WO2022062458A1 - 一种胎儿最优标准切面的确定方法及装置 - Google Patents

一种胎儿最优标准切面的确定方法及装置 Download PDF

Info

Publication number
WO2022062458A1
WO2022062458A1 PCT/CN2021/096821 CN2021096821W WO2022062458A1 WO 2022062458 A1 WO2022062458 A1 WO 2022062458A1 CN 2021096821 W CN2021096821 W CN 2021096821W WO 2022062458 A1 WO2022062458 A1 WO 2022062458A1
Authority
WO
WIPO (PCT)
Prior art keywords
standard
slice
structural feature
fetal ultrasound
ultrasound image
Prior art date
Application number
PCT/CN2021/096821
Other languages
English (en)
French (fr)
Inventor
谢红宁
汪南
冼建波
梁喆
杨燕淇
吴海涛
Original Assignee
广州爱孕记信息科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 广州爱孕记信息科技有限公司 filed Critical 广州爱孕记信息科技有限公司
Publication of WO2022062458A1 publication Critical patent/WO2022062458A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/52—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/5215—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data
    • A61B8/523—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data for generating planar views from image data in a user selectable plane not corresponding to the acquisition plane
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/08—Detecting organic movements or changes, e.g. tumours, cysts, swellings
    • A61B8/0866—Detecting organic movements or changes, e.g. tumours, cysts, swellings involving foetal diagnosis; pre-natal or peri-natal diagnosis of the baby
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/52—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/5215—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data
    • A61B8/5223—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data for extracting a diagnostic or physiological parameter from medical diagnostic data

Definitions

  • the invention relates to the technical field of image processing, in particular to a method and device for determining an optimal standard slice of a fetus.
  • the optimal fetal standard section becomes the key point for accurate determination of fetal growth and development.
  • the method for determining the optimal standard fetal section is as follows: by analyzing a single fetal ultrasound image, a preliminary fetal standard section is obtained. Further, after obtaining the preliminary fetal standard section, an experienced staff member analyzes the preliminary fetal standard section, thereby Complete the final determination of the optimal standard section of the fetus.
  • the technical problem to be solved by the present invention is to provide a method and device for determining the optimal standard section of the fetus, which can obtain the accurate optimal standard section of the fetus, thereby realizing accurate determination of the growth and development of the fetus.
  • a first aspect of the present invention discloses a method for determining an optimal standard section of a fetus, the method comprising:
  • the standard slice corresponding to the highest slice score is determined from all the standard slices as the optimal standard slice of all the fetal ultrasound images.
  • the method further includes:
  • the standard slices of all the fetal ultrasound images are executed according to a preset classification method.
  • a classification operation is performed to obtain at least two standard slice sets, each of the standard slice sets includes at least one standard slice of the fetal ultrasound image, and all the standard slices included in each of the standard slice sets are of the same category standard cut;
  • the standard section corresponding to the highest section score is determined from all the standard sections, as the optimal standard section, including:
  • the standard slice corresponding to the highest slice score is determined from all the standard slices included in each standard slice set, as the corresponding standard slice set for each standard slice set The optimal standard section of .
  • the method further includes:
  • the standard slice corresponding to the highest normalized slice score is selected from all the standard slices as the optimal standard slice corresponding to all the fetal ultrasound images.
  • At least one structural feature exists in the standard slice of each frame of the fetal ultrasound image, and each structural feature has a corresponding weight value;
  • determining the slice score of the standard slice of each frame of the fetal ultrasound image including:
  • the slice score of the standard slice of each frame of the fetal ultrasound image is calculated.
  • the determining a weight value corresponding to each structural feature of the standard slice of the fetal ultrasound image in each frame includes:
  • each of the key weight value influence factors corresponding to each of the structural features determine the sub-weight value corresponding to each of the key weight value influence factors, and calculate all the sub-weights corresponding to each of the structural features The sum of the values is used as the weight value corresponding to each of the structural features.
  • a second aspect of the present invention discloses a device for determining an optimal standard section of a fetus, the device comprising:
  • an acquisition module configured to acquire a standard section corresponding to each frame of the fetal ultrasound image in the multiple frames of fetal ultrasound images
  • a first determining module for determining the slice score of the standard slice of the fetal ultrasound image in each frame
  • the second determining module is configured to determine, according to the slice scores of all the standard slices, the standard slice corresponding to the highest slice score from all the standard slices, as the optimal standard slice of all the fetal ultrasound images.
  • the device further includes:
  • the first judging module is used to judge whether all the standard sectional planes belong to the same category of standard sectional planes after the first judging module determines the sectional plane score of the standard sectional plane of each frame of the fetal ultrasound image.
  • trigger the second determination module to execute the cut plane scores according to all the standard cut planes, and determine the standard cut plane corresponding to the highest cut plane score from all the standard cut planes, Operation as the optimal standard view for all of the fetal ultrasound images.
  • the device further includes:
  • the classification module is configured to perform a classification operation on all the standard slices of the fetal ultrasound image according to a preset classification method when the first judgment module determines that all the standard slices do not belong to the standard slices of the same category, and obtain at least two standard slices.
  • standard slice sets each standard slice set includes at least one standard slice of the fetal ultrasound image, and all the standard slices included in each standard slice set are standard slices of the same category;
  • the second determination module determines the standard cut plane corresponding to the highest cut plane score from all the standard cut planes according to the cut plane scores of all the standard cut planes, and the method as the optimal standard cut plane is specifically:
  • the standard slice corresponding to the highest slice score is determined from all the standard slices included in each standard slice set, as the corresponding standard slice set for each standard slice set The optimal standard section of .
  • the device further includes:
  • a normalization module configured to determine the highest slice score from all the standard slices included in each standard slice set according to all slice scores corresponding to each of the standard slice sets in the second determination module After the corresponding standard cut planes are taken as the optimal standard cut planes corresponding to each of the standard cut plane sets, perform a normalization operation on the cut plane scores of the optimal standard cut planes corresponding to each of the standard cut plane sets to obtain each of the standard cut planes.
  • the screening module is configured to screen the standard section corresponding to the highest normalized section score from all the standard sections according to the normalized scores of all the sections, as the highest corresponding to all the fetal ultrasound images. Excellent standard cut.
  • At least one structural feature exists in the standard slice of each frame of the fetal ultrasound image, and each structural feature has a corresponding weight value;
  • the first determining module includes:
  • Determining submodule for determining the weight value corresponding to each of the structural features of the standard slice of the fetal ultrasound image in each frame
  • the calculation submodule is configured to calculate the slice score of the standard slice of each frame of the fetal ultrasound image based on the weight value corresponding to each of the structural features of each of the standard slices and the characteristic parameters of the structural feature.
  • the determining submodule includes:
  • a determining unit configured to determine a key weight value influencing factor corresponding to each of the structural features of the standard slice of the fetal ultrasound image in each frame, and the number of key weight value influencing factors corresponding to each of the structural features is greater than or equal to 1, And each of the key weight value influencing factors has a corresponding sub-weight value;
  • the determining unit is further configured to determine the sub-weight value corresponding to each of the key weight value influence factors according to each of the key weight value influence factors corresponding to each of the structural features;
  • a calculation unit configured to calculate the sum of all the sub-weight values corresponding to each of the structural features as a weight value corresponding to each of the structural features.
  • a method and device for determining an optimal standard section of a fetus includes acquiring a standard section of each frame of fetal ultrasound images in multiple frames of fetal ultrasound images, and determining the standard section of each frame of fetal ultrasound images.
  • Slice score According to the score of all slices, the standard slice corresponding to the highest slice score is determined from all standard slices as the optimal standard slice of all fetal ultrasound images. It can be seen that the implementation of the present invention does not require manual analysis to determine the optimal standard slice of the fetal ultrasound image after acquiring the standard slice of the fetal ultrasound image, and can automatically determine the slice score of the standard slice of the fetal ultrasound image, and intelligently analyze all the standard slices.
  • the slice score selects the standard slice with the highest slice score, and realizes the automatic determination of the optimal standard slice, which can improve the accuracy and efficiency of determining the optimal standard slice of the fetal ultrasound image, so as to accurately obtain the growth and development of the fetus.
  • FIG. 1 is a schematic flowchart of a method for determining an optimal standard section of a fetus disclosed in an embodiment of the present invention
  • FIG. 2 is a schematic flowchart of another method for determining an optimal standard section of a fetus disclosed in an embodiment of the present invention
  • FIG. 3 is a schematic flowchart of a method for determining a section score of a fetal standard section disclosed in an embodiment of the present invention
  • FIG. 4 is a schematic structural diagram of a device for determining an optimal standard section of a fetus disclosed in an embodiment of the present invention
  • FIG. 5 is a schematic structural diagram of another device for determining an optimal standard section of a fetus disclosed in an embodiment of the present invention.
  • FIG. 6 is a schematic structural diagram of a first determination module disclosed in an embodiment of the present invention.
  • FIG. 7 is a schematic structural diagram of another first determination module disclosed in an embodiment of the present invention.
  • FIG. 8 is a schematic structural diagram of another device for determining an optimal standard section of a fetus disclosed in an embodiment of the present invention.
  • the invention discloses a method and a device for determining an optimal standard slice of a fetus, which can automatically determine the optimal standard slice of a fetal ultrasonic image without manual analysis after acquiring the standard slice of an ultrasonic image of the fetus.
  • the slice score of the standard slice and intelligently select the standard slice with the highest slice score from all slice scores to realize the automatic determination of the optimal standard slice, which can improve the accuracy and efficiency of the determination of the optimal standard slice of fetal ultrasound images. , so as to accurately obtain the growth and development of the fetus.
  • FIG. 1 is a schematic flowchart of a method for determining an optimal standard section of a fetus disclosed in an embodiment of the present invention.
  • the method for determining the optimal standard aspect of the fetus described in FIG. 1 can be applied to a standard aspect determination server (service device), wherein the standard aspect determination server may include a local standard aspect determination server or a cloud standard aspect determination server.
  • the standard aspect determination server may include a local standard aspect determination server or a cloud standard aspect determination server.
  • the method for determining the optimal standard section of the fetus may include the following operations:
  • acquiring a standard slice of each frame of fetal ultrasound images in the multiple frames of fetal ultrasound images may include:
  • the analysis results output by the feature detection model in turn are obtained as the feature information of each frame of fetal ultrasound image
  • the feature information of each frame of fetal ultrasound image includes the part feature information of the fetal ultrasound image and the structural feature information of the fetal ultrasound image.
  • the part feature information of the ultrasound image at least includes the category of the part feature of the fetal ultrasound image
  • the structural feature information of each frame of the fetal ultrasound image at least includes the category of the structural feature of the fetal ultrasound image
  • the structural feature of each fetal ultrasound image at least includes the Key structural features of fetal ultrasound images
  • the standard slice corresponding to the fetal ultrasound image is determined according to the category of the site feature of each frame of the fetal ultrasound image and the category of the structural feature of the fetal ultrasound image.
  • multiple frames of fetal ultrasound images may be continuously acquired according to a predetermined frame rate, wherein the predetermined frame rate is related to the standard slice of the fetal ultrasound image to be acquired, that is, the acquired fetal ultrasound images are acquired according to the required frame rate.
  • the frame rate can be selected from the standard slice of the fetal ultrasound image. For example, if the abdominal slice is to be acquired, the frame rate can be 30 frames per second; if the four-chamber slice is to be acquired, the frame rate can be 60 frames. /Second. In this way, the corresponding frame rate is selected according to the standard slice of the fetal ultrasound image to be acquired, which is beneficial to improve the efficiency and accuracy of the standard slice of the required fetal ultrasound image.
  • each frame of fetal ultrasound image has a unique corresponding frame sequence number.
  • a unique frame serial number for each frame of fetal ultrasound image it is possible to clearly distinguish each frame of fetal ultrasound image during the process of acquiring the standard section of the fetal ultrasound image, and it is beneficial to the information of the fetal ultrasound image and its standard section. manage.
  • the feature detection model may include at least one of a target detection model, an instance segmentation model, and a semantic segmentation model, which can obtain at least one of the part feature information and structural feature information of the fetal ultrasound image, which is not limited in the embodiment of the present invention. .
  • this optional embodiment determines the standard slice of the fetal ultrasound image by acquiring the site features and structural features of the continuous multiple frames of fetal ultrasound images, and combining the site features and structural features of the fetal ultrasound images, without manual participation in the fetal ultrasound image.
  • the determination of the standard section can improve the accuracy of determining the standard section of the fetal ultrasound image; and by inputting the fetal ultrasound image into the feature detection model for analysis, the efficiency of determining the standard section of the fetal ultrasound image can also be improved.
  • the standard slice of the fetal ultrasound image can be obtained by receiving the standard slice of each frame of the fetal ultrasound image in the multiple frames of fetal ultrasound images sent by the authorized terminal device.
  • the standard slices of the fetal ultrasound image can be obtained through various ways, which can enrich the way of obtaining the standard slices and improve the possibility of obtaining the standard slices.
  • the method may further include the following operations:
  • the proportion of the target feature of each frame of fetal ultrasound image is used to represent the display ratio of the target feature and the display device where it is located, and the target feature of each frame of fetal ultrasound image includes the standard section of the fetal ultrasound image or the Structural features in standard slices of fetal ultrasound images;
  • the method may further include the following operations:
  • the ratio optionally, by calculating the area enclosed by the contour of the standard cutting plane and/or the distance value between the two most distant endpoints on the contour of the standard cutting plane, the ratio, which can improve the calculation accuracy and reliability of the ratio of standard slices.
  • Priority is given to the area enclosed by the outline of the standard section to calculate the proportion of structural features. For example, when the area enclosed by the outline of the abdominal section accounts for two-thirds of the display screen area, the proportion of the abdominal section corresponds to The score correction factor is 1.
  • the slice is further updated according to the score coefficient corresponding to the ratio of the obtained standard slice of the fetal ultrasound image to the display area of the current display device.
  • the score is beneficial to improve the accuracy and reliability of the determination of the slice score of the standard slice of the fetal ultrasound image, thereby improving the accuracy and reliability of the determination of the optimal standard slice of the fetal ultrasound image.
  • determining the slice score of the standard slice of each frame of fetal ultrasound image may include:
  • the slice score of the standard slice of each frame of fetal ultrasound image is calculated.
  • each structural feature exists in the standard slice of each frame of fetal ultrasound image, and each structural feature has a corresponding weight value.
  • the structural features in each standard section include at least the key structural features (also known as basic structural features or main structural features) of the standard section, and further, the structural features in each standard section may also include other than key structural features. other structural features.
  • the standard view of the thalamus includes at least one key structural feature of the septum pellucidum, the thalamus and the lateral ventricle, and further, the standard view of the thalamus may also include at least one other structural feature of the choroid and the sylvian fissure.
  • the key structural feature of each standard slice is a structural feature that can represent the standard slice, that is, when the key structural feature of the fetal ultrasound image is acquired, the standard slice corresponding to the key structural feature can be determined.
  • the standard section of the fetal ultrasound image is the abdominal circumference section. In this way, the standard section of the fetal ultrasound image is determined by the key structural features, which can improve the determination efficiency of the standard section while ensuring the correct determination of the standard section.
  • this optional embodiment can realize the automatic calculation of the section score of the standard section by combining the weight value of each structural feature of the standard section with the feature parameter of the structural feature, and improve the calculation of the section score of the standard section. accuracy and efficiency.
  • determining the weight value corresponding to each structural feature of the standard slice of each frame of fetal ultrasound image may include:
  • each key weight value influencing factor corresponding to each structural feature determines the sub-weight value corresponding to each key weight value influencing factor, and calculate the sum of all sub-weight values corresponding to each structural feature as the corresponding value of each structural feature. weight value.
  • the impact factor of the key weight value corresponding to each structural feature of each standard section may or may not be the same.
  • the key weight value of the skull halo structural feature in the lateral ventricle section include the size of the head circumference corresponding to the contour of the skull halo structural feature, the integrity of the skull halo structural feature, and the area enclosed by the skull halo structural feature.
  • the influencing factors include the length corresponding to the outline of the femoral structural feature, the area enclosed by the outline of the femoral structural feature, and the area enclosed by the outline of the femoral structural feature. The relative position of the region to the midline of the brain.
  • this optional embodiment can improve the structure by determining the key weight value influencing factor corresponding to each structural feature in a targeted manner, and determining the sub-weight values corresponding to all key weight value influencing factors as the weight value corresponding to the structural feature.
  • the calculation accuracy of the weight value of the feature can improve the calculation accuracy of the slice score corresponding to the standard slice, thereby improving the determination accuracy of the optimal standard slice.
  • determining the sub-weight value corresponding to each key weight value influencing factor may include:
  • the key weight value influencing factor corresponding to the structural feature includes the geometric parameters of the contour of the structural feature
  • the sub-weight value corresponding to the geometric parameters of the contour of the structural feature is determined according to the geometric parameters of the contour of the structural feature
  • the geometric parameters of the outline of the structural feature include the size and/or area of the outline of the structural feature
  • the key weight value influencing factor corresponding to the structural feature includes the clarity of the structural feature
  • the sub-weight value corresponding to the clarity of the structural feature
  • the geometric parameters corresponding to the structural feature are calculated according to the outline of the structural feature, and the structure is determined according to the geometric parameters corresponding to the structural feature.
  • the sub-weight value corresponding to the integrity of the feature is calculated according to the outline of the structural feature, and the structure is determined according to the geometric parameters corresponding to the structural feature.
  • the key weight value influencing factor corresponding to the structural feature includes the position of the structural feature in the standard section, the relative positional relationship between the midline of the brain corresponding to the structural feature and the area enclosed by the contour of the structural feature , to determine the sub-weight value corresponding to the position of the structural feature in the standard section;
  • a sub-weight value matching the proportion of the structural feature is determined according to the proportion of the structural feature.
  • the ratio is calculated by calculating the area enclosed by the contour of the structural feature and/or the distance value between the two farthest endpoints on the contour of the structural feature, calculate the ratio, which can improve the calculation accuracy and reliability of the ratio of structural features.
  • the area enclosed by the outline of the structural feature is preferentially selected to calculate the proportion of the structural feature. For example, if the area enclosed by the outline of the structural feature of the left atrium accounts for one-seventh of the area of the display screen, then the proportion of the structural feature of the left atrium The corresponding sub-weight value is 0.8.
  • the size of the contour of the structural feature may include the perimeter of the contour of the structural feature and/or the length corresponding to the contour of the structural feature (for example, the length of the humerus structural feature).
  • the geometric parameters of the contour of the structural feature after obtaining the geometric parameters of the contour of the structural feature, it is further judged whether the geometric parameters of the contour of the structural feature are within the range of the geometric parameters corresponding to the determined gestational age of the fetal ultrasound image. When it is not within the range of geometric parameters, multiply the sub-weight value corresponding to the geometric parameter of the outline of the structural feature by the determined weight correction coefficient (for example: 0.8) to obtain the corrected sub-weight value; when the judgment result is If yes, trigger the execution of the above operation of calculating the sum of all sub-weight values corresponding to each structural feature as the weight value corresponding to each structural feature.
  • the determined weight correction coefficient for example: 0.8
  • the calculated subweight value ( 0.7) remains unchanged.
  • the calculated sub-weight value (0.7) is multiplied by the weight correction coefficient (0.9) to obtain the corrected sub-weight value (0.63).
  • the higher the weight value the more obvious the corresponding structural feature.
  • the calculation accuracy of the section score of the corresponding standard section can be improved.
  • the sub-weight value corresponding to the position of the standard section where the structural feature is located is determined, specifically: when When there is an intersection between the midline of the brain corresponding to the structural feature and the area enclosed by the contour of the structural feature, the sub-weight value corresponding to the position of the standard section of the structural feature is determined to be the first sub-weight value; when the midline of the brain corresponding to the structural feature When the distance from the contour of the structural feature is within the predetermined distance range value, the sub-weight value corresponding to the position of the standard section of the structural feature is determined to be the second sub-weight value; when the midline of the brain corresponding to the structural feature and the When the distance between the contours of the structural feature is greater than the maximum distance value in the predetermined distance range value, it is determined that the sub-weight value corresponding to the position of the standard section where the structural feature is located is the third sub-weight
  • the sub-weight value is 1, indicating that the midline capsule structure feature does not deviate from the midline; If there is no intersection point between the midline of the brain and the deviation distance is 1mm, the sub-weight value is 0.8; when the deviation distance is 5mm, the sub-weight value is 0.
  • the weight value of the corresponding structural feature is equal to the sum of the sub-weight values corresponding to each key weight value influencing factor.
  • the key weight value of the femoral structural feature of the femur measurement section include the length corresponding to the outline of the femoral structural feature, the area enclosed by the outline of the femoral structural feature, and the distance between the area enclosed by the outline of the femoral structural feature and the midline of the brain.
  • the relative position, and the sub-weight value of the length corresponding to the contour of the femoral structural feature is 0.7, the sub-weight value corresponding to the area surrounded by the contour of the femoral structural feature is 0.6, and the area surrounded by the contour of the femoral structural feature and the midline of the brain
  • this optional embodiment can not only achieve the acquisition of the sub-weight value corresponding to the key weight value influencing factor, but also improve the sub-weight value. Obtain efficiency and accuracy, thereby improving the calculation accuracy and efficiency of the weight value corresponding to the structural feature, thereby improving the calculation accuracy and efficiency of the section score corresponding to the standard section.
  • the contour of each structural feature corresponds to multiple nodes, and the contour of the structural feature is fitted based on the determined fitting method to obtain the target contour of the structural feature, which may include:
  • the arc radius corresponding to the contour of each structural feature is greater than or equal to the determined arc radius threshold (for example: 5mm)
  • select a preset number targets from all nodes corresponding to the structural feature nodes, and connect all target nodes corresponding to each structural feature in turn according to the way that every two adjacent nodes are connected to obtain the target contour of the structural feature;
  • the contour of the structural feature when the contour of the structural feature has multiple arcs and/or the curvature of the contour is greater than or equal to the determined curvature threshold, the contour of the structural feature is segmented to perform a fitting operation. Specifically: when there are multiple arcs in the contour of the structural feature, a fitting operation will be performed on each of the multiple arcs of the structural feature; when the curvature of the contour of the structural feature is greater than or equal to the curvature threshold, the The contour of the structural feature is divided into multiple segments at equal or unequal intervals, and the fitting operation is performed on each segment of the contour separately.
  • the contour of the structural feature has multiple arcs and/or the curvature of the contour is relatively large, by performing the fitting operation on the contour segment of the structural feature, the fitting efficiency and accuracy of the contour of the structural feature can be improved, so that there is a It is beneficial to further improve the measurement accuracy and reliability of the geometric parameters of the structural features of the fetal ultrasound image.
  • a fitting operation may also be performed on the contour of each structural feature based on the determined B-spline curve fitting method and/or ellipse fitting method to obtain the target contour of the structural feature.
  • the implementation manner is not limited.
  • this optional embodiment can not only realize the fitting of structural features, but also improve the fitting efficiency and accuracy of structural features by selecting different fitting methods according to the size of the arc radius of the structural features of the fetal ultrasound image. , so as to improve the calculation accuracy of the geometric parameters of the structural features.
  • the method further includes:
  • the method further includes:
  • this optional embodiment can enrich the acquisition methods of the geometric parameters corresponding to the structural features by providing multiple ways to determine the geometric parameters corresponding to the structural features, and improve the possibility of obtaining the geometric parameters corresponding to the structural features; One of the length of the contour, the central angle corresponding to the contour of the structural feature, the length of the contour of the overlapping part of the contour of the structural feature and the fitted contour, and the center angle corresponding to the contour of the overlapping part, or a combination as the geometry corresponding to the structural feature The parameters can improve the accuracy of obtaining the geometric parameters corresponding to the structural features, thereby improving the calculation accuracy of the weight values corresponding to the structural features.
  • the slice score of the standard slice of each frame of fetal ultrasound image is calculated based on the weight value corresponding to each structural feature of each standard slice and the characteristic parameter of the structural feature. value, which can include:
  • the structural score corresponding to each structural feature of each standard section is calculated;
  • the sum of the structural scores corresponding to all structural features of each standard slice is calculated as the slice score of the standard slice of each frame of fetal ultrasound image.
  • the feature parameter of each structural feature of each standard slice includes the class probability of the structural feature and the position probability of the structural feature.
  • the calculation formula of the slice score of the standard slice of each frame of fetal ultrasound image is as follows:
  • H i P i ⁇ Q i ⁇ O i ;
  • S is the section score of each standard section
  • H i is the structural score of the i-th structural feature in each standard section
  • M is the total number of structural features in each standard section
  • P i is each The class probability (also known as confidence) of the ith structural feature in the standard section
  • Q i is the position probability of the ith structural feature in each standard section
  • O i is the weight of the ith structural feature in each standard section value
  • N is the total number of key weight value influencing factors of the ith structural feature
  • O ij is the sub-weight value corresponding to the jth key weight value influencing factor in the ith structural feature in each standard section.
  • the parameters included in the structural feature in each standard section also include the probability of the position where the structural feature is located.
  • the structural score of the i-th structural feature in each standard section is The calculation formula is:
  • H i P i ⁇ Q i ⁇ O i ⁇ C i ;
  • C i is the parameter included in the structural feature in each standard section and also includes the probability of the position where the structural feature is located.
  • this optional embodiment can realize the calculation of the section score of the standard section by separately calculating the structural score corresponding to each structural feature of the standard section, which is beneficial to improve the accuracy and efficiency of the calculation of the section score of the standard section; And selecting different parameters according to different structural features can improve the calculation accuracy and efficiency of the structural score corresponding to the structural feature, thereby further improving the calculation accuracy and efficiency of the section score of the standard section.
  • the slice scores of all the above standard slices determine the standard slice corresponding to the highest slice score from all the above standard slices, as the optimal standard slice of all fetal ultrasound images.
  • the method for determining the optimal standard section of the fetus may further include the following operations:
  • the pixel value of the positive fetal ultrasound image sample is greater than the pixel value of the negative fetal ultrasound image sample, each of the positive fetal ultrasound image and the negative fetal ultrasound image in the positive fetal ultrasound image sample.
  • the key weight value impact factor of the structural feature of each negative sample fetal ultrasound image in the sample includes the clarity of the structural feature;
  • the determined initial classification model is trained, and the trained initial classification model is obtained as the determined classification model.
  • the initial classification model includes KNN, Bayesian, Neural Network, Ensemble-Stacking, Ensemble-Boosting, and Ensemble-Bagging, etc., which can realize image classification or a combination of classification models.
  • KNN Bayesian, Neural Network, Ensemble-Stacking, Ensemble-Boosting, and Ensemble-Bagging, etc.
  • the sample fetal ultrasound images included in the positive fetal ultrasound image sample and the negative fetal ultrasound image sample may be selected by the device terminal, or may be selected by relevant personnel based on experience, or determined by both. of.
  • the positive fetal ultrasound image sample is composed of a plurality of sub-positive fetal ultrasound image samples, and a negative fetal ultrasound image sample Consists of multiple sub-negative fetal ultrasound image samples.
  • each sub-positive fetal ultrasound image sample corresponds to a sub-negative fetal ultrasound image sample.
  • each sample fetal ultrasound image has a corresponding sample weight value.
  • a positive fetal ultrasound image sample includes a sub-positive fetal ultrasound image sample that includes structural features of the transparent compartment and a sub-positive fetal ultrasound image sample that includes structural features of the ductus arteriosus
  • a negative fetal ultrasound image sample includes a sub-negative fetal ultrasound image that includes the structural features of the transparent compartment.
  • the determined initial classification model is trained based on the positive fetal ultrasound image sample, the negative fetal ultrasound image sample and the weight value corresponding to each sample fetal ultrasound image, and the trained initial classification model is obtained as the determined classification model . In this way, the training accuracy of the classification model can be improved, so that a classification model with high accuracy can be obtained.
  • this optional embodiment can obtain a required and accurate classification model by performing a training operation on the initial classification model based on the sample fetal ultrasound image in advance, thereby improving the sub-weight value of the key weight value influencing factors including the clarity of structural features The accuracy and reliability of the analysis can be improved, so as to improve the calculation accuracy and efficiency of the weight value corresponding to the structural feature.
  • the method for determining the optimal standard slice of the fetus may further include the following operations:
  • step 103 is triggered.
  • step of judging whether all standard slices belong to the same category of standard slices may occur simultaneously with step 102, or may occur before step 102, which is not limited in this optional embodiment .
  • judging whether all standard slices belong to the same category of standard slices may include:
  • all standard slices are standard slices of the same category, and are standard slices of the abdominal circumference category. For another example, if the first several digits in the slice number of each standard slice are 0001, all standard slices are standard slices of the same category.
  • the method for determining the optimal standard section of the fetus may further include the following operations:
  • the classification operation is performed on the standard slices of all fetal ultrasound images according to the preset classification method to obtain at least two standard slice sets, each standard slice set includes at least one frame of the fetus Standard slices of ultrasound images, and all standard slices included in each standard slice set are standard slices of the same category;
  • the standard section corresponding to the highest section score is determined from all standard sections as the optimal standard section, including:
  • the standard slice corresponding to the highest slice score is determined from all the standard slices included in each standard slice set, as the optimal standard slice corresponding to each standard slice set.
  • slice scores of all standard slices included in each standard slice set may be determined.
  • the classification operation is performed on all the standard slices, and the standard slices of different categories can be obtained, thereby helping to reduce the occurrence of different slice types. Reduce the probability of obtaining non-optimal standard slices and the occurrence and reliability. To improve the accuracy of determining the optimal standard section corresponding to different types of standard sections
  • the standard sectional corresponding to the highest sectional score is determined from all the standard sectionals included in each standard sectional set, as each standard After the optimal standard slice corresponding to the slice set, the method for determining the optimal standard slice of the fetus may further include the following operations:
  • the standard slice corresponding to the highest normalized slice score is selected from all the standard slices as the optimal standard slice corresponding to all fetal ultrasound images.
  • the slice score of the optimal standard slice of the femur measurement standard slice is 20 points
  • the slice score of the optimal standard slice of the standard slice of the skull is 100 points
  • the gallbladder-umbilical vein standard The section score of the optimal standard section of the section is 60 points, then the section score of the optimal standard section of the standard section of the femur, the section score of the optimal standard section of the standard section of the skull, and the standard section of the gallbladder-umbilical vein are measured respectively.
  • the slice scores of the optimal standard slices of the The optimal standard slice of fetal ultrasound images is the optimal standard slice of the standard slice of the gallbladder-umbilical vein.
  • this optional embodiment further performs a normalization operation on the slice scores of the optimal standard slices corresponding to different types of standard slices, which can make different
  • the slice scores of the optimal standard slices corresponding to the standard slices of the categories are comparable, thereby improving the accuracy and efficiency of determining the optimal standard slices corresponding to all fetal ultrasound images, thereby further facilitating accurate acquisition of fetal growth and development.
  • implementing the method for determining the optimal standard section of the fetus described in FIG. 1 can automatically determine the standard section of the fetal ultrasound image without manual analysis to determine the optimal standard section of the fetal ultrasound image after obtaining the standard section of the fetal ultrasound image. It can intelligently select the standard slice with the highest slice score from all the slice scores, and realize the automatic determination of the optimal standard slice, which can improve the accuracy and efficiency of the determination of the optimal standard slice of fetal ultrasound images. In order to achieve accurate acquisition of fetal growth and development.
  • FIG. 2 is a schematic flowchart of another method for determining an optimal standard section of a fetus disclosed in an embodiment of the present invention.
  • the method for determining the optimal standard aspect of the fetus described in FIG. 2 can be applied to a standard aspect determination server (service device), wherein the standard aspect determination server may include a local standard aspect determination server or a cloud standard aspect determination server.
  • the standard aspect determination server may include a local standard aspect determination server or a cloud standard aspect determination server.
  • the method for determining the optimal standard section of the fetus may include the following operations:
  • step 205 determines whether there are abnormal standard sections whose structural features are abnormal structural features in all standard sections. When the judgment result is no, the execution of step 205 can be triggered; when the judgment result is yes, the execution of step 205 can be triggered. Step 204 is performed.
  • step 202 and step 203 may also occur simultaneously.
  • step 204 the execution of step 205 is triggered.
  • the slice scores of all standard slices in step 205 include the slice scores of all standard slices that do not need to be corrected and the slice scores that need to be corrected and have been corrected.
  • each abnormal standard section has a corresponding score correction coefficient.
  • the score correction coefficients corresponding to different abnormal standard slices may be the same or different, which is not limited in the embodiment of the present invention.
  • the correction coefficient of the score corresponding to the standard section of the abnormal lateral ventricle is 10
  • the correction coefficient of the score corresponding to the standard section of the abnormal thalamus is 8.
  • the score correction coefficient includes at least one of a section score correction coefficient, a structure score correction coefficient, and a weight value correction coefficient.
  • correcting the slice score of the abnormal standard slice based on the score correction coefficient corresponding to each abnormal standard slice may include:
  • the score correction coefficient is the section score correction coefficient
  • the score correction coefficient corresponding to the abnormal standard section by the section score of the abnormal standard section to obtain the corrected abnormal standard section. slice score;
  • the score correction coefficient is the structural score correction coefficient
  • the score correction coefficient corresponding to the abnormal standard section is multiplied by the structural score corresponding to the abnormal structural feature to obtain the corrected abnormal structural feature corresponding to the Structural score, and adding up the structural scores of each structural feature (including normal structural features and abnormal structural features) of the abnormal standard section to obtain the revised section score of the abnormal standard section;
  • the score correction coefficient is the weight value correction coefficient
  • the score correction coefficient corresponding to the abnormal standard section is multiplied by the weight value corresponding to the abnormal structural feature, and the weight value corresponding to the abnormal structural feature after modification is obtained.
  • calculate the structural score corresponding to the abnormal structural feature and add the structural scores of each structural feature (including normal structural features and abnormal structural features) of the abnormal standard section to obtain the modified section of the abnormal standard section points.
  • the mean value of the slice scores of the abnormal standard slices corrected by all the correction methods is obtained as the corrected abnormal standard slice. slice score. In this way, the correction accuracy of the section score of the abnormal standard section can be improved, and the accurate section score of the abnormal standard section can be further obtained.
  • this optional embodiment corrects the section score of the abnormal standard section by providing at least one of section score correction, structural score correction and weight value correction, which can not only enrich the section score of the abnormal standard section.
  • the correction method can also improve the correction accuracy of the slice score of the abnormal standard slice, so that the accurate slice score of the abnormal standard slice can be obtained, which is beneficial to improve the accuracy and reliability of the determination of the optimal standard slice.
  • the slice scores of all the above standard slices determine the standard slice corresponding to the highest slice score from all the above standard slices, as the optimal standard slice of all fetal ultrasound images.
  • the slice scores of the standard slices of the fetal ultrasound image after obtaining the slice scores of the standard slices of the fetal ultrasound image, it is further determined whether there are abnormal standard slices in all the standard slices, and if so, the slice scores of the abnormal standard slices are determined based on the score correction coefficient. Performing the correction operation can improve the accuracy of the determination of the slice score of the abnormal standard slice, so as to reduce the occurrence of obtaining non-optimal standard slices by continuing to perform the acquisition of the optimal standard slice when an abnormal standard slice occurs, thereby improving the current situation. The accuracy and reliability of the determination of the optimal standard cut plane when abnormal standard cut planes occur.
  • step 201 for other descriptions of step 201, step 202, and step 205, please refer to the detailed description of step 101 to step 103 in Embodiment 1, which is not repeated in this embodiment of the present invention.
  • judging whether there is an abnormal standard cut surface whose structural feature is an abnormal structural feature in all standard cut surfaces may include:
  • each structural feature of each standard section determine whether each structural feature matches the standard section where it is located;
  • this optional embodiment can realize the determination of abnormal standard slices by acquiring the target information of each structural feature of the standard slice, and judging whether each structural characteristic matches the corresponding standard slice according to the target information of each structural characteristic .
  • judging whether each structural feature matches the standard section where it is located may include:
  • the representation type corresponding to each structural feature includes numerical representation type and/or feature morphological representation type
  • the representation type of the structural feature is a numerical representation type, obtain the target geometric parameter value corresponding to the structural feature, and judge whether the target geometric parameter value corresponding to the structural feature is within the predetermined normal parameter value range, and when the judgment result is no When , it is determined that the structural feature does not match the standard section where it is located;
  • the representation type of the structural feature is the feature morphological representation type, it is judged whether the structural feature is located in the detection area of the part feature corresponding to the structural feature, and when the judgment result is no, it is determined that the structural feature does not match the standard section where it is located. .
  • the detection area of each part feature can be selected by a detection frame, such as a polygon frame or an oval frame.
  • the target geometric parameter value corresponding to each structural feature includes the transverse diameter corresponding to the structural feature and/or the perimeter corresponding to the structural feature, so that the more content the geometric parameter value includes, the more beneficial it is Improve the judgment accuracy that the structural features match the standard section where they are located.
  • different structural features have corresponding normal parameter value ranges, wherein the normal parameter value ranges corresponding to different structural features may be the same or different.
  • different geometric parameter values of the same structural feature correspond to different normal parameter value ranges.
  • the geometric parameter values corresponding to each structural feature may include proportional dimensions and/or actual dimensions.
  • the actual size corresponding to the structural feature is further obtained, and it is judged whether the actual size is within the predetermined normal size range.
  • the judgment result is no, it is determined that the structural feature does not match the standard section where it is located.
  • the accuracy of determining whether the structural feature matches the standard section where it is located can be improved, thereby reducing the error correction of the section score of the abnormal standard section. If the situation occurs, improve the accuracy and reliability of the correction of the section score of the abnormal standard section.
  • the structural feature when it is judged that the structural feature is located in the detection area of the part feature corresponding to the structural feature, it is judged whether the structural feature exists in all multiple frames of fetal ultrasound images, and when the judgment result is yes, the structural feature is determined
  • the feature does not match the standard slice it is in.
  • the multiple frames of fetal ultrasound images may be fetal ultrasound images that appear continuously or intermittently after the first frame of fetal ultrasound images with structural features as the first frame of fetal ultrasound images. In this way, when it is judged that the structural feature is in the detection area of the corresponding part feature, it is further judged whether there are multiple frames of fetal ultrasound images with the structural feature.
  • the accuracy of determining whether the feature matches the standard slice where it is located can reduce the occurrence of incorrect correction of the slice score of the abnormal standard slice, and improve the accuracy and reliability of the correction of the slice score of the abnormal standard slice.
  • Numerical representation type when the detected structural feature is the characteristic of critical enlargement of the lateral ventricle, the outline information of the characteristic of the critical enlargement of the lateral ventricle is input into the measurement module for measurement, and the transverse dimension of the characteristic of the critical enlargement of the lateral ventricle is obtained. Diameter (proportional size), and determine whether the transverse diameter is greater than or equal to 12 pixels, if the judgment result is yes, then the critical enlargement feature of the lateral ventricle is an abnormal structural feature, that is, the critical enlargement feature of the lateral ventricle is located in the standard section. does not match.
  • the characteristic of the critical enlargement of the lateral ventricle is an abnormal structural feature, that is, the characteristic of the critical enlargement of the lateral ventricle does not match the standard view.
  • the critical enlargement feature of the lateral ventricle is determined as a normal structural feature,
  • the characteristic of critical enlargement of lateral ventricle is modified to the characteristic of normal lateral ventricle, that is, the characteristic of critical enlargement of lateral ventricle matches the standard slice.
  • Type of feature morphology representation when the detected structural feature is the structural feature of the choroidal sub-cyst, it is detected whether the structural feature of the choroidal sub-cyst appears in the detection area of the lateral ventricle, and when it appears in the detection area of the lateral ventricle, it is determined Cyst structural features are abnormal structural features, that is, it is determined that the choroid does not match the standard section where the cystic structural features are located. Further, when it is detected that the choroidal secondary cyst structure feature appears in the detection area of the lateral ventricle, it is determined whether there is the choroidal secondary cyst structure feature in all four frames of fetal ultrasound images, and when the judgment result is yes, determine the choroidal secondary cyst structure. The feature does not match the standard slice it is in.
  • the fetal ultrasound image when judging that the fetal ultrasound image has structural features, it can determine whether the structural features match the standard section where the structural features are located by using the obtained geometric parameter values of the structural features, or whether the structural features are located in the corresponding In the detection area of the part feature, to realize the judgment of whether the structural feature matches the standard section where it is located, it can improve the possibility, accuracy and efficiency of determining whether the structural feature matches the standard section where it is located.
  • implementing the method for determining the optimal standard section of the fetus described in FIG. 2 can automatically determine the standard section of the fetal ultrasound image without manual analysis to determine the optimal standard section of the fetal ultrasound image after obtaining the standard section of the fetal ultrasound image. It can intelligently select the standard slice with the highest slice score from all the slice scores, and realize the automatic determination of the optimal standard slice, which can improve the accuracy and efficiency of the determination of the optimal standard slice of fetal ultrasound images. In order to achieve accurate acquisition of fetal growth and development.
  • FIG. 3 is a schematic flowchart of a method for determining a slice score of a fetal standard slice disclosed in an embodiment of the present invention.
  • the method for determining the slice score of the fetal standard slice described in FIG. 3 can be applied to a standard slice determination server (service device), wherein the standard slice determination server can include a local standard slice determination server or a cloud standard slice determination server , the embodiments of the present invention are not limited.
  • the method for determining the slice score of the fetal standard slice may include the following operations:
  • At least one structural feature exists in the standard slice of each frame of fetal ultrasound image, and each structural feature has a corresponding weight value.
  • step 301 and step 302 for other related descriptions of step 301 and step 302 and related descriptions of other solutions expanded on the basis of step 301 and step 302, please refer to the detailed description of the related content in Embodiment 1 and Embodiment 2, This embodiment of the present invention will not be described repeatedly.
  • FIG. 4 is a schematic structural diagram of a device for determining an optimal standard section of a fetus disclosed in an embodiment of the present invention.
  • the device for determining the optimal standard aspect of the fetus described in FIG. 4 can be applied to a standard aspect determination server (service device), wherein the standard aspect determination server may include a local standard aspect determination server or a cloud standard aspect determination server.
  • the standard aspect determination server may include a local standard aspect determination server or a cloud standard aspect determination server.
  • the device for determining the optimal standard section of the fetus may include an acquisition module 401, a first determination module 402 and a second determination module 403, wherein:
  • the acquiring module 401 is configured to acquire a standard slice corresponding to each frame of fetal ultrasound images in the multiple frames of fetal ultrasound images.
  • the first determination module 402 is configured to determine the slice score of the standard slice of each frame of fetal ultrasound image.
  • the second determination module 403 is configured to determine, according to the slice scores of all standard slices, the standard slice corresponding to the highest slice score from all the standard slices, as the optimal standard slice of all fetal ultrasound images.
  • the device for determining the optimal standard section of the fetus described in FIG. 4 can automatically determine the standard section of the fetal ultrasound image without manual analysis to determine the optimal standard section of the fetal ultrasound image after acquiring the standard section of the fetal ultrasound image. It can intelligently select the standard slice with the highest slice score from all the slice scores, and realize the automatic determination of the optimal standard slice, which can improve the accuracy and efficiency of the determination of the optimal standard slice of fetal ultrasound images. In order to achieve accurate acquisition of fetal growth and development.
  • the apparatus further includes a first judgment module 404, wherein:
  • the first determination module 404 is used to determine whether all the standard slices belong to the same category of standard slices after the first determination module 402 determines the slice score of the standard slices of each frame of fetal ultrasound images, and when it is determined that all the standard slices belong to the same category
  • the second determination module 403 is triggered to perform the above-mentioned operation according to the slice scores of all standard slices, and the standard slice corresponding to the highest slice score is determined from all the standard slices as the operation of the optimal standard slice of all fetal ultrasound images .
  • implementing the determination device described in FIG. 5 can further determine that all standard slices belong to the same category of standard slices after acquiring the standard slices of all fetal ultrasound images. Accuracy of determination of standard cut planes; and determining whether all standard cut planes belong to the same category by providing categories based on structural features of standard planes and/or facet identifications of standard planes, which can enrich the way of determining standard planes that all standard planes belong to the same category, It is also possible to improve the accuracy of the determination of standard facets in which all standard facets belong to the same category.
  • the apparatus further includes a classification module 405, wherein:
  • the classification module 405 is configured to perform a classification operation on the standard slices of all fetal ultrasound images according to a preset classification method when the first determination module 404 determines that all the standard slices do not belong to the standard slices of the same category, to obtain at least two sets of standard slices , each standard slice set includes standard slices of at least one frame of fetal ultrasound images, and all standard slices included in each standard slice set are standard slices of the same category.
  • the second determination module 403 determines the standard cut plane corresponding to the highest cut plane score from all the standard cut planes according to the cut plane scores of all standard cut planes, and the method as the optimal standard cut plane is specifically:
  • the standard slice corresponding to the highest slice score is determined from all the standard slices included in each standard slice set, as the optimal standard slice corresponding to each standard slice set.
  • the implementation of the determining device described in FIG. 5 can perform a classification operation on all standard slices when it is determined that the standard slices of all fetal ultrasound images belong to different categories, and can obtain standard slices of different categories, thereby helping to reduce the number of different slice types.
  • the difference leads to the reduction of the acquisition of non-optimal standard slices, thereby improving the accuracy and reliability of the determination of the optimal standard slices corresponding to different types of standard slices.
  • the apparatus may further include a normalization module 406 and a screening module 407, wherein:
  • the normalization module 406 is configured to, in the second determination module 403, determine the standard cut plane corresponding to the highest cut plane score from all the standard cut planes included in each standard cut plane set according to all the cut plane scores corresponding to each standard cut plane set, as: After the optimal standard slice corresponding to each standard slice set, perform a normalization operation on the slice score of the optimal standard slice corresponding to each standard slice set, and obtain the normalized optimal standard slice corresponding to each standard slice set. Later slice scores.
  • the screening module 407 is configured to screen the standard section corresponding to the highest normalized section score from all the standard sections according to the normalized scores of all sections, as the optimal standard section corresponding to all fetal ultrasound images.
  • implementing the determination device described in FIG. 5 can further perform normalization operations on the slice scores of the optimal standard slices corresponding to different types of standard slices after obtaining the optimal standard slices corresponding to different types of standard slices, It can make the slice scores of the optimal standard slices corresponding to different types of standard slices comparable, thereby improving the accuracy and efficiency of determining the optimal standard slices corresponding to all fetal ultrasound images, thereby further facilitating accurate acquisition of fetal growth and development. condition.
  • the apparatus further includes a training module 408, wherein:
  • the acquiring module 401 is further configured to acquire a positive fetal ultrasound image sample and a negative fetal ultrasound image sample, the pixel value of the positive fetal ultrasound image sample is greater than the pixel value of the negative fetal ultrasound image sample, and each positive fetal ultrasound image sample in the positive fetal ultrasound image sample.
  • the key weight value influencing factors of the structural feature of each negative fetal ultrasound image in the ultrasound image and the negative fetal ultrasound image sample include the clarity of the structural feature.
  • the training module 408 is configured to train the determined initial classification model based on the positive fetal ultrasound image samples and the negative fetal ultrasound image samples.
  • the obtaining module 401 is further configured to obtain the initial classification model after training as the determined classification model.
  • implementing the determination device described in FIG. 5 can perform training operations on the initial classification model based on the sample fetal ultrasound images in advance, and can obtain a classification model that meets the requirements and is accurate, thereby improving the clarity of key weight value influencing factors including structural features.
  • the analysis accuracy and reliability of the sub-weight values can improve the calculation accuracy and efficiency of the weight values corresponding to the structural features.
  • the apparatus further includes a second judgment module 409, wherein:
  • the second determination module 409 is configured to, after the first determination module 402 determines the slice score of the standard slices of each frame of fetal ultrasound images, according to the structural features of all the standard slices, determine whether there are structural features in all the standard slices that are abnormal structural features When it is judged that there is no abnormal standard section in all the standard sections, trigger the second determination module 403 to execute the above-mentioned according to all the section scores, from all the standard sections to determine the standard section corresponding to the highest section score, as Optimal standard slice operations.
  • implementing the determination device described in FIG. 5 can further determine whether there are abnormal standard slices in all standard slices after acquiring the slice scores of the standard slices of the fetal ultrasound image, and if not, continue to perform the optimal standard slices.
  • the determination operation can improve the determination accuracy of executing the optimal standard slice, thereby improving the determination accuracy of the optimal standard slice.
  • the apparatus further includes a correction module 410, wherein:
  • the second determination module 403 is further configured to determine a score correction coefficient corresponding to each abnormal standard slice when the second determination module 409 determines that there is at least one abnormal standard slice in all the standard slices.
  • the correction module 410 is used to correct the slice score of the abnormal standard slice based on the score correction coefficient corresponding to each abnormal standard slice, and trigger the second determination module 403 to perform the above-mentioned determination from all the standard slices according to the scores of all the slices
  • the standard slice corresponding to the highest slice score is used as the operation of the optimal standard slice.
  • the implementation of the determination device described in FIG. 5 can further determine whether there are abnormal standard slices in all standard slices after acquiring the slice scores of the standard slices of the fetal ultrasound image. Performing the correction operation on the slice score of the slice can improve the accuracy of the determination of the slice score of the abnormal standard slice, so as to reduce the situation of obtaining the non-optimal standard slice by continuing to obtain the optimal standard slice when an abnormal standard slice occurs. occurs, thereby improving the accuracy and reliability of the determination of the optimal standard slice when abnormal standard slices occur.
  • the second judgment module 409 judges, according to the structural features of all standard cut planes, whether there is an abnormal standard cut plane whose structural feature is an abnormal structural feature in all standard cut planes, specifically: :
  • each structural feature of each standard section determine whether each structural feature matches the standard section where it is located;
  • the first determining module 402 may include determining Sub-module 4021 and calculation sub-module 4022, wherein:
  • the determination sub-module 4021 is configured to determine the weight value corresponding to each structural feature of the standard slice of each frame of fetal ultrasound image.
  • the calculation sub-module 4022 is configured to calculate the slice score of the standard slice of each frame of fetal ultrasound image based on the weight value corresponding to each structural feature of each standard slice and the characteristic parameter of the structural characteristic.
  • the implementation of the determination device described in FIG. 6 can realize the automatic calculation of the section score of the standard section by combining the weight value of each structural feature of the standard section with the feature parameter of the structural feature, and improve the section score of the standard section.
  • the calculation accuracy and efficiency of the value are conducive to the automatic determination of the optimal standard section, so as to accurately obtain the growth and development of the fetus.
  • the determination sub-module 4021 includes a determination unit 40211 and a calculation unit 40212, wherein:
  • the determining unit 40211 is used to determine the key weight value influencing factor corresponding to each structural feature of the standard slice of each frame of fetal ultrasound image, the number of key weight value influencing factors corresponding to each structural feature is greater than or equal to 1, and each key weight There is a corresponding sub-weight value for the value influence factor.
  • the determining unit 40211 is further configured to determine, according to each key weight value influencing factor corresponding to each structural feature, a sub-weight value corresponding to each key weight value influencing factor.
  • the calculation unit 40212 is configured to calculate the sum of all sub-weight values corresponding to each structural feature as the weight value corresponding to each structural feature.
  • implementing the determination device described in FIG. 7 can determine the key weight value influencing factor corresponding to each structural feature in a targeted manner, and determine the sub-weight values corresponding to all key weight value influencing factors as the weight value corresponding to the structural feature, The calculation accuracy of the weight value of the structural feature can be improved, thereby improving the calculation accuracy of the section score corresponding to the standard section, and further improving the determination accuracy of the optimal standard section.
  • the determining unit 40211 determines, according to each key weight value influencing factor corresponding to each structural feature, the specific manner of determining the sub-weight value corresponding to each key weight value influencing factor for:
  • the key weight value influencing factor corresponding to the structural feature includes the geometric parameters of the contour of the structural feature
  • the sub-weight value corresponding to the geometric parameters of the contour of the structural feature is determined according to the geometric parameters of the contour of the structural feature
  • the geometric parameters of the outline of the structural feature include the size and/or area of the outline of the structural feature
  • the key weight value influencing factor corresponding to the structural feature includes the clarity of the structural feature
  • the sub-weight value corresponding to the clarity of the structural feature
  • the geometric parameters corresponding to the structural feature are calculated according to the outline of the structural feature, and the structure is determined according to the geometric parameters corresponding to the structural feature.
  • the sub-weight value corresponding to the integrity of the feature is calculated according to the outline of the structural feature, and the structure is determined according to the geometric parameters corresponding to the structural feature.
  • the impact factor of the key weight value corresponding to the structural feature includes the position of the structural feature on the standard slice
  • the area of the area enclosed by the contour of the structural feature is calculated, and based on the midline and midline of the brain corresponding to the structural feature
  • the relative positional relationship of the area of the area enclosed by the contour of the structural feature determines the sub-weight value corresponding to the position of the structural feature in the standard section.
  • implementing the determination device described in FIG. 7 can select the corresponding sub-weight value determination method according to different key weight value influence factors, which can not only realize the acquisition of the sub-weight value corresponding to the key weight value influence factor, but also improve the sub-weight value.
  • the acquisition efficiency and accuracy of the weight value can improve the calculation accuracy and efficiency of the weight value corresponding to the structural feature, thereby improving the calculation accuracy and efficiency of the section score corresponding to the standard section.
  • the second determination module 403 is further configured to determine the proportion of the target features of each frame of fetal ultrasound images, and the proportion of the target features is used to indicate the relationship between the target features and the location of the target features.
  • the display scale of the display device, the target feature of each frame of fetal ultrasound image includes the standard slice of the fetal ultrasound image or the structural feature in the standard slice of the fetal ultrasound image.
  • the second determination module 403 is further configured to, when the target feature of each frame of fetal ultrasound image is the standard slice of the fetal ultrasound image and after the first determination module 402 determines the slice score of the standard slice of each frame of fetal ultrasound image, determine The score coefficient corresponding to the proportion of standard slices in each frame of fetal ultrasound images.
  • the acquisition module 401 is further configured to correct the slice score of the standard slice of the fetal ultrasound image based on the fractional coefficient corresponding to the proportion of the standard slice of each frame of fetal ultrasound image, to obtain the corrected standard slice of each frame of the fetal ultrasound image.
  • the slice score, and triggering the second determination module 403 to execute the above-mentioned slice score according to all standard slices, and determine the standard slice corresponding to the highest slice score from all the standard slices, as the operation of the optimal standard slice of all fetal ultrasound images .
  • the determination device described in FIG. 5 is further based on the obtained standard slice of the fetal ultrasound image. Updating the slice score is beneficial to improve the accuracy and reliability of the determination of the slice score of the standard slice of the fetal ultrasound image, thereby improving the accuracy and reliability of the determination of the optimal standard slice of the fetal ultrasound image.
  • the determining unit 40211 determines, according to each key weight value influencing factor corresponding to each structural feature, the specific manner of determining the sub-weight value corresponding to each key weight value influencing factor for:
  • a sub-weight value matching the proportion of the structural feature is determined according to the proportion of the structural feature.
  • implementing the determination device described in FIG. 7 can determine the weight value corresponding to the structural feature by calculating the proportion of the structural feature in the standard section of the fetal ultrasound image, and by determining the sub-weight value corresponding to the proportion, and Increasing the calculation dimension of the weight value corresponding to the structural feature can further improve the calculation accuracy and reliability of the weight value corresponding to the structural feature, thereby improving the accuracy and reliability of determining the slice score of the standard slice of the fetal ultrasound image, thereby improving the fetal ultrasound image. Accuracy and reliability of optimal standard slice determination of ultrasound images.
  • the manner in which the determining unit 40211 calculates the geometric parameters corresponding to the structural feature according to the outline of the structural feature is specifically:
  • implementing the determination device described in FIG. 7 can select different fitting methods according to the size of the arc radius of the structural features of the fetal ultrasound image, which can not only realize the fitting of the structural features, but also improve the fitting efficiency of the structural features. and accuracy, thereby improving the computational accuracy of geometric parameters of structural features.
  • the feature parameter of each structural feature of each standard slice includes a class probability of the structural feature and a position probability of the structural feature.
  • the calculation sub-module 4022 calculates the slice score of the standard slice of each frame of fetal ultrasound image based on the corresponding weight value of each structural feature of each standard slice and the characteristic parameter of the structural characteristic. Specifically for:
  • the structural score corresponding to each structural feature of each standard section is calculated;
  • the sum of the structural scores corresponding to all structural features of each standard slice is calculated as the slice score of the standard slice of each frame of fetal ultrasound image.
  • the implementation of the determination device described in FIG. 6 can realize the calculation of the section score of the standard section by separately calculating the structural score corresponding to each structural feature of the standard section, which is beneficial to improve the accuracy of the calculation of the section score of the standard section. and efficiency; and selecting different parameters according to different structural features can improve the calculation accuracy and efficiency of the structural score corresponding to the structural feature, thereby further improving the accuracy and efficiency of the calculation of the section score of the standard section.
  • FIG. 8 is another device for determining an optimal standard slice of a fetus disclosed in an embodiment of the present invention.
  • the device for determining the optimal standard aspect of the fetus described in FIG. 8 can be applied to a standard aspect determination server (service device), wherein the standard aspect determination server may include a local standard aspect determination server or a cloud standard aspect determination server.
  • the standard aspect determination server may include a local standard aspect determination server or a cloud standard aspect determination server.
  • the device for determining the optimal standard section of the fetus may include:
  • a memory 801 storing executable program code
  • processor 802 coupled to the memory 801;
  • an input interface 803 coupled with the processor 802 and an output interface 804;
  • the processor 802 calls the executable program code stored in the memory 801 to execute some or all of the steps in the method for determining the optimal standard slice of the fetus described in Embodiment 1 or Embodiment 2.
  • An embodiment of the present invention discloses a computer-readable storage medium, which stores a computer program for electronic data exchange, wherein the computer program enables a computer to perform the determination of the optimal standard slice of the fetus described in the first embodiment or the second embodiment some or all of the steps in the method.
  • An embodiment of the present invention discloses a computer program product.
  • the computer program product includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to cause a computer to execute the description in the first embodiment or the second embodiment. Part or all of the steps in the method for determining the optimal standard section of the fetus.
  • modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in One place, or it can be distributed over multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution in this embodiment. Those of ordinary skill in the art can understand and implement it without creative effort.
  • Read-Only Memory ROM
  • Random Access Memory Random Access Memory
  • PROM Programmable Read-only Memory
  • EPROM Erasable Programmable Read Only Memory
  • OTPROM One-time Programmable Read-Only Memory
  • EEPROM Electronically Erasable Programmable Read-Only Memory
  • CD-ROM Compact Disc Read -Only Memory
  • the method for determining the optimal standard section of the fetus disclosed in the embodiments of the present invention and the method and device for determining the optimal standard section of the fetus are only the preferred embodiments of the present invention, and are only used to illustrate the present invention.
  • the technical solution of the invention is not intended to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that it is still possible to modify the technical solutions recorded in the foregoing embodiments, Or equivalently replace some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Molecular Biology (AREA)
  • Biophysics (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Pathology (AREA)
  • Radiology & Medical Imaging (AREA)
  • Biomedical Technology (AREA)
  • Veterinary Medicine (AREA)
  • Medical Informatics (AREA)
  • Physics & Mathematics (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physiology (AREA)
  • Gynecology & Obstetrics (AREA)
  • Pregnancy & Childbirth (AREA)
  • Ultra Sonic Daignosis Equipment (AREA)

Abstract

一种胎儿最优标准切面的确定方法及装置，方法包括获取多帧胎儿超声图像中每帧胎儿超声图像的标准切面（101），确定每帧胎儿超声图像的标准切面的切面分值（102）；根据所有切面分值，从所有标准切面中确定最高切面分值对应的标准切面，作为最优标准切面（103）。在获取到胎儿超声图像的标准切面之后，无需人工分析以确定胎儿超声图像的最优标准切面，能够自动确定胎儿超声图像的标准切面的切面分值，并智能化地从所有切面分值选取最高切面分值的标准切面，实现最优标准切面的自动确定，能够提高胎儿超声图像的最优标准切面的确定准确性以及效率，从而实现准确获取胎儿的生长发育情况。

Description

一种胎儿最优标准切面的确定方法及装置 技术领域
本发明涉及图像处理技术领域，尤其涉及一种胎儿最优标准切面的确定方法及装置。
背景技术
由于可以从胎儿标准切面，尤其是胎儿最优标准切面，知晓胎儿的发育情况，因此，胎儿最优标准切面成为胎儿生长发育情况准确确定的关键点。目前胎儿最优标准切面的确定方法为：通过分析单张胎儿超声图片，得到初步胎儿标准切面，进一步的，在得到初步胎儿标准切面之后，由具有经验的工作人员分析该初步胎儿标准切面，从而完成胎儿最优标准切面的最终确定。
然而，实践发现，由于是直接从数据量较少的单张超声图片确定初步胎儿标准切面以及由于工作人员的经验有限和/或疲劳工作，这很容易导致确定出来的胎儿最优标准切面的准确性较低，从而无法准确确定胎儿的生长发育情况。因此，如何获取到准确的最优胎儿标准切面，从而实现胎儿的生长发育情况的准确确定显得尤为重要。
发明内容
本发明所要解决的技术问题在于，提供一种胎儿最优标准切面的确定方法及装置，能够获取到准确的最优胎儿标准切面，从而实现胎儿的生长发育情况的准确确定。
为了解决上述技术问题，本发明第一方面公开了一种胎儿最优标准切面的确定方法，所述方法包括：
获取多帧胎儿超声图像中每帧所述胎儿超声图像的标准切面，并确定每帧所述胎儿超声图像的标准切面的切面分值；
根据所有所述标准切面的切面分值，从所有所述标准切面中确定最高切面分值对应的标准切面，作为所有所述胎儿超声图像的最优标准切面。
作为一种可选的实施方式，在本发明第一方面中，所述确定每帧所述胎儿超声图像的标准切面的切面分值之后，所述方法还包括：
判断所有所述标准切面是否属于同一类别的标准切面；
当判断出所有所述标准切面属于同一类别的标准切面时，触发执行所述的根据所有所述标准切面的切面分值，从所有所述标准切面中确定最高切面分值对应的标准切面，作为所有所述胎儿超声图像的最优标准切面的操作。
作为一种可选的实施方式，在本发明第一方面中，当判断出所有所述标准切面不属于同一类别的标准切面时，按照预设分类方式对所有所述胎儿超声图像的标准切面执行分类操作，得到至少两个标准切面集合，每个所述标准切面集合包括至少一帧所述胎儿超声图像的标准切面，且每个所述标准切面集合包括的所有所述标准切面为同一类别的标准切面；
其中，所述根据所有所述标准切面的切面分值，从所有所述标准切面中确定最高切面分值对应的标准切面，作为最优标准切面，包括：
根据每个所述标准切面集合对应的所有切面分值，从每个所述标准切面集合包括的所有所述标准切面中确定最高切面分值对应的标准切面，作为每个所述标准切面集合对应的最优标准切面。
作为一种可选的实施方式，在本发明第一方面中，所述根据每个所述标准切面集合对应的所有切面分值，从每个所述标准切面集合包括的所有 所述标准切面中确定最高切面分值对应的标准切面，作为每个所述标准切面集合对应的最优标准切面之后，所述方法还包括：
对每个所述标准切面集合对应的最优标准切面的切面分值执行归一化操作，得到每个所述标准切面集合对应的最优标准切面归一化后的切面分值；
根据归一化后的所有所述切面分值，从所有所述标准切面中筛选最高归一化后的切面分值对应的标准切面，作为所有所述胎儿超声图像对应的最优标准切面。
作为一种可选的实施方式，在本发明第一方面中，每帧所述胎儿超声图像的标准切面内存在至少一个结构特征，每个所述结构特征均存在对应的权重值；
以及，所述确定每帧所述胎儿超声图像的标准切面的切面分值，包括：
确定每帧所述胎儿超声图像的标准切面的每个所述结构特征对应的权重值；
基于每个所述标准切面的每个所述结构特征对应的权重值以及该结构特征的特征参数，计算每帧所述胎儿超声图像的标准切面的切面分值。
作为一种可选的实施方式，在本发明第一方面中，所述确定每帧所述胎儿超声图像的标准切面的每个所述结构特征对应的权重值，包括：
确定每帧所述胎儿超声图像的标准切面的每个所述结构特征对应的关键权重值影响因子，每个所述结构特征对应的关键权重值影响因子的数量大于等于1，且每个所述关键权重值影响因子存在对应的子权重值；
根据每个所述结构特征对应的每个所述关键权重值影响因子，确定每个所述关键权重值影响因子对应的子权重值，并计算每个所述结构特征对应的所有所述子权重值之和，作为每个所述结构特征对应的权重值。
本发明第二方面公开了一种胎儿最优标准切面的确定装置，所述装置包括：
获取模块，用于获取多帧胎儿超声图像中每帧所述胎儿超声图像对应的标准切面；
第一确定模块，用于确定每帧所述胎儿超声图像的标准切面的切面分值；
第二确定模块，用于根据所有所述标准切面的切面分值，从所有所述标准切面中确定最高切面分值对应的标准切面，作为所有所述胎儿超声图像的最优标准切面。
作为一种可选的实施方式，在本发明第二方面中，所述装置还包括：
第一判断模块，用于在所述第一确定模块确定每帧所述胎儿超声图像的标准切面的切面分值之后，判断所有所述标准切面是否属于同一类别的标准切面，当判断出所有所述标准切面属于同一类别的标准切面时，触发所述第二确定模块执行所述的根据所有所述标准切面的切面分值，从所有所述标准切面中确定最高切面分值对应的标准切面，作为所有所述胎儿超声图像的最优标准切面的操作。
作为一种可选的实施方式，在本发明第二方面中，所述装置还包括：
分类模块，用于当所述第一判断模块判断出所有所述标准切面不属于同一类别的标准切面时，按照预设分类方式对所有所述胎儿超声图像的标准切面执行分类操作，得到至少两个标准切面集合，每个所述标准切面集合包括至少一帧所述胎儿超声图像的标准切面，且每个所述标准切面集合包括的所有所述标准切面为同一类别的标准切面；
其中，所述第二确定模块根据所有所述标准切面的切面分值，从所有所述标准切面中确定最高切面分值对应的标准切面，作为最优标准切面的方式具体为：
根据每个所述标准切面集合对应的所有切面分值，从每个所述标准切 面集合包括的所有所述标准切面中确定最高切面分值对应的标准切面，作为每个所述标准切面集合对应的最优标准切面。
作为一种可选的实施方式，在本发明第二方面中，所述装置还包括：
归一化模块，用于在所述第二确定模块根据每个所述标准切面集合对应的所有切面分值，从每个所述标准切面集合包括的所有所述标准切面中确定最高切面分值对应的标准切面，作为每个所述标准切面集合对应的最优标准切面之后，对每个所述标准切面集合对应的最优标准切面的切面分值执行归一化操作，得到每个所述标准切面集合对应的最优标准切面归一化后的切面分值；
筛选模块，用于根据归一化后的所有所述切面分值，从所有所述标准切面中筛选最高归一化后的切面分值对应的标准切面，作为所有所述胎儿超声图像对应的最优标准切面。
作为一种可选的实施方式，在本发明第二方面中，每帧所述胎儿超声图像的标准切面内存在至少一个结构特征，每个所述结构特征均存在对应的权重值；
以及，所述第一确定模块包括：
确定子模块，用于确定每帧所述胎儿超声图像的标准切面的每个所述结构特征对应的权重值；
计算子模块，用于基于每个所述标准切面的每个所述结构特征对应的权重值以及该结构特征的特征参数，计算每帧所述胎儿超声图像的标准切面的切面分值。
作为一种可选的实施方式，在本发明第二方面中，所述确定子模块包括：
确定单元，用于确定每帧所述胎儿超声图像的标准切面的每个所述结构特征对应的关键权重值影响因子，每个所述结构特征对应的关键权重值影响因子的数量大于等于1，且每个所述关键权重值影响因子存在对应的子权重值；
所述确定单元，还用于根据每个所述结构特征对应的每个所述关键权重值影响因子，确定每个所述关键权重值影响因子对应的子权重值；
计算单元，用于计算每个所述结构特征对应的所有所述子权重值之和，作为每个所述结构特征对应的权重值。
与现有技术相比，本发明实施例具有以下有益效果：
本发明实施例中，提供了一种胎儿最优标准切面的确定方法及装置，该方法包括获取多帧胎儿超声图像中每帧胎儿超声图像的标准切面，确定每帧胎儿超声图像的标准切面的切面分值；根据所有切面分值，从所有标准切面中确定最高切面分值对应的标准切面，作为所有胎儿超声图像的最优标准切面。可见，实施本发明在获取到胎儿超声图像的标准切面之后，无需人工分析以确定胎儿超声图像的最优标准切面，能够自动确定胎儿超声图像的标准切面的切面分值，并智能化地从所有切面分值选取最高切面分值的标准切面，实现最优标准切面的自动确定，能够提高胎儿超声图像的最优标准切面的确定准确性以及效率，从而实现准确获取胎儿的生长发育情况。
附图说明
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是本发明实施例公开的一种胎儿最优标准切面的确定方法的流程示意图；
图2是本发明实施例公开的另一种胎儿最优标准切面的确定方法的流程示意图；
图3是本发明实施例公开的一种胎儿标准切面的切面分值的确定方法的流程示意图；
图4是本发明实施例公开的一种胎儿最优标准切面的确定装置的结构示意图；
图5是本发明实施例公开的另一种胎儿最优标准切面的确定装置的结构示意图；
图6是本发明实施例公开的一种第一确定模块的结构示意图；
图7是本发明实施例公开的另一种第一确定模块的结构示意图；
图8是本发明实施例公开的又一种胎儿最优标准切面的确定装置的结构示意图。
具体实施方式
为了使本技术领域的人员更好地理解本发明方案，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
本发明的说明书和权利要求书及上述附图中的术语“第一”、“第二”等是用于区别不同对象，而不是用于描述特定顺序。此外，术语“包括”和“具有”以及它们任何变形，意图在于覆盖不排他的包含。例如包含了一系列步骤或单元的过程、方法、装置、产品或设备没有限定于已列出的步骤或单元，而是可选地还包括没有列出的步骤或单元，或可选地还包括对于这些过程、方法、产品或设备固有的其他步骤或单元。
在本文中提及“实施例”意味着，结合实施例描述的特定特征、结构或特性可以包含在本发明的至少一个实施例中。在说明书中的各个位置出现该短语并不一定均是指相同的实施例，也不是与其它实施例互斥的独立的或备选的实施例。本领域技术人员显式地和隐式地理解的是，本文所描述的实施例可以与其它实施例相结合。
本发明公开了一种胎儿最优标准切面的确定方法及装置，能够在获取到胎儿超声图像的标准切面之后，无需人工分析以确定胎儿超声图像的最优标准切面，能够自动确定胎儿超声图像的标准切面的切面分值，并智能化地从所有切面分值选取最高切面分值的标准切面，实现最优标准切面的自动确定，能够提高胎儿超声图像的最优标准切面的确定准确性以及效率，从而实现准确获取胎儿的生长发育情况。以下分别进行详细说明。
实施例一
请参阅图1，图1是本发明实施例公开的一种胎儿最优标准切面的确定方法的流程示意图。其中，图1所描述的胎儿最优标准切面的确定方法可以应用于标准切面确定服务器(服务设备)中，其中，该标准切面确定服务器可以包括本地标准切面确定服务器或云标准切面确定服务器，本发明实施例不做限定。如图1所示，该胎儿最优标准切面的确定方法可以包括以下操作：
101、获取多帧胎儿超声图像中每帧胎儿超声图像的标准切面。
本发明实施例中，作为一种可选的实施方式，获取多帧胎儿超声图像中每帧胎儿超声图像的标准切面，可以包括：
将获取到的连续多帧胎儿超声图像中每帧胎儿超声图像依次输入预先确定出的特征检测模型中进行分析；
获取特征检测模型依次输出的分析结果，作为每帧胎儿超声图像的特征信息，每帧胎儿超声图像的特征信息包括该胎儿超声图像的部位特征信 息以及该胎儿超声图像的结构特征信息，每帧胎儿超声图像的部位特征信息至少包括该胎儿超声图像的部位特征的类别，每帧胎儿超声图像的结构特征信息至少包括该胎儿超声图像的结构特征的类别，每个胎儿超声图像的结构特征至少包括该胎儿超声图像的关键结构特征；
根据每帧胎儿超声图像的部位特征的类别以及该胎儿超声图像的结构特征的类别确定该胎儿超声图像对应的标准切面。
该可选的实施方式中，可以按照预先确定出的帧率连续获取多帧胎儿超声图像，其中，预先确定出的帧率与所需获取的胎儿超声图像的标准切面有关，即根据所需获取的胎儿超声图像的标准切面来选择帧率，例如：若需要获取的是腹围切面，则帧率可以为30帧/秒；若需要获取的是四腔心切面，则帧率可以为60帧/秒。这样根据所需获取的胎儿超声图像的标准切面选择对应的帧率，有利于提高所需胎儿超声图像的标准切面的获取效率以及准确性。
本发明实施例中，每帧胎儿超声图像均存在唯一对应的帧序号。这样通过为每帧胎儿超声图像设定唯一的帧序号，能够在胎儿超声图像的标准切面的获取过程中，清楚区分每帧胎儿超声图像，以及有利于对胎儿超声图像及其标准切面的信息的管理。
本发明实施例中，特征检测模型可以包括目标检测模型、实例分割模型以及语义分割模型等能够获取到胎儿超声图像的部位特征信息以及结构特征信息中的至少一种，本发明实施例不做限定。
可见，该可选的实施方式通过获取连续多帧胎儿超声图像的部位特征以及结构特征，并结合胎儿超声图像的部位特征以及结构特征，确定胎儿超声图像的标准切面，无需人工参与胎儿超声图像的标准切面的确定，能够提高胎儿超声图像的标准切面的确定准确性；以及通过将胎儿超声图像输入特征检测模型进行分析，还能够提高胎儿超声图像的标准切面的确定效率。
本发明实施例中，进一步可选的，也可以通过接收授权终端设备发送的多帧胎儿超声图像中每帧胎儿超声图像的标准切面，来实现胎儿超声图像的标准切面的获取。这样通过多种途径获取胎儿超声图像的标准切面，能够丰富标准切面的获取方式，提高标准切面的获取可能性。
102、确定每帧胎儿超声图像的标准切面的切面分值。
在一个可选的实施例中，该方法还可以包括以下操作：
确定每帧胎儿超声图像的目标特征的占比，该目标特征的占比用于表示目标特征与所在显示装置的显示比例，每帧胎儿超声图像的目标特征包括该胎儿超声图像的标准切面或该胎儿超声图像的标准切面中的结构特征；
以及，当每帧胎儿超声图像的目标特征为该胎儿超声图像的标准切面时，在执行完毕步骤102之后，该方法还可以包括以下操作：
确定每帧胎儿超声图像的标准切面的占比对应的分值系数，并基于每帧胎儿超声图像的标准切面的占比对应的分值系数更正该胎儿超声图像的标准切面的切面分值，得到更正后的每帧胎儿超声图像的标准切面的切面分值，以及触发执行步骤103。
该可选的实施例中，可选的，可以通过计算标准切面的轮廓所围成的面积和/或该标准切面的轮廓上距离最远的两个端点之间的距离值，计算标准切面的占比，这样能够提高标准切面的占比的计算准确性以及可靠性。优先选择标准切面的轮廓所围成的面积计算结构特征的占比，例如：腹围切面的轮廓所围成的面积占所在显示屏幕面积的三分之二时，则腹围切面的占比对应的分值修正系数为1。
可见，该可选的实施例在得到胎儿超声图像的标准切面的切面分值之后，进一步根据获取到胎儿超声图像的标准切面与当前显示装置的显示区域的占比对应的分值系数来更新切面分值，有利于提高胎儿超声图像的标 准切面的切面分值的确定准确性以及可靠性，进而提高胎儿超声图像的最优标准切面确定准确性以及可靠性。
本发明实施例中，作为一种可选的实施方式，确定每帧胎儿超声图像的标准切面的切面分值，可以包括：
确定每帧胎儿超声图像的标准切面的每个结构特征对应的权重值；
基于每个标准切面的每个结构特征对应的权重值以及该结构特征的特征参数，计算每帧胎儿超声图像的标准切面的切面分值。
在该可选的实施方式中，每帧胎儿超声图像的标准切面内存在至少一个结构特征，每个结构特征均存在对应的权重值。其中，每个标准切面内的结构特征至少包括该标准切面的关键结构特征(又称基础结构特征或者主要结构特征)，进一步的，每个标准切面内的结构特征还可以包括除关键结构特征之外的其他结构特。例如：丘脑标准切面至少包括透明隔腔、丘脑和侧脑室中的至少一个关键结构特征，进一步的，丘脑标准切面还可以包括脉络膜从和大脑外侧裂中的至少一个其他结构特征。这样标准切面内的结构特征越多，越有利于提高标准切面的切面分值的计算准确性以及可靠性，从而有利于提高最优标准切面的确定准确性以及可靠性。其中，每个标准切面的关键结构特征为能够表示该标准切面的结构特征，即当获取到胎儿超声图像的关键结构特征时，即可确定关键结构特征对应的标准切面。例如：当胎儿超生图像的结构特征为胃泡、脐静脉时，则胎儿超生图像的标准切面为腹围切面。这样通过关键结构特征确定胎儿超生图像的标准切面，能够在保证正确确定标准切面的同时，提高标准切面的确定效率。
可见，该可选的实施方式通过将标准切面的每个结构特征的权重值与该结构特征的特征参数结合，能够实现标准切面的切面分值的自动计算，提高标准切面的切面分值的计算准确性以及效率。
在该可选的实施方式中，进一步可选的，确定每帧胎儿超声图像的标准切面的每个结构特征对应的权重值，可以包括：
确定每帧胎儿超声图像的标准切面的每个结构特征对应的关键权重值影响因子，每个结构特征对应的关键权重值影响因子的数量大于等于1，且每个关键权重值影响因子存在对应的子权重值；
根据每个结构特征对应的每个关键权重值影响因子，确定每个关键权重值影响因子对应的子权重值，并计算每个结构特征对应的所有子权重值之和，作为每个结构特征对应的权重值。
该可选的实施方式中，每个标准切面的每个结构特征对应的关键权重值影响因子可能相同，也可能不相同。例如：侧脑室切面的颅骨光环结构特征的关键权重值影响因子包括颅骨光环结构特征的轮廓对应的头围尺寸、颅骨光环结构特征的轮廓的完整度以及颅骨光环结构特征的轮廓所围成的区域与脑中线的相对位置；股骨测量切面的股骨结构特征的关键权重值影响因子包括股骨结构特征的轮廓对应的长度、股骨结构特征的轮廓所围成的区域以及股骨结构特征的轮廓所围成的区域与脑中线的相对位置。
可见，该可选的实施方式通过针对性确定每个结构特征对应的关键权重值影响因子，并将所有关键权重值影响因子对应的子权重值确定为该结构特征对应的权重值，能够提高结构特征的权重值的计算准确性，从而提高对应标准切面的切面分值的计算准确性，进而提高最优标准切面的确定准确性。
在该进一步可选的实施方式中，进一步可选的，根据每个结构特征对应的每个关键权重值影响因子，确定每个关键权重值影响因子对应的子权重值，可以包括：
对于任一结构特征，当结构特征对应的关键权重值影响因子包括该结构特征的轮廓的几何参数时，根据结构特征的轮廓的几何参数，确定结构 特征的轮廓的几何参数对应的子权重值，结构特征的轮廓的几何参数包括该结构特征的轮廓的尺寸和/或面积；
对于任一结构特征，当结构特征对应的关键权重值影响因子包括该结构特征的清晰度时，将结构特征对应的胎儿超声图像输入确定出的分类模型中进行分析，并获取分类模型输出的分析结果，作为结构特征的清晰度对应的子权重值；
对于任一结构特征，当结构特征对应的关键权重值影响因子包括该结构特征的完整度时，根据结构特征的轮廓，计算结构特征对应的几何参数，并根据结构特征对应的几何参数，确定结构特征的完整度对应的子权重值；
对于任一结构特征，当结构特征对应的关键权重值影响因子包括该结构特征在所在标准切面的位置时，基于结构特征对应的脑中线与该结构特征的轮廓所围成的区域的相对位置关系，确定结构特征在所在标准切面的位置对应的子权重值；
当每帧胎儿超声图像的目标特征为该胎儿超声图像的标准切面中的结构特征时，对于任一结构特征，根据结构特征的占比，确定与结构特征的占比相匹配的子权重值。
该可选的实施方式中，可选的，可以通过计算结构特征的轮廓所围成的面积和/或该结构特征的轮廓上距离最远的两个端点之间的距离值，计算结构特征的占比，这样能够提高结构特征的占比的计算准确性以及可靠性。优先选择结构特征的轮廓所围成的面积计算结构特征的占比，例如：左心房结构特征的轮廓所围成的面积占所在显示屏幕面积的七分之一，则左心房结构特征的占比对应的子权重值为0.8。
该可选的实施方式中，该结构特征的轮廓的尺寸可以包括该结构特征的轮廓的周长和/或该结构特征的轮廓对应的长度(例如：肱骨结构特征的长度)。
该可选的实施方式中，在获取到结构特征的轮廓的几何参数之后，进一步判断该结构特征的轮廓的几何参数是否在确定出的胎儿超声图像的孕周对应的几何参数范围内，当判断出不在几何参数范围内时，则将该结构特征的轮廓的几何参数对应的子权重值乘以确定出的权重修正系数(例如：0.8)，以得到修正后的子权重值；当判断结果为是时，触发执行上述的计算每个结构特征对应的所有子权重值之和，作为每个结构特征对应的权重值的操作。例如：若当前胎儿孕周为第20周，且第20周时，胎儿的股骨长度正常为10cm-15cm，当确定出的股骨结构特征的长度为13cm时，则保持计算出的子权重值(0.7)不变，当股骨结构特征的长度为8cm或者20cm时，则将计算出的子权重值(0.7)乘以权重修正系数(0.9)，得到修正后的子权重值(0.63)。其中，权重值越高，对应的结构特征表现越明显。这样通过对不在孕周对应的正常参数范围内的结构特征的几何参数对应的子权重值，执行修正操作，能够提高对应的标准切面的切面分值的计算准确性。
该可选的实施例中，基于结构特征对应的脑中线与该结构特征的轮廓所围成的区域的相对位置关系，确定结构特征在所在标准切面的位置对应的子权重值，具体的：当结构特征对应的脑中线与该结构特征的轮廓所围成的区域存在交点时，确定该结构特征在所在标准切面的位置对应的子权重值为第一子权重值；当结构特征对应的脑中线与该结构特征的轮廓的距离处于预先确定出的距离范围值内时，确定该结构特征在所在标准切面的位置对应的子权重值为第二子权重值；当结构特征对应的脑中线与该结构特征的轮廓之间的距离大于预先确定出的距离范围值中的最大距离值时，确定该结构特征在所在标准切面的位置对应的子权重值为第三子权重值，且第一子权重值、第二子权重值以及第三子权重值依次减小。例如：当脑中线穿过脑中线小囊结构特征的轮廓所围成的区域时，子权重值为1，表示 脑中线小囊结构特征没有偏离脑中线；当脑中线小囊结构特征的轮廓与脑中线不存在交点，且偏离距离为1mm，则子权重值为0.8；当偏离距离为5mm时，则子权重值为0。
该可选的实施方式中，当结构特征对应的关键权重值影响因子的数量有多个时，对应的结构特征的权重值等于每个关键权重值影响因子对应的子权重值的和。例如：股骨测量切面的股骨结构特征的关键权重值影响因子包括股骨结构特征的轮廓对应的长度、股骨结构特征的轮廓所围成的区域以及股骨结构特征的轮廓所围成的区域与脑中线的相对位置，且股骨结构特征的轮廓对应的长度的子权重值为0.7，股骨结构特征的轮廓所围成的区域对应的子权重值为0.6，股骨结构特征的轮廓所围成的区域与脑中线的相对位置对应的子权重值为0.8，则股骨结构特征的权重值为0.7+0.6+0.8＝2.1。
可见，该可选的实施方式通过根据不同的关键权重值影响因子，选择对应的子权重值确定方式，既能够实现关键权重值影响因子对应的子权重值的获取，又能够提高子权重值的获取效率以及精准性，从而提高结构特征对应的权重值的计算精准性以及效率，进而提高对应标准切面的切面分值的计算精准性以及效率。
在该进一步可选的实施方式中，进一步可选的，根据结构特征的轮廓，计算结构特征对应的几何参数，包括：
计算结构特征的轮廓的长度，作为结构特征对应的几何参数；和/或，
确定结构特征的轮廓对应的中心点，并基于结构特征的轮廓对应的中心点以及该结构特征的轮廓，确定结构特征的轮廓对应的中心角，作为结构特征对应的几何参数；和/或，
基于确定出的拟合方法拟合结构特征的轮廓，得到结构特征的目标轮廓；
计算结构特征的轮廓与该结构特征的目标轮廓的重叠部分轮廓的长度，作为结构特征对应的几何参数，和/或，确定结构特征的目标轮廓对应的中心点，并基于结构特征的目标轮廓对应的中心点以及重叠部分轮廓，确定结构特征的轮廓对应的中心角，作为结构特征对应的几何参数。
该可选的实施方式中，可选的，每个结构特征的轮廓均对应多个节点，基于确定出的拟合方法拟合结构特征的轮廓，得到结构特征的目标轮廓，可以包括：
获取每个结构特征的轮廓对应的圆弧半径；
当每个结构特征的轮廓对应的圆弧半径大于等于确定出的圆弧半径阈值(例如：5mm)时，从该结构特征对应的所有节点中选取预设数量(例如：50个等)的目标节点，并按照每相邻两个节点进行连接的方式将每个结构特征对应的所有目标节点依次连接起来，得到结构特征的目标轮廓；
当每个结构特征的轮廓对应的圆弧半径不大于等于确定出的圆弧半径阈值时，按照每相邻两个节点进行连接的方式将每个结构特征对应的所有节点依次连接起来，得到结构特征的目标轮廓。
该可选的实施例中，当结构特征的轮廓存在多个圆弧和/或轮廓的曲率大于等于确定出的曲率阈值时，对该结构特征的轮廓进行分段执行拟合操作。具体的：当结构特征的轮廓存在多个圆弧时，将分别对该结构特征的多个圆弧中每个圆弧执行拟合操作；当结构特征的轮廓的曲率大于等于曲率阈值时，将结构特征的轮廓等间隔或非等间隔分成多段，并分别对每个每段轮廓执行拟合操作。这样在结构特征的轮廓存在多个圆弧和/或轮廓的曲率较大时，通过对结构特征的轮廓分段执行拟合操作，能够提高结构特征的轮廓的拟合效率以及准确性，从而有利于进一步提高胎儿超声图像的结构特征的几何参数的测量准确性以及可靠性。
该可选的实施方式中，还可以基于确定出的B样条曲线拟合方式和/或 椭圆拟合方式对每个结构特征的轮廓执行拟合操作，得到结构特征的目标轮廓，该可选的实施方式不做限定。
可见，该可选的实施方式通过根据胎儿超声图像的结构特征的圆弧半径的大小选择不同的拟合方式，不仅能够实现结构特征的拟合，还能够提高结构特征的拟合效率以及准确性，从而提高结构特征的几何参数的计算准确性。
该可选的实施方式中，计算结构特征的轮廓与该结构特征的目标轮廓的重叠部分轮廓的长度，作为结构特征对应的几何参数之后，该方法还包括：
计算结构特征的目标轮廓的重叠部分轮廓的长度与目标轮廓的周长的比值，并将结构特征对应的几何参数更新为该比值。其中，不同的比值对应不同的子权重值，例如：当比值大于等于0.8时，则对应的子权重值为1；当比值小于0.8时，则对应的子权重值为0.8。这样通过将结构特征对应的几何参数更新为结构特征的目标轮廓的重叠部分轮廓的长度与目标轮廓的周长的比值，有利于提高子权重值的确定准确性，从而提高结构特征的权重值的计算准确性。
该可选的实施方式中，确定结构特征的轮廓对应的中心角，作为结构特征对应的几何参数之后，该方法还包括：
计算结构特征的轮廓对应的中心角与360°圆心角的比值，并将结构特征对应的几何参数更新为结构特征的轮廓对应的中心角与360°圆心角的比值。
可见，该可选的实施方式通过提供多种方式确定结构特征对应的几何参数，能够丰富结构特征对应的几何参数的获取方式，提高结构特征对应的几何参数的获取可能性；以及通过将结构特征的轮廓的长度、结构特征的轮廓对应的中心角、结构特征的轮廓与拟合后的轮廓的重叠部分轮廓的长度以及重叠部分轮廓对应的中心角中的一种或者组合作为结构特征对应的几何参数，能够提高结构特征对应的几何参数的获取准确性，从而提高结构特征对应的权重值的计算精准性。
在该进一步可选的实施方式中，又进一步可选的，基于每个标准切面的每个结构特征对应的权重值以及该结构特征的特征参数，计算每帧胎儿超声图像的标准切面的切面分值，可以包括：
基于每个标准切面的每个结构特征对应的权重值、该结构特征的类别概率以及该结构特征的位置概率，计算每个标准切面的每个结构特征对应的结构分值；
计算每个标准切面的所有结构特征对应的结构分值之和，作为每帧胎儿超声图像的标准切面的切面分值。
该可选的实施方式中，每个标准切面的每个结构特征的特征参数包括该结构特征的类别概率以及该结构特征的位置概率。
该可选的实施方式中，每帧胎儿超声图像的标准切面的切面分值的计算公式如下：
Figure PCTCN2021096821-appb-000001
H i＝P i×Q i×O i；
Figure PCTCN2021096821-appb-000002
式中，S为每个标准切面的切面分值，H i为每个标准切面中第i个结构特征的结构分值，M为每个标准切面中结构特征的总数量，P i为每个标准切面中第i个结构特征的类别概率(又称置信度)，Q i为每个标准切面中第 i个结构特征的位置概率，O i为每个标准切面中第i个结构特征的权重值，N为第i个结构特征的关键权重值影响因子的总数量，O ij为每个标准切面中第i个结构特征中第j个关键权重值影响因子对应的子权重值。
该可选的实施方式中，进一步的，每个标准切面内结构特征包括的参数还包括该结构特征所在的部位的概率，此时，每个标准切面中第i个结构特征的结构分值的计算公式为：
H i＝P i×Q i×O i×C i；
式中，C i为每个标准切面中结构特征包括的参数还包括该结构特征所在的部位的概率。这样结构特征的参数越多，有利于提高结构特征的结构分值的计算准确性，从而提高标准切面的切面分值的计算准确性，进而有利于提高最优标准切面的确定精准性以及可靠性。
可见，该可选的实施方式通过分别计算标准切面的每个结构特征对应的结构分值，能够实现标准切面的切面分值的计算，有利于提高标准切面的切面分值计算精准性以及效率；以及根据不同的结构特征，选取不同的参数，能够提高结构特征对应的结构分值的计算精准性以及效率，从而进一步提高标准切面的切面分值计算精准性以及效率。
103、根据上述所有标准切面的切面分值，从上述所有标准切面中确定最高切面分值对应的标准切面，作为所有胎儿超声图像的最优标准切面。
在一个可选的实施例中，该胎儿最优标准切面的确定方法还可以包括以下操作：
获取正胎儿超声图像样本以及负胎儿超声图像样本，正胎儿超声图像样本的像素值大于负胎儿超声图像样本的像素值，正胎儿超声图像样本中的每个正样本胎儿超声图像以及负胎儿超声图像样本中每个负样本胎儿超声图像的结构特征的关键权重值影响因子包括该结构特征的清晰度；
基于正胎儿超声图像样本以及负胎儿超声图像样本，训练确定出的初始分类模型，并获取训练后的初始分类模型，作为确定出的分类模型。
该可选的实施例中，初始分类模型包括KNN、Bayesian、Neural Network、Ensemble-Stacking、Ensemble-Boosting以及Ensemble-Bagging等能够实现图像分类的一个或者组合形成的分类模型，该可选的实施例不做限定。
该可选的实施例中，正胎儿超声图像样本以及负胎儿超声图像样本包括的样本胎儿超声图像可以为设备终端筛选出来的，也可以为相关人员根据经验挑选出来的，或者两者共同确定出的。
该可选的实施例中，由于关键权重值影响因子包括对应结构特征的清晰度的结构特征有多种，因此正胎儿超声图像样本由多个子正胎儿超声图像样本组成，以及负胎儿超声图像样本由多个子负胎儿超声图像样本组成。其中，每个子正胎儿超声图像样本对应一个子负胎儿超声图像样本。进一步的，每个样本胎儿超声图像均存在对应的样本权重值。例如：正胎儿超声图像样本包括包含透明隔腔结构特征的子正胎儿超声图像样本和动脉导管结构特征的子正胎儿超声图像样本，负胎儿超声图像样本包括包含透明隔腔结构特征的子负胎儿超声图像样本和动脉导管结构特征的子负胎儿超声图像样本。此时，基于正胎儿超声图像样本、负胎儿超声图像样本以及每个样本胎儿超声图像对应的权重值，训练确定出的初始分类模型，并获取训练后的初始分类模型，作为确定出的分类模型。这样能够提高分类模型的训练精准性，从而得到精准度高的分类模型。
可见，该可选的实施例通过预先基于样本胎儿超声图像对初始分类模型执行训练操作，能够获取符合要求且精准的分类模型，从而提高关键权 重值影响因子包括结构特征的清晰度的子权重值的分析准确性以及可靠性，从而提高结构特征对应的权重值的计算精准性以及效率。
在另一个可选的实施例中，在执行步骤102之后，该胎儿最优标准切面的确定方法还可以包括以下操作：
判断所有标准切面是否属于同一类别的标准切面；
当判断出所有标准切面属于同一类别的标准切面时，触发执行步骤103。
该可选的实施例中，需要说明的是，判断所有标准切面是否属于同一类别的标准切面的步骤可以和步骤102同时发生，也可以发生在步骤102之前，该可选的实施例不做限定。
该可选的实施例中，作为一种可选的实施方式，判断所有标准切面是否属于同一类别的标准切面，可以包括：
判断每个标准切面的结构特征的类别是否相匹配，当判断结果为是时，确定所有标准切面属于同一类别的标准切面；当判断结果为否时，确定所有标准切面不属于同一类别的标准切面；或者，
获取每个标准切面的切面标识，并判断每个标准切面的切面标识是否相匹配，当判断结果为是时，确定所有标准切面属于同一类别的标准切面；当判断结果为否时，确定所有标准切面不属于同一类别的标准切面，其中，每个标准切面的切面标识包括切面编号和/或切面图标。
举例来说，当每个标准切面均包括胃泡结构特征时，则所有标准切面为同一类别的标准切面，且为腹围类别的标准切面。又举例来说，每个标准切面的切面编号中的前若干位均为0001，则所有标准切面为同一类别的标准切面。
该可选的实施方式，进一步的，当两个判断结果均为是时，才确定所有标准切面属于同一类别的标准切面，这样不仅能够丰富所有标准切面属于同一类别的标准切面的确定方式，还能够提高所有标准切面属于同一类别的标准切面的确定精准性。
可见，该可选的实施例在获取到所有胎儿超声图像的标准切面之后，进一步判断所有标准切面属于同一类别的标准切面，若是，则执行后续操作，能够提高胎儿超声图像的最优标准切面的确定精准性；以及通过提供基于标准切面的结构特征的类别和/或标准切面的切面标识确定所有标准切面是否属于同一类别，能够丰富所有标准切面属于同一类别的标准切面的确定方式，还能够提高所有标准切面属于同一类别的标准切面的确定精准性。
在又一个可选的实施例中，该胎儿最优标准切面的确定方法还可以包括以下操作：
当判断出所有标准切面不属于同一类别的标准切面时，按照预设分类方式对所有胎儿超声图像的标准切面执行分类操作，得到至少两个标准切面集合，每个标准切面集合包括至少一帧胎儿超声图像的标准切面，且每个标准切面集合包括的所有标准切面为同一类别的标准切面；
其中，根据所有标准切面的切面分值，从所有标准切面中确定最高切面分值对应的标准切面，作为最优标准切面，包括：
根据每个标准切面集合对应的所有切面分值，从每个标准切面集合包括的所有标准切面中确定最高切面分值对应的标准切面，作为每个标准切面集合对应的最优标准切面。
该可选的实施例中，可选的，在对所有胎儿超声图像的标准切面执行分类操作的同时，可以确定每个标准切面集合包括的所有标准切面的切面分值。
可见，该可选的实施例在判断出所有胎儿超声图像的标准切面属于不同类别时，对所有标准切面执行分类操作，能够获取到不同类别的标准切面，从而有利于减少因切面类别不同而导致降低获取到非最优标准切面的 性情况以及发可生，靠性进。而提高不同类别的标准切面对应的最优标准切面的确定精准
在又一个可选的实施例中，在根据每个标准切面集合对应的所有切面分值，从每个标准切面集合包括的所有标准切面中确定最高切面分值对应的标准切面，作为每个标准切面集合对应的最优标准切面之后，该胎儿最优标准切面的确定方法还可以包括以下操作：
对每个标准切面集合对应的最优标准切面的切面分值执行归一化操作，得到每个标准切面集合对应的最优标准切面归一化后的切面分值；
根据归一化后的所有切面分值，从所有标准切面中筛选最高归一化后的切面分值对应的标准切面，作为所有胎儿超声图像对应的最优标准切面。
该可选的实施例中，举例来说，股骨测量标准切面的最优标准切面的切面分值为20分，头颅正中标准切面的最优标准切面的切面分值为100分，胆囊脐静脉标准切面的最优标准切面的切面分值为60分，则分别对股骨测量标准切面的最优标准切面的切面分值、头颅正中标准切面的最优标准切面的切面分值以及胆囊脐静脉标准切面的最优标准切面的切面分值分别执行归一化操作，以使得每个切面分值落入0-1的范围，则归一化后的切面分值依次为0.8、0.5、0.9，则所有胎儿超声图像的最优标准切面为胆囊脐静脉标准切面的最优标准切面。
可见，该可选的实施例在获取到不同类别的标准切面对应的最优标准切面之后，进一步对不同类别的标准切面对应的最优标准切面的切面分值执行归一化操作，能够使得不同类别的标准切面对应的最优标准切面的切面分值具有可比性，从而提高所有胎儿超声图像对应的最优标准切面的确定准确性以及效率，从而进一步有利于准确获取胎儿的生长发育情况。
可见，实施图1所描述的胎儿最优标准切面的确定方法能够在获取到胎儿超声图像的标准切面之后，无需人工分析以确定胎儿超声图像的最优标准切面，能够自动确定胎儿超声图像的标准切面的切面分值，并智能化地从所有切面分值选取最高切面分值的标准切面，实现最优标准切面的自动确定，能够提高胎儿超声图像的最优标准切面的确定准确性以及效率，从而实现准确获取胎儿的生长发育情况。
实施例二
请参阅图2，图2是本发明实施例公开的另一种胎儿最优标准切面的确定方法的流程示意图。其中，图2所描述的胎儿最优标准切面的确定方法可以应用于标准切面确定服务器(服务设备)中，其中，该标准切面确定服务器可以包括本地标准切面确定服务器或云标准切面确定服务器，本发明实施例不做限定。如图2所示，该胎儿最优标准切面的确定方法可以包括以下操作：
201、获取多帧胎儿超声图像中每帧胎儿超声图像的标准切面。
202、确定每帧胎儿超声图像的标准切面的切面分值。
203、根据所有标准切面的结构特征，判断所有标准切面中是否存在结构特征为异常结构特征的异常标准切面，当判断结果为否时，可以触发执行步骤205；当判断结果为是时，可以触发执行步骤204。
本发明实施例中，步骤202和步骤203也可以同时发生。
204、基于每个异常标准切面对应的分值修正系数修正该异常标准切面的切面分值。
本发明实施例中，当执行完毕步骤204之后，触发执行步骤205，此时，步骤205中所有标准切面的切面分值包括不需要修正的所有标准切面的切面分值以及需要修正且已经修正后的所有标准切面的切面分值。
本发明实施例中，每个异常标准切面均存在对应的分值修正系数。进一步的，不同异常标准切面对应的分值修正系数可以相同，也可以不同，本发明实施例不做限定。例如：异常侧脑室标准切面对应的分值修正系数 为10，异常丘脑标准切面对应的分值修正系数为8。
本发明实施例中，可选的，分值修正系数包括切面分值修正系数、结构分值修正系数以及权重值修正系数中的至少一种。以及，作为可选的实施方式，基于每个异常标准切面对应的分值修正系数修正该异常标准切面的切面分值，可以包括：
当分值修正系数为切面分值修正系数时，对于任一异常标准切面，将异常标准切面对应的分值修正系数乘以该异常标准切面的切面分值，得到该异常标准切面的修正后的切面分值；
当分值修正系数为结构分值修正系数时，对于任一异常标准切面，将异常标准切面对应的分值修正系数乘以异常结构特征对应的结构分值，得到修正后的异常结构特征对应的结构分值，并将异常标准切面的每个结构特征(包括正常结构特征以及异常结构特征)的结构分值相加，得到该异常标准切面的修正后的切面分值；
当分值修正系数为权重值修正系数时，对于任一异常标准切面，将异常标准切面对应的分值修正系数乘以异常结构特征对应的权重值，得到修正后的异常结构特征对应的权重值，并计算该异常结构特征对应的结构分值，以及将异常标准切面的每个结构特征(包括正常结构特征以及异常结构特征)的结构分值相加，得到该异常标准切面的修正后的切面分值。
该可选的实施方式中，当存在至少两种修正方式修正异常标准切面的切面分值时，获取所有修正方式修正后的异常标准切面的切面分值的均值，作为该异常标准切面修正后的切面分值。这样能够提高异常标准切面的切面分值的修正准确性，进一步获取到精准的异常标准切面的切面分值。
可见，该可选的实施方式通过提供切面分值修正、结构分值修正以及权重值修正中的至少一种修正方式修正异常标准切面的切面分值，不仅能够丰富异常标准切面的切面分值的修正方式，还能够提高异常标准切面的切面分值的修正精准性，从而能够获取到精准的异常标准切面的切面分值，进而有利于提高最优标准切面的确定准确性以及可靠性。
205、根据上述所有标准切面的切面分值，从上述所有标准切面中确定最高切面分值对应的标准切面，作为所有胎儿超声图像的最优标准切面。
可见，本发明实施例在获取到胎儿超声图像的标准切面的切面分值之后，进一步判断所有标准切面中是否存在异常标准切面，若存在，则基于分值修正系数对异常标准切面的切面分值执行修正操作，能够提高异常标准切面的切面分值的确定准确性，以减少出现异常标准切面时却继续执行最优标准切面的获取而导致获取到非最优标准切面的情况发生，从而提高当出现异常标准切面时的最优标准切面的确定准确性以及可靠性。
本发明实施例中，针对步骤201、步骤202以及步骤205的其他描述请参阅实施例一中针对步骤101-步骤103的详细描述，本发明实施例不再赘述。
在一个可选的实施例中，根据所有标准切面的结构特征，判断所有标准切面中是否存在结构特征为异常结构特征的异常标准切面，可以包括：
获取每个标准切面的每个结构特征的目标信息，每个结构特征的目标信息用于确定该结构特征所在的标准切面是否为异常标准切面；
根据每个标准切面的每个结构特征的目标信息，判断每个结构特征是否与所在的标准切面相匹配；
当判断出所有结构特征中存在与所在的标准切面不匹配的非匹配结构特征时，确定所有标准切面中存在结构特征为异常结构特征的异常标准切面，且异常标准切面为非匹配结构特征所在的标准切面。
可见，该可选的实施例通过获取标准切面的每个结构特征的目标信息，并根据每个结构特征的目标信息判断每个结构特征是否与对应的标准切面匹配，能够实现异常标准切面的确定。
在另一个可选的实施例中，根据每个标准切面的每个结构特征的目标信息，判断每个结构特征是否与所在的标准切面相匹配，可以包括：
根据每个标准切面的每个结构特征的目标信息，确定每个结构特征对应的表示类型，该表示类型包括数值表示类型和/或特征形态表示类型；
当结构特征的表示类型为数值表示类型时，获取该结构特征对应的目标几何参数值，并判断该结构特征对应的目标几何参数值是否在预先确定出的正常参数值范围内，当判断结果否时，确定该结构特征与所在的标准切面不相匹配；
当结构特征的表示类型为特征形态表示类型时，判断该结构特征是否位于该结构特征对应的部位特征的检测区域内，当判断结果为否时，确定该结构特征与所在的标准切面不相匹配。
该可选的实施例中，每个部位特征的检测区域可以用检测框，例如：多边形框或椭圆形框，框选出来。
该可选的实施例中，每个结构特征对应的目标几何参数值包括该结构特征对应的横径和/或该结构特征对应的周长，这样几何参数值包括的内容越多，越有利于提高结构特征与所在的标准切面相匹配的判断准确性。其中，不同结构特征均存在对应的正常参数值范围，其中，不同结构特征对应的正常参数值范围可以是相同的也可以是不同的。进一步的，同一结构特征的不同几何参数值对应不同的正常参数值范围。又进一步的，每个结构特征对应的几何参数值可以包括比例尺寸和/或实际尺寸。具体的，在判断出结构特征对应的比例尺寸在预先确定出的正常参数值范围内之后，进一步获取结构特征对应的实际尺寸，并判断该实际尺寸是否在预先确定出的正常尺寸范围内，当判断结果否时，确定该结构特征与所在的标准切面不相匹配。这样通过同时将结构特征的比例尺寸以及实际尺寸与各自的正常值进行对比，能够提高结构特征与所在的标准切面是否相匹配的确定准确性，从而减少异常标准切面的切面分值的错误修正的情况发生，提高异常标准切面的切面分值的修正准确性以及可靠性。
该可选的实施例中，在判断出结构特征位于该结构特征对应的部位特征的检测区域内时，判断是否多帧胎儿超声图像均存在该结构特征，当判断结果为是时，确定该结构特征与所在的标准切面不相匹配。其中，该多帧胎儿超声图像可以为以结构特征所在胎儿超声图像为首帧胎儿超声图像往后连续或者断续出现的胎儿超声图像。这样在判断出结构特征在对应部位特征的检测区域内时，进一步判断是否存在多帧胎儿超声图像存在该结构特征，若存在，则确定该结构特征与所在的标准切面不相匹配，能够提高结构特征与所在的标准切面是否相匹配的确定准确性，从而减少异常标准切面的切面分值的错误修正的情况发生，提高异常标准切面的切面分值的修正准确性以及可靠性。
现分别对数值表示类型以及特征形态表示类型的结构特征进行举例说明：
(一)数值表示类型：当检测到的结构特征为侧脑室临界性增宽特征时，将侧脑室临界性增宽特征的轮廓信息输入测量模块进行测量，得到侧脑室临界性增宽特征的横径(比例尺寸)，并判断该横径是否大于等于12个像素，若判断结果为是，则侧脑室临界性增宽特征为异常结构特征，即侧脑室临界性增宽特征与所在的标准切面不相匹配。进一步的，在得到侧脑室临界性增宽特征的横径之后，根据该横径以及胎儿超声图像的比例尺，计算侧脑室临界性增宽特征的实际尺寸，并判断该实际尺寸是否大于等于10mm，若判断结果为是，侧脑室临界性增宽特征为异常结构特征，即侧脑室临界性增宽特征与所在的标准切面不相匹配。又进一步的，当判断出侧脑室临界性增宽特征的横径小于12个像素和/或侧脑室临界性增宽特征的实际尺寸小于10mm，确定侧脑室临界性增宽特征为正常结构特征，并将侧 脑室临界性增宽特征修改为正常侧脑室特征，即侧脑室临界性增宽特征与所在的标准切面相匹配。
(二)特征形态表示类型：当检测到的结构特征为脉络膜从囊肿结构特征时，检测脉络膜从囊肿结构特征是否出现在侧脑室的检测区域，当出现在侧脑室的检测区域时，确定脉络膜从囊肿结构特征为异常结构特征，即确定脉络膜从囊肿结构特征与所在的标准切面不相匹配。进一步的，在检测到脉络膜从囊肿结构特征出现在侧脑室的检测区域时，判断是否存在4帧胎儿超声图像中均存在该脉络膜从囊肿结构特征，当判断结果为是时，确定脉络膜从囊肿结构特征与所在的标准切面不相匹配。
可见，该可选的实施例在判断出胎儿超声图像出现结构特征时，通过获取到的结构特征的几何参数值，判断结构特征与所在的标准切面是否相匹配，或者通过判断结构特征是否位于对应部位特征的检测区域内，来实现结构特征与所在的标准切面是否相匹配的判断，能够提高结构特征与所在的标准切面是否相匹配的确定可能性、准确性以及确定效率。
可见，实施图2所描述的胎儿最优标准切面的确定方法能够在获取到胎儿超声图像的标准切面之后，无需人工分析以确定胎儿超声图像的最优标准切面，能够自动确定胎儿超声图像的标准切面的切面分值，并智能化地从所有切面分值选取最高切面分值的标准切面，实现最优标准切面的自动确定，能够提高胎儿超声图像的最优标准切面的确定准确性以及效率，从而实现准确获取胎儿的生长发育情况。
实施例三
请参阅图3，图3是本发明实施例公开的一种胎儿标准切面的切面分值的确定方法的流程示意图。其中，图3所描述的胎儿标准切面的切面分值的确定方法可以应用于标准切面确定服务器(服务设备)中，其中，该标准切面确定服务器可以包括本地标准切面确定服务器或云标准切面确定服务器，本发明实施例不做限定。如图3所示，该胎儿标准切面的切面分值的确定方法可以包括以下操作：
301、确定每帧胎儿超声图像的标准切面的每个结构特征对应的权重值。
本发明实施例中，每帧胎儿超声图像的标准切面内存在至少一个结构特征，每个结构特征均存在对应的权重值。
302、基于每个标准切面的每个结构特征对应的权重值以及该结构特征的特征参数，计算每帧胎儿超声图像的标准切面的切面分值。
本发明实施例中，针对步骤301以及步骤302的其他相关描述以及在步骤301以及步骤302的基础上所拓展的其他方案的相关描述请参阅实施例一以及实施例二中相关内容的详细描述，本发明实施例不在赘述。
可见，实施图3所描述的胎儿最优标准切面的确定方法能够通过将标准切面的每个结构特征的权重值与该结构特征的特征参数结合，能够实现标准切面的切面分值的自动计算，提高标准切面的切面分值的计算准确性以及效率，有利于实现最优标准切面的自动确定，从而实现准确获取胎儿的生长发育情况。
实施例四
请参阅图4，图4是本发明实施例公开的一种胎儿最优标准切面的确定装置的结构示意图。其中，图4所描述的胎儿最优标准切面的确定装置可以应用于标准切面确定服务器(服务设备)中，其中，该标准切面确定服务器可以包括本地标准切面确定服务器或云标准切面确定服务器，本发明实施例不做限定。如图4所示，该胎儿最优标准切面的确定装置可以包括获取模块401、第一确定模块402以及第二确定模块403，其中：
获取模块401，用于获取多帧胎儿超声图像中每帧胎儿超声图像对应的标准切面。
第一确定模块402，用于确定每帧胎儿超声图像的标准切面的切面分值。
第二确定模块403，用于根据所有标准切面的切面分值，从所有标准切面中确定最高切面分值对应的标准切面，作为所有胎儿超声图像的最优标准切面。
可见，实施图4所描述的胎儿最优标准切面的确定装置能够在获取到胎儿超声图像的标准切面之后，无需人工分析以确定胎儿超声图像的最优标准切面，能够自动确定胎儿超声图像的标准切面的切面分值，并智能化地从所有切面分值选取最高切面分值的标准切面，实现最优标准切面的自动确定，能够提高胎儿超声图像的最优标准切面的确定准确性以及效率，从而实现准确获取胎儿的生长发育情况。
在一个可选的实施例中，如图5所示，该装置还包括第一判断模块404，其中：
第一判断模块404，用于在第一确定模块402确定每帧胎儿超声图像的标准切面的切面分值之后，判断所有标准切面是否属于同一类别的标准切面，当判断出所有标准切面属于同一类别的标准切面时，触发第二确定模块403执行上述的根据所有标准切面的切面分值，从所有标准切面中确定最高切面分值对应的标准切面，作为所有胎儿超声图像的最优标准切面的操作。
可见，实施图5所描述的确定装置能够在获取到所有胎儿超声图像的标准切面之后，进一步判断所有标准切面属于同一类别的标准切面，若是，则执行后续操作，能够提高胎儿超声图像的最优标准切面的确定精准性；以及通过提供基于标准切面的结构特征的类别和/或标准切面的切面标识确定所有标准切面是否属于同一类别，能够丰富所有标准切面属于同一类别的标准切面的确定方式，还能够提高所有标准切面属于同一类别的标准切面的确定精准性。
在另一个可选的实施例中，如图5所示，该装置还包括分类模块405，其中：
分类模块405，用于当第一判断模块404判断出所有标准切面不属于同一类别的标准切面时，按照预设分类方式对所有胎儿超声图像的标准切面执行分类操作，得到至少两个标准切面集合，每个标准切面集合包括至少一帧胎儿超声图像的标准切面，且每个标准切面集合包括的所有标准切面为同一类别的标准切面。
其中，第二确定模块403根据所有标准切面的切面分值，从所有标准切面中确定最高切面分值对应的标准切面，作为最优标准切面的方式具体为：
根据每个标准切面集合对应的所有切面分值，从每个标准切面集合包括的所有标准切面中确定最高切面分值对应的标准切面，作为每个标准切面集合对应的最优标准切面。
可见，实施图5所描述的确定装置能够在判断出所有胎儿超声图像的标准切面属于不同类别时，对所有标准切面执行分类操作，能够获取到不同类别的标准切面，从而有利于减少因切面类别不同而导致降低获取到非最优标准切面的情况发生，进而提高不同类别的标准切面对应的最优标准切面的确定精准性以及可靠性。
在又一个可选的实施例中，如图5所示，该装置还可以包括归一化模块406以及筛选模块407，其中：
归一化模块406，用于在第二确定模块403根据每个标准切面集合对应的所有切面分值，从每个标准切面集合包括的所有标准切面中确定最高切面分值对应的标准切面，作为每个标准切面集合对应的最优标准切面之后，对每个标准切面集合对应的最优标准切面的切面分值执行归一化操作，得到每个标准切面集合对应的最优标准切面归一化后的切面分值。
筛选模块407，用于根据归一化后的所有切面分值，从所有标准切面中 筛选最高归一化后的切面分值对应的标准切面，作为所有胎儿超声图像对应的最优标准切面。
可见，实施图5所描述的确定装置能够在获取到不同类别的标准切面对应的最优标准切面之后，进一步对不同类别的标准切面对应的最优标准切面的切面分值执行归一化操作，能够使得不同类别的标准切面对应的最优标准切面的切面分值具有可比性，从而提高所有胎儿超声图像对应的最优标准切面的确定准确性以及效率，从而进一步有利于准确获取胎儿的生长发育情况。
在又一个可选的实施例中，如图5所示，该装置还包括训练模块408，其中：
获取模块401，还用于获取正胎儿超声图像样本以及负胎儿超声图像样本，正胎儿超声图像样本的像素值大于负胎儿超声图像样本的像素值，正胎儿超声图像样本中的每个正样本胎儿超声图像以及负胎儿超声图像样本中每个负样本胎儿超声图像的结构特征的关键权重值影响因子包括该结构特征的清晰度。
训练模块408，用于基于正胎儿超声图像样本以及负胎儿超声图像样本，训练确定出的初始分类模型。
获取模块401，还用于获取训练后的初始分类模型，作为确定出的分类模型。
可见，实施图5所描述的确定装置能够通过预先基于样本胎儿超声图像对初始分类模型执行训练操作，能够获取符合要求且精准的分类模型，从而提高关键权重值影响因子包括结构特征的清晰度的子权重值的分析准确性以及可靠性，从而提高结构特征对应的权重值的计算精准性以及效率。
在又一个可选的实施例中，如图5所示，该装置还包括第二判断模块409，其中：
第二判断模块409，用于在第一确定模块402确定每帧胎儿超声图像的标准切面的切面分值之后，根据所有标准切面的结构特征，判断所有标准切面中是否存在结构特征为异常结构特征的异常标准切面；当判断出所有标准切面中不存在异常标准切面时，触发第二确定模块403执行上述的根据所有切面分值，从所有标准切面中确定最高切面分值对应的标准切面，作为最优标准切面的操作。
可见，实施图5所描述的确定装置能够在获取到胎儿超声图像的标准切面的切面分值之后，进一步判断所有标准切面中是否存在异常标准切面，若不存在，则继续执行最优标准切面的确定操作，能够提高执行最优标准切面的确定准确性，从而提高最优标准切面的确定准确性。
在又一个可选的实施例中，如图5所示，该装置还包括修正模块410，其中：
第二确定模块403，还用于当第二判断模块409判断出所有标准切面中存在至少一个异常标准切面时，确定与每个异常标准切面对应的分值修正系数。
修正模块410，用于基于每个异常标准切面对应的分值修正系数修正该异常标准切面的切面分值，以及触发第二确定模块403执行上述的根据所有切面分值，从所有标准切面中确定最高切面分值对应的标准切面，作为最优标准切面的操作。
可见，实施图5所描述的确定装置能够在获取到胎儿超声图像的标准切面的切面分值之后，进一步判断所有标准切面中是否存在异常标准切面，若存在，则基于分值修正系数对异常标准切面的切面分值执行修正操作，能够提高异常标准切面的切面分值的确定准确性，以减少出现异常标准切面时却继续执行最优标准切面的获取而导致获取到非最优标准切面的情况发生，从而提高当出现异常标准切面时的最优标准切面的确定准确性以及 可靠性。
在又一个可选的实施例中，如图5所示，第二判断模块409根据所有标准切面的结构特征，判断所有标准切面中是否存在结构特征为异常结构特征的异常标准切面的方式具体为：
获取每个标准切面的每个结构特征的目标信息，每个结构特征的目标信息用于确定该结构特征所在的标准切面是否为异常标准切面；
根据每个标准切面的每个结构特征的目标信息，判断每个结构特征是否与所在的标准切面相匹配；
当判断出所有结构特征中存在与所在的标准切面不匹配的非匹配结构特征时，确定所有标准切面中存在结构特征为异常结构特征的异常标准切面，且异常标准切面为非匹配结构特征所在的标准切面。
可见，实施图5所描述的确定装置能够通过获取标准切面的每个结构特征的目标信息，并根据每个结构特征的目标信息判断每个结构特征是否与对应的标准切面匹配，能够实现异常标准切面的确定。
在又一个可选的实施例中，每帧胎儿超声图像的标准切面内存在至少一个结构特征，每个结构特征均存在对应的权重值；如图6所示，第一确定模块402可以包括确定子模块4021以及计算子模块4022，其中：
确定子模块4021，用于确定每帧胎儿超声图像的标准切面的每个结构特征对应的权重值。
计算子模块4022，用于基于每个标准切面的每个结构特征对应的权重值以及该结构特征的特征参数，计算每帧胎儿超声图像的标准切面的切面分值。
可见，实施图6所描述的确定装置能够通过将标准切面的每个结构特征的权重值与该结构特征的特征参数结合，能够实现标准切面的切面分值的自动计算，提高标准切面的切面分值的计算准确性以及效率，有利于实现最优标准切面的自动确定，从而实现准确获取胎儿的生长发育情况。
在又一个可选的实施例中，如图7所示，确定子模块4021包括确定单元40211以及计算单元40212，其中：
确定单元40211，用于确定每帧胎儿超声图像的标准切面的每个结构特征对应的关键权重值影响因子，每个结构特征对应的关键权重值影响因子的数量大于等于1，且每个关键权重值影响因子存在对应的子权重值。
确定单元40211，还用于根据每个结构特征对应的每个关键权重值影响因子，确定每个关键权重值影响因子对应的子权重值。
计算单元40212，用于计算每个结构特征对应的所有子权重值之和，作为每个结构特征对应的权重值。
可见，实施图7所描述的确定装置能够通过针对性确定每个结构特征对应的关键权重值影响因子，并将所有关键权重值影响因子对应的子权重值确定为该结构特征对应的权重值，能够提高结构特征的权重值的计算准确性，从而提高对应标准切面的切面分值的计算准确性，进而提高最优标准切面的确定准确性。
在又一个可选的实施例中，如图7所示，确定单元40211根据每个结构特征对应的每个关键权重值影响因子，确定每个关键权重值影响因子对应的子权重值的方式具体为：
对于任一结构特征，当结构特征对应的关键权重值影响因子包括该结构特征的轮廓的几何参数时，根据结构特征的轮廓的几何参数，确定结构特征的轮廓的几何参数对应的子权重值，结构特征的轮廓的几何参数包括该结构特征的轮廓的尺寸和/或面积；
对于任一结构特征，当结构特征对应的关键权重值影响因子包括该结构特征的清晰度时，将结构特征对应的胎儿超声图像输入确定出的分类模型中进行分析，并获取分类模型输出的分析结果，作为结构特征的清晰度 对应的子权重值；
对于任一结构特征，当结构特征对应的关键权重值影响因子包括该结构特征的完整度时，根据结构特征的轮廓，计算结构特征对应的几何参数，并根据结构特征对应的几何参数，确定结构特征的完整度对应的子权重值；
对于任一结构特征，当结构特征对应的关键权重值影响因子包括该结构特征在所在标准切面的位置时，计算结构特征的轮廓所围成的区域的面积，并基于结构特征对应的脑中线与该结构特征的轮廓所围成的区域的面积的相对位置关系，确定结构特征在所在标准切面的位置对应的子权重值。
可见，实施图7所描述的确定装置能够通过根据不同的关键权重值影响因子，选择对应的子权重值确定方式，既能够实现关键权重值影响因子对应的子权重值的获取，又能够提高子权重值的获取效率以及精准性，从而提高结构特征对应的权重值的计算精准性以及效率，进而提高对应标准切面的切面分值的计算精准性以及效率。
在又一个可选的实施例中，如图5所示，第二确定模块403，还用于确定每帧胎儿超声图像的目标特征的占比，目标特征的占比用于表示目标特征与所在显示装置的显示比例，每帧胎儿超声图像的目标特征包括该胎儿超声图像的标准切面或该胎儿超声图像的标准切面中的结构特征。
第二确定模块403，还用于当每帧胎儿超声图像的目标特征为该胎儿超声图像的标准切面时且在第一确定模块402确定每帧胎儿超声图像的标准切面的切面分值之后，确定每帧胎儿超声图像的标准切面的占比对应的分值系数。
获取模块401，还用于基于每帧胎儿超声图像的标准切面的占比对应的分值系数更正该胎儿超声图像的标准切面的切面分值，得到更正后的每帧胎儿超声图像的标准切面的切面分值，以及触发第二确定模块403执行上述的根据所有标准切面的切面分值，从所有标准切面中确定最高切面分值对应的标准切面，作为所有胎儿超声图像的最优标准切面的操作。
可见，实施图5所描述的确定装置在得到胎儿超声图像的标准切面的切面分值之后，进一步根据获取到胎儿超声图像的标准切面与当前显示装置的显示区域的占比对应的分值系数来更新切面分值，有利于提高胎儿超声图像的标准切面的切面分值的确定准确性以及可靠性，进而提高胎儿超声图像的最优标准切面确定准确性以及可靠性。
在又一个可选的实施例中，如图7所示，确定单元40211根据每个结构特征对应的每个关键权重值影响因子，确定每个关键权重值影响因子对应的子权重值的方式具体为：
当每帧胎儿超声图像的目标特征为该胎儿超声图像的标准切面中的结构特征时，对于任一结构特征，根据结构特征的占比，确定与结构特征的占比相匹配的子权重值。
可见，实施图7所描述的确定装置能够通过计算胎儿超声图像的标准切面内结构特征的占比，并通过确定该占比对应的子权重值，能够实现结构特征对应的权重值的确定，以及增加结构特征对应的权重值的计算维度，能够进一步提高结构特征对应的权重值计算准确性以及可靠性，进而提高胎儿超声图像的标准切面的切面分值的确定准确性以及可靠性，进而提高胎儿超声图像的最优标准切面确定准确性以及可靠性。
在又一个可选的实施例中，如图7所示，确定单元40211根据结构特征的轮廓，计算结构特征对应的几何参数的方式具体为：
计算结构特征的轮廓的长度，作为结构特征对应的几何参数；和/或，
确定结构特征的轮廓对应的中心点，并基于结构特征的轮廓对应的中心点以及该结构特征的轮廓，确定结构特征的轮廓对应的中心角，作为结构特征对应的几何参数；和/或，
基于确定出的拟合方法拟合结构特征的轮廓，得到结构特征的目标轮 廓；
计算结构特征的轮廓与该结构特征的目标轮廓的重叠部分轮廓的长度，作为结构特征对应的几何参数，和/或，确定结构特征的目标轮廓对应的中心点，并基于结构特征的目标轮廓对应的中心点以及重叠部分轮廓，确定重叠部分轮廓对应的中心角，作为结构特征对应的几何参数。
可见，实施图7所描述的确定装置能够通过根据胎儿超声图像的结构特征的圆弧半径的大小选择不同的拟合方式，不仅能够实现结构特征的拟合，还能够提高结构特征的拟合效率以及准确性，从而提高结构特征的几何参数的计算准确性。
在又一个可选的实施例中，每个标准切面的每个结构特征的特征参数包括该结构特征的类别概率以及该结构特征的位置概率。以及，如图6所示，计算子模块4022基于每个标准切面的每个结构特征对应的权重值以及该结构特征的特征参数，计算每帧胎儿超声图像的标准切面的切面分值的方式具体为：
基于每个标准切面的每个结构特征对应的权重值、该结构特征的类别概率以及该结构特征的位置概率，计算每个标准切面的每个结构特征对应的结构分值；
计算每个标准切面的所有结构特征对应的结构分值之和，作为每帧胎儿超声图像的标准切面的切面分值。
可见，实施图6所描述的确定装置能够通过分别计算标准切面的每个结构特征对应的结构分值，能够实现标准切面的切面分值的计算，有利于提高标准切面的切面分值计算精准性以及效率；以及根据不同的结构特征，选取不同的参数，能够提高结构特征对应的结构分值的计算精准性以及效率，从而进一步提高标准切面的切面分值计算精准性以及效率。
实施例五
请参阅图8，图8是本发明实施例公开的又一种胎儿最优标准切面的确定装置。其中，图8所描述的胎儿最优标准切面的确定装置可以应用于标准切面确定服务器(服务设备)中，其中，该标准切面确定服务器可以包括本地标准切面确定服务器或云标准切面确定服务器，本发明实施例不做限定。如图8所示，该胎儿最优标准切面的确定装置可以包括：
存储有可执行程序代码的存储器801；
与存储器801耦合的处理器802；
进一步的，还可以包括与处理器802耦合的输入接口803以及输出接口804；
其中，处理器802调用存储器801中存储的可执行程序代码，用于执行实施例一或实施例二所描述的胎儿最优标准切面的确定方法中部分或者全部的步骤。
实施例六
本发明实施例公开了一种计算机可读存储介质，其存储用于电子数据交换的计算机程序，其中，该计算机程序使得计算机执行实施例一或实施例二所描述的胎儿最优标准切面的确定方法中部分或者全部的步骤。
实施例七
本发明实施例公开了一种计算机程序产品，该计算机程序产品包括存储了计算机程序的非瞬时性计算机可读存储介质，且该计算机程序可操作来使计算机执行实施例一或实施例二所描述的胎儿最优标准切面的确定方法中部分或者全部的步骤。
以上所描述的装置实施例仅是示意性的，其中所述作为分离部件说明的模块可以是或者也可以不是物理上分开的，作为模块显示的部件可以是或者也可以不是物理模块，即可以位于一个地方，或者也可以分布到多个网络模块上。可以根据实际的需要选择其中的部分或者全部模块来实现本 实施例方案的目的。本领域普通技术人员在不付出创造性的劳动的情况下，即可以理解并实施。
通过以上的实施例的具体描述，本领域的技术人员可以清楚地了解到各实施方式可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件。基于这样的理解，上述技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品可以存储在计算机可读存储介质中,存储介质包括只读存储器(Read-Only Memory，ROM)、随机存储器(Random Access Memory，RAM)、可编程只读存储器(Programmable Read-only Memory，PROM)、可擦除可编程只读存储器(Erasable Programmable Read Only Memory，EPROM)、一次可编程只读存储器(One-time Programmable Read-Only Memory，OTPROM)、电子抹除式可复写只读存储器(Electrically-Erasable Programmable Read-Only Memory，EEPROM)、只读光盘(Compact Disc Read-Only Memory，CD-ROM)或其他光盘存储器、磁盘存储器、磁带存储器、或者能够用于携带或存储数据的计算机可读的任何其他介质。
最后应说明的是：本发明实施例公开的一种胎儿最优标准切面的确定方法胎儿最优标准切面的确定方法及装置所揭露的仅为本发明较佳实施例而已，仅用于说明本发明的技术方案，而非对其限制；尽管参照前述实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解；其依然可以对前述各项实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或替换，并不使相应的技术方案的本质脱离本发明各项实施例技术方案的精神和范围。

Claims (16)

  1. 一种胎儿最优标准切面的确定方法，其特征在于，所述方法包括：
    获取多帧胎儿超声图像中每帧所述胎儿超声图像的标准切面，并确定每帧所述胎儿超声图像的标准切面的切面分值；
    根据所有所述标准切面的切面分值，从所有所述标准切面中确定最高切面分值对应的标准切面，作为所有所述胎儿超声图像的最优标准切面。
  2. 根据权利要求1所述的胎儿最优标准切面的确定方法，其特征在于，所述确定每帧所述胎儿超声图像的标准切面的切面分值之后，所述方法还包括：
    判断所有所述标准切面是否属于同一类别的标准切面；
    当判断出所有所述标准切面属于同一类别的标准切面时，触发执行所述的根据所有所述标准切面的切面分值，从所有所述标准切面中确定最高切面分值对应的标准切面，作为所有所述胎儿超声图像的最优标准切面的操作。
  3. 根据权利要求2所述的胎儿最优标准切面的确定方法，其特征在于，所述方法还包括：
    当判断出所有所述标准切面不属于同一类别的标准切面时，按照预设分类方式对所有所述胎儿超声图像的标准切面执行分类操作，得到至少两个标准切面集合，每个所述标准切面集合包括至少一帧所述胎儿超声图像的标准切面，且每个所述标准切面集合包括的所有所述标准切面为同一类别的标准切面；
    其中，所述根据所有所述标准切面的切面分值，从所有所述标准切面中确定最高切面分值对应的标准切面，作为最优标准切面，包括：
    根据每个所述标准切面集合对应的所有切面分值，从每个所述标准切面集合包括的所有所述标准切面中确定最高切面分值对应的标准切面，作为每个所述标准切面集合对应的最优标准切面。
  4. 根据权利要求3所述的胎儿最优标准切面的确定方法，其特征在于，所述根据每个所述标准切面集合对应的所有切面分值，从每个所述标准切面集合包括的所有所述标准切面中确定最高切面分值对应的标准切面，作为每个所述标准切面集合对应的最优标准切面之后，所述方法还包括：
    对每个所述标准切面集合对应的最优标准切面的切面分值执行归一化操作，得到每个所述标准切面集合对应的最优标准切面归一化后的切面分值；
    根据归一化后的所有所述切面分值，从所有所述标准切面中筛选最高归一化后的切面分值对应的标准切面，作为所有所述胎儿超声图像对应的最优标准切面。
  5. 根据权利要求1-4任一项所述的胎儿最优标准切面的确定方法，其特征在于，每帧所述胎儿超声图像的标准切面内存在至少一个结构特征，每个所述结构特征均存在对应的权重值；
    以及，所述确定每帧所述胎儿超声图像的标准切面的切面分值，包括：
    确定每帧所述胎儿超声图像的标准切面的每个所述结构特征对应的权重值；
    基于每个所述标准切面的每个所述结构特征对应的权重值以及该结构特征的特征参数，计算每帧所述胎儿超声图像的标准切面的切面分值。
  6. 根据权利要求5所述的胎儿最优标准切面的确定方法，其特征在于，所述确定每帧所述胎儿超声图像的标准切面的每个所述结构特征对应的权重值，包括：
    确定每帧所述胎儿超声图像的标准切面的每个所述结构特征对应的关键权重值影响因子，每个所述结构特征对应的关键权重值影响因子的数量大于等于1，且每个所述关键权重值影响因子存在对应的子权重值；
    根据每个所述结构特征对应的每个所述关键权重值影响因子，确定每个所述关键权重值影响因子对应的子权重值，并计算每个所述结构特征对应的所有所述子权重值之和，作为每个所述结构特征对应的权重值。
  7. 根据权利要求6所述的胎儿最优标准切面的确定方法，其特征在于，所述根据每个所述结构特征对应的每个所述关键权重值影响因子，确定每个所述关键权重值影响因子对应的子权重值，包括：
    对于任一所述结构特征，当所述结构特征对应的所述关键权重值影响因子包括该结构特征的轮廓的几何参数时，根据所述结构特征的轮廓的几何参数，确定所述结构特征的轮廓的几何参数对应的子权重值，所述结构特征的轮廓的几何参数包括该结构特征的轮廓的尺寸和/或面积；
    对于任一所述结构特征，当所述结构特征对应的所述关键权重值影响因子包括该结构特征的清晰度时，将所述结构特征对应的胎儿超声图像输入确定出的分类模型中进行分析，并获取所述分类模型输出的分析结果，作为所述结构特征的清晰度对应的子权重值；
    对于任一所述结构特征，当所述结构特征对应的所述关键权重值影响因子包括该结构特征的完整度时，根据所述结构特征的轮廓，计算所述结构特征对应的几何参数，并根据所述结构特征对应的几何参数，确定所述结构特征的完整度对应的子权重值；
    对于任一所述结构特征，当所述结构特征对应的所述关键权重值影响因子包括该结构特征在所在标准切面的位置时，基于所述结构特征对应的脑中线与该结构特征的轮廓所围成的区域的相对位置关系，确定所述结构特征在所在标准切面的位置对应的子权重值。
  8. 根据权利要求6所述的胎儿最优标准切面的确定方法，其特征在于，所述方法还包括：
    确定每帧所述胎儿超声图像的目标特征的占比，所述目标特征的占比用于表示所述目标特征与所在显示装置的显示比例，每帧所述胎儿超声图像的目标特征包括该胎儿超声图像的标准切面或该胎儿超声图像的标准切面中的结构特征；
    以及，当每帧所述胎儿超声图像的目标特征为该胎儿超声图像的标准切面时，所述确定每帧所述胎儿超声图像的标准切面的切面分值之后，所述方法还包括：
    确定每帧所述胎儿超声图像的标准切面的占比对应的分值系数，并基于每帧所述胎儿超声图像的标准切面的占比对应的分值系数更正该胎儿超声图像的标准切面的切面分值，得到更正后的每帧所述胎儿超声图像的标准切面的切面分值，以及触发执行所述的根据所有所述标准切面的切面分值，从所有所述标准切面中确定最高切面分值对应的标准切面，作为所有所述胎儿超声图像的最优标准切面的操作；
    当每帧所述胎儿超声图像的目标特征为该胎儿超声图像的标准切面中的结构特征时，所述根据每个所述结构特征对应的每个所述关键权重值影响因子，确定每个所述关键权重值影响因子对应的子权重值，包括：
    对于任一所述结构特征，根据所述结构特征的占比，确定与所述结构特征的占比相匹配的子权重值。
  9. 根据权利要求7所述的胎儿最优标准切面的确定方法，其特征在于，所述根据所述结构特征的轮廓，计算所述结构特征对应的几何参数，包括：
    计算所述结构特征的轮廓的长度，作为所述结构特征对应的几何参数；和/或，
    确定所述结构特征的轮廓对应的中心点，并基于所述结构特征的轮廓对应的中心点以及该结构特征的轮廓，确定所述结构特征的轮廓对应的中心角，作为所述结构特征对应的几何参数；和/或，
    基于确定出的拟合方法拟合所述结构特征的轮廓，得到所述结构特征的目标轮廓；
    计算所述结构特征的轮廓与该结构特征的目标轮廓的重叠部分轮廓的长度，作为所述结构特征对应的几何参数，和/或，确定所述结构特征的目标轮廓对应的中心点，并基于所述结构特征的目标轮廓对应的中心点以及所述重叠部分轮廓，确定所述重叠部分轮廓对应的中心角，作为所述结构特征对应的几何参数。
  10. 根据权利要求8或9所述的胎儿最优标准切面的确定方法，其特征在于，所述方法包括：
    获取正胎儿超声图像样本以及负胎儿超声图像样本，所述正胎儿超声图像样本的像素值大于所述负胎儿超声图像样本的像素值，所述正胎儿超声图像样本中的每个正样本胎儿超声图像以及所述负胎儿超声图像样本中每个负样本胎儿超声图像的结构特征的关键权重值影响因子包括该结构特征的清晰度；
    基于所述正胎儿超声图像样本以及所述负胎儿超声图像样本，训练确定出的初始分类模型，并获取训练后的初始分类模型，作为确定出的分类模型。
  11. 根据权利要求5-10任一项所述的胎儿最优标准切面的确定方法，其特征在于，每个所述标准切面的每个所述结构特征的特征参数包括该结构特征的类别概率以及该结构特征的位置概率；
    其中，所述基于每个所述标准切面的每个所述结构特征对应的权重值以及该结构特征的特征参数，计算每帧所述胎儿超声图像的标准切面的切面分值，包括：
    基于每个所述标准切面的每个所述结构特征对应的权重值、该结构特征的类别概率以及该结构特征的位置概率，计算每个所述标准切面的每个所述结构特征对应的结构分值；
    计算每个所述标准切面的所有所述结构特征对应的结构分值之和，作为每帧所述胎儿超声图像的标准切面的切面分值。
  12. 根据权利要求1-9任一项所述的胎儿最优标准切面的确定方法，其特征在于，所述确定每帧所述胎儿超声图像的标准切面的切面分值之后，所述方法还包括：
    根据所有所述标准切面的结构特征，判断所有所述标准切面中是否存在结构特征为异常结构特征的异常标准切面；
    当判断出所有所述标准切面中不存在所述异常标准切面时，触发执行所述的根据所有所述切面分值，从所有所述标准切面中确定最高切面分值对应的标准切面，作为最优标准切面的操作。
  13. 根据权利要求12所述的胎儿最优标准切面的确定方法，其特征在于，所述方法包括：
    当判断出所有所述标准切面中存在至少一个所述异常标准切面时，确定与每个所述异常标准切面对应的分值修正系数；
    基于每个所述异常标准切面对应的分值修正系数修正该异常标准切面的切面分值，以及触发执行所述的根据所有所述切面分值，从所有所述标准切面中确定最高切面分值对应的标准切面，作为最优标准切面的操作。
  14. 根据权利要求13所述的胎儿最优标准切面的确定方法，其特征在于，所述根据所有所述标准切面的结构特征，判断所有所述标准切面中是否存在结构特征为异常结构特征的异常标准切面，包括：
    获取每个所述标准切面的每个所述结构特征的目标信息，每个所述结构特征的目标信息用于确定该结构特征所在的标准切面是否为异常标准切面；
    根据每个所述标准切面的每个所述结构特征的目标信息，判断每个所述结构特征是否与所在的标准切面相匹配；
    当判断出所有所述结构特征中存在与所在的标准切面不匹配的非匹配结构特征时，确定所有所述标准切面中存在结构特征为异常结构特征的异常标准切面，且所述异常标准切面为所述非匹配结构特征所在的标准切面。
  15. 一种胎儿最优标准切面的确定装置，其特征在于，所述装置包括：
    获取模块，用于获取多帧胎儿超声图像中每帧所述胎儿超声图像对应的标准切面；
    第一确定模块，用于确定每帧所述胎儿超声图像的标准切面的切面分值；
    第二确定模块，用于根据所有所述标准切面的切面分值，从所有所述标准切面中确定最高切面分值对应的标准切面，作为所有所述胎儿超声图像的最优标准切面。
  16. 一种胎儿最优标准切面的确定装置，其特征在于，所述装置包括：
    存储有可执行程序代码的存储器；
    与所述存储器耦合的处理器；
    所述处理器调用所述存储器中存储的所述可执行程序代码，执行如权利要求1-14任一项所述的胎儿最优标准切面的确定方法。
PCT/CN2021/096821 2020-09-24 2021-05-28 一种胎儿最优标准切面的确定方法及装置 WO2022062458A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011015507.6 2020-09-24
CN202011015507.6A CN112155602B (zh) 2020-09-24 2020-09-24 一种胎儿最优标准切面的确定方法及装置

Publications (1)

Publication Number Publication Date
WO2022062458A1 true WO2022062458A1 (zh) 2022-03-31

Family

ID=73863773

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/096821 WO2022062458A1 (zh) 2020-09-24 2021-05-28 一种胎儿最优标准切面的确定方法及装置

Country Status (2)

Country Link
CN (1) CN112155602B (zh)
WO (1) WO2022062458A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115919367A (zh) * 2022-12-09 2023-04-07 开立生物医疗科技(武汉)有限公司 超声图像处理方法及装置、电子设备和存储介质

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112155602B (zh) * 2020-09-24 2023-05-05 广州爱孕记信息科技有限公司 一种胎儿最优标准切面的确定方法及装置
CN112155603B (zh) * 2020-09-24 2023-06-09 广州爱孕记信息科技有限公司 胎儿结构特征的权重值确定方法及装置
CN112991289B (zh) * 2021-03-10 2024-03-26 深圳市鹭鸣科技有限公司 图像标准切面的处理方法及装置
CN115760851B (zh) * 2023-01-06 2023-05-09 首都儿科研究所附属儿童医院 一种基于机器学习的超声图像数据处理设备、系统及计算机可读存储介质

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3510933A1 (en) * 2018-01-10 2019-07-17 Samsung Medison Co., Ltd. Ultrasound diagnosis apparatus and method for controlling the same
CN110033020A (zh) * 2019-03-07 2019-07-19 李胜利 基于深度学习的胎儿超声图像中标准切面图像识别方法及识别系统
CN110555836A (zh) * 2019-09-05 2019-12-10 李肯立 一种超声图像中胎儿标准切面的自动识别方法和系统
CN111223092A (zh) * 2020-02-28 2020-06-02 长沙大端信息科技有限公司 胎儿超声切面图像自动质控系统及检测方法
CN111402244A (zh) * 2020-03-20 2020-07-10 华侨大学 一种胎儿心脏标准切面自动分类法
CN112155602A (zh) * 2020-09-24 2021-01-01 广州爱孕记信息科技有限公司 一种胎儿最优标准切面的确定方法及装置
CN112184656A (zh) * 2020-09-24 2021-01-05 广州爱孕记信息科技有限公司 基于超声动态图像的胎儿切面的确定方法及装置

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110604592B (zh) * 2019-03-04 2024-04-02 北京大学第三医院 一种髋关节的成像方法以及髋关节成像系统

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3510933A1 (en) * 2018-01-10 2019-07-17 Samsung Medison Co., Ltd. Ultrasound diagnosis apparatus and method for controlling the same
CN110033020A (zh) * 2019-03-07 2019-07-19 李胜利 基于深度学习的胎儿超声图像中标准切面图像识别方法及识别系统
CN110555836A (zh) * 2019-09-05 2019-12-10 李肯立 一种超声图像中胎儿标准切面的自动识别方法和系统
CN111223092A (zh) * 2020-02-28 2020-06-02 长沙大端信息科技有限公司 胎儿超声切面图像自动质控系统及检测方法
CN111402244A (zh) * 2020-03-20 2020-07-10 华侨大学 一种胎儿心脏标准切面自动分类法
CN112155602A (zh) * 2020-09-24 2021-01-01 广州爱孕记信息科技有限公司 一种胎儿最优标准切面的确定方法及装置
CN112184656A (zh) * 2020-09-24 2021-01-05 广州爱孕记信息科技有限公司 基于超声动态图像的胎儿切面的确定方法及装置

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115919367A (zh) * 2022-12-09 2023-04-07 开立生物医疗科技(武汉)有限公司 超声图像处理方法及装置、电子设备和存储介质

Also Published As

Publication number Publication date
CN112155602B (zh) 2023-05-05
CN112155602A (zh) 2021-01-01

Similar Documents

Publication Publication Date Title
WO2022062458A1 (zh) 一种胎儿最优标准切面的确定方法及装置
WO2022062461A1 (zh) 基于超声动态图像的胎儿切面的确定方法及装置
US11961227B2 (en) Method and device for detecting and locating lesion in medical image, equipment and storage medium
CN109907732B (zh) 一种颅内动脉瘤破裂风险的评估方法及系统
WO2022062460A1 (zh) 一种胎儿超声图像的成像质量控制的确定方法及装置
US20190102603A1 (en) Method and apparatus for determining image quality
WO2020181805A1 (zh) 糖尿病的预测方法及装置、存储介质、计算机设备
WO2019119505A1 (zh) 人脸识别的方法和装置、计算机装置及存储介质
WO2022062456A1 (zh) 胎儿超声影像的检测结果的输出控制方法、装置及系统
EP3625696A1 (en) Systems and methods for searching images
Oloumi et al. Parabolic modeling of the major temporal arcade in retinal fundus images
WO2022062459A1 (zh) 一种胎儿结构特征的自动测量方法及装置
CN109924956B (zh) 一种颅内动脉瘤图像的形态学参数的测量方法及装置
WO2018076495A1 (zh) 检索人脸图像的方法和系统
CN110522449B (zh) 一种斑块分型方法、装置、电子设备及可读存储介质
WO2022062457A1 (zh) 胎儿结构特征的权重值确定方法及装置
CN111326243B (zh) 分诊推荐方法、装置、电子设备和存储介质
CN111968750A (zh) 一种服务器、数据处理方法、装置及可读存储介质
CN112155604A (zh) 基于胎儿超声图像的胎儿严重畸形检测方法及装置
CN110969617B (zh) 视杯视盘图像识别方法、装置、设备及存储介质
CN111383353B (zh) 基于高斯混合模型和轮廓描述子的断骨模型配准方法
US11232571B2 (en) Method and device for quick segmentation of optical coherence tomography image
CN111968087A (zh) 一种植物病害区域检测方法
CN116758607A (zh) 基于眼部与面部特征的人员识别方法
CN114612967B (zh) 一种人脸聚类的方法、装置、设备及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21870845

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21870845

Country of ref document: EP

Kind code of ref document: A1