initial review annotation end.

[0180] Among them, the second review horizontal difference is the horizontal difference between the second marked contour and the third marked contour, and the second review vertical difference is the vertical difference between the second marked contour and the third marked contour. The specific implementation method of obtaining it is consistent with the above-mentioned steps of obtaining the preliminary review horizontal difference and the preliminary review vertical difference, and will not be repeated here.

[0181] S177, when it is determined that the first review transverse difference and the first review longitudinal difference are less than or equal to a preset difference, or the second review transverse difference and the second review longitudinal difference are less than or equal to a preset difference, the review annotated image is used as a plaque annotated image.

[0182] It can be understood that when the first review lateral difference and the first review longitudinal difference are less than or equal to the preset difference, it can be said that the first annotation contour is close to the third annotation contour, and the review annotation image can be used as a plaque annotation image to facilitate the subsequent fusion of the first annotation contour and the third annotation contour.

[0183] Among them, when the second review horizontal difference and the second review vertical difference are less than or equal to the preset difference, it can be said that the second annotation contour is close to the third annotation contour, and the review annotation image can be used as a plaque annotation image to facilitate the subsequent fusion of the second annotation contour and the third annotation contour.

[0184] It should be noted that, in some embodiments, it is possible that the first review horizontal difference or the first review vertical difference is greater than the preset difference, and the second review horizontal difference or the second review vertical difference is greater than the preset difference, which means that the annotation contour after the review annotation end performs the annotation processing is not similar to any of the initial review annotation combinations. Therefore, in order to judge the accuracy of the annotation information corresponding to the review, the corresponding review annotation image can be sent to the final review annotation end corresponding to the current final review node, so that the terminal end performs the third annotation judgment. Therefore, it also includes:

[0185] B1, when it is determined that the first review horizontal difference or the first review vertical difference is greater than a preset difference, and the second review horizontal difference or the second review vertical difference is greater than a preset difference, the review annotation image is sent to the final review annotation end corresponding to the current final review node.

[0186] It can be understood that when it is determined that the first review horizontal difference or the first review vertical difference is greater than the preset difference, and the second review horizontal difference or the second review vertical difference is greater than the preset difference, the review annotation image will be sent to the final review annotation end corresponding to the current final review node, so that the terminal annotation end can perform plaque annotation.

[0187] B2, receiving the annotation processing of the review annotation image by the final review annotation end of the current final review node, and obtaining the final review annotation image corresponding to the final review annotation end.

[0188] It can be understood that after the final review annotation end has annotated the review annotation image, the server can obtain the final review annotation image corresponding to the final review annotation end, so as to facilitate the subsequent comparison of the contours in the final review annotation image, thereby determining the plaque annotation image, so as to send it to the annotation request end for diagnosis and review.

[0189] The final review annotated image is an image with the annotation information of the final review annotated end.

[0190] B3, obtaining the marked patch contour marked at the final marking end in the final marking image as the fourth marked contour, determining a fourth contour coordinate set of the fourth marked contour, and obtaining a first final horizontal difference and a first final vertical difference based on comparison information between the fourth contour coordinate set and the first contour coordinate set.

[0191] It can be understood that the fourth annotation contour is the annotation patch contour annotated by the final annotation end in the final annotation image, and the fourth contour coordinate set is the coordinate set corresponding to the fourth annotation contour.

[0192] It is not difficult to understand that the fourth contour coordinate set and the first contour coordinate set can be compared to determine the difference between the fourth annotation contour marked at the final review annotation end and the first annotation contour marked at the preliminary review annotation end, thereby obtaining the first final review horizontal difference and the first final review vertical difference.

[0193] Among them, the first final horizontal difference is the difference between the first marked contour and the fourth marked contour in the horizontal direction, and the first final vertical difference is the difference between the first marked contour and the fourth marked contour in the vertical direction. The specific implementation method of obtaining it is consistent with the steps of obtaining the preliminary horizontal difference and the preliminary vertical difference mentioned above, and will not be repeated here.

[0194] B4. Obtain a second final horizontal difference and a second final vertical difference according to the comparison information of the fourth contour coordinate set and the second contour coordinate set.

[0195] It can be understood that the fourth contour coordinate set and the second contour coordinate set can be compared to determine the difference between the fourth annotation contour marked at the final review annotation end and the second annotation contour marked at the preliminary review annotation end, thereby obtaining the second final review horizontal difference and the second final review vertical difference.

[0196] Among them, the second final horizontal difference is the difference between the second marked contour and the fourth marked contour in the horizontal direction, and the second final vertical difference is the difference between the second marked contour and the fourth marked contour in the vertical direction. The specific implementation method of obtaining it is consistent with the steps of obtaining the preliminary horizontal difference and the preliminary vertical difference mentioned above, and will not be repeated here.

[0197] B5. Obtain a third final horizontal difference and a third final vertical difference according to the comparison information of the fourth contour coordinate set and the third contour coordinate set.

[0198] It can be understood that the fourth contour coordinate set and the third contour coordinate set can be compared to determine the difference between the fourth annotation contour marked at the final annotation end and the third annotation contour marked at the review annotation end, so as to obtain the third final horizontal difference and the third final vertical difference.

[0199] Among them, the third final horizontal difference is the difference between the third marked contour and the fourth marked contour in the horizontal direction, and the third final vertical difference is the difference between the third marked contour and the fourth marked contour in the vertical direction. The specific implementation method of obtaining it is consistent with the steps of obtaining the preliminary horizontal difference and the preliminary vertical difference mentioned above, and will not be repeated here.

[0200] B6, when it is determined that the first final horizontal difference and the first final vertical difference, the second final horizontal difference and the second final vertical difference, or the third final horizontal difference and the third final vertical difference are less than or equal to a preset difference, the final annotated image is used as a plaque annotated image.

[0201] It can be understood that when the first final horizontal difference and the first final longitudinal difference are less than or equal to the preset difference, it can be said that the first marked contour is close to the fourth marked contour, and the final marked image can be used as a plaque marked image to facilitate the subsequent fusion of the first marked contour and the fourth marked contour.

[0202] Similarly, when the second final horizontal difference and the second final vertical difference are less than or equal to the preset difference, it can be said that the second annotation contour is close to the fourth annotation contour, and the final annotation image can be used as a plaque annotation image to facilitate the subsequent fusion of the second annotation contour and the fourth annotation contour. When the third final horizontal difference and the third final vertical difference are less than or equal to the preset difference, it can be said that the third annotation contour is close to the fourth annotation contour, and the final annotation image can be used as a plaque annotation image to facilitate the subsequent fusion of the third annotation contour and the fourth annotation contour.

[0203] S2, based on the plaque fusion strategy, the annotated plaque contours in the plaque annotated image are fused to obtain a fused plaque image and send it to the annotation request end.

[0204] It is understandable that in order to prevent circling deviations when manually marking plaque contours, the labeled plaque contours in the plaque annotation image can be fused to obtain accurate plate contours, which can then be sent to the annotation request end to facilitate accurate diagnosis and review by the corresponding medical staff at the annotation request end.

[0205] The fused plaque image is an image displayed after fusing the outlines of the labeled plaques.

[0206] In some embodiments, the specific implementation of step S2 (the fusion processing of the annotated plaque contours in the plaque annotated image based on the plaque fusion strategy to obtain a fused plaque image and send it to the annotation request end) includes:

[0207] S21, when it is determined that the plaque annotated image is a preliminary annotated image, two annotated plaque contours in the corresponding plaque annotated image are used as contours to be fused.

[0208] It can be understood that when the plaque annotation image is a preliminary annotation image, it can be said that the corresponding plaque annotation image has plaque contours annotated by two preliminary annotation personnel, so that the two annotated plaque contours in the plaque annotation image can be used as contours to be fused, so as to facilitate the subsequent fusion processing of the two contours to be fused, so as to obtain the corresponding fused plaque image.

[0209] S22, when it is determined that the plaque annotated image is a review annotated image, a first review difference value is obtained according to the sum of the first review lateral difference value and the first review longitudinal difference value in the corresponding plaque annotated image, and a second review difference value is obtained according to the sum of the second review lateral difference value and the second review longitudinal difference value.

[0210] It can be understood that when the plaque annotation image is a review annotation image, it can be explained that in the corresponding plaque annotation image, the two plaque contours marked by the two preliminary review annotation ends in the preliminary review annotation combination are not similar, that is, it is necessary to judge the third annotation contour marked by the review annotation end and compare it with the first annotation contour or the second annotation contour marked by the preliminary review annotation end in order to determine the corresponding contour to be fused.

[0211] Therefore, when it is determined that the plaque-annotated image is a review-annotated image, the first review difference value and the second review difference value in the corresponding plaque-annotated image can be obtained, so as to subsequently determine the corresponding contour to be fused.

[0212] Among them, the first review difference value is the difference value between the third marked contour and the first marked contour, that is, the sum of the first review horizontal difference and the first review vertical difference, and the second review difference value is the difference value between the third marked contour and the second marked contour, that is, the sum of the second review horizontal difference and the second review vertical difference.

[0213] S23, selecting the minimum value between the first review difference value and the second review difference value as the screening review difference value, and determining two annotated plaque contours in the plaque annotated image as contours to be fused based on the screening review difference value.

[0214] It can be understood that when the difference value is smaller, it means that the similarity between the two corresponding contours is higher, and the accuracy of the plaque contour displayed after the fusion processing is higher. When the difference value is larger, it means that the difference between the two corresponding contours is greater, and the fusion processing cannot be performed. Therefore, the smallest one between the first review difference value and the second review difference value can be selected as the screening review difference value to determine the annotated plaque contour to be fused in the plaque annotation image.

[0215] It is not difficult to understand that when the first review difference value is smaller than the second review difference value, the first review difference value can be used as a screening review difference value, that is, the first marked contour and the third marked contour can be determined as contours to be fused; when the first review difference value is greater than the second review difference value, the second review difference value can be used as a screening review difference value, that is, the second marked contour and the third marked contour can be determined as contours to be fused.

[0216] S24, when it is determined that the plaque annotated image is the final annotated image, a first final difference value is obtained according to the sum of the first final horizontal difference value and the first final vertical difference in the corresponding plaque annotated image, a second final difference value is obtained according to the sum of the second final horizontal difference and the second final vertical difference, and a third final difference value is obtained according to the sum of the third final horizontal difference and the third final vertical difference.

[0217] It can be understood that when the patch annotated image is the final annotated image, the first final difference value, the second final difference value and the third final difference value can be determined, so as to facilitate the subsequent determination of the similarity between the fourth annotated contour annotated at the final annotated end and the remaining three contours, thereby determining the contour to be fused with the fourth annotated contour.

[0218] Among them, the first final difference value is the difference value between the fourth marked contour marked by the final marking end and the first marked contour, that is, the sum of the first final horizontal difference and the first final vertical difference; the second final difference value is the difference value between the fourth marked contour marked by the final marking end and the second marked contour, that is, the sum of the second final horizontal difference and the second final vertical difference; the third final difference value is the difference value between the fourth marked contour marked by the final marking end and the third marked contour, that is, the sum of the third final horizontal difference and the third final vertical difference.

[0219] S25, selecting the minimum value among the first final difference value, the second final difference value and the third final difference value as the screening final difference value, and determining two annotated plaque contours in the plaque annotated image as contours to be fused based on the screening final difference value.

[0220] It can be understood that the final difference value is screened as the minimum value among the first final difference value, the second final difference value and the third final difference value, so as to determine the contour to be fused.

[0221] Among them, the method of determining the contour to be fused according to the final screening difference value is consistent with the method of determining the contour to be fused according to the above screening review difference value, and will not be repeated here.

[0222] S26, obtaining the positional relationship of the contours to be fused, where the positional relationship includes a containment relationship and a non-containment relationship.

[0223] It is understandable that, according to the positional relationship of the contours to be fused, different strategies may be adopted to perform fusion processing on the contours to be fused, so as to improve the fusion speed and accuracy of the contours to be fused.

[0224] Among them, the positional relationship is a containment relationship and a non-containment relationship. The containment relationship is that one contour contains another contour, and the non-containment relationship is that the positions of two contours are not a relationship in which one contour contains another contour.

[0225] Therefore, the corresponding fusion strategy can be determined according to the positional relationship of the contours to be fused, so that the contour fusion can be performed quickly.

[0226] S27, when it is determined that the positional relationship is a containment relationship, a fusion process is performed on the contours to be fused in the patch annotated image according to a nested fusion strategy, and a fused patch image is obtained and sent to the annotation request end.

[0227] It is understandable that when the positional relationship is an inclusion relationship, the contours to be fused in the patch annotation image can be fused according to the nested fusion strategy to obtain a fused patch image so as to be sent to the annotation request end, thereby facilitating the diagnosis and viewing of the annotation request end.

[0228] In some embodiments, the specific implementation of step S27 (when determining that the positional relationship is a containment relationship, fusing the contours to be fused in the patch annotated image according to the nested fusion strategy to obtain a fused patch image and send it to the annotation request end) includes:

[0229] S271 , when it is determined that the positional relationship is a containing relationship, the contour to be fused located on the outside is obtained as the containing contour, and the contour to be fused located on the inside is obtained as the contained contour.

[0230] It can be understood that when the positional relationship is determined to be an inclusion relationship, the contour to be fused located on the outside can be taken as the including contour, and the contour to be fused located on the inside can be taken as the included contour, so as to facilitate the subsequent determination of the corresponding contour coordinates, thereby calculating the coordinate set of the fused contour, and thus determining the fused contour.

[0231] S272, obtaining the center point of the containing contour as the positioning center point, constructing a positioning straight line based on a preset angle and the positioning center point, and determining the outer intersection point of the positioning straight line with the containing contour, and the inner intersection point of the positioning straight line with the contained contour.

[0232] It can be understood that in order to make the subsequently determined contour fusion coordinates more accurate, the corresponding fused contour is closer to the actual patch image contour. Therefore, multiple positioning straight lines can be constructed based on the positioning center point, and there is a preset angle between each positioning straight line, so that an intersection can be generated with the containing contour and the included contour, thereby facilitating the subsequent acquisition of the corresponding contour fusion coordinates.

[0233] Among them, the positioning center point is the center point of the included contour. The center horizontal coordinate can be obtained by averaging the minimum horizontal coordinate and the maximum horizontal coordinate of the included contour, and the center vertical coordinate can be obtained by averaging the minimum vertical coordinate and the maximum vertical coordinate of the included contour, so as to obtain the corresponding positioning center point.

[0234] It is not difficult to understand that the positioning straight line is a straight line that passes through the positioning center point and intersects the containing contour and the contained contour, and the angle between adjacent positioning straight lines is a preset angle, wherein the preset angle is a preset angle value, which can be 90°, 45°, etc. The outer intersection point is the intersection point of the positioning straight line and the containing contour, and the inner intersection point is the intersection point of the positioning straight line and the contained contour.

[0235] For example: Figure 3 As shown, when there are two positioning straight lines, they have 4 outer intersection points with the containing contour and 4 inner intersection points with the contained contour, so that when they are located on the same positioning straight line and on the same side of the positioning center point, the contour fusion coordinates can be determined to obtain the fused intermediate contour.

[0236] S273, calculating according to the outer coordinates of the adjacent outer intersection points and the inner coordinates of the inner intersection points on the same side of the positioning straight line to obtain the contour fusion coordinates.

[0237] It can be understood that in order to calculate the contour coordinates between the containing contour and the included contour, the outer coordinates and inner coordinates corresponding to the adjacent outer intersection points and inner intersection points on the same side of the positioning straight line can be used for calculation, so as to obtain the contour fusion coordinates, so as to facilitate the subsequent determination of the corresponding fused contour based on the contour fusion coordinates.

[0238] Among them, the outer coordinates are the coordinates corresponding to the outer intersection points, the inner coordinates are the coordinates corresponding to the inner intersection points, and the contour fusion coordinates are the coordinates of the fused contour.

[0239] S274, sequentially connecting the adjacent contour fusion coordinates to obtain a fused intermediate contour, and deleting the annotated plaque contour in the plaque annotated image to obtain a fused plaque image and send it to the annotation request end.

[0240] It can be understood that the fused middle contour is the contour located between the two contours to be fused, that is, Figure 3 As shown, the contours after connecting the adjacent contour fusion coordinates are connected in sequence. In order to prevent the labeled plaque contour from misleading the medical staff at the labeling request end, the labeled plaque contour in the plaque labeling image is deleted, and then the obtained fused plaque image is sent to the labeling request end so that the corresponding medical staff can view it intuitively.

[0241] S28, when it is determined that the positional relationship is a non-inclusion relationship, the contours to be fused in the patch annotated image are fused according to an overlapping fusion strategy to obtain a fused patch image and send it to the annotation request end.

[0242] It can be understood that when it is determined that the positional relationship between two contours to be fused is a non-inclusion relationship, the contours to be fused in the patch annotation image can be fused according to the overlapping fusion strategy, so that the obtained fused patch image is sent to the annotation request end for easy viewing by personnel.

[0243] In some embodiments, the specific implementation of step S28 (when determining that the positional relationship is a non-inclusion relationship, fusing the contour to be fused in the patch annotated image according to the overlapping fusion strategy, obtaining a fused patch image and sending it to the annotation request end) includes:

[0244] S281, when it is determined that the positional relationship is a non-inclusion relationship, the intersection points of the two contours to be fused are obtained, the intersection points are connected to obtain an intersection connection line, and the midpoint of the intersection connection line is obtained as the positioning midpoint.

[0245] It can be understood that when two contours to be fused are not in an inclusion relationship, the intersection points of the contours to be fused can be connected to obtain an intersection connection line, and then the midpoint of the intersection connection line is obtained as the positioning midpoint, so as to facilitate the subsequent construction of the corresponding overlapping positioning line, thereby determining the corresponding fused contour.

[0246] The intersection connection line is the intersection connection line connecting the two contours to be fused, and the positioning midpoint is the midpoint of the intersection connection line.

[0247] S282, taking any one of the intersection points as a first contour intersection point, and taking another intersection point as a second contour intersection point.

[0248] It can be understood that, in order to facilitate the subsequent determination of corresponding contour coordinates, one intersection point can be used as the first contour intersection point, and the other intersection point can be used as the second contour intersection point.

[0249] S283: taking any one of the contours to be fused as a first intersecting contour, and taking another contour to be fused as a second intersecting contour.

[0250] It can be understood that, in order to facilitate the subsequent determination of the corresponding fusion contour, any one of the contours to be fused can be used as the first intersection contour, and another contour to be fused can be used as the second intersection contour.

[0251] S284, constructing an overlapping positioning line perpendicular to the intersection connection line based on the positioning midpoint, and determining a first positioning intersection point between the overlapping positioning line and the first intersecting contour, and a second positioning intersection point between the overlapping positioning line and the second intersecting contour.

[0252] It can be understood that the overlapping positioning line is a line segment passing through the positioning midpoint and perpendicular to the intersection connection line, the first positioning intersection is the intersection of the overlapping positioning line and the first intersecting contour, and the second positioning intersection is the intersection of the overlapping positioning line and the second intersecting contour.

[0253] Through the above implementation, the present invention can obtain the first positioning intersection point and the second positioning intersection point, so as to facilitate the subsequent determination of the center positioning coordinates and further determine the fusion contour.

[0254] S285, performing mean calculation based on the first coordinates of the first positioning intersection point and the second coordinates of the second positioning intersection point adjacent to each other on the same side of the overlapping positioning line to obtain the center positioning coordinates corresponding to the center positioning point.

[0255] It can be understood that since the overlapping positioning line and the intersection connection line are perpendicular to each other, a coordinate system can be constructed with the positioning midpoint as the origin, and then the first coordinates of the first positioning intersection and the second coordinates of the second positioning intersection adjacent to each other on the same side of the overlapping positioning line can be averaged to obtain the center positioning coordinates corresponding to the center positioning point.

[0256] Among them, the first coordinate is the coordinate of the first positioning intersection, the second coordinate is the coordinate of the second positioning intersection, the center positioning point is the center point between the first positioning intersection and the second positioning intersection on the same side of the overlapping positioning line, and the center positioning coordinate is the coordinate of the center positioning point.

[0257] S286, connecting the center positioning points corresponding to the two center positioning coordinates to obtain a fused long axis.

[0258] It can be understood that the fusion long axis is a line connecting two central positioning points.

[0259] Through the above implementation, the present invention can obtain the fused major axis, so as to facilitate the subsequent determination of the fused ellipse contours corresponding to different intersection points.

[0260] S287, performing mean calculation on the center positioning coordinates to obtain fused midpoint coordinates, and performing calculation based on the fused midpoint coordinates and the coordinates of the first contour intersection point to obtain a first fused semi-minor axis distance.

[0261] It can be understood that the fusion midpoint coordinate is the midpoint of the fusion major axis, that is, the midpoint of the two central positioning points, so that the first fusion semi-minor axis distance can be calculated with the intersection of the first contour, wherein the first fusion semi-minor axis distance is the semi-minor axis distance of the fused ellipse contour corresponding to the intersection of the first contour.

[0262] Through the above implementation, the present invention can obtain the first fusion semi-minor axis distance, so as to facilitate the subsequent determination of the corresponding ellipse contour.

[0263] S288: Calculate based on the coordinates of the fusion midpoint and the coordinates of the second contour intersection point to obtain a second fusion semi-minor axis distance.

[0264] It can be understood that the calculation is performed based on the coordinates of the fusion midpoint and the coordinates of the intersection of the second contour, that is, the difference between the horizontal coordinates and the vertical coordinates of the two corresponding points is calculated respectively, so as to calculate the second fusion semi-minor axis distance according to the Pythagorean theorem.

[0265] Through the above implementation, the present invention can obtain the second fused semi-minor axis distance, so as to subsequently determine the ellipse contour corresponding to the class minor axis distance.

[0266] S289, generating a special-shaped fused ellipse contour according to the fused midpoint coordinates, the first fused semi-minor axis distance, the second fused semi-minor axis distance and the fused major axis, and deleting the annotated plaque contour in the plaque annotated image to obtain a fused plaque image and send it to the annotation request end.

[0267] It can be understood that the fusion midpoint coordinates, the first fusion semi-minor axis distance, the second fusion semi-minor axis distance and the fusion major axis are obtained, so that two semi-ellipses with inconsistent minor axis distances can be generated according to the ellipse construction method, and the irregular fused ellipse contour can be obtained, so that the annotated plaque contour in the plaque annotated image can be deleted, and the fused plaque image can be obtained and sent to the annotation request end.

[0268] Among them, the irregular fused ellipse contour is the contour displayed by fusing the contours to be fused that are not in a containment relationship, so as to annotate and display the patch image for easy viewing by personnel.

[0269] S3, determining the basic number of ultrasound plaque screening images received by the dynamic annotation combination in the dynamic ultrasound annotation tree within a preset rotation period, and retrieving the fused number of fused plaque images corresponding to the dynamic annotation combination within the preset rotation period.

[0270] It should be noted that in order to calculate the annotation coordination of the corresponding personnel for each annotation combination, the basic number of plaque screening images and the fused number of fused plaque images received by the dynamic annotation combination can be calculated. When the ratio of the generated fusion number to the basic number is larger, it can be explained that the corresponding dynamic annotation combination has a higher degree of coordination for annotation, so as to facilitate the subsequent determination of annotation combinations with higher coordination, thereby improving the annotation efficiency.

[0271] It can be understood that the dynamic annotation combination is a combination of annotation end nodes that annotate the same patch screening map, for example, a dynamic annotation combination consisting of personnel A and personnel B corresponding to the preliminary review annotation end, personnel 1 at the review annotation end, and final review annotation end A.

[0272] Among them, the basic number is the number of ultrasound plaque screening images received by the dynamic annotation combination, and the fusion number is the number of fused plaque images generated by comparison of the corresponding dynamic annotation combination.

[0273] It is not difficult to understand that, when the number of fused patch images generated is greater, it can be explained that the number of images with contours of two labeled ends for patch annotation in the corresponding combination is greater, which can be explained that the degree of coordination of the corresponding two labeled ends is also higher.

[0274] Before the step of determining the basic number of ultrasound plaque screening images received by the dynamic annotation combination in the dynamic ultrasound annotation tree within the preset rotation period, and calling the fused number of fused plaque images corresponding to the dynamic annotation combination within the preset rotation period, the method includes:

[0275] C1, obtaining the preliminary review annotation combination corresponding to the preliminary review associated node in the dynamic ultrasound annotation tree within the preset rotation period.

[0276] It can be understood that in order to determine the annotation ends corresponding to each dynamic annotation combination at different nodes, so that they can be obtained from the preliminary review associated nodes, the preliminary review annotation end, the review annotation end and the final review annotation end are determined in turn, so as to obtain the corresponding dynamic annotation combination subsequently.

[0277] For example: when the corresponding preliminary review association node corresponds to person A, the corresponding preliminary review annotation combination of person A and person B can be retrieved.

[0278] C2, determining the review annotation end corresponding to the review annotation node connected to the corresponding preliminary review associated node in the dynamic ultrasonic annotation tree, and the final review annotation end corresponding to the final review annotation node connected to the corresponding review annotation node.

[0279] It can be understood that after the preliminary review associated node in the dynamic ultrasonic annotation tree is determined, the review annotation end corresponding to the review annotation node connected to the corresponding preliminary review associated node and the corresponding final review annotation end are determined. For example, when the review annotation node connected to the preliminary review associated node corresponds to person 1, and the final review annotation end is person A.

[0280] C3, adding the review annotation end to the preliminary review annotation combination to obtain a review annotation combination, and adding the final review annotation end to the review annotation combination to obtain a final review annotation combination.

[0281] It can be understood that the review annotation combination is a combination of the preliminary review annotation combination with the review annotation end added. For example, when the preliminary review annotation combination is personnel A and personnel B, and the review annotation end is personnel 1, then personnel 1 can be added to the preliminary review annotation combination, and the review annotation combination is personnel A, personnel B and personnel 1. The final review annotation combination is a combination of the review annotation combination with the final review annotation end added. For example, when the review annotation combination is personnel A, personnel B and personnel 1, and the final review annotation end is personnel A, then the final review annotation combination is personnel A, personnel B, personnel 1 and personnel A.

[0282] C4, obtaining a dynamic annotation combination according to the initial review annotation combination, the review annotation combination and the final review annotation combination within the preset rotation period.

[0283] It can be understood that by presetting the initial review annotation combination, the review annotation combination and the final review annotation combination within the rotation period, multiple different dynamic annotation combinations corresponding to each time period can be obtained, so as to facilitate the subsequent calculation of the corresponding annotation end score value.

[0284] It is not difficult to understand that the dynamic annotation combination includes the preliminary review annotation combination, the review annotation combination and the final review annotation combination.

[0285] S4, performing calculation according to the fusion quantity and the basic quantity to obtain a score value corresponding to each annotation end.

[0286] It should be noted that in order to determine the annotation situation corresponding to each annotation end, it is necessary to calculate the score value corresponding to each annotation end and evaluate the accuracy of the annotation of each annotation end according to the score value.

[0287] It is understandable that the obtained fusion quantity and basic quantity are calculated, so as to determine the score value corresponding to each annotation end.

[0288] The score value is a numerical value for evaluating the annotation at the annotation end.

[0289] In some embodiments, the specific implementation of step S4 (the calculation based on the fusion quantity and the basic quantity to obtain the score value corresponding to each annotation end) includes:

[0290] The score corresponding to each annotation end is obtained by the following formula:

[0291] in, is the score value corresponding to each annotation end, is the number of fusions of the initial review annotation combination corresponding to the annotation end, is the basic number of initial review annotation combinations corresponding to the annotation end. is the initial review weight, is the number of fusions of the corresponding review annotation combination on the annotation side, is the basic number of review annotation combinations corresponding to the annotation end, For review weight, is the number of fusions of the final annotation combination corresponding to the annotation end, is the basic number of final review annotation combinations corresponding to the annotation end. The final weight.

[0292] It can be understood that by comparing the fusion quantity of the corresponding annotation end in different combinations with the basic quantity, and then multiplying by the corresponding weight value and adding them together, the score value of each annotation end in the corresponding rotation time period can be obtained.

[0293] Among them, when the annotation end corresponds to the number of fusions of the initial annotation combination The larger the value, the corresponding The larger the value of, the higher the score of the annotation end in the preliminary annotation combination. The larger the The value of will also become larger, which means that the higher the similarity between the annotation end and the review annotation personnel in the review annotation combination, the larger the corresponding score value will be. The larger the value, the corresponding The value of will also become larger, which means that the higher the similarity between the annotation end and the final review annotation personnel in the final review annotation combination, the higher the accuracy of the plaque annotation by the corresponding preliminary review annotation personnel, so it can be multiplied with the corresponding weight value to obtain the sum and get the score value of the corresponding annotation end.

[0294] It is not difficult to understand that the initial review weight is the weight ratio in the initial review annotation combination, the review weight is the weight ratio in the review annotation combination, and the final review weight is the weight ratio in the final review annotation combination. Since the fused image obtained in the initial review annotation combination is a comparison of the annotation contours between the initial review annotation end and the other initial review annotation end, and the fused image obtained in the review annotation combination is a comparison of the annotation contours between the initial review annotation end and the review annotation end, and the fused image obtained in the final review annotation combination is a comparison of the annotation contours between the initial review annotation end and the final review annotation end, due to the different diagnostic levels of the corresponding other annotation end, the reference of the corresponding fused image is also different. Therefore, the corresponding weight values ​​​​can be artificially determined for different combinations in order to accurately judge the corresponding score values ​​of each annotation end.

[0295] It is worth mentioning that due to the inconsistency of the diagnostic levels of the doctors corresponding to different annotation ends, the determination of the fused image may be affected. Therefore, according to the corresponding score values ​​of each annotation end in multiple preset rotation time periods, another annotation end that makes the score value of each annotation end the largest can be selected for combination to improve the efficiency of plaque annotation.

[0296] See also Figure 4 , is a schematic diagram of the structure of a carotid artery plaque identification device provided by an embodiment of the present invention, the carotid artery plaque identification device comprising:

[0297] The sending module is used to generate an ultrasound screening queue according to the ultrasound plaque screening image sent by the annotation request end, and sequentially send the ultrasound screening set in the ultrasound screening queue to the annotation end of the annotation node in the dynamic ultrasound annotation tree for processing to obtain a plaque annotation image.

[0298] The fusion module is used to fuse the annotated plaque contours in the plaque annotated image based on the plaque fusion strategy, obtain a fused plaque image and send it to the annotation request end.

[0299] The determination module is used to determine the basic number of ultrasound plaque screening images received by the dynamic annotation combination in the dynamic ultrasound annotation tree within a preset rotation period, and to retrieve the fused number of fused plaque images corresponding to the dynamic annotation combination within the preset rotation period.

[0300] The calculation module is used to calculate according to the fusion quantity and the basic quantity to obtain the score value corresponding to each annotation end.

[0301] See also Figure 5 , is a schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present invention, the electronic device 50 includes: a processor 51, a memory 52 and a computer program; wherein

[0302] The memory 52 is used to store the computer program, which may also be a flash memory. The computer program is, for example, an application program, a functional module, etc. for implementing the above method.

[0303] The processor 51 is used to execute the computer program stored in the memory to implement each step performed by the device in the above method. For details, please refer to the relevant description in the above method embodiment.

[0304] Optionally, the memory 52 may be independent or integrated with the processor 51 .

[0305] When the memory 52 is a device independent of the processor 51, the device may further include:

[0306] The bus 53 is used to connect the memory 52 and the processor 51 .

[0307] The present invention also provides a readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, it is used to implement the methods provided by the various embodiments described above.

[0308] Among them, the readable storage medium can be a computer storage medium or a communication medium. The communication medium includes any medium that facilitates the transmission of a computer program from one place to another. The computer storage medium can be any available medium that can be accessed by a general-purpose or special-purpose computer. For example, a readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application-specific integrated circuit (Application Specific Integrated Circuits, referred to as: ASIC). In addition, the ASIC can be located in a user device. Of course, the processor and the readable storage medium can also exist in a communication device as discrete components. The readable storage medium can be a read-only memory (ROM), a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0309] The present invention also provides a program product, which includes an execution instruction, which is stored in a readable storage medium. At least one processor of a device can read the execution instruction from the readable storage medium, and at least one processor executes the execution instruction so that the device implements the methods provided in the above various embodiments.

[0310] In the embodiments of the above-mentioned devices, it should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the present invention may be directly implemented as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor.

[0311] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying carotid artery plaque, characterized in that: include: Generate an ultrasound screening queue according to the ultrasound plaque screening image sent by the annotation request end, and send the ultrasound screening set in the ultrasound screening queue to the annotation end of the annotation node in the dynamic ultrasound annotation tree for processing in sequence to obtain a plaque annotation image; Based on the plaque fusion strategy, the annotated plaque contours in the plaque annotated image are fused to obtain a fused plaque image and send it to the annotation request end; Determine the basic number of ultrasound plaque screening images received by the dynamic annotation combination in the dynamic ultrasound annotation tree within a preset rotation period, and retrieve the fused number of fused plaque images corresponding to the dynamic annotation combination within the preset rotation period; Calculate the score corresponding to each annotation end according to the fusion quantity and the basic quantity; The following steps are used to construct a dynamic ultrasound annotation tree, including: Counting all preliminary review annotation ends to obtain a preliminary review annotation set, selecting any two preliminary review annotation ends in the preliminary review annotation set based on a preset rotation period to combine them to obtain multiple preliminary review annotation combinations, and constructing preliminary review association nodes based on the preliminary review annotation ends in the preliminary review annotation combinations; Select any review annotation terminal according to the preset rotation period to construct a review annotation node, and connect the initial review association node corresponding to the initial review annotation combination with the review annotation node; Constructing a final review annotation node based on the final review annotation end, connecting the review annotation node with the final review annotation node to generate a dynamic ultrasonic annotation tree; The method generates an ultrasound screening queue according to the ultrasound plaque screening image sent by the annotation request end, and sequentially sends the ultrasound screening set in the ultrasound screening queue to the annotation end of the annotation node in the dynamic ultrasound annotation tree for processing to obtain a plaque annotation image, including: Counting the ultrasonic plaque screening images sent by each annotation request end, and obtaining the ultrasonic screening set corresponding to each annotation request end; Determine the sending time of the ultrasonic screening set, and sort the ultrasonic screening set in ascending order based on the sending time to obtain an ultrasonic screening queue; Acquire an idle primary review associated node in the dynamic ultrasonic annotation tree as the current primary review node, and take the review annotation node connected to the current primary review node as the current review node, and the final review annotation node as the current final review node; Sending the ultrasound screening set in the ultrasound screening queue to the preliminary review annotation end corresponding to the current preliminary review node in sequence; Receiving the annotation processing of the ultrasound plaque screening image in the ultrasound screening set by the preliminary review annotation end of the current preliminary review node, and obtaining the initial annotation image corresponding to each of the preliminary review annotation ends; Using the initial annotation image corresponding to any one of the preliminary review annotation ends in the preliminary review annotation combination as the first annotation image, and using the initial annotation images corresponding to the remaining preliminary review annotation ends in the preliminary review annotation combination as the second annotation image; Obtaining a plaque annotated image according to the first annotated image and the second annotated image; The step of obtaining a plaque annotated image according to the first annotated image and the second annotated image includes: Updating the second annotation contour in the second annotation image to the first annotation image to obtain a preliminary annotation image; Obtaining a preliminary horizontal difference value and a preliminary vertical difference value corresponding to the preliminary annotated image according to comparison information of the first contour coordinate set and the second contour coordinate set in the preliminary annotated image; When it is determined that both the preliminary review transverse difference and the preliminary review longitudinal difference are less than or equal to the preset difference, the preliminary review annotated image is used as a plaque annotated image; When it is determined that the preliminary review horizontal difference or the preliminary review vertical difference is greater than a preset difference, the preliminary review annotated image is sent to the review annotated end corresponding to the current review node, and the review annotated end of the current review node receives the annotation processing of the preliminary review annotated image, so as to obtain the review annotated image corresponding to the review annotated end; Acquire the contour of the marked patch marked at the review mark end in the review mark image as a third marked contour, determine a third contour coordinate set of the third marked contour, and obtain a first review transverse difference and a first review longitudinal difference according to comparison information of the third contour coordinate set and the first contour coordinate set; Based on the comparison information of the third contour coordinate set and the second contour coordinate set, obtaining a second review transverse difference and a second review longitudinal difference; When it is determined that the first review transverse difference and the first review longitudinal difference are less than or equal to a preset difference, or the second review transverse difference and the second review longitudinal difference are less than or equal to a preset difference, the review annotated image is used as a plaque annotated image; Also includes: When it is determined that the first review horizontal difference or the first review vertical difference is greater than a preset difference, and the second review horizontal difference or the second review vertical difference is greater than a preset difference, sending the review annotated image to the final review annotated end corresponding to the current final review node; Receiving the annotation processing of the review annotation image by the final review annotation end of the current final review node, and obtaining the final review annotation image corresponding to the final review annotation end; Acquire the contour of the marked patch marked at the final marking end in the final marking image as a fourth marked contour, determine a fourth contour coordinate set of the fourth marked contour, and obtain a first final horizontal difference and a first final vertical difference according to comparison information of the fourth contour coordinate set and the first contour coordinate set; Obtaining a second final horizontal difference and a second final vertical difference according to comparison information of the fourth contour coordinate set and the second contour coordinate set; Obtaining a third final horizontal difference and a third final vertical difference according to comparison information of the fourth contour coordinate set and the third contour coordinate set; When it is determined that the first final horizontal difference and the first final vertical difference, the second final horizontal difference and the second final vertical difference, or the third final horizontal difference and the third final vertical difference are less than or equal to a preset difference, the final annotated image is used as a plaque annotated image; The method of fusing the annotated plaque contours in the plaque annotated image based on the plaque fusion strategy to obtain a fused plaque image and sending it to the annotation request end includes: When it is determined that the plaque annotated image is a preliminary annotated image, two annotated plaque contours in the corresponding plaque annotated image are used as contours to be fused; When the plaque annotated image is determined to be a review annotated image, a first review difference value is obtained according to the sum of the first review transverse difference value and the first review longitudinal difference value in the corresponding plaque annotated image, and a second review difference value is obtained according to the sum of the second review transverse difference value and the second review longitudinal difference value; Selecting the minimum value of the first review difference value and the second review difference value as the screening review difference value, and determining two annotated plaque contours in the plaque annotated image as contours to be fused based on the screening review difference value; When the plaque-annotated image is determined to be a final-reviewed annotated image, a first final-reviewed difference value is obtained according to the sum of the first final-reviewed transverse difference value and the first final-reviewed longitudinal difference value in the corresponding plaque-annotated image, a second final-reviewed difference value is obtained according to the sum of the second final-reviewed transverse difference value and the second final-reviewed longitudinal difference value, and a third final-reviewed difference value is obtained according to the sum of the third final-reviewed transverse difference value and the third final-reviewed longitudinal difference value; Selecting the minimum value among the first final difference value, the second final difference value and the third final difference value as the screening final difference value, and determining two annotated plaque contours in the plaque annotated image as contours to be fused based on the screening final difference value; Acquire the positional relationship of the contours to be fused, where the positional relationship includes a containment relationship and a non-containment relationship; When it is determined that the positional relationship is a containment relationship, the contours to be fused in the patch annotation image are fused according to a nested fusion strategy to obtain a fused patch image and send it to the annotation request end; When it is determined that the positional relationship is a non-inclusion relationship, the contour to be fused in the patch annotation image is fused according to an overlapping fusion strategy to obtain a fused patch image and send it to the annotation request end; Before the step of determining the basic number of ultrasound plaque screening images received by the dynamic annotation combination in the dynamic ultrasound annotation tree within the preset rotation period, and calling the fused number of fused plaque images corresponding to the dynamic annotation combination within the preset rotation period, the method includes: Obtaining a preliminary review annotation combination corresponding to a preliminary review associated node in the dynamic ultrasonic annotation tree within the preset rotation period; Determine a review annotation end corresponding to a review annotation node connected to the corresponding preliminary review associated node in the dynamic ultrasonic annotation tree, and a final review annotation end corresponding to a final review annotation node connected to the corresponding review annotation node; Adding the review annotation end to the preliminary review annotation combination to obtain a review annotation combination, and adding the final review annotation end to the review annotation combination to obtain a final review annotation combination; Obtaining a dynamic annotation combination according to the initial review annotation combination, the review annotation combination and the final review annotation combination within the preset rotation period; The calculation is performed according to the fusion quantity and the basic quantity to obtain the score value corresponding to each annotation end, including: The score corresponding to each annotation end is obtained by the following formula: in, is the score value corresponding to each annotation end, is the number of fusions of the initial review annotation combination corresponding to the annotation end, is the basic number of initial review annotation combinations corresponding to the annotation end. is the initial review weight, is the number of fusions of the corresponding review annotation combination on the annotation side, is the basic number of review annotation combinations corresponding to the annotation end, For review weight, is the number of fusions of the final annotation combination corresponding to the annotation end, is the basic number of final review annotation combinations corresponding to the annotation end. The final weight.

2. The method according to claim 1, characterized in that The updating of the second annotation contour in the second annotation image to the first annotation image to obtain a preliminary annotation image includes: Acquire a first center point of the first annotated image and a second center point of the second annotated image; Constructing a first updated coordinate system based on the first center point, and constructing a second updated coordinate system based on the second center point; Performing coordinate processing on the first annotated image according to the first updated coordinate system, taking the annotated patch contour in the first annotated image as a first annotated contour, and counting the coordinates of the first annotated contour to obtain a first contour coordinate set; Performing coordinate processing on the second annotated image based on the second updated coordinate system, taking the annotated patch contour in the second annotated image as a second annotated contour, and obtaining a second contour coordinate set by counting the coordinates of the second annotated contour; The second annotated contour is updated into the first annotated image according to the second contour coordinate set to obtain a preliminary annotated image.

3. The method according to claim 1, characterized in that The obtaining, according to the comparison information of the first contour coordinate set and the second contour coordinate set in the preliminary review annotated image, a preliminary review horizontal difference value and a preliminary review vertical difference value corresponding to the preliminary review annotated image comprises: Acquire a first minimum abscissa, a first maximum abscissa, a first minimum ordinate and a first maximum ordinate of a first contour coordinate set in the preliminary annotated image, and a second minimum abscissa, a second maximum abscissa, a second minimum ordinate and a second maximum ordinate of a second contour coordinate set; Obtain a left difference value based on an absolute value of a difference between the first minimum horizontal coordinate and the second minimum horizontal coordinate, and obtain a right difference value based on an absolute value of a difference between the first maximum horizontal coordinate and the second maximum horizontal coordinate; Based on the absolute value of the difference between the first minimum ordinate and the second minimum ordinate, a lower difference is obtained, and based on the absolute value of the difference between the first maximum ordinate and the second maximum ordinate, an upper difference is obtained; A preliminary lateral difference is obtained based on the sum of the left side difference and the right side difference, and a preliminary longitudinal difference is obtained based on the sum of the lower side difference and the upper side difference.

4. The method according to claim 1, characterized in that When it is determined that the positional relationship is a containment relationship, the contour to be fused in the patch annotated image is fused according to a nested fusion strategy to obtain a fused patch image and send it to the annotation request end, including: When it is determined that the positional relationship is a containing relationship, the contour to be fused located on the outside is obtained as the containing contour, and the contour to be fused located on the inside is obtained as the contained contour; Acquire the center point of the included contour as the positioning center point, construct a positioning straight line based on a preset angle and the positioning center point, and determine the outer intersection point of the positioning straight line with the included contour, and the inner intersection point of the positioning straight line with the included contour; Calculate the contour fusion coordinates based on the outer coordinates of the adjacent outer intersection points and the inner coordinates of the inner intersection points on the same side of the positioning straight line; The adjacent contour fusion coordinates are sequentially connected to each other to obtain a fused intermediate contour, and the annotated plaque contour in the plaque annotated image is deleted to obtain a fused plaque image and send it to the annotation request end; When it is determined that the positional relationship is a non-inclusion relationship, the contour to be fused in the patch annotated image is fused according to an overlapping fusion strategy to obtain a fused patch image and send it to the annotation request end, including: When it is determined that the positional relationship is a non-inclusion relationship, the intersection points of the two contours to be fused are obtained, the intersection points are connected to obtain an intersection connection line, and the midpoint of the intersection connection line is obtained as the positioning midpoint; Take any one intersection point as the first contour intersection point, and take the other intersection point as the second contour intersection point; Taking any one of the contours to be fused as the first intersecting contour, and taking another contour to be fused as the second intersecting contour; Constructing an overlapping positioning line perpendicular to the intersection connection line based on the positioning midpoint, determining a first positioning intersection point between the overlapping positioning line and the first intersection contour, and a second positioning intersection point between the overlapping positioning line and the second intersection contour; The center positioning coordinates corresponding to the center positioning point are obtained by performing mean calculation based on the first coordinates of the first positioning intersection point and the second coordinates of the second positioning intersection point adjacent to each other on the same side of the overlapping positioning line; Connecting the center positioning points corresponding to the two center positioning coordinates to obtain a fused long axis; Performing mean calculation on the center positioning coordinates to obtain fusion midpoint coordinates, and performing calculation based on the fusion midpoint coordinates and the coordinates of the first contour intersection point to obtain a first fusion semi-minor axis distance; Calculating based on the coordinates of the fusion midpoint and the coordinates of the second contour intersection point to obtain a second fusion semi-minor axis distance; A special-shaped fused ellipse contour is generated according to the fused midpoint coordinates, the first fused semi-minor axis distance, the second fused semi-minor axis distance and the fused major axis, and the annotated plaque contour in the plaque annotated image is deleted to obtain a fused plaque image and send it to the annotation request end.

5. A carotid artery plaque identification device, characterized in that: include: A sending module is used to generate an ultrasound screening queue according to the ultrasound plaque screening image sent by the annotation request end, and sequentially send the ultrasound screening set in the ultrasound screening queue to the annotation end of the annotation node in the dynamic ultrasound annotation tree for processing to obtain a plaque annotation image; A fusion module is used to fuse the annotated plaque contours in the plaque annotated image based on a plaque fusion strategy, and obtain a fused plaque image to send to the annotation request end; A determination module, used to determine the basic number of ultrasound plaque screening images received by the dynamic annotation combination in the dynamic ultrasound annotation tree within a preset rotation period, and to retrieve the fused number of fused plaque images corresponding to the dynamic annotation combination within the preset rotation period; A calculation module, used to calculate according to the fusion quantity and the basic quantity to obtain a score value corresponding to each annotation end; The following steps are used to construct a dynamic ultrasound annotation tree, including: Counting all preliminary review annotation ends to obtain a preliminary review annotation set, selecting any two preliminary review annotation ends in the preliminary review annotation set based on a preset rotation period to combine them to obtain multiple preliminary review annotation combinations, and constructing preliminary review association nodes based on the preliminary review annotation ends in the preliminary review annotation combinations; Select any review annotation terminal according to the preset rotation period to construct a review annotation node, and connect the initial review association node corresponding to the initial review annotation combination with the review annotation node; Constructing a final review annotation node based on the final review annotation end, connecting the review annotation node with the final review annotation node to generate a dynamic ultrasonic annotation tree; The method generates an ultrasound screening queue according to the ultrasound plaque screening image sent by the annotation request end, and sequentially sends the ultrasound screening set in the ultrasound screening queue to the annotation end of the annotation node in the dynamic ultrasound annotation tree for processing to obtain a plaque annotation image, including: Counting the ultrasonic plaque screening images sent by each annotation request end, and obtaining the ultrasonic screening set corresponding to each annotation request end; Determine the sending time of the ultrasonic screening set, and sort the ultrasonic screening set in ascending order based on the sending time to obtain an ultrasonic screening queue; Acquire an idle primary review associated node in the dynamic ultrasonic annotation tree as the current primary review node, and take the review annotation node connected to the current primary review node as the current review node, and the final review annotation node as the current final review node; Sending the ultrasound screening set in the ultrasound screening queue to the preliminary review annotation end corresponding to the current preliminary review node in sequence; Receiving the annotation processing of the ultrasound plaque screening image in the ultrasound screening set by the preliminary review annotation end of the current preliminary review node, and obtaining the initial annotation image corresponding to each of the preliminary review annotation ends; Using the initial annotation image corresponding to any one of the preliminary review annotation ends in the preliminary review annotation combination as the first annotation image, and using the initial annotation images corresponding to the remaining preliminary review annotation ends in the preliminary review annotation combination as the second annotation image; Obtaining a plaque annotated image according to the first annotated image and the second annotated image; The step of obtaining a plaque annotated image according to the first annotated image and the second annotated image includes: Updating the second annotation contour in the second annotation image to the first annotation image to obtain a preliminary annotation image; Obtaining a preliminary horizontal difference value and a preliminary vertical difference value corresponding to the preliminary annotated image according to comparison information of the first contour coordinate set and the second contour coordinate set in the preliminary annotated image; When it is determined that both the preliminary review transverse difference and the preliminary review longitudinal difference are less than or equal to the preset difference, the preliminary review annotated image is used as a plaque annotated image; When it is determined that the preliminary review horizontal difference or the preliminary review vertical difference is greater than a preset difference, the preliminary review annotated image is sent to the review annotated end corresponding to the current review node, and the review annotated end of the current review node receives the annotation processing of the preliminary review annotated image, so as to obtain the review annotated image corresponding to the review annotated end; Acquire the contour of the marked patch marked at the review mark end in the review mark image as a third marked contour, determine a third contour coordinate set of the third marked contour, and obtain a first review transverse difference and a first review longitudinal difference according to comparison information of the third contour coordinate set and the first contour coordinate set; Based on the comparison information of the third contour coordinate set and the second contour coordinate set, obtaining a second review transverse difference and a second review longitudinal difference; When it is determined that the first review transverse difference and the first review longitudinal difference are less than or equal to a preset difference, or the second review transverse difference and the second review longitudinal difference are less than or equal to a preset difference, the review annotated image is used as a plaque annotated image; Also includes: When it is determined that the first review horizontal difference or the first review vertical difference is greater than a preset difference, and the second review horizontal difference or the second review vertical difference is greater than a preset difference, sending the review annotated image to the final review annotated end corresponding to the current final review node; Receiving the annotation processing of the review annotation image by the final review annotation end of the current final review node, and obtaining the final review annotation image corresponding to the final review annotation end; Acquire the contour of the marked patch marked at the final marking end in the final marking image as a fourth marked contour, determine a fourth contour coordinate set of the fourth marked contour, and obtain a first final horizontal difference and a first final vertical difference according to comparison information of the fourth contour coordinate set and the first contour coordinate set; Obtaining a second final horizontal difference and a second final vertical difference according to comparison information of the fourth contour coordinate set and the second contour coordinate set; Obtaining a third final horizontal difference and a third final vertical difference according to comparison information of the fourth contour coordinate set and the third contour coordinate set; When it is determined that the first final horizontal difference and the first final vertical difference, the second final horizontal difference and the second final vertical difference, or the third final horizontal difference and the third final vertical difference are less than or equal to a preset difference, the final annotated image is used as a plaque annotated image; The method of fusing the annotated plaque contours in the plaque annotated image based on the plaque fusion strategy to obtain a fused plaque image and sending it to the annotation request end includes: When it is determined that the plaque annotated image is a preliminary annotated image, two annotated plaque contours in the corresponding plaque annotated image are used as contours to be fused; When the plaque annotated image is determined to be a review annotated image, a first review difference value is obtained according to the sum of the first review transverse difference value and the first review longitudinal difference value in the corresponding plaque annotated image, and a second review difference value is obtained according to the sum of the second review transverse difference value and the second review longitudinal difference value; Selecting the minimum value of the first review difference value and the second review difference value as the screening review difference value, and determining two annotated plaque contours in the plaque annotated image as contours to be fused based on the screening review difference value; When the plaque-annotated image is determined to be a final-reviewed annotated image, a first final-reviewed difference value is obtained according to the sum of the first final-reviewed transverse difference value and the first final-reviewed longitudinal difference value in the corresponding plaque-annotated image, a second final-reviewed difference value is obtained according to the sum of the second final-reviewed transverse difference value and the second final-reviewed longitudinal difference value, and a third final-reviewed difference value is obtained according to the sum of the third final-reviewed transverse difference value and the third final-reviewed longitudinal difference value; Selecting the minimum value among the first final difference value, the second final difference value and the third final difference value as the screening final difference value, and determining two annotated plaque contours in the plaque annotated image as contours to be fused based on the screening final difference value; Acquire the positional relationship of the contours to be fused, where the positional relationship includes a containment relationship and a non-containment relationship; When it is determined that the positional relationship is a containment relationship, the contours to be fused in the patch annotation image are fused according to a nested fusion strategy to obtain a fused patch image and send it to the annotation request end; When it is determined that the positional relationship is a non-inclusion relationship, the contour to be fused in the patch annotation image is fused according to an overlapping fusion strategy to obtain a fused patch image and send it to the annotation request end; Before the step of determining the basic number of ultrasound plaque screening images received by the dynamic annotation combination in the dynamic ultrasound annotation tree within the preset rotation period, and calling the fused number of fused plaque images corresponding to the dynamic annotation combination within the preset rotation period, the method includes: Obtaining a preliminary review annotation combination corresponding to a preliminary review associated node in the dynamic ultrasonic annotation tree within the preset rotation period; Determine a review annotation end corresponding to a review annotation node connected to the corresponding preliminary review associated node in the dynamic ultrasonic annotation tree, and a final review annotation end corresponding to a final review annotation node connected to the corresponding review annotation node; Adding the review annotation end to the preliminary review annotation combination to obtain a review annotation combination, and adding the final review annotation end to the review annotation combination to obtain a final review annotation combination; Obtaining a dynamic annotation combination according to the initial review annotation combination, the review annotation combination and the final review annotation combination within the preset rotation period; The calculation is performed according to the fusion quantity and the basic quantity to obtain the score value corresponding to each annotation end, including: The score corresponding to each annotation end is obtained by the following formula: in, is the score value corresponding to each annotation end, is the number of fusions of the initial review annotation combination corresponding to the annotation end, is the basic number of initial review annotation combinations corresponding to the annotation end. is the initial review weight, is the number of fusions of the corresponding review annotation combination on the annotation side, is the basic number of review annotation combinations corresponding to the annotation end, For review weight, is the number of fusions of the final annotation combination corresponding to the annotation end, is the basic number of final review annotation combinations corresponding to the annotation end. The final weight.

Images