WO2022062455A1 - 基于胎儿超声图像的胎儿严重畸形检测方法及装置 - Google Patents

基于胎儿超声图像的胎儿严重畸形检测方法及装置 Download PDF

Info

Publication number
WO2022062455A1
WO2022062455A1 PCT/CN2021/096813 CN2021096813W WO2022062455A1 WO 2022062455 A1 WO2022062455 A1 WO 2022062455A1 CN 2021096813 W CN2021096813 W CN 2021096813W WO 2022062455 A1 WO2022062455 A1 WO 2022062455A1
Authority
WO
WIPO (PCT)
Prior art keywords
ultrasound image
slice
fetal ultrasound
section
fetal
Prior art date
Application number
PCT/CN2021/096813
Other languages
English (en)
French (fr)
Inventor
谢红宁
汪南
冼建波
梁喆
吴海涛
杨燕淇
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 WO2022062455A1 publication Critical patent/WO2022062455A1/zh

Links

Images

Classifications

    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/08Detecting organic movements or changes, e.g. tumours, cysts, swellings
    • A61B8/0866Detecting organic movements or changes, e.g. tumours, cysts, swellings involving foetal diagnosis; pre-natal or peri-natal diagnosis of the baby
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/46Ultrasonic, sonic or infrasonic diagnostic devices with special arrangements for interfacing with the operator or the patient
    • A61B8/467Ultrasonic, sonic or infrasonic diagnostic devices with special arrangements for interfacing with the operator or the patient characterised by special input means
    • A61B8/468Ultrasonic, sonic or infrasonic diagnostic devices with special arrangements for interfacing with the operator or the patient characterised by special input means allowing annotation or message recording
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/52Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/5215Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data
    • A61B8/5223Devices 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
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/52Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/5215Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data
    • A61B8/523Devices 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
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/52Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/5215Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data
    • A61B8/5238Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data for combining image data of patient, e.g. merging several images from different acquisition modes into one image
    • A61B8/5246Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data for combining image data of patient, e.g. merging several images from different acquisition modes into one image combining images from the same or different imaging techniques, e.g. color Doppler and B-mode

Definitions

  • the invention relates to the technical field of image processing, in particular to a method and device for detecting severe fetal malformation based on fetal ultrasound images.
  • fetal malformations may occur during the development process, especially in the early gestational weeks of the fetus, such as anencephaly, severe meningocele, and no lobe.
  • the prenatal detection method of fetal malformation is as follows: by obtaining the ultrasound image of the fetus, and analyzing the ultrasound image of the fetus by the staff with relevant experience combined with their own experience, so as to determine the fetal malformation, and then determine the growth and development of the fetus.
  • the technical problem to be solved by the present invention is to provide a method and device for detecting severe fetal malformation based on fetal ultrasound images, which can accurately detect fetal malformation, thereby realizing accurate determination of fetal growth and development.
  • a first aspect of the present invention discloses a method for detecting severe fetal malformation based on fetal ultrasound images, the method comprising:
  • the target information corresponding to the slice of the fetal ultrasound image is used to determine the developmental situation of the fetus corresponding to the fetal ultrasound image;
  • the target information corresponding to the section of the fetal ultrasound image determine whether the fetus corresponding to the fetal ultrasound image has a deformity, and when the judgment result is yes, determine the fetal ultrasound according to the target information corresponding to the section of the fetal ultrasound image.
  • the malformation condition of the fetus corresponding to the image where the malformation condition includes the type of malformation of the fetus corresponding to the fetal ultrasound image.
  • the slices of the fetal ultrasound image include craniocerebral slices, limb slices, abdominal slices, spinal cord slices, cardiac slices, humerus long-diameter slices, and femur long-diameter slices
  • the craniocerebral section of the fetal ultrasound image includes a horizontal section and/or a craniocerebral sagittal section
  • the limb section of the fetal ultrasound image includes a double upper limb section or a double lower limb section
  • the abdominal slice of the image includes a horizontal slice of the abdomen and/or a sagittal slice of the abdomen
  • the slice of the spinal cord of the fetal ultrasound image includes a horizontal slice of the spinal cord and/or a sagittal slice of the spinal cord.
  • judging whether the fetus corresponding to the fetal ultrasound image has a deformity includes:
  • the target information corresponding to the section of the fetal ultrasound image includes the cranial structure feature of the section
  • the contour and the buttocks structure feature contour according to the craniocerebral structure feature contour and the buttocks structure feature contour of the section of the fetal ultrasound image, measure the head-rump length of the fetal ultrasound image, and determine whether the head-rump length is within the preset head. Within the range of the buttock length, when it is judged that they do not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity;
  • the target information corresponding to the section of the fetal ultrasound image includes the section of the fetal ultrasound image
  • the geometric parameters of the cranial structure feature of the slice are determined, whether the geometric parameters of the craniocerebral structure feature of the section match the preset geometric parameters, and when it is judged that they do not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity
  • the craniocerebral structural feature of the section includes at least one of the calvarial structural feature, the cerebral hemisphere structural feature and the midbrain structural feature in the section, and the geometric parameters of the craniocerebral structural feature include the At least one of shape, size, position and area, the position of the cranial structure feature is the position of the cranial structure feature in the slice of the fetal ultrasound image;
  • the target information corresponding to the slice of the fetal ultrasound image includes the shape of the slice of the fetal ultrasound image. Whether the shape of the cut plane matches the shape of the preset cut plane, and when it is judged that it does not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity;
  • the target information corresponding to the slice of the fetal ultrasound image includes the characteristic parameters corresponding to the target structural features of the slice, according to The feature parameter corresponding to the target structural feature of the cut plane determines whether the target structural feature matches the cut plane, and when it is judged that it does not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity;
  • the target information corresponding to the slice of the fetal ultrasound image includes the size of the slice of the fetal ultrasound image.
  • the inner contour of the cranial structure feature and the outer contour of the cranial structure feature determine the target of the cranial structure feature of the section geometric parameters, and determine whether the target geometric parameters of the craniocerebral structural features are within the range of the preset geometric parameters, and when it is determined that the target geometric parameters are not within the range of the preset geometric parameters, determine that the fetus corresponding to the fetal ultrasound image has a deformity,
  • the target geometric parameters of the craniocerebral structural feature of the slice include the head circumference parameter and/or the biparietal diameter parameter of the craniocerebral structural feature;
  • the target information corresponding to the slice of the fetal ultrasound image includes the left thalamus contour and the right side of the slice.
  • the thalamus contour obtains the first degree of fit between the contour of the left thalamus and the contour of the right thalamus, and determines whether the first degree of fit is greater than or equal to a first preset degree of fit threshold, and when the determination result is yes, determining that the fetus corresponding to the fetal ultrasound image has a deformity;
  • the target information corresponding to the slice of the fetal ultrasound image includes the position of the falx of the slice, According to the position of the sickle cerebrum, determine whether the appearance of the sickle cerebellum at the midline position of the fetal ultrasound image satisfies a preset appearance situation, and when it is determined that the preset appearance situation is not satisfied, determine whether the fetal ultrasound The fetus corresponding to the image has an abnormality;
  • the target information corresponding to the slice of the fetal ultrasound image includes the contour of the choroid plexus and the thalamus of the slice. contour, obtain a second degree of fit between the contour of the choroid plexus and the contour of the thalamus, and determine whether the second degree of fit is greater than or equal to a second preset degree of fit threshold, when the determination result is yes, It is determined that the fetus corresponding to the fetal ultrasound image has a deformity.
  • the cranial structure feature of the section is determined according to the inner contour of the cranial structure feature of the section and the outer contour of the cranial structure feature
  • the target geometry parameters including:
  • a biparietal diameter parameter corresponding to the cranial brain structural feature is determined.
  • the method further includes:
  • the section of the fetal ultrasound image includes the craniocerebral section, determine whether the section of the fetal ultrasound image matches the standard cranial section, and when the determination result is yes, trigger the execution of the determining of the fetal ultrasound The operation of the target information corresponding to the slice of the image;
  • the cranial standard section when the craniocerebral section is the cranial horizontal section, the cranial standard section includes a lateral ventricle section, and when the cranial section is the cranial sagittal section, the cranial standard section Including the midsagittal section of the cranial brain.
  • the target information corresponding to the slice of the fetal ultrasound image includes characteristic parameters of the limb structure feature of the slice, wherein, when the slice of the fetal ultrasound image includes The limb section, and when the limb section is the double upper limb section, the limb structural features of the section include at least one of the hand, forearm and upper arm of the section; when the fetal ultrasound image section includes all Described limb section, and when described limb section is described double lower limb section, the limb structure feature of described section includes at least one of foot, thigh and calf of this section;
  • judging whether the fetus corresponding to the fetal ultrasound image has a deformity including:
  • the characteristic parameters of the limb structure features of the section of the fetal ultrasound image it is determined whether the limb structure feature matches the section of the fetal ultrasound image, and when it is judged that they do not match, determine the corresponding section of the fetal ultrasound image.
  • the fetus is deformed.
  • judging whether the fetus corresponding to the fetal ultrasound image has a deformity includes:
  • the target information corresponding to the slice of the fetal ultrasound image includes the feature of the bone structure feature of the slice.
  • the feature parameter of the feature includes at least one of the contour, length, area, shape and position corresponding to the bone structure feature.
  • judging whether the fetus corresponding to the fetal ultrasound image has a deformity includes:
  • the target information corresponding to the slice of the fetal ultrasound image includes the characteristic parameter of the slice, and it is determined whether the slice characteristic parameter matches the preset slice parameter. When no match is found, it is determined that the corresponding fetus of the fetal ultrasound image has a deformity, and the characteristic parameters of the slice include Doppler blood flow parameters and/or contour parameters of the slice;
  • the target information corresponding to the slice of the fetal ultrasound image includes the characteristic parameter of the cardiac structural feature of the slice, and the target information is determined according to the characteristic parameter of the cardiac structural feature of the slice.
  • the cardiac structural feature of the slice matches the standard cardiac structural feature of the slice, and when it is judged that it does not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity, and the characteristic parameters of the cardiac structural feature of the slice include all the parameters. the number of the structural features of the heart and/or the area corresponding to the structural features of the heart;
  • judging whether the cardiac structure feature of the section matches the standard cardiac structure feature of the section according to the characteristic parameter of the cardiac structure feature of the section includes:
  • the feature parameter of the cardiac structural feature of the slice is the number of the cardiac structural feature, determine whether the number of the cardiac structural feature of the slice is less than or equal to a preset number, and when the determination result is yes, determine the number of the cardiac structural feature of the slice.
  • Cardiac structural features do not match standard cardiac structural features for this slice;
  • the feature parameter of the cardiac structural feature of the slice is the area corresponding to the cardiac structural feature, determine whether the area corresponding to the cardiac structural feature of the slice is greater than or equal to a preset area threshold, and when the determination result is yes, determine the The cardiac structural features of the slice do not match the standard cardiac structural characteristics of the slice.
  • the method further includes:
  • the structural feature information of the fetal ultrasound image in each frame includes the structural feature information of the fetal ultrasound image and the fetal ultrasound image.
  • the structural feature information of the image, the structural feature information of each frame of the fetal ultrasound image at least includes the type of the structural feature of the fetal ultrasound image, and the structural feature information of each frame of the fetal ultrasound image at least includes the structure of the fetal ultrasound image. the type of feature;
  • the slice of the fetal ultrasound image is determined according to the category of the structural feature of the fetal ultrasound image and the category of the structural feature of the fetal ultrasound image in each frame.
  • the method further includes:
  • a corresponding structural feature label is set for each abnormal structural feature of the fetal ultrasound image, and the structural feature label corresponding to each abnormal structural feature is used to indicate the A type of malformation with abnormal structural features.
  • a second aspect of the present invention discloses a device for detecting severe fetal malformations based on fetal ultrasound images, the device comprising:
  • a determination module configured to determine target information corresponding to the slice of the fetal ultrasound image after acquiring the slice of the fetal ultrasound image, and the target information corresponding to the slice of the fetal ultrasound image is used to determine the fetus corresponding to the fetal ultrasound image development;
  • a first judgment module configured to judge whether the fetus corresponding to the fetal ultrasound image has a deformity according to the target information corresponding to the slice of the fetal ultrasound image
  • the determining module is further configured to determine, according to the target information corresponding to the slice of the fetal ultrasound image, the deformity of the fetus corresponding to the fetal ultrasound image, when the first judgment module determines that the result is yes, the deformity condition Including the fetal malformation type corresponding to the fetal ultrasound image.
  • the slices of the fetal ultrasound image include a craniocerebral slice, a limb slice, an abdominal slice, a spinal cord slice, a heart slice, a long-diameter slice of the humerus, and a long-diameter slice of the femur
  • the craniocerebral section of the fetal ultrasound image includes a horizontal section and/or a craniocerebral sagittal section
  • the limb section of the fetal ultrasound image includes a double upper limb section or a double lower limb section
  • the abdominal slice of the image includes a horizontal slice of the abdomen and/or a sagittal slice of the abdomen
  • the slice of the spinal cord of the fetal ultrasound image includes a horizontal slice of the spinal cord and/or a sagittal slice of the spinal cord.
  • the first determination module determines whether the fetus corresponding to the fetal ultrasound image has a deformity according to the target information corresponding to the slice of the fetal ultrasound image. Specifically:
  • the target information corresponding to the section of the fetal ultrasound image includes the cranial structure feature of the section
  • the contour and the buttocks structure feature contour according to the craniocerebral structure feature contour and the buttocks structure feature contour of the section of the fetal ultrasound image, measure the head-rump length of the fetal ultrasound image, and determine whether the head-rump length is within the preset head. Within the range of the buttock length, when it is judged that they do not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity;
  • the target information corresponding to the section of the fetal ultrasound image includes the section of the fetal ultrasound image
  • the geometric parameters of the cranial structure feature of the slice are determined, whether the geometric parameters of the craniocerebral structure feature of the section match the preset geometric parameters, and when it is judged that they do not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity
  • the craniocerebral structural feature of the section includes at least one of the calvarial structural feature, the cerebral hemisphere structural feature and the midbrain structural feature in the section, and the geometric parameters of the craniocerebral structural feature include the At least one of shape, size, position and area, the position of the cranial structure feature is the position of the cranial structure feature in the slice of the fetal ultrasound image;
  • the target information corresponding to the slice of the fetal ultrasound image includes the shape of the slice of the fetal ultrasound image. Whether the shape of the cut plane matches the shape of the preset cut plane, and when it is judged that it does not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity;
  • the target information corresponding to the slice of the fetal ultrasound image includes the characteristic parameters corresponding to the target structural features of the slice, according to The feature parameter corresponding to the target structural feature of the cut plane determines whether the target structural feature matches the cut plane, and when it is judged that it does not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity;
  • the target information corresponding to the slice of the fetal ultrasound image includes the size of the slice of the fetal ultrasound image.
  • the inner contour of the cranial structure feature and the outer contour of the cranial structure feature determine the target of the cranial structure feature of the section geometric parameters, and determine whether the target geometric parameters of the craniocerebral structural features are within the range of the preset geometric parameters, and when it is determined that the target geometric parameters are not within the range of the preset geometric parameters, determine that the fetus corresponding to the fetal ultrasound image has a deformity,
  • the target geometric parameters of the craniocerebral structural feature of the slice include the head circumference parameter and/or the biparietal diameter parameter of the craniocerebral structural feature;
  • the target information corresponding to the slice of the fetal ultrasound image includes the left thalamus contour and the right side of the slice.
  • the thalamus contour obtains the first degree of fit between the contour of the left thalamus and the contour of the right thalamus, and determines whether the first degree of fit is greater than or equal to a first preset degree of fit threshold, and when the determination result is yes, determining that the fetus corresponding to the fetal ultrasound image has a deformity;
  • the target information corresponding to the slice of the fetal ultrasound image includes the position of the falx of the slice, According to the position of the sickle cerebrum, determine whether the appearance of the sickle cerebellum at the midline position of the fetal ultrasound image satisfies a preset appearance situation, and when it is determined that the preset appearance situation is not satisfied, determine whether the fetal ultrasound The fetus corresponding to the image has an abnormality;
  • the target information corresponding to the slice of the fetal ultrasound image includes the contour of the choroid plexus and the thalamus of the slice. contour, obtain a second degree of fit between the contour of the choroid plexus and the contour of the thalamus, and determine whether the second degree of fit is greater than or equal to a second preset degree of fit threshold, when the determination result is yes, It is determined that the fetus corresponding to the fetal ultrasound image has a deformity.
  • the first judging module determines the cut surface according to the inner contour of the skull structure feature of the cut plane and the outer contour of the skull structure feature.
  • the method of the target geometric parameters of the skull structure feature is as follows:
  • a biparietal diameter parameter corresponding to the cranial brain structural feature is determined.
  • the device further includes:
  • the second judgment module is configured to, when the slice of the fetal ultrasound image includes the craniocerebral slice, after the slice of the fetal ultrasound image is acquired, and the determining module determines the target corresponding to the slice of the fetal ultrasound image Before the information, determine whether the slice of the fetal ultrasound image matches the standard cranial slice, and when the judgment result is yes, trigger the determination module to perform the operation of determining the target information corresponding to the slice of the fetal ultrasound image ;
  • a correction module configured to correct the slice of the fetal ultrasound image based on the acquired structural features when the second determination module determines that it does not match, so that the slice matches the standard slice of the cranial brain, and triggers the
  • the determining module performs the operation of determining the target information corresponding to the slice of the fetal ultrasound image
  • the cranial standard section when the craniocerebral section is the cranial horizontal section, the cranial standard section includes a lateral ventricle section, and when the cranial section is the cranial sagittal section, the cranial standard section Including the midsagittal section of the cranial brain.
  • the target information corresponding to the slice of the fetal ultrasound image includes characteristic parameters of the limb structure feature of the slice, wherein, when the slice of the fetal ultrasound image includes The limb section, and when the limb section is the double upper limb section, the limb structural features of the section include at least one of the hand, forearm and upper arm of the section; when the fetal ultrasound image section includes all Described limb section, and when described limb section is described double lower limb section, the limb structure feature of described section includes at least one of foot, thigh and calf of this section;
  • the first judgment module determines whether the fetus corresponding to the fetal ultrasound image has a deformity. Specifically:
  • the characteristic parameters of the limb structure features of the section of the fetal ultrasound image it is determined whether the limb structure feature matches the section of the fetal ultrasound image, and when it is judged that they do not match, determine the corresponding section of the fetal ultrasound image.
  • the fetus is deformed.
  • the first determination module determines whether the fetus corresponding to the fetal ultrasound image has a deformity according to the target information corresponding to the slice of the fetal ultrasound image. Specifically:
  • the target information corresponding to the slice of the fetal ultrasound image includes the feature of the bone structure feature of the slice.
  • the feature parameter of the feature includes at least one of the contour, length, area, shape and position corresponding to the bone structure feature.
  • the first determination module determines whether the fetus corresponding to the fetal ultrasound image has a deformity according to the target information corresponding to the slice of the fetal ultrasound image. Specifically:
  • the target information corresponding to the slice of the fetal ultrasound image includes the characteristic parameter of the slice, and it is determined whether the slice characteristic parameter matches the preset slice parameter. When no match is found, it is determined that the corresponding fetus of the fetal ultrasound image has a deformity, and the characteristic parameters of the slice include Doppler blood flow parameters and/or contour parameters of the slice;
  • the target information corresponding to the slice of the fetal ultrasound image includes the characteristic parameter of the cardiac structural feature of the slice, and the target information is determined according to the characteristic parameter of the cardiac structural feature of the slice.
  • the cardiac structural feature of the slice matches the standard cardiac structural feature of the slice, and when it is judged that it does not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity, and the characteristic parameters of the cardiac structural feature of the slice include all the parameters. the number of the structural features of the heart and/or the area corresponding to the structural features of the heart;
  • the manner in which the first judgment module judges whether the cardiac structure feature of the section matches the standard cardiac structure feature of the section according to the characteristic parameter of the cardiac structure feature of the section is specifically:
  • the feature parameter of the cardiac structural feature of the slice is the number of the cardiac structural feature, determine whether the number of the cardiac structural feature of the slice is less than or equal to a preset number, and when the determination result is yes, determine the number of the cardiac structural feature of the slice.
  • Cardiac structural features do not match standard cardiac structural features for this slice;
  • the feature parameter of the cardiac structural feature of the slice is the area corresponding to the cardiac structural feature, determine whether the area corresponding to the cardiac structural feature of the slice is greater than or equal to a preset area threshold, and when the determination result is yes, determine the The cardiac structural features of the slice do not match the standard cardiac structural characteristics of the slice.
  • the device further includes:
  • the analysis module is used for inputting the acquired continuous multi-frame fetal ultrasound images into the determined structural feature detection model for analysis;
  • the acquisition module is used to acquire the analysis result output by the structural feature detection model, as the structural feature information of each frame of the fetal ultrasound image, and the structural feature information of each frame of the fetal ultrasound image includes the structural feature of the part of the fetal ultrasound image.
  • information and structural feature information of the fetal ultrasound image the structural feature information of each frame of the fetal ultrasound image includes at least the type of the structural feature of the fetal ultrasound image, and the structural feature information of each frame of the fetal ultrasound image at least includes the categories of structural features of fetal ultrasound images;
  • the determining module is further configured to determine the slice of the fetal ultrasound image according to the category of the structural feature of the fetal ultrasound image and the category of the structural feature of the fetal ultrasound image in each frame.
  • the device further includes:
  • a setting module configured to set a corresponding structural feature label for each abnormal structural feature of the fetal ultrasound image when the first judgment module determines that the fetus corresponding to the fetal ultrasound image has a deformity, and each abnormality
  • the structural feature label corresponding to the structural feature is used to indicate the malformation type of the abnormal structural feature.
  • a third aspect of the present invention discloses another device for detecting severe fetal malformation based on fetal ultrasound images, the determining device comprising:
  • a processor coupled to the memory
  • the processor invokes the executable program code stored in the memory to execute the method for detecting severe fetal malformation based on a fetal ultrasound image disclosed in the first aspect of the present invention.
  • a fourth aspect of the present invention discloses a computer storage medium, where the computer storage medium stores computer instructions, and when the computer instructions are invoked, is used to perform the fetal ultrasound image-based severe fetal malformation detection disclosed in the first aspect of the present invention method.
  • a method and device for detecting severe fetal malformation based on fetal ultrasound images are provided.
  • the target information corresponding to the section of the ultrasound image is used to determine the development of the fetus corresponding to the fetal ultrasound image; according to the target information corresponding to the section of the fetal ultrasound image, it is determined whether the fetus corresponding to the fetal ultrasound image has deformity, and when the judgment result is yes,
  • the malformation of the fetus corresponding to the fetal ultrasound image is determined according to the target information corresponding to the slice of the fetal ultrasound image, where the malformation condition includes the fetal malformation type corresponding to the fetal ultrasound image.
  • Fetal malformation such as anencephaly, gastroschisis, etc., is conducive to accurate detection of fetal malformation, so as to achieve accurate determination of fetal growth and development; and comprehensive detection and analysis of fetal growth and development through multiple aspects It is beneficial to improve the detection accuracy of fetal malformation.
  • FIG. 1 is a schematic flowchart of a method for detecting severe fetal malformations based on fetal ultrasound images disclosed in an embodiment of the present invention
  • FIG. 2 is a schematic flowchart of another method for detecting severe fetal malformation based on fetal ultrasound images disclosed in an embodiment of the present invention
  • FIG. 3 is a schematic structural diagram of a device for detecting severe fetal malformation based on fetal ultrasound images disclosed in an embodiment of the present invention
  • FIG. 4 is a schematic structural diagram of another device for detecting severe fetal malformation based on fetal ultrasound images disclosed in an embodiment of the present invention
  • FIG. 5 is a schematic structural diagram of another device for detecting severe fetal malformation based on fetal ultrasound images disclosed in an embodiment of the present invention.
  • the invention discloses a method and device for detecting severe fetal malformation based on fetal ultrasound images. After acquiring the slices of the fetal ultrasound images, it can automatically determine whether the fetus is abnormal according to the information of the determined slices of the fetal ultrasound images. When there is an abnormality, the abnormality of the fetus is automatically determined according to the information of the section of the fetal ultrasound image, such as whether it is anencephaly, gastroschisis, etc., which is conducive to the accurate detection of the abnormality of the fetus, so as to realize the accurate growth and development of the fetus. It is helpful to improve the detection accuracy of fetal malformation by comprehensively detecting and analyzing the growth and development of the fetus. Each of them will be described in detail below.
  • FIG. 1 is a schematic flowchart of a method for detecting severe fetal malformation based on fetal ultrasound images disclosed in an embodiment of the present invention.
  • the method for detecting severe fetal malformation based on fetal ultrasound images described in FIG. 1 may be applied to a detection server (service equipment/service system), wherein the detection server may include a local detection server or a cloud detection server.
  • the detection server may include a local detection server or a cloud detection server.
  • the embodiment of the present invention Not limited.
  • the method for detecting severe fetal malformation based on fetal ultrasound images may include the following operations:
  • the target information corresponding to the slices of the fetal ultrasound image is used to determine the development of the fetus corresponding to the fetal ultrasound image, and different slices correspond to different fetal ultrasound images.
  • the fetal ultrasound image may be an image obtained continuously, that is, the fetal ultrasound image in step 101 represents a single frame image, and the fetal ultrasound image may also be a multi-frame image.
  • the slice of the fetal ultrasound image in step 101 is: Section of a single-frame fetal ultrasound image.
  • the section of the fetal ultrasound image includes one of a craniocerebral section, a limb section, an abdominal section, a spinal cord section, a heart section, a long-diameter section of the humerus, and a long-diameter section of the femur.
  • the brain section includes the craniocerebral horizontal section and/or the craniocerebral sagittal section
  • the limb section of the fetal ultrasound image includes the double upper limb section or the double lower limb section
  • the abdominal section of the fetal ultrasound image includes the abdominal horizontal section and/or the abdominal sagittal section
  • the spinal cord section of the fetal ultrasound image includes the spinal cord horizontal section and/or the spinal cord sagittal section.
  • the target information corresponding to the slice of the fetal ultrasound image it is determined whether the fetus corresponding to the fetal ultrasound image has a deformity, including:
  • the target information corresponding to the section of the fetal ultrasound image includes the craniocerebral structural feature contour and the buttocks structural feature contour of the section.
  • the characteristic contour of the skull structure and the structural characteristic contour of the buttocks of the section of the ultrasound image measure the head-rump length of the fetal ultrasound image, and determine whether the head-rump length is within the preset head-rump length range, and when it is judged that they do not match, determine There is an abnormality in the fetus corresponding to the fetal ultrasound image;
  • the target information corresponding to the section of the fetal ultrasound image includes the geometric parameters of the craniocerebral structural features of the section of the fetal ultrasound image. Whether the geometric parameters of the craniocerebral structural features of the section match the preset geometric parameters, and when it is judged that they do not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity.
  • the craniocerebral structural feature of the slice of the fetal ultrasound image includes at least one of the calvarial structural feature, the cerebral hemisphere structural feature, and the midbrain structural feature in the slice, and the craniocerebral structural feature is
  • the geometric parameters include at least one of the shape, size (eg, length), location, and area of the cranial structural feature, where the cranial structural feature is located in the slice of the fetal ultrasound image.
  • different pregnancy periods correspond to different preset head-rump length ranges and preset geometric parameters
  • the preset head-rump length range is 2mm-5mm
  • the preset head-butt-length range is 5cm-10cm.
  • the preset geometric parameters include a preset shape, a preset size, a preset position, and a preset area.
  • the characteristic contour of the cranial brain structure and the characteristic contour of the buttocks structure of the slice of the fetal ultrasound image, or the geometric parameters of the cranial structure feature determine the deformity of the fetus corresponding to the fetal ultrasound image, and the deformity includes the fetus corresponding to the fetal ultrasound image.
  • the types of malformations include the anencephaly type.
  • the head-rump length of the fetal ultrasound image is not within the preset head-rump length range, and it is determined that the geometric parameters of the craniocerebral structural features of the slice do not match the preset geometric parameters
  • the fetus is determined to have an abnormality in the fetus corresponding to the fetal ultrasound image. In this way, the accuracy and reliability of detecting that the fetus corresponding to the fetal ultrasound image is anencephaly can be improved.
  • the target information corresponding to the slice of the fetal ultrasound image it is determined whether the fetus corresponding to the fetal ultrasound image has a deformity, including:
  • the target information corresponding to the slice of the fetal ultrasound image includes the shape (contour) of the slice of the fetal ultrasound image, and it is determined whether the shape of the slice is consistent with the shape of the preset slice. Matching, when it is judged that there is no match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity.
  • different gestational periods correspond to different preset section shapes.
  • the shape of the slice of the fetal ultrasound image may include the shape of the horizontal slice and/or the shape of the sagittal slice, which can enrich the way of judging whether the fetus is abnormal, and when the shape of the horizontal slice and the sagittal slice are included
  • the shape of the horizontal section and the shape of the sagittal section are compared with the shape of the corresponding standard section, which is beneficial to improve the accuracy and reliability of fetal abnormality determination.
  • the section of the fetal ultrasound image is a craniocerebral horizontal section
  • the brain of the fetus is normal; when it is not a standard ellipse, that is, an ellipse with different degrees of convexity, it is determined that the structural characteristics of the brain are abnormal, that is, the brain of the fetus in the fetal ultrasound image is abnormal, it is determined that the fetus has meningocele. exception.
  • the section of the fetal ultrasound image is a craniocerebral sagittal section
  • the shape of the craniocerebral sagittal section (cranial parenchyma image or edge contour) to determine (for example: pre-trained) image classification model, and obtain the analysis result output by the image classification model, as the judgment result of the shape of the craniocerebral sagittal section, when the judgment result is used to indicate that the cranial meninges are in the bulging shape, It indicates that the fetus has abnormal meningocele.
  • the judgment result is used to indicate that the cranial meninges are in the determined normal shape, it indicates that the fetal brain is normal.
  • the degree of meningocele is determined according to the concave and convex shape of the meningocele on the oval shape, and the severity level corresponding to the meningocele is determined according to the degree of concave and convex (for example: grade 3, where , the higher the severity level, the more serious the meningocele is), so that after determining the presence of meningocele in the fetus, the severity level of the meningocele can be further determined, which is beneficial to clearly and accurately know the malformation of the meningocele in the fetus.
  • the concave-convex area is further obtained, and the severity level of the bulge is determined according to the concave-convex area, which is beneficial to improve the accuracy and reliability of the determination of the level of meningocele. It is beneficial to further improve the accuracy and reliability of determination of fetal (meningeal) bulge.
  • the target information corresponding to the slice of the fetal ultrasound image it is determined whether the fetus corresponding to the fetal ultrasound image has a deformity, including:
  • the target information corresponding to the slice of the fetal ultrasound image includes the characteristic parameters corresponding to the target structural features of the slice, and the judgment is based on the characteristic parameters corresponding to the target structural characteristics of the slice. Whether the target structural feature matches the section, and when it is judged that it does not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity.
  • the slices of different fetal ultrasound images correspond to the target structural features of different slices, specifically: the target structural characteristics corresponding to the cranial slices include transparent compartment, lateral ventricle, choroid plexus, thalamus or cerebellum, etc.
  • the target structural features corresponding to the abdominal section include gastric vesicle, umbilical vein and gallbladder, etc.; the target structural features corresponding to the spinal cord section include spine, meninges, etc.
  • the feature parameter corresponding to the target structural feature includes at least one of the contour, size, position, area, type, etc. of the target structural feature.
  • the target structural feature when it is determined according to the type of the target structural feature that the target structural feature is not a structural feature in the slice of the fetal ultrasound image, it is determined that the target structural feature does not match the slice, for example: when the cranial When the structural features of hydrocephalus appear in the brain section, it is determined that the cranial brain is an abnormal cranial brain, that is, the fetus in the fetal ultrasound image has an abnormality, and the abnormality is a meningocele.
  • the specific judgment method for determining whether the target structural feature matches the slice by the feature parameter corresponding to the target structural feature in the abdominal section or the spinal cord section is the same as that for the craniocerebral section, and will not be repeated here.
  • the target information corresponding to the slice of the fetal ultrasound image it is determined whether the fetus corresponding to the fetal ultrasound image has a deformity, including:
  • the target information corresponding to the section of the fetal ultrasound image includes the inner contour of the cranial structure feature of the section of the fetal ultrasound image and the cranial structure
  • the outer contour of the feature determine the target geometric parameters of the cranial structural feature of the section, and determine whether the target geometric parameter of the cranial structural feature is in the prediction.
  • the target geometric parameters of the craniocerebral structural feature of the slice include the head circumference parameter of the craniocerebral structural feature and/or or double parietal diameter parameter.
  • the slice of the fetal ultrasound image is input to determine the slice shape determination model for analysis, and the analysis result output by the slice shape determination model is obtained.
  • the tangent shape determination model may include any one or more combined models such as instance-based segmentation model, semantic segmentation model, etc., and the optional implementation is not limited.
  • the target geometric parameters of the cranial structure feature of the section are determined, including:
  • the calculation method of the head circumference parameter of the cranial structure feature is as follows:
  • C is the head circumference parameter of the skull structure feature, that is, the circumference of the head;
  • C 1 is the second circumference of the outer contour of the skull structure feature;
  • C 2 is the first circumference of the inner contour of the skull structure feature .
  • the first intersection includes a first sub-intersection and a second sub-intersection
  • the second intersection includes a third sub-intersection and a fourth sub-intersection.
  • the distance between the first sub-intersection and the third sub-intersection is smaller than the distance from the fourth sub-intersection.
  • the geometric parameter of the parietal diameter that is, the length of the biparietal diameter; or, the second line segment formed by connecting the second sub-intersection and the third sub-intersection is used as the biparietal diameter parameter of the structural features of the brain; or, the first line segment and the first line segment are obtained.
  • the mean of the two line segments is used as the biparietal diameter parameter corresponding to the structural features of the brain. In this way, the possibility and accuracy of obtaining the length of the biparietal diameter can be improved by providing ways of obtaining the length of the biparietal diameter of various craniocerebral structural features.
  • the perimeter of the inner and outer contours of the structural features of the skull, and the intersection of the midline corresponding to the midline of the brain and the inner and outer contours of the structural features of the skull can realize the features of the skull structure.
  • the head circumference and biparietal diameter length were obtained, so as to achieve the determination of fetal craniocele.
  • the ratio of the biparietal diameter parameter and the head circumference parameter is further obtained, and it is judged whether the ratio is greater than or equal to the predetermined value.
  • the value for example: the ratio of the circumference of the head to the length of the biparietal diameter is greater than or equal to 6
  • the judgment result is yes, the fetal craniocele can be determined, which is beneficial to improve the accuracy of the determination of fetal craniocele. performance and reliability.
  • this optional embodiment can realize the abnormality that the fetus corresponding to the fetal ultrasound image is anencephaly through the feature contour of the skull structure and the feature contour of the buttocks structure of the slice of the fetal ultrasound image and/or the geometric parameters of the skull structure feature. Determining, and determining the corresponding fetal ultrasound image by determining that the head-rump length of the fetal ultrasound image is not within the preset head-rump length range and determining that the geometric parameters of the craniocerebral structural features of the slice do not match the preset geometric parameters.
  • the presence of malformations in the fetus can improve the accuracy and reliability of the detection of anencephaly in the fetus corresponding to the fetal ultrasound image.
  • the target information corresponding to the slice of the fetal ultrasound image it is determined whether the fetus corresponding to the fetal ultrasound image has a deformity, including:
  • the target information corresponding to the section of the fetal ultrasound image includes the contour of the left thalamus and the contour of the right thalamus.
  • the contour of the left thalamus and the contour of the right thalamus is determined, and whether the first fitting degree is greater than or equal to a first preset fitting degree threshold is determined. When the determination result is yes, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity.
  • the target information corresponding to the section of the fetal ultrasound image includes the position of the falx cerebellum of the section, and the falx cerebellum is determined according to the position of the falx in the fetus. Whether the appearance of the midline position of the brain in the ultrasound image satisfies the preset appearance, and when it is determined that the appearance does not meet the preset appearance, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity.
  • the target information corresponding to the section of the fetal ultrasound image includes the contour of the choroid plexus and the contour of the thalamus, and the difference between the contour of the choroid plexus and the contour of the thalamus is obtained. and determining whether the second fitting degree is greater than or equal to a second preset fitting degree threshold, and when the determination result is yes, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity.
  • the contour of the choroid plexus includes the contour of the left choroid plexus and the contour of the right choroid plexus
  • the contour of the thalamus includes the contour of the left thalamus and the contour of the right thalamus.
  • the second degree of fit between the contour of the choroid plexus and the contour of the thalamus is obtained. Specifically, the degree of fit between the contour of the left choroid plexus and the contour of the left thalamus and/or the contour of the right choroid plexus and the contour of the right thalamus are respectively obtained.
  • the second degree of fit between the contour of the choroid plexus and the contour of the thalamus so that by obtaining the degree of fit of the left and right choroid plexus and the left and right thalamus respectively, it is beneficial to improve the accuracy and reliability of the determination that the choroid plexus and the thalamus are not separated. It is beneficial to improve the accuracy and reliability of determining that the fetus in the fetal ultrasound image is a lobeless whole forebrain fetus.
  • this optional embodiment can not only realize that the fetus in the fetal ultrasound image is a lobeless whole forebrain fetus by determining at least one of the fusion of the left and right thalamus in the cranial brain, the absence of the falx cerebrum, and the non-separation of the choroid plexus and the thalamus.
  • the determination of lobeless whole forebrain fetuses can also enrich the determination methods of lobarless whole forebrain fetuses; and by providing at least two determination methods of lobeless whole forebrain fetuses, the accuracy and reliability of the determination of lobeless whole forebrain fetuses can be improved.
  • the thalamus of the fetus corresponding to the fetal ultrasound image is not fused, that is, the thalamus of the fetus corresponding to the fetal ultrasound image is normal;
  • the coincidence threshold it is determined that the choroid plexus of the fetus corresponding to the fetal ultrasound image is separated from the thalamus.
  • the thalamus of the fetus corresponding to the fetal ultrasound image is not fused, the falx cerebral is normal, and the choroid plexus is separated from the thalamus, it is determined that the fetus corresponding to the fetal ultrasound image is not a lobulated whole forebrain fetus.
  • the target information corresponding to the slice of the fetal ultrasound image includes characteristic parameters of the limb structural features of the slice, wherein, when the slice of the fetal ultrasound image includes a limb slice, and the limb slice is both upper limbs
  • the limb structural features of the section include at least one of the hand, forearm, and upper arm of the section; when the section of the fetal ultrasound image includes a limb section, and the limb section is a section of both lower limbs, the limb structure features of the section include this section. At least one of the feet, thighs, and calves of the cut surface.
  • the target information corresponding to the slice of the fetal ultrasound image to determine whether the fetus corresponding to the fetal ultrasound image has a deformity, may include:
  • the characteristic parameters of the limb structure features of the section of the fetal ultrasound image it is determined whether the limb structure feature matches the section of the fetal ultrasound image, and when it is judged that they do not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity.
  • the feature parameter of the limb structure feature includes at least one of the number, shape, size (for example: length) and position of the limb structure feature, so that the more content the feature parameter of the limb structure feature includes, It is more beneficial to improve the accuracy and reliability of the judgment of whether the structural features of the limbs match the section of the fetal ultrasound image, which is beneficial to the accuracy and reliability of the determination of the fetal malformation.
  • the structural feature of the limb when it is determined that the structural feature of the limb includes the hand, it can be determined that the structural feature of the limb matches the section of the fetal ultrasound image only when it is determined that the corresponding section has the wrist, the palm and at least one finger. ;
  • the structural features of the limbs include feet, only when it is judged that there are ankles, soles of feet and at least one toe in the corresponding cutting planes, it can be determined that the structural features of the limbs match the cutting planes of the fetal ultrasound image; Only when the position is in the position corresponding to the human body can it be determined that the structural features of the limbs match the slices of the fetal ultrasound image, which can further improve the accuracy and reliability of determining the matching of the structural features of the limbs to the slices of the fetal ultrasound image.
  • this optional embodiment can determine whether the fetus corresponding to the fetal ultrasound image is a fetus with missing limbs through the characteristic parameters of limb structural features (eg, hands and feet) in the slice of the fetal ultrasound image.
  • limb structural features eg, hands and feet
  • judging whether the fetus corresponding to the fetal ultrasound image has a deformity may include:
  • the target information corresponding to the slice of the fetal ultrasound image includes the characteristic parameters of the bone structural features of the slice.
  • the parameter judges whether the bone structure feature matches the cut plane, and when it is judged that it does not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity.
  • the characteristic parameter of the bone structure feature includes at least one of the contour, length, area, shape, and position corresponding to the bone structure feature.
  • the target information corresponding to the slice of the fetal ultrasound image is determined, including:
  • the characteristic parameters include the cranial thickness that characterizes the cranial structure.
  • the thickness of the cranial brain is within the determined range of the thickness of the cranial brain, and when it is determined that it is within the range of the thickness of the cranial brain, it is determined that the cranial brain of the fetal ultrasound image is normal, and when it is determined that it is not within the range of the thickness of the cranial brain , judging whether the cranial thickness is greater than the maximum thickness value of the cranial thickness range, and when it is judged that it is greater than the maximum thickness value, determine the craniocerebral abnormality in the fetal ultrasound image. Further, the thickness grade of the brain is determined according to the thickness of the brain.
  • the characteristic parameter of the bone structure feature also includes the skull shape of the skull structure feature, and it is determined whether the skull shape matches the determined skull shape, and when it is determined that there is no match, it is determined that the skull shape of the fetal ultrasound image is abnormal. .
  • the target information corresponding to the section of the fetal ultrasound image includes parameters of structural features of the skull forehead, wherein the forehead of the skull is
  • the parameters of the structural feature include the area and shape of the area enclosed by the contour of the skull forehead structural feature, and the value of the target vertical distance between the most distal end of the contour of the skull forehead structural feature and the outer contour of the skull.
  • the degree of protrusion of the forehead of the skull is determined according to the protrusion shape of the forehead of the skull, and the severity level corresponding to the protrusion of the forehead of the skull is determined according to the degree of protrusion (for example, : Level 3, where the higher the severity level, the more severe the forward protrusion of the skull forehead is), so that after it is determined that the fetus has forward protrusion of the skull forehead, the severity level of the forward protrusion of the skull forehead is further determined, which is conducive to clear And accurately know the deformity of the forward protrusion of the skull of the fetus, which is conducive to improving the accuracy and reliability of the level of the forward protrusion of the skull, which is conducive to further improving the fetal ultrasound image of the fetus is a lethal bone dysplasia fetus. Determine accuracy and reliability.
  • the characteristic parameter of the bone structure feature is the length of the bone structure feature
  • it is judged whether the length of the bone structure feature is within the preset length range and when it is judged that it is not within the preset length range, determine whether the bone structure feature is within the preset length range.
  • the structural features do not match the section, wherein different gestational weeks correspond to different preset length ranges, for example: the 5th gestational week, the preset length range is 2mm-5mm, the 10th gestational week, the preset length range 4cm-8cm.
  • this optional embodiment can realize that the fetus in the fetal ultrasound image has lethal bone dysplasia by matching the characteristic parameter (for example: length) of the bone structure feature of the measured slice of the fetal ultrasound image with the corresponding slice. Determination and analysis of fetal lethal skeletal dysplasia through the dimensions of the skull, humerus, and femur can determine the accuracy and reliability of fetal lethal skeletal dysplasia.
  • the characteristic parameter for example: length
  • judging whether the fetus corresponding to the fetal ultrasound image has a deformity may include:
  • the target information corresponding to the slice of the fetal ultrasound image includes the characteristic parameters of the slice, and it is determined whether the slice characteristic parameters match the preset slice parameters.
  • the corresponding fetus in the image has a deformity, and the characteristic parameters of the slice include Doppler blood flow parameters and/or contour parameters of the slice;
  • the target information corresponding to the slice of the fetal ultrasound image includes the characteristic parameter of the cardiac structural feature of the slice, and it is determined whether the cardiac structural feature of the slice is consistent with the slice according to the characteristic parameter of the cardiac structural feature of the slice.
  • the standard cardiac structural features of the fetus are matched, and when it is judged that they do not match, it is determined that the corresponding fetus in the fetal ultrasound image has an abnormality;
  • the characteristic parameter of the cardiac structure feature of the section it is determined whether the cardiac structure feature of the section matches the standard cardiac structure feature of the section, including:
  • the characteristic parameter of the cardiac structural features of the slice is the number of cardiac structural features, determine whether the number of cardiac structural features of the slice is less than or equal to the preset number, and when the determination result is yes, determine the cardiac structural feature of the slice and the standard of the slice Cardiac structural features do not match;
  • the feature parameter of the cardiac structural feature of the slice is the area corresponding to the cardiac structural feature, determine whether the area corresponding to the cardiac structural feature of the slice is greater than or equal to a preset area threshold, and when the determination result is yes, determine whether the cardiac structural feature of the slice is the same as the area.
  • the standard cardiac structural features of the slices do not match.
  • the characteristic parameter of the cardiac structural feature of the slice includes the number of the cardiac structural feature and/or the area corresponding to the cardiac structural feature.
  • the standard cardiac structural features of the slices of the fetal ultrasound images corresponding to different gestational weeks are also different, for example: in the second gestational week, the standard cardiac structural features include the endocardial cushion and the area of the endocardial cushion, and the first At 14 weeks of gestation, standard cardiac structural features included left atrium, left ventricle, right atrium, right ventricle, endocardial cushion, and the area of each standard cardiac structural feature.
  • the cardiac section when it is determined that the number and area of the cardiac structural features of the section match the standard cardiac structural features, the cardiac section is a normal cardiac section, that is, the fetal heart is a normal heart; If the number of cardiac structural features in the outgoing section is less than or equal to a preset number (for example: 2), the cardiac section is determined to be an abnormal cardiac section. Specifically, when it is determined that the number of cardiac structural features in the slice is one (for example, only the left ventricle exists), and/or the ratio of the area of the cardiac structural feature to the area of the cardiac slice is greater than or equal to a preset ratio value , then it is determined that the cardiac section is a single-ventricular cardiac section.
  • a preset number for example: 2
  • the cardiac section is It is a single ventricle heart section, that is, the fetal heart is a single ventricle state.
  • this optional embodiment determines whether the fetal heart is abnormal in the fetal ultrasound image by using the Doppler blood flow parameters, contour parameters, the number of cardiac structural features, and the area of each cardiac structural feature of the cardiac section, which can improve the fetal heart rate.
  • the accuracy and reliability of determining the abnormality of the heart of the fetus is beneficial to improve the accuracy and reliability of determining the fetal cardiac malformation.
  • the method for detecting severe fetal malformation based on the fetal ultrasound image may further include the following operations:
  • the section of the fetal ultrasound image includes the cranial section
  • the standard section of the cranial brain when the cranial section is the horizontal section of the brain, the standard section of the cranial brain includes the lateral ventricle section, and when the cranial section is the sagittal section of the brain, the standard section of the cranial brain includes the mid-sagittal section of the brain.
  • the craniocerebral section is a craniocerebral sagittal section
  • the acquired structural features include one or more of nasal bone structural features, maxillary structural features, mandibular structural features, neck transparent layer structural features, and fetal torso, etc.
  • the obtained structural features include one or more of the structural features of the transparent compartment, the structural features of the lateral ventricle, the structural features of the choroid plexus, the structural features of the thalamus, and the structural features of the cerebellum. kind of combination.
  • this optional embodiment first determines whether the obtained slice of the fetal ultrasound image matches the standard slice of the cranial brain, and if so, continues to perform the subsequent operation of obtaining the target information of the slice, and if not, obtains the corresponding slice. It is helpful to improve the accuracy and reliability of fetal malformation determination by matching the slices with the standard slices of the brain and re-acquiring the target information of the slices.
  • Step 101 According to the target information corresponding to the section of the fetal ultrasound image, determine whether the fetus corresponding to the fetal ultrasound image has a deformity. When the determination result is yes, trigger the execution of step 103, and when the determination result is no, this process can be ended, and the Step 101 may be triggered when the above-mentioned fetal ultrasound image is a single-frame fetal ultrasound image.
  • the fetus corresponding to the fetal ultrasound image has a deformity
  • the degree (and/or grade) of conjoined conjoined deformity of the fetus corresponding to the fetal ultrasound image is determined according to the target information corresponding to the slice of the fetal ultrasound image.
  • the fetus has a deformity
  • the accuracy and reliability of determining the presence of conjoined fetal malformation in the fetus corresponding to the fetal ultrasound image can be further improved.
  • 103 Determine, according to the target information corresponding to the slice of the fetal ultrasound image, a fetal malformation condition corresponding to the fetal ultrasound image, where the malformation condition includes the fetal malformation type corresponding to the fetal ultrasound image.
  • the fetal malformation types corresponding to the fetal ultrasound images include anencephaly malformation type, meningocele malformation type, lobar holoprosencephaly malformation type, cystic myelomeningocele type , at least one of the type of limb (one or both) missing deformity, the type of single ventricle malformation, the type of gastroschisis with visceral valgus deformity, and the type of fatal bone dysplasia, and further, it can also include conjoined fetuses Type of deformity.
  • conjoined fetuses Type of deformity In this way, by comprehensively detecting and analyzing fetal ultrasound images, a comprehensive fetal developmental malformation situation can be obtained, and the accuracy and reliability of the fetal developmental malformation situation can be further improved.
  • the method for detecting severe fetal malformation based on fetal ultrasound images may further include the following operations:
  • a corresponding structural feature label is set for each abnormal structural feature of the fetal ultrasound image, and the structural feature label corresponding to each abnormal structural feature is used to indicate the abnormal structural feature.
  • Type of deformity is set for each abnormal structural feature of the fetal ultrasound image, and the structural feature label corresponding to each abnormal structural feature is used to indicate the abnormal structural feature.
  • a corresponding structural feature label is set for the abnormal structural feature, so that the medical staff can clearly and quickly know the deformity of the fetus according to the structural feature label.
  • the method for detecting severe fetal malformation based on fetal ultrasound images may further include the following operations:
  • a detection result of the fetus corresponding to the fetal ultrasound image is also generated, wherein the detection result is used to indicate that the fetus corresponding to the fetal ultrasound image has normal growth and development.
  • the detection result of the fetus is also generated, which further facilitates the relevant personnel to know the growth and development of the fetus.
  • the implementation of the method for detecting severe fetal malformation based on fetal ultrasound images described in FIG. 1 can automatically determine whether there is an abnormality in the fetus according to the information of the determined slices of the fetal ultrasound image after acquiring the slices of the fetal ultrasound images.
  • the abnormality of the fetus is automatically determined according to the information of the section of the fetal ultrasound image, such as whether it is anencephaly, gastroschisis, etc., which is conducive to the accurate detection of the abnormality of the fetus, so as to realize the accurate determination of the growth and development of the fetus. ;
  • FIG. 2 is a schematic flowchart of another method for detecting severe fetal malformation based on fetal ultrasound images disclosed in an embodiment of the present invention.
  • the method for detecting severe fetal malformation based on fetal ultrasound images described in FIG. 2 may be applied to a detection server (service device/service system), wherein the detection server may include a local detection server or a cloud detection server, the embodiment of the present invention Not limited.
  • the method for detecting severe fetal malformation based on fetal ultrasound images may include the following operations:
  • the structural feature information of each frame of fetal ultrasound image includes the site structure feature information of the fetal ultrasound image and the structure feature information of the fetal ultrasound image
  • the site structure feature information of each fetal ultrasound image includes at least the fetal ultrasound image.
  • the category of the structural feature of the part of the image, and the structural feature information of each frame of fetal ultrasound image at least includes the category of the structural feature of the fetal ultrasound image.
  • the specific implementation manner of determining the shape of the slice of the fetal ultrasound image in the first embodiment may also be that when the fetal ultrasound image is input into the structural feature detection model for analysis, the fetal ultrasound image can be analyzed simultaneously.
  • the shape of the slice of the ultrasound image can improve the acquisition efficiency of the shape of the slice of the fetal ultrasound image while ensuring the accuracy of the acquisition of the shape of the slice of the fetal ultrasound image, thereby helping to improve the efficiency of judging whether the fetus has abnormality.
  • multiple frames of fetal ultrasound images can be continuously acquired according to a predetermined frame rate, wherein the predetermined frame rate is related to the slice of the fetal ultrasound image to be acquired, that is, according to the fetal ultrasound image to be acquired
  • the frame rate is selected according to the slice of the image. For example, if the abdominal slice is to be acquired, the frame rate can be 30 frames/second; if the four-chamber slice is to be acquired, the frame rate can be 60 frames/second.
  • the corresponding frame rate is selected according to the slice of the fetal ultrasound image to be obtained, which is beneficial to improve the efficiency and accuracy of the slice of the fetal ultrasound image, thereby helping to improve the efficiency of the acquisition of the target information of the slice of the fetal ultrasound image. It is beneficial to improve the judgment efficiency of abnormal fetal ultrasound images.
  • each frame of fetal ultrasound image has a unique corresponding frame sequence number. In this way, by setting a unique frame number for each frame of fetal ultrasound images, each frame of fetal ultrasound images can be clearly distinguished during the acquisition of the slices of the fetal ultrasound images, which is beneficial to the management of the information of the fetal ultrasound images and their slices.
  • the structural feature detection model may include at least one of a target detection model, an instance segmentation model, a semantic segmentation model, etc., which can obtain the structural feature information and structural feature information of the fetal ultrasound image.
  • This embodiment of the present invention does not Do limit.
  • the embodiment of the present invention determines the standard section of the fetal ultrasound image by acquiring the part structure features and structural features of consecutive multiple frames of fetal ultrasound images, and combining the site structure features and structural features of the fetal ultrasound image, without manual participation in the fetal ultrasound image.
  • the determination of the standard slice can improve the accuracy of the determination of the standard slice of the fetal ultrasound image; and by inputting the fetal ultrasound image into the structural feature detection model for analysis, it can also improve the efficiency of the determination of the standard slice of the fetal ultrasound image.
  • the accuracy and reliability of the deformity detection is determining the standard section of the fetal ultrasound image by acquiring the part structure features and structural features of consecutive multiple frames of fetal ultrasound images, and combining the site structure features and structural features of the fetal ultrasound image, without manual participation in the fetal ultrasound image.
  • the slice of the fetal ultrasound image may also be obtained by receiving slices of each frame of the fetal ultrasound image in the multiple frames of fetal ultrasound images sent by the authorized terminal device.
  • the slices of the fetal ultrasound image can be obtained through various ways, which can enrich the ways of obtaining slices and improve the possibility of obtaining slices.
  • step 204 Determine whether the fetus corresponding to the fetal ultrasound image has a deformity according to the target information corresponding to the slice of the fetal ultrasound image.
  • the determination result is yes, trigger the execution of step 103, and when the determination result is no, end this process, or you can Trigger to execute step 203 .
  • 205 Determine, according to the target information corresponding to the slices of the fetal ultrasound image, a fetal malformation condition corresponding to the fetal ultrasound image, where the malformation condition includes the fetal malformation type corresponding to the fetal ultrasound image.
  • the implementation of the method for detecting severe fetal malformation based on fetal ultrasound images described in FIG. 2 can automatically determine whether there is an abnormality in the fetus according to the information of the determined slices of the fetal ultrasound image after acquiring the slices of the fetal ultrasound images.
  • the abnormality of the fetus is automatically determined according to the information of the section of the fetal ultrasound image, such as whether it is anencephaly, gastroschisis, etc., which is conducive to the accurate detection of the abnormality of the fetus, so as to realize the accurate determination of the growth and development of the fetus.
  • the accuracy of determining the standard section of the fetal ultrasound image can be improved; and by inputting the fetal ultrasound image into the structural feature detection model for analysis, the efficiency of determining the standard section of the fetal ultrasound image can also be improved, thereby helping to improve fetal malformation.
  • the detection accuracy and reliability of the situation can be improved.
  • FIG. 3 is a schematic structural diagram of a device for detecting severe fetal malformations based on fetal ultrasound images disclosed in an embodiment of the present invention.
  • the device for detecting severe fetal malformation based on fetal ultrasound images described in FIG. 3 may be applied to a detection server (service equipment/service system), wherein the detection server may include a local detection server or a cloud detection server, an embodiment of the present invention Not limited.
  • the device for detecting severe fetal malformation based on fetal ultrasound images may include a determination module 301 and a first determination module 302, wherein:
  • the determination module 301 is configured to determine target information corresponding to the slice of the fetal ultrasound image after acquiring the slice of the fetal ultrasound image, and the target information corresponding to the slice of the fetal ultrasound image is used to determine the development of the fetus corresponding to the fetal ultrasound image ;
  • the first judgment module 302 is configured to judge whether the fetus corresponding to the fetal ultrasound image has a deformity according to the target information corresponding to the slice of the fetal ultrasound image.
  • the determining module 301 is further configured to, when the first determining module 302 determines that the result is yes, determine the deformity of the fetus corresponding to the fetal ultrasound image according to the target information corresponding to the slice of the fetal ultrasound image, where the deformity includes the fetus corresponding to the fetal ultrasound image. type of deformity.
  • the slice of the fetal ultrasound image includes one of a craniocerebral slice, a limb slice, an abdominal slice, a spinal cord slice, a cardiac slice, a long-diameter slice of the humerus, and a long-diameter slice of the femur.
  • the cranial section of the image includes the horizontal section of the brain and/or the sagittal section of the cranial brain
  • the limb section of the fetal ultrasound image includes the double upper limb section or the double lower limb section
  • the abdominal section of the fetal ultrasound image includes the abdomen horizontal section and/or abdomen Sagittal section
  • the spinal cord section of the fetal ultrasound image includes the spinal cord horizontal section and/or the spinal cord sagittal section.
  • the device for detecting severe fetal malformation based on the fetal ultrasound image described in FIG. 3 can automatically determine whether there is an abnormality in the fetus according to the information of the determined slice of the fetal ultrasound image after acquiring the slice of the fetal ultrasound image.
  • the abnormality of the fetus is automatically determined according to the information of the section of the fetal ultrasound image, such as whether it is anencephaly, gastroschisis, etc., which is conducive to the accurate detection of the abnormality of the fetus, so as to realize the accurate determination of the growth and development of the fetus. ;
  • the first judging module 302 determines whether the fetus corresponding to the fetal ultrasound image has an abnormality, specifically:
  • the target information corresponding to the section of the fetal ultrasound image includes the craniocerebral structural feature contour and the buttocks structural feature contour of the section.
  • the characteristic contour of the skull structure and the structural characteristic contour of the buttocks of the section of the ultrasound image measure the length of the head and buttocks of the fetal ultrasound image, and determine whether the length of the head and buttocks is within the preset range of the head and buttocks. When it is judged that they do not match, the fetus is determined.
  • the fetus corresponding to the ultrasound image has an abnormality;
  • the target information corresponding to the section of the fetal ultrasound image includes the geometric parameters of the craniocerebral structural features of the section of the fetal ultrasound image, and the section is determined.
  • the geometric parameters of the craniocerebral structural features of the section match the preset geometric parameters, when it is judged that they do not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity, and the cranial structural features of the section include the calvaria in the section At least one of structural features, cerebral hemisphere structural features, and midbrain structural features, the geometric parameters of the skull structural features include at least one of the shape, size, location, and area of the skull structural features, the skull structural features is the position of the cranial structure feature in the slice of the fetal ultrasound image;
  • the target information corresponding to the section of the fetal ultrasound image includes the shape of the section of the fetal ultrasound image, and it is determined whether the shape of the section matches the shape of the preset section. When it is judged that they do not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity;
  • the target information corresponding to the slice of the fetal ultrasound image includes the feature parameters corresponding to the target structural features of the slice, according to the feature parameters corresponding to the target structural features of the slice. Judging whether the target structural feature matches the section, and when it is judged that it does not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity;
  • the target information corresponding to the section of the fetal ultrasound image includes the inner contour of the craniocerebral structural feature of the section of the fetal ultrasound image and the craniocerebral section.
  • the outer contour of the structural feature determine the target geometric parameters of the cranial structural feature of the section, and determine whether the target geometric parameters of the cranial structural feature are within the Within the range of the preset geometric parameters, when it is determined that it is not within the range of the preset geometric parameters, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity, and the target geometric parameters of the cranial structure feature of the slice include the head circumference parameter of the cranial structure feature. and/or biparietal diameter parameters;
  • the target information corresponding to the section of the fetal ultrasound image includes the left thalamus contour and the right thalamus contour of the section, and the left thalamus contour and the right thalamus contour are obtained.
  • the target information corresponding to the section of the fetal ultrasound image includes the position of the falx of the section, and the falx is determined according to the position of the falx. Whether the occurrence of the midline position of the brain in the fetal ultrasound image meets the preset occurrence situation, and when it is determined that the preset occurrence situation is not met, determine that the fetus corresponding to the fetal ultrasound image has a deformity;
  • the target information corresponding to the section of the fetal ultrasound image includes the contour of the choroid plexus and the contour of the thalamus, and the contour of the choroid plexus and the contour of the thalamus are obtained. and determine whether the second fit degree is greater than or equal to a second preset fit degree threshold, and when the determination result is yes, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity.
  • the determination device described in FIG. 3 by judging whether the structural features in the section of the fetal ultrasound image match the section, it is possible to achieve craniocerebral meningocele, cystic myelomeningocele, and gastroschisis with extra-visceral and extra-visceral malformations. It is helpful to improve the accuracy of judging the fetal malformation; and the fetal ultrasound image can be realized through the feature contour of the skull structure and the feature contour of the buttocks structure and/or the geometric parameters of the feature of the brain structure in the slice of the fetal ultrasound image.
  • the corresponding abnormal determination of the fetus as anencephaly and by determining at least one of the fusion of the left and right thalamus in the cranial brain, the absence of the falx cerebral, and the non-separation of the choroid plexus and the thalamus, not only can the fetus of the fetal ultrasound image be lobulated.
  • the determination of the lobeless whole forebrain fetus can also enrich the way of determining the lobeless whole forebrain fetus; and by providing at least two ways of determining the lobeless whole forebrain fetus, the accuracy and reliability of the determination of the lobeless whole forebrain fetus can be improved. sex.
  • the first determination module 302 determines the craniocerebral structural feature of the section according to the inner contour of the cranial structural feature of the section and the outer contour of the cranial structural feature of the section.
  • the method of target geometric parameters is as follows:
  • intersection point is used to determine the biparietal diameter parameters corresponding to the structural features of the skull.
  • the implementation of the determination device described in FIG. 3 can realize the cranial brain structure by obtaining the perimeters of the inner and outer contours of the cranial structural features, and the intersection of the mid-perpendicular line and the inner and outer contours corresponding to the brain midline of the craniocerebral structural features.
  • the apparatus may further include a second judgment module 303 and a correction module 304, wherein:
  • the second judging module 303 is configured to, when the slice of the fetal ultrasound image includes a craniocerebral slice, after acquiring the slice of the fetal ultrasound image and before the determining module 301 determines the target information corresponding to the slice of the fetal ultrasound image, determine the fetal ultrasound Whether the slice of the image matches the standard slice of the cranial brain, when the judgment result is yes, trigger the determination module 301 to perform the above-mentioned operation of determining the target information corresponding to the slice of the fetal ultrasound image;
  • the correction module 304 is used to correct the slice of the fetal ultrasound image based on the acquired structural features when the second determination module 303 determines that it does not match, so that the slice matches the standard slice of the cranial brain, and triggers the determination module 301 to execute the above-mentioned The operation of determining the target information corresponding to the slice of the fetal ultrasound image;
  • the standard section of the cranial brain when the cranial section is the horizontal section of the brain, the standard section of the cranial brain includes the lateral ventricle section, and when the cranial section is the sagittal section of the brain, the standard section of the cranial brain includes the mid-sagittal section of the brain.
  • the determination device described in FIG. 4 is implemented by first judging whether the obtained slice of the fetal ultrasound image matches the standard slice of the cranial brain. Obtaining a slice that matches the standard slice of the brain and re-acquiring the target information of the slice is beneficial to improve the accuracy and reliability of fetal malformation determination.
  • the target information corresponding to the slice of the fetal ultrasound image includes the characteristic parameter of the limb structure feature of the slice, wherein, when the slice of the fetal ultrasound image includes the limb slice, And when the limb section is a double upper limb section, the limb structural features of the section include at least one of the hand, forearm and upper arm of the section; when the section of the fetal ultrasound image includes a limb section, and the limb section is a double lower limb section, the section's
  • the limb structural feature includes at least one of the foot, the thigh and the lower leg of the section;
  • the first judgment module 302 determines whether the fetus corresponding to the fetal ultrasound image has an abnormality, specifically:
  • the characteristic parameters of the limb structure features of the section of the fetal ultrasound image it is determined whether the limb structure feature matches the section of the fetal ultrasound image, and when it is judged that they do not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity.
  • the first judgment module 302 determines whether the fetus corresponding to the fetal ultrasound image has a deformity according to the target information corresponding to the slice of the fetal ultrasound image, specifically:
  • the target information corresponding to the slice of the fetal ultrasound image includes the characteristic parameters of the bone structural features of the slice.
  • the parameters determine whether the bone structure feature matches the section. When it is judged that it does not match, it is determined that the fetus corresponding to the fetal ultrasound image has a deformity.
  • the characteristic parameters of the bone structure feature include the contour, length, area, shape and shape corresponding to the bone structure feature. at least one of the locations.
  • the determination device described in FIG. 3 or 4 can realize that the fetus in the fetal ultrasound image is lethal by matching the characteristic parameter (for example: length) of the bone structure feature of the measured slice of the fetal ultrasound image with the corresponding slice. Determination of skeletal dysplasia.
  • the first judgment module 302 judges whether the fetus corresponding to the fetal ultrasound image is abnormal according to the target information corresponding to the slice of the fetal ultrasound image, specifically:
  • the target information corresponding to the slice of the fetal ultrasound image includes the characteristic parameters of the slice, and it is determined whether the slice characteristic parameters match the preset slice parameters.
  • the corresponding fetus in the image has a deformity, and the characteristic parameters of the slice include Doppler blood flow parameters and/or contour parameters of the slice;
  • the target information corresponding to the slice of the fetal ultrasound image includes the characteristic parameter of the cardiac structural feature of the slice, and it is determined whether the cardiac structural feature of the slice is consistent with the slice according to the characteristic parameter of the cardiac structural feature of the slice.
  • the standard cardiac structural features of the slices match, and when it is judged that they do not match, it is determined that the corresponding fetus in the fetal ultrasound image has a deformity, and the characteristic parameters of the cardiac structural features of the slice include the number of cardiac structural features and/or the area corresponding to the cardiac structural features;
  • the manner in which the first judgment module 302 judges whether the cardiac structural feature of the sectional plane matches the standard cardiac structural feature of the sectional plane according to the characteristic parameter of the cardiac structural feature of the sectional plane is specifically:
  • the characteristic parameter of the cardiac structural features of the slice is the number of cardiac structural features, determine whether the number of cardiac structural features of the slice is less than or equal to the preset number, and when the determination result is yes, determine the cardiac structural feature of the slice and the standard of the slice Cardiac structural features do not match;
  • the feature parameter of the cardiac structural feature of the slice is the area corresponding to the cardiac structural feature, determine whether the area corresponding to the cardiac structural feature of the slice is greater than or equal to a preset area threshold, and when the determination result is yes, determine whether the cardiac structural feature of the slice is the same as the area.
  • the standard cardiac structural features of the slices do not match.
  • implementing the determination device described in FIG. 3 or 4 can determine whether the fetal heart in the fetal ultrasound image exists through the Doppler blood flow parameters, contour parameters, the number of cardiac structural features, and the area of each cardiac structural feature of the cardiac section.
  • the abnormality can improve the accuracy and reliability of determining the abnormality of the heart of the fetus, and is beneficial to improve the accuracy and reliability of determining the abnormality of the heart of the fetus.
  • the apparatus further includes an analysis module 305 and an acquisition module 306, wherein:
  • the analysis module 305 is configured to input the acquired continuous multiple frames of fetal ultrasound images into the determined structural feature detection model for analysis.
  • the obtaining module 306 is used to obtain the analysis result output by the structural feature detection model, as the structural feature information of each frame of fetal ultrasound image, and the structural feature information of each frame of fetal ultrasound image includes the part structure feature information of the fetal ultrasound image and the fetal ultrasound image.
  • the determining module 301 is further configured to determine the slice of the fetal ultrasound image according to the category of the structural features of each frame of the fetal ultrasound image and the category of the structural features of the fetal ultrasound image.
  • implementing the determination device described in FIG. 4 can determine the standard section of the fetal ultrasound image by acquiring the structural features and structural features of successive multiple frames of fetal ultrasound images, and combining the structural features and structural features of the fetal ultrasound images, without manual labor. Participating in the determination of the standard section of the fetal ultrasound image can improve the accuracy of the determination of the standard section of the fetal ultrasound image; and by inputting the fetal ultrasound image into the structural feature detection model for analysis, it can also improve the efficiency of the determination of the standard section of the fetal ultrasound image. Thereby, it is beneficial to improve the detection accuracy and reliability of fetal malformation.
  • the apparatus further includes a setting module 307, wherein:
  • the setting module 307 is used to set a corresponding structural feature label for each abnormal structural feature of the fetal ultrasonic image when the first judgment module 302 judges that the fetus corresponding to the fetal ultrasound image has a deformity, and the structural feature corresponding to each abnormal structural feature Labels are used to indicate the type of malformation characteristic of the abnormal structure.
  • FIG. 5 is another device for detecting severe fetal malformation based on fetal ultrasound images disclosed in an embodiment of the present invention.
  • the device for detecting severe fetal malformation based on fetal ultrasound images described in FIG. 5 may be applied to a detection server (service equipment/service system), wherein the detection server may include a local detection server or a cloud detection server, an embodiment of the present invention Not limited.
  • the device for detecting severe fetal malformation based on fetal ultrasound images may include:
  • a memory 501 storing executable program code
  • processor 502 coupled to the memory 501;
  • an input interface 503 coupled with the processor 502 and an output interface 504;
  • the processor 502 calls the executable program code stored in the memory 501 to execute some or all of the steps in the method for detecting severe fetal malformation based on fetal ultrasound images 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 execute the fetal ultrasound image-based fetal diagnosis described in Embodiment 1 or Embodiment 2 Some or all of the steps in a deformity detection 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 detecting severe fetal malformations based on fetal ultrasound images.
  • 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 may 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 and device for detecting severe fetal malformation based on fetal ultrasound images disclosed in the embodiments of the present invention are only preferred embodiments of the present invention, and only It is used to illustrate the technical solution of the present invention, but not 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; Modifications are made to the solutions, or equivalent replacements are made to some of the technical structural 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)
  • Medical Informatics (AREA)
  • Surgery (AREA)
  • Pathology (AREA)
  • Radiology & Medical Imaging (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Physics & Mathematics (AREA)
  • Molecular Biology (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physiology (AREA)
  • Gynecology & Obstetrics (AREA)
  • Pregnancy & Childbirth (AREA)
  • Ultra Sonic Daignosis Equipment (AREA)

Abstract

一种基于胎儿超声图像的胎儿严重畸形检测方法及装置,基于胎儿超声图像的胎儿严重畸形检测方法包括在获取到胎儿超声图像的切面后,确定胎儿超声图像的切面对应的目标信息(S101);根据胎儿超声图像的切面对应的目标信息,判断胎儿超声图像对应的胎儿是否存在畸形(S102),若是,则根据胎儿超声图像的切面对应的目标信息确定胎儿超声图像对应的胎儿的畸形情况,该畸形情况包括胎儿的畸形类型(S103)。可见,实施基于胎儿超声图像的胎儿严重畸形检测方法及装置能够自动根据确定出的胎儿超声图像的切面的信息判断胎儿是否存在异常,当存在异常时,自动确定胎儿的畸形情况,如:是否为无脑儿等,有利于准确检测胎儿的畸形情况,从而实现胎儿的生长发育情况的准确确定;及通过多方面综合检测及分析胎儿的生长发育情况,有利于提高胎儿畸形情况的检测准确性。

Description

基于胎儿超声图像的胎儿严重畸形检测方法及装置 技术领域
本发明涉及图像处理技术领域,尤其涉及一种基于胎儿超声图像的胎儿严重畸形检测方法及装置。
背景技术
由于各种各样因素(例如:遗传因素、环境因素等)的影响,胎儿在发育的过程中可能会发生畸形,尤其在胎儿的早孕周,如无脑儿、严重脑膜膨出、无叶全前脑、囊状脊髓脊膜膨出、一侧或双侧肢体完全缺失、单心室、腹裂畸形并内脏外翻、致死性骨发育不良以及连体胎儿畸形。目前胎儿畸形情况的产前检测方法为:通过获取胎儿的超声图像,并由具有相关经验的工作人员结合自身经验分析胎儿的超声图像,从而确定胎儿的畸形情况,进而确定胎儿的生长发育情况。
然而,实践发现,由于胎儿的发育过程相当复杂以及工作人员的经验有限且长时间工作容易疲劳,这很容易导致无法检测到准确的胎儿畸形情况,从而无法准确确定胎儿的生长发育情况。因此,如何准确检测胎儿畸形情况,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,提供一种基于胎儿超声图像的胎儿严重畸形检测方法及装置,能够准确检测胎儿的畸形情况,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基于胎儿超声图像的胎儿严重畸形检测方法,所述方法包括:
在获取到胎儿超声图像的切面之后,确定所述胎儿超声图像的切面对应的目标信息,所述胎儿超声图像的切面对应的目标信息用于确定所述胎儿超声图像对应的胎儿的发育情况;
根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形,当判断结果为是时,根据所述胎儿超声图像的切面对应的目标信息确定所述胎儿超声图像对应的胎儿的畸形情况,所述畸形情况包括所述胎儿超声图像对应的胎儿的畸形类型。
作为一种可选的实施方式,在本发明第一方面中,所述胎儿超声图像的切面包括颅脑切面、肢体切面、腹部切面、脊髓切面、心脏切面、肱骨长径切面以及股骨长径切面中的其中一种,所述胎儿超声图像的颅脑切面包括颅脑水平切面和/或颅脑矢状切面,所述胎儿超声图像的肢体切面包括双上肢切面或双下肢切面,所述胎儿超声图像的腹部切面包括腹部水平切面和/或腹部矢状切面,所述胎儿超声图像的脊髓切面包括脊髓水平切面和/或脊髓矢状切面。
作为一种可选的实施方式,在本发明第一方面中,所述根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形,包括:
当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑矢状切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的颅脑结构特征轮廓以及臀部结构特征轮廓,根据所述胎儿超声图像的切面的颅脑结构特征轮廓以及臀部结构特征轮廓,测量所述胎儿超声图像的头臀长度,并判断所述头臀长度是否在预设头臀长度范围内,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形;
当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑矢状切面时,所述胎儿超声图像的切面对应的目标信息包括所述胎儿超声图像的切面的颅脑结构特征的几何参数,判断所述切面的颅脑结构特征的几何参 数与预设几何参数是否相匹配,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形,所述切面的颅脑结构特征包括该切面内的颅盖骨结构特征、大脑半球结构特征以及中脑结构特征中的至少一种,所述颅脑结构特征的几何参数包括该颅脑结构特征的形状、尺寸、位置以及面积中的至少一种,所述颅脑结构特征的位置为该颅脑结构特征在所述胎儿超声图像的切面中的位置;
当所述胎儿超声图像的切面包括所述颅脑切面或所述脊髓切面或所述腹部切面时,所述胎儿超声图像的切面对应的目标信息包括所述胎儿超声图像的切面的形状,判断所述切面的形状是否与预设切面形状相匹配,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形;
当所述胎儿超声图像的切面为所述颅脑切面或所述腹部切面或所述脊髓切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的目标结构特征对应的特征参数,根据所述切面的目标结构特征对应的特征参数判断所述目标结构特征是否与该切面相匹配,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形;
当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑水平切面时,所述胎儿超声图像的切面对应的目标信息包括所述胎儿超声图像的切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,根据所述切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,确定所述切面的颅脑结构特征的目标几何参数,并判断所述颅脑结构特征的目标几何参数是否在预设几何参数范围内,当判断出不在所述预设几何参数范围内时,确定所述胎儿超声图像对应的胎儿存在畸形,所述切面的颅脑结构特征的目标几何参数包括该颅脑结构特征的头围参数和/或双顶径参数;
当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑水平切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的左丘脑轮廓以及右丘脑轮廓,获取所述左丘脑轮廓与所述右丘脑轮廓的第一拟合度,并判断所述第一拟合度是否大于等于第一预设拟合度阈值,当判断结果为是时,确定所述胎儿超声图像对应的胎儿存在畸形;
当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑水平切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的大脑镰的位置,根据所述大脑镰的位置判断所述大脑镰在所述胎儿超声图像的脑中线位置的出现情况是否满足预设出现情况,当判断出不满足所述预设出现情况时,确定所述胎儿超声图像对应的胎儿存在畸形;
当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑水平切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的脉络膜丛轮廓以及丘脑轮廓,获取所述脉络膜丛轮廓与所述丘脑轮廓之间的第二拟合度,并判断所述第二拟合度是否大于等于第二预设拟合度阈值,当判断结果为是时,确定所述胎儿超声图像对应的胎儿存在畸形。
作为一种可选的实施方式,在本发明第一方面中,所述根据所述切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,确定所述切面的颅脑结构特征的目标几何参数,包括:
获取所述切面的颅脑结构特征的内轮廓的第一周长与所述颅脑结构特征的外轮廓的第二周长,并基于所述第一周长和所述第二周长,确定所述颅脑结构特征对应的头围参数;
确定所述颅脑结构特征的脑中线对应的中垂线与所述颅脑结构特征的外轮廓的第一交点以及所述中垂线与所述颅脑结构特征的内轮廓的第二交点,并基于所述第一交点与所述第二交点,确定所述颅脑结构特征对应的双顶径参数。
作为一种可选的实施方式,在本发明第一方面中,在获取到胎儿超声图像的切面之后,以及所述确定所述胎儿超声图像的切面对应的目标信息之前,所述方法还包括:
当所述胎儿超声图像的切面包括所述颅脑切面时,判断所述胎儿超声图像的切面与颅脑标准切面是否相匹配,当判断结果为是时,触发执行所述的确定所述胎儿超声图像的切面对应的目标信息的操作;
当判断出不匹配时,基于获取到的结构特征校正所述胎儿超声图像的切面以使该切面与所述颅脑标准切面相匹配,并触发执行所述的确定所述胎儿超声图像的切面对应的目标信息的操作;
其中,当所述颅脑切面为所述颅脑水平切面时,所述颅脑标准切面包括侧脑室切面,当所述颅脑切面为所述颅脑矢状切面时,所述颅脑标准切面包括颅脑正中矢状切面。
作为一种可选的实施方式,在本发明第一方面中,所述胎儿超声图像的切面对应的目标信息包括该切面的肢体结构特征的特征参数,其中,当所述胎儿超声图像的切面包括所述肢体切面,且所述肢体切面为所述双上肢切面时,所述切面的肢体结构特征包括该切面的手、前臂以及上臂中的至少一种;当所述胎儿超声图像的切面包括所述肢体切面,且所述肢体切面为所述双下肢切面时,所述切面的肢体结构特征包括该切面的足、大腿以及小腿中的至少一种;
以及,所述根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形,包括:
根据所述胎儿超声图像的切面的肢体结构特征的特征参数,判断所述肢体结构特征是否与所述胎儿超声图像的切面相匹配,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形。
作为一种可选的实施方式,在本发明第一方面中,所述根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形,包括:
当所述胎儿超声图像的切面为所述颅脑切面或所述肱骨长径切面或所述股骨长径切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的骨结构特征的特征参数,根据所述切面的骨结构特征的特征参数判断所述骨结构特征是否与该切面相匹配,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形,所述骨结构特征的特征参数包括所述骨结构特征对应的轮廓、长度、面积、形状以及位置中的至少一种。
作为一种可选的实施方式,在本发明第一方面中,所述根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形,包括:
当所述胎儿超声图像的切面为所述心脏切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的特征参数,判断所述切面特征参数与预设切面参数是否相匹配,当判断出不匹配时,确定所述胎儿超声图像的对应的胎儿存在畸形,所述切面的特征参数包括所述切面的多普勒血流参数和/或轮廓参数;
当所述胎儿超声图像的切面为所述心脏切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的心脏结构特征的特征参数,根据所述切面的心脏结构特征的特征参数判断所述切面的心脏结构特征是否与该切面的标准心脏结构特征相匹配,当判断出不匹配时,确定所述胎儿超声图像的对应的胎儿存在畸形,所述切面的心脏结构特征的特征参数包括所述心脏结构特征的数量和/或所述心脏结构特征对应的面积;
其中,所述根据所述切面的心脏结构特征的特征参数判断所述切面的心脏结构特征是否与该切面的标准心脏结构特征相匹配,包括:
当所述切面的心脏结构特征的特征参数为所述心脏结构特征的数量时,判断所述切面的心脏结构特征的数量是否小于等于预设数量,当判断结果为是时,确定所述切面的心脏结构特征与该切面的标准心脏结构特征不相匹配;
当所述切面的心脏结构特征的特征参数为所述心脏结构特征对应的面积时,判断所述切面的心脏结构特征对应的面积是否大于等于预设面积阈值,当判断结果为是时,确定所述切面的心脏结构特征与该切面的标准心脏结构特征不相匹配。
作为一种可选的实施方式,在本发明第一方面中,所述方法还包括:
将获取到的连续多帧胎儿超声图像输入确定出的结构特征检测模型中进行分析;
获取所述结构特征检测模型输出的分析结果,作为每帧所述胎儿超声图像 的结构特征信息,每帧所述胎儿超声图像的结构特征信息包括该胎儿超声图像的部位结构特征信息以及该胎儿超声图像的结构特征信息,每帧所述胎儿超声图像的部位结构特征信息至少包括该胎儿超声图像的部位结构特征的类别,每帧所述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;
根据每帧所述胎儿超声图像的部位结构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的切面。
作为一种可选的实施方式,在本发明第一方面中,所述方法还包括:
当判断出所述胎儿超声图像对应的胎儿存在畸形时,为所述胎儿超声图像的每个异常结构特征设置对应的结构特征标签,每个所述异常结构特征对应的结构特征标签用于表示该异常结构特征的畸形类型。
本发明第二方面公开了一种基于胎儿超声图像的胎儿严重畸形检测装置,所述装置包括:
确定模块,用于在获取到胎儿超声图像的切面之后,确定所述胎儿超声图像的切面对应的目标信息,所述胎儿超声图像的切面对应的目标信息用于确定所述胎儿超声图像对应的胎儿的发育情况;
第一判断模块,用于根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形;
所述确定模块,还用于当所述第一判断模块判断结果为是时,根据所述胎儿超声图像的切面对应的目标信息确定所述胎儿超声图像对应的胎儿的畸形情况,所述畸形情况包括所述胎儿超声图像对应的胎儿的畸形类型。
作为一种可选的实施方式,在本发明第二方面中,所述胎儿超声图像的切面包括颅脑切面、肢体切面、腹部切面、脊髓切面、心脏切面、肱骨长径切面以及股骨长径切面中的其中一种,所述胎儿超声图像的颅脑切面包括颅脑水平切面和/或颅脑矢状切面,所述胎儿超声图像的肢体切面包括双上肢切面或双下肢切面,所述胎儿超声图像的腹部切面包括腹部水平切面和/或腹部矢状切面,所述胎儿超声图像的脊髓切面包括脊髓水平切面和/或脊髓矢状切面。
作为一种可选的实施方式,在本发明第二方面中,所述第一判断模块根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形的方式具体为:
当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑矢状切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的颅脑结构特征轮廓以及臀部结构特征轮廓,根据所述胎儿超声图像的切面的颅脑结构特征轮廓以及臀部结构特征轮廓,测量所述胎儿超声图像的头臀长度,并判断所述头臀长度是否在预设头臀长度范围内,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形;
当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑矢状切面时,所述胎儿超声图像的切面对应的目标信息包括所述胎儿超声图像的切面的颅脑结构特征的几何参数,判断所述切面的颅脑结构特征的几何参数与预设几何参数是否相匹配,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形,所述切面的颅脑结构特征包括该切面内的颅盖骨结构特征、大脑半球结构特征以及中脑结构特征中的至少一种,所述颅脑结构特征的几何参数包括该颅脑结构特征的形状、尺寸、位置以及面积中的至少一种,所述颅脑结构特征的位置为该颅脑结构特征在所述胎儿超声图像的切面中的位置;
当所述胎儿超声图像的切面包括所述颅脑切面或所述脊髓切面或所述腹部切面时,所述胎儿超声图像的切面对应的目标信息包括所述胎儿超声图像的切面的形状,判断所述切面的形状是否与预设切面形状相匹配,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形;
当所述胎儿超声图像的切面为所述颅脑切面或所述腹部切面或所述脊髓切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的目标结构特征对应的特征参数,根据所述切面的目标结构特征对应的特征参数判断所述目标结构特征是否与该切面相匹配,当判断出不相匹配时,确定所述胎儿超声图像对 应的胎儿存在畸形;
当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑水平切面时,所述胎儿超声图像的切面对应的目标信息包括所述胎儿超声图像的切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,根据所述切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,确定所述切面的颅脑结构特征的目标几何参数,并判断所述颅脑结构特征的目标几何参数是否在预设几何参数范围内,当判断出不在所述预设几何参数范围内时,确定所述胎儿超声图像对应的胎儿存在畸形,所述切面的颅脑结构特征的目标几何参数包括该颅脑结构特征的头围参数和/或双顶径参数;
当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑水平切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的左丘脑轮廓以及右丘脑轮廓,获取所述左丘脑轮廓与所述右丘脑轮廓的第一拟合度,并判断所述第一拟合度是否大于等于第一预设拟合度阈值,当判断结果为是时,确定所述胎儿超声图像对应的胎儿存在畸形;
当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑水平切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的大脑镰的位置,根据所述大脑镰的位置判断所述大脑镰在所述胎儿超声图像的脑中线位置的出现情况是否满足预设出现情况,当判断出不满足所述预设出现情况时,确定所述胎儿超声图像对应的胎儿存在畸形;
当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑水平切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的脉络膜丛轮廓以及丘脑轮廓,获取所述脉络膜丛轮廓与所述丘脑轮廓之间的第二拟合度,并判断所述第二拟合度是否大于等于第二预设拟合度阈值,当判断结果为是时,确定所述胎儿超声图像对应的胎儿存在畸形。
作为一种可选的实施方式,在本发明第二方面中,所述第一判断模块根据所述切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,确定所述切面的颅脑结构特征的目标几何参数的方式具体为:
获取所述切面的颅脑结构特征的内轮廓的第一周长与所述颅脑结构特征的外轮廓的第二周长,并基于所述第一周长和所述第二周长,确定所述颅脑结构特征对应的头围参数;
确定所述颅脑结构特征的脑中线对应的中垂线与所述颅脑结构特征的外轮廓的第一交点以及所述中垂线与所述颅脑结构特征的内轮廓的第二交点,并基于所述第一交点与所述第二交点,确定所述颅脑结构特征对应的双顶径参数。
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:
第二判断模块,用于当所述胎儿超声图像的切面包括所述颅脑切面时,在获取到胎儿超声图像的切面之后,以及在所述确定模块确定所述胎儿超声图像的切面对应的目标信息之前,判断所述胎儿超声图像的切面与颅脑标准切面是否相匹配,当判断结果为是时,触发所述确定模块执行所述的确定所述胎儿超声图像的切面对应的目标信息的操作;
校正模块,用于当所述第二判断模块判断出不匹配时,基于获取到的结构特征校正所述胎儿超声图像的切面以使该切面与所述颅脑标准切面相匹配,并触发所述确定模块执行所述的确定所述胎儿超声图像的切面对应的目标信息的操作;
其中,当所述颅脑切面为所述颅脑水平切面时,所述颅脑标准切面包括侧脑室切面,当所述颅脑切面为所述颅脑矢状切面时,所述颅脑标准切面包括颅脑正中矢状切面。
作为一种可选的实施方式,在本发明第二方面中,所述胎儿超声图像的切面对应的目标信息包括该切面的肢体结构特征的特征参数,其中,当所述胎儿超声图像的切面包括所述肢体切面,且所述肢体切面为所述双上肢切面时,所述切面的肢体结构特征包括该切面的手、前臂以及上臂中的至少一种;当所述胎儿超声图像的切面包括所述肢体切面,且所述肢体切面为所述双下肢切面时,所述切面的肢体结构特征包括该切面的足、大腿以及小腿中的至少一种;
以及,所述第一判断模块根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形的方式具体为:
根据所述胎儿超声图像的切面的肢体结构特征的特征参数,判断所述肢体结构特征是否与所述胎儿超声图像的切面相匹配,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形。
作为一种可选的实施方式,在本发明第二方面中,所述第一判断模块根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形的方式具体为:
当所述胎儿超声图像的切面为所述颅脑切面或所述肱骨长径切面或所述股骨长径切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的骨结构特征的特征参数,根据所述切面的骨结构特征的特征参数判断所述骨结构特征是否与该切面相匹配,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形,所述骨结构特征的特征参数包括所述骨结构特征对应的轮廓、长度、面积、形状以及位置中的至少一种。
作为一种可选的实施方式,在本发明第二方面中,所述第一判断模块根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形的方式具体为:
当所述胎儿超声图像的切面为所述心脏切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的特征参数,判断所述切面特征参数与预设切面参数是否相匹配,当判断出不匹配时,确定所述胎儿超声图像的对应的胎儿存在畸形,所述切面的特征参数包括所述切面的多普勒血流参数和/或轮廓参数;
当所述胎儿超声图像的切面为所述心脏切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的心脏结构特征的特征参数,根据所述切面的心脏结构特征的特征参数判断所述切面的心脏结构特征是否与该切面的标准心脏结构特征相匹配,当判断出不匹配时,确定所述胎儿超声图像的对应的胎儿存在畸形,所述切面的心脏结构特征的特征参数包括所述心脏结构特征的数量和/或所述心脏结构特征对应的面积;
其中,所述第一判断模块根据所述切面的心脏结构特征的特征参数判断所述切面的心脏结构特征是否与该切面的标准心脏结构特征相匹配的方式具体为:
当所述切面的心脏结构特征的特征参数为所述心脏结构特征的数量时,判断所述切面的心脏结构特征的数量是否小于等于预设数量,当判断结果为是时,确定所述切面的心脏结构特征与该切面的标准心脏结构特征不相匹配;
当所述切面的心脏结构特征的特征参数为所述心脏结构特征对应的面积时,判断所述切面的心脏结构特征对应的面积是否大于等于预设面积阈值,当判断结果为是时,确定所述切面的心脏结构特征与该切面的标准心脏结构特征不相匹配。
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:
分析模块,用于将获取到的连续多帧胎儿超声图像输入确定出的结构特征检测模型中进行分析;
获取模块,用于获取所述结构特征检测模型输出的分析结果,作为每帧所述胎儿超声图像的结构特征信息,每帧所述胎儿超声图像的结构特征信息包括该胎儿超声图像的部位结构特征信息以及该胎儿超声图像的结构特征信息,每帧所述胎儿超声图像的部位结构特征信息至少包括该胎儿超声图像的部位结构特征的类别,每帧所述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;
所述确定模块,还用于根据每帧所述胎儿超声图像的部位结构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的切面。
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:
设置模块,用于当所述第一判断模块判断出所述胎儿超声图像对应的胎儿存在畸形时,为所述胎儿超声图像的每个异常结构特征设置对应的结构特征标签,每个所述异常结构特征对应的结构特征标签用于表示该异常结构特征的畸形类型。
本发明第三方面公开了另一种基于胎儿超声图像的胎儿严重畸形检测装置,所述确定装置包括:
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,执行本发明第一方面公开的基于胎儿超声图像的胎儿严重畸形检测方法。
本发明第四方面公开了一种计算机存储介质,所述计算机存储介质存储有计算机指令,所述计算机指令被调用时,用于执行本发明第一方面公开的基于胎儿超声图像的胎儿严重畸形检测方法。
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,提供了一种基于胎儿超声图像的胎儿严重畸形检测方法及装置,该方法包括在获取到胎儿超声图像的切面之后,确定该胎儿超声图像的切面对应的目标信息,该胎儿超声图像的切面对应的目标信息用于确定胎儿超声图像对应的胎儿的发育情况;根据胎儿超声图像的切面对应的目标信息,判断胎儿超声图像对应的胎儿是否存在畸形,当判断结果为是时,根据胎儿超声图像的切面对应的目标信息确定该胎儿超声图像对应的胎儿的畸形情况,该畸形情况包括胎儿超声图像对应的胎儿的畸形类型。可见,实施本发明通过在获取到胎儿超声图像的切面之后,能够自动根据确定出的胎儿超声图像的切面的信息判断胎儿是否存在异常,当存在异常时,自动根据胎儿超声图像的切面的信息确定胎儿的畸形情况,例如:是否为无脑儿、腹裂畸形等,有利于准确检测胎儿的畸形情况,从而实现胎儿的生长发育情况的准确确定;以及通过多方面综合检测及分析胎儿的生长发育情况,有利于提高胎儿的畸形情况的检测准确性。
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基于胎儿超声图像的胎儿严重畸形检测方法的流程示意图;
图2是本发明实施例公开的另一种基于胎儿超声图像的胎儿严重畸形检测方法的流程示意图;
图3是本发明实施例公开的一种基于胎儿超声图像的胎儿严重畸形检测装置的结构示意图;
图4是本发明实施例公开的另一种基于胎儿超声图像的胎儿严重畸形检测装置的结构示意图;
图5是本发明实施例公开的又一种基于胎儿超声图像的胎儿严重畸形检测装置的结构示意图。
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结构特征、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。
本发明公开了一种基于胎儿超声图像的胎儿严重畸形检测方法及装置,通过在获取到胎儿超声图像的切面之后,能够自动根据确定出的胎儿超声图像的切面的信息判断胎儿是否存在异常,当存在异常时,自动根据胎儿超声图像的切面的信息确定胎儿的畸形情况,例如:是否为无脑儿、腹裂畸形等,有利于准确检测胎儿的畸形情况,从而实现胎儿的生长发育情况的准确确定;以及通过多方面综合检测及分析胎儿的生长发育情况,有利于提高胎儿的畸形情况的检测准确性。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基于胎儿超声图像的胎儿严重畸形检测方法的流程示意图。其中,图1所描述的基于胎儿超声图像的胎儿严重畸形检测方法可以应用于检测服务器(服务设备/服务***)中,其中,该检测服务器可以包括本地检测服务器或云检测服务器,本发明实施例不做限定。如图1所示,该基于胎儿超声图像的胎儿严重畸形检测方法可以包括以下操作:
101、在获取到胎儿超声图像的切面之后,确定胎儿超声图像的切面对应的目标信息。
本发明实施例中,胎儿超声图像的切面对应的目标信息用于确定胎儿超声图像对应的胎儿的发育情况,且不同的切面对应不同的胎儿超声图像。进一步的,胎儿超声图像可以是连续获取到的图像,即步骤101中的胎儿超声图像代表单帧图像,胎儿超声图像也可以是多帧图像,此时,步骤101中的胎儿超声图像的切面为单帧胎儿超声图像的切面。
本发明实施例中,该胎儿超声图像的切面包括颅脑切面、肢体切面、腹部切面、脊髓切面、心脏切面、肱骨长径切面以及股骨长径切面中的其中一种,该胎儿超声图像的颅脑切面包括颅脑水平切面和/或颅脑矢状切面,该胎儿超声图像的肢体切面包括双上肢切面或双下肢切面,该胎儿超声图像的腹部切面包括腹部水平切面和/或腹部矢状切面,该胎儿超声图像的脊髓切面包括脊髓水平切面和/或脊髓矢状切面。
作为一种可选的实施方式,根据所述胎儿超声图像的切面对应的目标信息,判断胎儿超声图像对应的胎儿是否存在畸形,包括:
当胎儿超声图像的切面包括颅脑切面,且该颅脑切面为颅脑矢状切面时,胎儿超声图像的切面对应的目标信息包括该切面的颅脑结构特征轮廓以及臀部结构特征轮廓,根据胎儿超声图像的切面的颅脑结构特征轮廓以及臀部结构特征轮廓,测量胎儿超声图像的头臀长度,并判断该头臀长度是否在预设头臀长度范围内,当判断出不相匹配时,确定胎儿超声图像对应的胎儿存在畸形;
当胎儿超声图像的切面包括颅脑切面,且该颅脑切面为颅脑矢状切面时,胎儿超声图像的切面对应的目标信息包括胎儿超声图像的切面的颅脑结构特征的几何参数,判断该切面的颅脑结构特征的几何参数与预设几何参数是否相匹配,当判断出不相匹配时,确定胎儿超声图像对应的胎儿存在畸形。
该可选的实施方式中,胎儿超声图像的切面的颅脑结构特征包括该切面内的颅盖骨结构特征、大脑半球结构特征以及中脑结构特征中的至少一种,该颅脑结构特征的几何参数包括该颅脑结构特征的形状、尺寸(例如:长度)、位置以及面积中的至少一种,该颅脑结构特征的位置为该颅脑结构特征在胎儿超声图像的切面中的位置。
该可选的实施方式中,不同的孕期(孕周)对应不同的预设头臀长度范围、预设几何参数,例如:第3个孕周,预设头臀长度范围为2mm-5mm,第13个孕周,预设头臀长度范围为5cm-10cm。预设几何参数包括预设形状、预设尺寸、预设位置以及预设面积。
该可选的实施方式中,进一步的,根据胎儿超声图像的切面对应的目标信 息确定胎儿超声图像对应的胎儿的畸形情况,包括:
根据胎儿超声图像的切面的颅脑结构特征轮廓以及臀部结构特征轮廓,或者,颅脑结构特征的几何参数,确定胎儿超声图像对应的胎儿的畸形情况,且该畸形情况包括胎儿超声图像对应的胎儿的畸形类型包括无脑儿畸形类型。
该可选的实施方式中,需要说明的是,在判断出胎儿超声图像的头臀长度不在预设头臀长度范围内且判断出切面的颅脑结构特征的几何参数与预设几何参数不匹配时,方确定胎儿超声图像对应的胎儿存在畸形。这样能够提高胎儿超声图像对应的胎儿为无脑儿的检测准确性以及可靠性。
该可选的实施方式中,进一步的,当判断相匹配时,确定胎儿超声图像对应的胎儿的头臀长度正常;当判断出切面的颅脑结构特征的几何参数与预设几何参数相匹配时,确定胎儿超声图像对应的胎儿不为无脑儿胎儿。
作为另一种可选的实施方式,根据所述胎儿超声图像的切面对应的目标信息,判断胎儿超声图像对应的胎儿是否存在畸形,包括:
当胎儿超声图像的切面包括颅脑切面或脊髓切面或腹部切面时,胎儿超声图像的切面对应的目标信息包括胎儿超声图像的切面的形状(轮廓),判断切面的形状是否与预设切面形状相匹配,当判断出不相匹配时,确定胎儿超声图像对应的胎儿存在畸形。
该可选的实施方式中,不同的孕期(孕周)对应不同的预设切面形状。
该可选的实施方式中,胎儿超声图像的切面的形状可以包括水平切面的形状和/或矢状切面的形状,这样能够丰富胎儿是否出现异常的判断方式,以及当包括水平切面的形状和矢状切面的形状时,通过分别将水平切面的形状和矢状切面的形状与其对应的标准切面的形状进行比较,有利于提高胎儿异常的确定准确性以及可靠性。
具体的,当胎儿超声图像的切面为颅脑水平切面时,判断颅脑水平切面的形状是否呈现标准的椭圆形,当呈椭圆形时,确定颅脑结构特征正常,也即胎儿超声图像的胎儿的颅脑正常;当不呈标准的椭圆形,即呈现不同程度凸起的椭圆形时,确定颅脑结构特征异常,也即胎儿超声图像的胎儿的颅脑存在异常,则确定胎儿存在脑膜膨出的异常。当胎儿超声图像的切面为颅脑矢状切面时,且在获取到颅脑矢状切面的形状之后,将颅脑矢状切面的形状(颅脑实质图像或边缘轮廓)输入确定出(例如:预先训练好)的图像分类模型中进行分析,并获取该图像分类模型输出的分析结果,作为颅脑矢状切面的形状的判断结果,当判断结果用于表示颅脑脑膜处于膨出形态时,表示胎儿存在脑膜膨出异常,当判断结果用于表示颅脑脑膜处于确定出的正常形态时,表示胎儿的颅脑正常。进一步的,当胎儿存在脑膜膨出时,根据脑膜膨出在椭圆形上的凹凸形态确定脑膜膨出的凹凸程度,并根据该凹凸程度确定脑膜膨出所对应的严重级别(例如:3级,其中,严重级别越高,代表脑膜膨出越严重),这样在确定出胎儿存在脑膜膨出之后,进一步确定脑膜膨出所处的严重级别,有利于清楚且准确知晓胎儿的脑膜膨出的畸形情况。进一步的,在得到胎儿超声图像的胎儿的切面的凹凸形态之后,进一步获取凹凸面积,并根据凹凸面积确定膨出的严重级别,这样有利于提高脑膜膨出的级别确定准确性以及可靠性,从而有利于进一步提高胎儿(脑膜)膨出的确定准确性以及可靠性。
作为又一种可选的实施方式,根据所述胎儿超声图像的切面对应的目标信息,判断胎儿超声图像对应的胎儿是否存在畸形,包括:
当胎儿超声图像的切面为颅脑切面或腹部切面或脊髓切面时,胎儿超声图像的切面对应的目标信息包括该切面的目标结构特征对应的特征参数,根据切面的目标结构特征对应的特征参数判断目标结构特征是否与该切面相匹配,当判断出不相匹配时,确定胎儿超声图像对应的胎儿存在畸形。
该可选的实施方式中,不同胎儿超声图像的切面,对应不同的切面的目标结构特征,具体的:颅脑切面对应的目标结构特征包括透明隔腔、侧脑室、脉络膜丛、丘脑或小脑等;腹部切面对应的目标结构特征包括胃泡、脐静脉以及胆囊等;脊髓切面对应的目标结构特征包括脊柱、脊膜等。其中,目标结构特征对应的特征参数包括目标结构特征的轮廓、尺寸、位置、面积、类型等中的 至少一种。
该可选的实施方式中,具体的,当根据目标结构特征的类型判断出目标结构特征不是胎儿超声图像的切面中的结构特征时,确定目标结构特征与该切面不相匹配,例如:当颅脑切面中出现脑积水结构特征时,确定颅脑为异常颅脑,也即胎儿超声图像的胎儿存在畸形,且该异常为脑膜膨出。其中,腹部切面或脊髓切面中的目标结构特征对应的特征参数判断目标结构特征是否与该切面相匹配的具体判断方式和颅脑切面的判断方式一样,在此不再赘述。
该可选的实施方式通过判断胎儿超声图像的切面中的结构特征是否与该切面相匹配,能够实现颅脑脑膜膨出、囊状脊髓脊膜膨出以及腹裂畸形并内脏外翻的确定,有利于提高胎儿的畸形情况的判断准确性。
作为又一种可选的实施方式,根据所述胎儿超声图像的切面对应的目标信息,判断胎儿超声图像对应的胎儿是否存在畸形,包括:
当胎儿超声图像的切面包括颅脑切面,且颅脑切面为颅脑水平切面时,胎儿超声图像的切面对应的目标信息包括胎儿超声图像的切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,根据切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,确定切面的颅脑结构特征的目标几何参数,并判断颅脑结构特征的目标几何参数是否在预设几何参数范围内,当判断出不在预设几何参数范围内时,确定胎儿超声图像对应的胎儿存在畸形,切面的颅脑结构特征的目标几何参数包括该颅脑结构特征的头围参数和/或双顶径参数。
该可选的实施方式中,进一步的,在获取到胎儿超声图像的切面之后,将胎儿超声图像的切面输入确定出切面形状确定模型中进行分析,并获取该切面形状确定模型输出的分析结果,作为该胎儿超声图像的切面的形状,这样能够提高胎儿超声图像的切面的形状的获取准确性以及效率,从而有利于提高胎儿是否存在异常的判断准确性以及效率。其中,切面形状确定模型可以包括基于实例分割模型、语义分割模型等任意一种或者多种组合模型,该可选的实施方式不做限定。
该可选的实施方式中,进一步可选的,根据切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,确定切面的颅脑结构特征的目标几何参数,包括:
获取切面的颅脑结构特征的内轮廓的第一周长与颅脑结构特征的外轮廓的第二周长,并基于第一周长和第二周长,确定颅脑结构特征的头围参数;
确定颅脑结构特征的脑中线对应的中垂线与颅脑结构特征的外轮廓的第一交点以及中垂线与颅脑结构特征的内轮廓的第二交点,并基于第一交点与第二交点,确定颅脑结构特征的双顶径参数。
该可选的实施例中,颅脑结构特征的头围参数的计算方式如下:
C=(C 1+C 2)/2;
式中,C为颅脑结构特征的头围参数,即头围周长;C 1为颅脑结构特征的外轮廓的第二周长;C 2颅脑结构特征的内轮廓的第一周长。
该可选的实施方式中,第一交点包括第一子交点和第二子交点,第二交点包括第三子交点和第四子交点。其中,第一子交点与第三子交点的距离小于与第四子交点的距离。基于该第一交点与该第二交点,确定颅脑结构特征的双顶径参数,具体的:连接第一子交点与第四子交点所形成的第一线段,作为颅脑结构特征的双顶径几何参数,即双顶径长度;或者,连接第二子交点与第三子交点所形成的第二线段,作为颅脑结构特征的双顶径参数;或者,获取第一线段与第二线段的均值,作为颅脑结构特征对应的双顶径参数。这样通过提供多种颅脑结构特征的双顶径长度的获取方式,能够提高双顶径长度的获取可能性以及准确性。
可见,该可选的实施方式通过获取颅脑结构特征的内、外轮廓的周长,以及颅脑结构特征的脑中线对应的中垂线与内、外轮廓的交点,能够实现颅脑结构特征的头围周长以及双顶径长度的获取,从而实现胎儿颅脑脑膜膨出的确定。
该可选的实施例中,进一步的,在获取到颅脑结构特征的双顶径参数、头 围参数之后,进一步取双顶径参数、头围参数的比值,并判断该比值是否大于等于预设值(例如:头围周长与双顶径长度的比值大于等于6),当判断结果为是时,方可确定胎儿颅脑脑膜膨出,有利于提高胎儿颅脑脑膜膨出的确定准确性以及可靠性。
可见,该可选的实施方式通过胎儿超声图像的切面的颅脑结构特征轮廓与臀部结构特征轮廓和/或颅脑结构特征的几何参数,能够实现胎儿超声图像对应的胎儿为无脑儿的异常确定,以及通过在判断出胎儿超声图像的头臀长度不在预设头臀长度范围内且判断出切面的颅脑结构特征的几何参数与预设几何参数不匹配时,方确定胎儿超声图像对应的胎儿存在畸形,能够提高胎儿超声图像对应的胎儿为无脑儿的检测准确性以及可靠性。
作为又一种可选的实施方式,根据所述胎儿超声图像的切面对应的目标信息,判断胎儿超声图像对应的胎儿是否存在畸形,包括:
当胎儿超声图像的切面包括颅脑切面,且颅脑切面为颅脑水平切面时,胎儿超声图像的切面对应的目标信息包括该切面的左丘脑轮廓以及右丘脑轮廓,获取左丘脑轮廓与右丘脑轮廓的第一拟合度,并判断第一拟合度是否大于等于第一预设拟合度阈值,当判断结果为是时,确定胎儿超声图像对应的胎儿存在畸形。
当胎儿超声图像的切面包括颅脑切面,且颅脑切面为颅脑水平切面时,胎儿超声图像的切面对应的目标信息包括该切面的大脑镰的位置,根据大脑镰的位置判断大脑镰在胎儿超声图像的脑中线位置的出现情况是否满足预设出现情况,当判断出不满足预设出现情况时,确定胎儿超声图像对应的胎儿存在畸形。
该可选的实施方式中,当大脑镰的部分(例如:一半)或完全出现在胎儿超声图像的脑中线位置时,表示大脑镰在胎儿超声图像的脑中线位置的出现情况满足预设出现情况。
当胎儿超声图像的切面包括颅脑切面,且颅脑切面为颅脑水平切面时,胎儿超声图像的切面对应的目标信息包括该切面的脉络膜丛轮廓以及丘脑轮廓,获取脉络膜丛轮廓与丘脑轮廓之间的第二拟合度,并判断第二拟合度是否大于等于第二预设拟合度阈值,当判断结果为是时,确定胎儿超声图像对应的胎儿存在畸形。
该可选的实施方式中,脉络膜丛轮廓包括左脉络膜丛轮廓以及右脉络膜丛轮廓,丘脑轮廓包括左丘脑轮廓以及右丘脑轮廓。其中,获取脉络膜丛轮廓与丘脑轮廓之间的第二拟合度,具体的,分别获取左脉络膜丛轮廓与左丘脑轮廓的拟合度和/或右脉络膜丛轮廓与右丘脑轮廓的拟合度,作为脉络膜丛轮廓与丘脑轮廓之间的第二拟合度,这样通过分别获取左右脉络膜丛与左右丘脑的拟合度,有利于提高脉络膜丛与丘脑未分离的确定准确性以及可靠性,从而有利于提高胎儿超声图像的胎儿为无叶全前脑胎儿的确定准确性以及可靠性。
可见,该可选的实施方式通过确定颅脑的左右丘脑融合、大脑镰缺如以及脉络膜丛与丘脑未分离中的至少一种情况,不仅能够实现胎儿超声图像的胎儿为无叶全前脑胎儿的确定,还能够丰富无叶全前脑胎儿的确定方式;以及通过提供至少两种无叶全前脑胎儿的确定方式,能够提高无叶全前脑胎儿的确定准确性以及可靠性。
该可选的实施方式中,当判断出第一拟合度小于第一预设拟合度阈值时,确定胎儿超声图像对应的胎儿的丘脑未融合,即胎儿超声图像对应的胎儿的丘脑正常;当判断出大脑镰在胎儿超声图像的脑中线位置的出现情况不满足预设出现情况时,确定胎儿超声图像对应的胎儿的大脑镰正常;当判断出第二拟合度小于第二预设拟合度阈值时,确定胎儿超声图像对应的胎儿的脉络膜丛与丘脑分离。进一步的,当确定胎儿超声图像对应的胎儿的丘脑未融合、大脑镰正常以及脉络膜丛与丘脑分离时,确定胎儿超声图像对应的胎儿不为无叶全前脑胎儿。
本发明实施例中,进一步可选的,上述胎儿超声图像的切面对应的目标信息包括该切面的肢体结构特征的特征参数,其中,当胎儿超声图像的切面包括肢体切面,且肢体切面为双上肢切面时,该切面的肢体结构特征包括该切面的 手、前臂以及上臂中的至少一种;当胎儿超声图像的切面包括肢体切面,且肢体切面为双下肢切面时,切面的肢体结构特征包括该切面的足、大腿以及小腿中的至少一种。以及,作为又一种可选的实施方式,根据胎儿超声图像的切面对应的目标信息,判断胎儿超声图像对应的胎儿是否存在畸形,可以包括:
根据胎儿超声图像的切面的肢体结构特征的特征参数,判断肢体结构特征是否与胎儿超声图像的切面相匹配,当判断出不相匹配时,确定胎儿超声图像对应的胎儿存在畸形。
该可选的实施方式中,肢体结构特征的特征参数包括肢体结构特征的数量、形状、尺寸(例如:长度)以及位置中的至少一种,这样肢体结构特征的特征参数包括的内容越多,越有利于提高肢体结构特征是否与胎儿超声图像的切面相匹配的判断准确性以及可靠性,从而有利于胎儿的畸形情况的确定准确性以及可靠性。
该可选的实施方式中,进一步的,当肢体结构特征包括手时,在判断出对应的切面存在手腕、手掌和至少一根手指时,方可确定肢体结构特征与胎儿超声图像的切面相匹配;当肢体结构特征包括足时,在判断出对应的切面存在脚腕、脚掌和至少一根脚趾时,方可确定肢体结构特征与胎儿超声图像的切面相匹配;进一步的,当手、足的位置处于人体对应的位置时,方可确定肢体结构特征与胎儿超声图像的切面相匹配,能够进一步提高肢体结构特征与胎儿超声图像的切面相匹配的确定准确性以及可靠性。
可见,该可选的实施方式通过胎儿超声图像的切面中的肢体结构特征(例如:手、足)的特征参数,能够实现胎儿超声图像对应的胎儿是否为肢体缺失的胎儿的确定。
作为又一种可选的实施方式,根据胎儿超声图像的切面对应的目标信息,判断胎儿超声图像对应的胎儿是否存在畸形,可以包括:
当胎儿超声图像的切面为颅脑切面或肱骨长径切面或股骨长径切面时,胎儿超声图像的切面对应的目标信息包括该切面的骨结构特征的特征参数,根据切面的骨结构特征的特征参数判断骨结构特征是否与该切面相匹配,当判断出不相匹配时,确定胎儿超声图像对应的胎儿存在畸形。
该可选的实施方式中,骨结构特征的特征参数包括骨结构特征对应的轮廓、长度、面积、形状以及位置中的至少一种。
该可选的实施方式中,当胎儿超声图像的切面为颅脑水平切面时,确定胎儿超声图像的切面对应的目标信息,包括:
确定胎儿超声图像的切面的颅脑结构特征的中心点,并获取颅脑结构特征的中心点与颅脑结构特征的内轮廓的第一垂直距离值以及中心点与颅脑结构特征的外轮廓的第二垂直距离值,并获取第一垂直距离值与第二垂直距离值的距离差值,得到颅脑厚度,其中,胎儿超声图像的切面对应的目标信息包括颅脑厚度,也即骨结构特征的特征参数包括颅脑结构特征的颅脑厚度。此时,可选的,判断颅脑厚度是否在确定出的颅脑厚度范围内,当判断出在颅脑厚度范围时,确定胎儿超声图像的颅脑正常,当判断出不在颅脑厚度范围时,判断颅脑厚度是否大于该颅脑厚度范围的最大厚度值,当判断出大于最大厚度值,确定胎儿超声图像的颅脑异常。进一步的,根据颅脑厚度确定颅脑的厚度等级。需要说明的是,厚度等级越高,即厚度越大,颅脑骨化增强越明显。其中,不同孕周的胎儿超声图像,对应不同的颅脑厚度范围,例如:第10周,颅脑厚度范围为5mm-10mm,第13周,颅脑厚度范围为8mm-12mm。又进一步可选的,骨结构特征的特征参数还包括颅脑结构特征的颅骨形态,判断颅骨形态与确定出的颅骨形态是否相匹配,当判断出不匹配时,确定胎儿超声图像的颅骨形态异常。
该可选的实施方式中,又进一步可选的,当胎儿超声图像的切面为颅脑正中矢状切面时,胎儿超声图像的切面对应的目标信息包括颅骨前额结构特征的参数,其中,颅骨前额结构特征的参数包括颅骨前额结构特征的轮廓所围成的区域面积、形状以及颅骨前额结构特征的轮廓最远端与颅脑外轮廓的目标垂直距离值。此时,判断颅骨前额结构特征的轮廓所围成的区域面积是否在确定出 的面积范围内,当判断出不在面积范围内时,则确定胎儿超声图像的颅骨前额向前突出;或者,判断颅骨前额结构特征的形状是否与确定出的形状相匹配,当判断出不匹配时,则确定胎儿超声图像的颅骨前额向前突出;或者,判断目标垂直距离值是否在确定出的垂直距离值范围内,当判断出不在时,则确定胎儿超声图像的颅骨前额向前突出,进一步的,当该三种判断结果至少两种判断结果为否时,确定胎儿超声图像的颅骨前额向前突出。又进一步的,当胎儿颅骨前额向前突出时,根据颅骨前额向前突出的突出形态确定颅骨前额向前突出的突出程度,并根据该突出程度确定颅骨前额向前突出所对应的严重级别(例如:3级,其中,严重级别越高,代表颅骨前额向前突出越严重),这样在确定出胎儿存在颅骨前额向前突出之后,进一步确定颅骨前额向前突出所处的严重级别,有利于清楚且准确知晓胎儿的颅骨前额向前突出的畸形情况,这样有利于提高颅骨前额向前突出的级别确定准确性以及可靠性,从而有利于进一步提高胎儿超声图像的胎儿为致死性骨发育不良胎儿的确定准确性以及可靠性。
该可选的实施方式中,当骨结构特征的特征参数为骨结构特征的长度时,判断骨结构特征的长度是否在预设长度范围内,当判断出不在预设长度范围内时,确定骨结构特征与该切面不相匹配,其中,不同的孕周对应不同的预设长度范围,例如:第5个孕周,预设长度范围为2mm-5mm,第10个孕周,预设长度范围为4cm-8cm。
可见,该可选的实施方式通过将测量到的胎儿超声图像的切面的骨结构特征的特征参数(例如:长度)与对应切面相匹配,能够实现胎儿超声图像的胎儿为致死性骨发育不良的确定,以及通过颅脑、肱骨以及股骨等维度分析胎儿致死性骨发育不良情况,能够胎儿致死性骨发育不良的确定准确性以及可靠性。
作为又一种可选的实施方式,根据胎儿超声图像的切面对应的目标信息,判断胎儿超声图像对应的胎儿是否存在畸形,可以包括:
当胎儿超声图像的切面为心脏切面时,胎儿超声图像的切面对应的目标信息包括该切面的特征参数,判断切面特征参数与预设切面参数是否相匹配,当判断出不匹配时,确定胎儿超声图像的对应的胎儿存在畸形,切面的特征参数包括切面的多普勒血流参数和/或轮廓参数;
当胎儿超声图像的切面为心脏切面时,胎儿超声图像的切面对应的目标信息包括该切面的心脏结构特征的特征参数,根据切面的心脏结构特征的特征参数判断切面的心脏结构特征是否与该切面的标准心脏结构特征相匹配,当判断出不匹配时,确定胎儿超声图像的对应的胎儿存在畸形;
该可选的实施方式中,进一步可选的,根据切面的心脏结构特征的特征参数判断切面的心脏结构特征是否与该切面的标准心脏结构特征相匹配,包括:
当切面的心脏结构特征的特征参数为心脏结构特征的数量时,判断切面的心脏结构特征的数量是否小于等于预设数量,当判断结果为是时,确定切面的心脏结构特征与该切面的标准心脏结构特征不相匹配;
当切面的心脏结构特征的特征参数为心脏结构特征对应的面积时,判断切面的心脏结构特征对应的面积是否大于等于预设面积阈值,当判断结果为是时,确定切面的心脏结构特征与该切面的标准心脏结构特征不相匹配。
该可选的实施方式中,可选的,切面的心脏结构特征的特征参数包括心脏结构特征的数量和/或心脏结构特征对应的面积。进一步可选的,不同孕周对应的胎儿超声图像的切面的标准心脏结构特征也不相同,例如:第2个孕周,标准心脏结构特征包括心内膜垫以及心内膜垫的面积,第14个孕周,标准心脏结构特征包括左心房、左心室、右心房、右心室、心内膜垫以及每个标准心脏结构特征的面积。
该可选的实施方式中,当判断出切面的心脏结构特征包括的数量及面积均与标准心脏结构特征相匹配时,则该心脏切面为正常心脏切面,即胎儿的心脏为正常心脏;当判断出切面的心脏结构特征包括的数量小于等于预设数量(例如:2个),则确定该心脏切面为异常心脏切面。具体的,当判断出切面的心脏结构特征包括的数量为1个(例如:仅存在左心室),和/或,该心脏结构特征的面积与心脏切面的面积的比例值大于等于预设比例值,则确定该心脏切面为 单心室心脏切面,例如:当判断出心脏切面仅存在左心室,和/或,左心室的面积与心脏切面的面积的比例值等于二分之一,则表示心脏切面为单心室心脏切面,也即胎儿的心脏为单心室状态。
可见,该可选的实施方式通过心脏切面的多普勒血流参数、轮廓参数、心脏结构特征数量以及每个心脏结构特征的面积,判断胎儿超声图像的胎儿的心脏是否存在异常,能够提高胎儿的心脏存在异常的确定准确性以及可靠性,有利于提高胎儿的心脏畸形情况的确定准确性以及可靠性。
在一个可选的实施例中,在获取到胎儿超声图像的切面之后,以及确定胎儿超声图像的切面对应的目标信息之前,该基于胎儿超声图像的胎儿严重畸形检测方法还可以包括以下操作:
当胎儿超声图像的切面包括颅脑切面时,判断胎儿超声图像的切面与颅脑标准切面是否相匹配,当判断结果为是时,触发执行上述的确定胎儿超声图像的切面对应的目标信息的操作;
当判断出不匹配时,基于获取到的结构特征校正胎儿超声图像的切面以使该切面与颅脑标准切面是否相匹配,并触发执行上述的确定胎儿超声图像的切面对应的目标信息的操作;
其中,当颅脑切面为颅脑水平切面时,颅脑标准切面包括侧脑室切面,当颅脑切面为颅脑矢状切面时,颅脑标准切面包括颅脑正中矢状切面。进一步的,当颅脑切面为颅脑矢状切面时,获取到的结构特征包括鼻骨结构特征、上颌骨结构特征、下颌骨结构特征、颈项透明层结构特征以及胎儿躯干等中的一种或多种组合;当颅脑切面为颅脑水平切面时,获取到的结构特征包括透明隔腔结构特征、侧脑室结构特征、脉络膜丛结构特征、丘脑结构特征以及小脑结构特征等中的一种或多种组合。
可见,该可选的实施例通过先判断获取到的胎儿超声图像的切面与颅脑标准切面是否相匹配,若匹配,则继续执行后续的获取切面的目标信息的操作,若否,则获取与其颅脑标准切面相匹配的切面,并重新获取切面的目标信息,有利于提高胎儿的畸形情况的确定准确性以及可靠性。
102、根据胎儿超声图像的切面对应的目标信息,判断胎儿超声图像对应的胎儿是否存在畸形,当判断结果为是时,触发执行步骤103,当判断结果为否时,可以结束本次流程,也可以在上述胎儿超声图像为单帧胎儿超声图像时,触发执行步骤101。
本发明实施例中,又进一步可选的,当判断出胎儿超声图像对应的胎儿存在畸形时,判断胎儿超声图像对应的胎儿是否为连体胎儿(例如:连体双胞胎),当判断结果为是时,根据上述胎儿超声图像的切面对应的目标信息确定胎儿超声图像对应的胎儿的连体畸形程度(和/或等级)。其中,当判断出胎儿超声图像存在胸部连胎、臀部连胎、腹部连胎、坐骨连胎、颅部连胎中的至少一种情况时,确定胎儿超声图像对应的胎儿为连体胎儿。这样在判断出胎儿存在畸形之后,进一步判断胎儿是否为连体胎儿,当为连体胎儿时,进一步确定连体畸形程度,此时,畸形情况还包括胎儿超声图像对应的胎儿的畸形程度,这样能够进一步提高胎儿超声图像对应的胎儿存在连体胎儿畸形的确定准确性以及可靠性。其中,针对根据上述胎儿超声图像的切面对应的目标信息确定胎儿超声图像对应的胎儿的连体畸形程度的有关描述,请参阅上述针对各个切面的相关描述,在此不再赘述。
103、根据胎儿超声图像的切面对应的目标信息确定胎儿超声图像对应的胎儿的畸形情况,该畸形情况包括胎儿超声图像对应的胎儿的畸形类型。
本发明实施例中,根据前述的分析,胎儿超声图像对应的胎儿的畸形类型包括无脑儿畸形类型、脑膜膨出畸形类型、无叶全前脑畸形类型、囊状脊髓脊膜膨出畸形类型、肢体(一侧或者双侧)缺失畸形类型、单心室畸形类型、腹裂畸形并内脏外翻畸形类型以及致死性骨发育不良畸形类型中的至少一种,进一步的,还可以包括连体胎儿畸形类型。这样通过综合检测、分析胎儿超声图像,能够获取到全面的胎儿的发育畸形情况以及有利于进一步提高胎儿的发育畸形情况的确定准确性以及可靠性。
在另一个可选的实施例中,该基于胎儿超声图像的胎儿严重畸形检测方法还可以包括以下操作:
当步骤102判断出胎儿超声图像对应的胎儿存在畸形时,为胎儿超声图像的每个异常结构特征设置对应的结构特征标签,每个异常结构特征对应的结构特征标签用于表示该异常结构特征的畸形类型。
可见,该可选的实施例在判断出胎儿存在畸形时,为异常结构特征设置对应的结构特征标签,便于医护人员根据结构特征标签清楚且快速知晓胎儿的畸形情况。
在又一个可选的实施例中,该基于胎儿超声图像的胎儿严重畸形检测方法还可以包括以下操作:
当步骤102判断出胎儿超声图像对应的胎儿不存在异常时,同样生成胎儿超声图像对应的胎儿的检测结果,其中,该检测结果用于表示胎儿超声图像对应的胎儿生长发育正常。
可见,该可选的实施例在判断出胎儿超声图像对应的胎儿不存在异常时,同样生成胎儿的检测结果,进一步便于相关人员知晓胎儿的生长发育情况。
可见,实施图1所描述的基于胎儿超声图像的胎儿严重畸形检测方法通过在获取到胎儿超声图像的切面之后,能够自动根据确定出的胎儿超声图像的切面的信息判断胎儿是否存在异常,当存在异常时,自动根据胎儿超声图像的切面的信息确定胎儿的畸形情况,例如:是否为无脑儿、腹裂畸形等,有利于准确检测胎儿的畸形情况,从而实现胎儿的生长发育情况的准确确定;以及通过多方面综合检测及分析胎儿的生长发育情况,有利于提高胎儿的畸形情况的检测准确性。
实施例二
请参阅图2,图2是本发明实施例公开的另一种基于胎儿超声图像的胎儿严重畸形检测方法的流程示意图。其中,图2所描述的基于胎儿超声图像的胎儿严重畸形检测方法可以应用于检测服务器(服务设备/服务***)中,其中,该检测服务器可以包括本地检测服务器或云检测服务器,本发明实施例不做限定。如图2所示,该基于胎儿超声图像的胎儿严重畸形检测方法可以包括以下操作:
201、将获取到的连续多帧胎儿超声图像输入确定出的结构特征检测模型中进行分析,并获取结构特征检测模型输出的分析结果,作为每帧胎儿超声图像的结构特征信息。
本发明实施例中,每帧胎儿超声图像的结构特征信息包括该胎儿超声图像的部位结构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结构特征信息至少包括该胎儿超声图像的部位结构特征的类别,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。
本发明实施例中,可选的,实施例一中的确定胎儿超声图像的切面的形状的具体实现方式还可以为在将胎儿超声图像输入结构特征检测模型进行分析时,即可同步分析出胎儿超声图像的切面的形状,这样能够在保证胎儿超声图像的切面的形状的获取准确性的同时,提高胎儿超声图像的切面的形状的获取效率,从而有利于提高胎儿是否存在异常的判断效率。
本发明实施例中,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切面有关,即根据所需获取的胎儿超声图像的切面来选择帧率,例如:若需要获取的是腹部切面,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切面选择对应的帧率,有利于提高胎儿超声图像的切面的获取效率以及准确性,从而有利于提高胎儿超声图像的切面的目标信息的获取效率,进而有利于提高胎儿超声图像存在异常的判断效率。
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切面的信息的管理。
本发明实施例中,结构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结构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。
202、根据每帧胎儿超声图像的部位结构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的切面。
可见,本发明实施例通过获取连续多帧胎儿超声图像的部位结构特征以及结构特征,并结合胎儿超声图像的部位结构特征以及结构特征,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结构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,从而有利于提高胎儿的畸形情况的检测准确性以及可靠性。
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切面,来实现胎儿超声图像的切面的获取。这样通过多种途径获取胎儿超声图像的切面,能够丰富切面的获取方式,提高切面的获取可能性。
203、在获取到胎儿超声图像的切面之后,确定胎儿超声图像的切面对应的目标信息。
204、根据胎儿超声图像的切面对应的目标信息,判断胎儿超声图像对应的胎儿是否存在畸形,当判断结果为是时,触发执行步骤103,当判断结果为否时,结束本次流程,也可以触发执行步骤203。
205、根据胎儿超声图像的切面对应的目标信息确定胎儿超声图像对应的胎儿的畸形情况,该畸形情况包括胎儿超声图像对应的胎儿的畸形类型。
本发明实施例中,针对步骤203-205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。
可见,实施图2所描述的基于胎儿超声图像的胎儿严重畸形检测方法通过在获取到胎儿超声图像的切面之后,能够自动根据确定出的胎儿超声图像的切面的信息判断胎儿是否存在异常,当存在异常时,自动根据胎儿超声图像的切面的信息确定胎儿的畸形情况,例如:是否为无脑儿、腹裂畸形等,有利于准确检测胎儿的畸形情况,从而实现胎儿的生长发育情况的准确确定;以及通过多方面综合检测及分析胎儿的生长发育情况,有利于提高胎儿的畸形情况的检测准确性。此外,还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结构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,从而有利于提高胎儿的畸形情况的检测准确性以及可靠性。
实施例三
请参阅图3,图3是本发明实施例公开的一种基于胎儿超声图像的胎儿严重畸形检测装置的结构示意图。其中,图3所描述的基于胎儿超声图像的胎儿严重畸形检测装置可以应用于检测服务器(服务设备/服务***)中,其中,该检测服务器可以包括本地检测服务器或云检测服务器,本发明实施例不做限定。如图3所示,该基于胎儿超声图像的胎儿严重畸形检测装置可以包括确定模块301、第一判断模块302,其中:
确定模块301,用于在获取到胎儿超声图像的切面之后,确定该胎儿超声图像的切面对应的目标信息,该胎儿超声图像的切面对应的目标信息用于确定胎儿超声图像对应的胎儿的发育情况;
第一判断模块302,用于根据胎儿超声图像的切面对应的目标信息,判断胎儿超声图像对应的胎儿是否存在畸形。
确定模块301,还用于当第一判断模块302判断结果为是时,根据胎儿超声图像的切面对应的目标信息确定胎儿超声图像对应的胎儿的畸形情况,该畸形情况包括胎儿超声图像对应的胎儿的畸形类型。
本发明实施例中,可选的,胎儿超声图像的切面包括颅脑切面、肢体切面、腹部切面、脊髓切面、心脏切面、肱骨长径切面以及股骨长径切面中的其中一种,该胎儿超声图像的颅脑切面包括颅脑水平切面和/或颅脑矢状切面,该胎儿超声图像的肢体切面包括双上肢切面或双下肢切面,该胎儿超声图像的腹部切 面包括腹部水平切面和/或腹部矢状切面,该胎儿超声图像的脊髓切面包括脊髓水平切面和/或脊髓矢状切面。
可见,实施图3所描述的基于胎儿超声图像的胎儿严重畸形检测装置通过在获取到胎儿超声图像的切面之后,能够自动根据确定出的胎儿超声图像的切面的信息判断胎儿是否存在异常,当存在异常时,自动根据胎儿超声图像的切面的信息确定胎儿的畸形情况,例如:是否为无脑儿、腹裂畸形等,有利于准确检测胎儿的畸形情况,从而实现胎儿的生长发育情况的准确确定;以及通过多方面综合检测及分析胎儿的生长发育情况,有利于提高胎儿的畸形情况的检测准确性。
在一个可选的实施例中,如图3所示,第一判断模块302根据胎儿超声图像的切面对应的目标信息,判断该胎儿超声图像对应的胎儿是否存在异常的方式具体为:
当胎儿超声图像的切面包括颅脑切面,且该颅脑切面为颅脑矢状切面时,胎儿超声图像的切面对应的目标信息包括该切面的颅脑结构特征轮廓以及臀部结构特征轮廓,根据胎儿超声图像的切面的颅脑结构特征轮廓以及臀部结构特征轮廓,测量胎儿超声图像的头臀长度,并判断头臀长度是否在预设头臀长度范围内,当判断出不相匹配时,确定胎儿超声图像对应的胎儿存在畸形;
当胎儿超声图像的切面包括颅脑切面,且该颅脑切面为颅脑矢状切面时,胎儿超声图像的切面对应的目标信息包括胎儿超声图像的切面的颅脑结构特征的几何参数,判断切面的颅脑结构特征的几何参数与预设几何参数是否相匹配,当判断出不相匹配时,确定胎儿超声图像对应的胎儿存在畸形,该切面的颅脑结构特征包括该切面内的颅盖骨结构特征、大脑半球结构特征以及中脑结构特征中的至少一种,颅脑结构特征的几何参数包括该颅脑结构特征的形状、尺寸、位置以及面积中的至少一种,该颅脑结构特征的位置为该颅脑结构特征在胎儿超声图像的切面中的位置;
当胎儿超声图像的切面包括颅脑切面或脊髓切面或腹部切面时,该胎儿超声图像的切面对应的目标信息包括胎儿超声图像的切面的形状,判断切面的形状是否与预设切面形状相匹配,当判断出不相匹配时,确定胎儿超声图像对应的胎儿存在畸形;
当胎儿超声图像的切面为颅脑切面或腹部切面或脊髓切面时,该胎儿超声图像的切面对应的目标信息包括该切面的目标结构特征对应的特征参数,根据切面的目标结构特征对应的特征参数判断目标结构特征是否与该切面相匹配,当判断出不相匹配时,确定胎儿超声图像对应的胎儿存在畸形;
当胎儿超声图像的切面包括颅脑切面,且该颅脑切面为颅脑水平切面时,胎儿超声图像的切面对应的目标信息包括胎儿超声图像的切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,根据切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,确定切面的颅脑结构特征的目标几何参数,并判断颅脑结构特征的目标几何参数是否在预设几何参数范围内,当判断出不在预设几何参数范围内时,确定胎儿超声图像对应的胎儿存在畸形,该切面的颅脑结构特征的目标几何参数包括该颅脑结构特征的头围参数和/或双顶径参数;
当胎儿超声图像的切面包括颅脑切面,且该颅脑切面为颅脑水平切面时,胎儿超声图像的切面对应的目标信息包括该切面的左丘脑轮廓以及右丘脑轮廓,获取左丘脑轮廓与右丘脑轮廓的第一拟合度,并判断第一拟合度是否大于等于第一预设拟合度阈值,当判断结果为是时,确定胎儿超声图像对应的胎儿存在畸形;
当胎儿超声图像的切面包括颅脑切面,且该颅脑切面为颅脑水平切面时,胎儿超声图像的切面对应的目标信息包括该切面的大脑镰的位置,根据大脑镰的位置判断大脑镰在胎儿超声图像的脑中线位置的出现情况是否满足预设出现情况,当判断出不满足预设出现情况时,确定胎儿超声图像对应的胎儿存在畸形;
当胎儿超声图像的切面包颅脑切面,且该颅脑切面为颅脑水平切面时,胎儿超声图像的切面对应的目标信息包括该切面的脉络膜丛轮廓以及丘脑轮廓, 获取脉络膜丛轮廓与丘脑轮廓之间的第二拟合度,并判断第二拟合度是否大于等于第二预设拟合度阈值,当判断结果为是时,确定胎儿超声图像对应的胎儿存在畸形。
可见,实施图3所描述的确定装置通过判断胎儿超声图像的切面中的结构特征是否与该切面相匹配,能够实现颅脑脑膜膨出、囊状脊髓脊膜膨出以及腹裂畸形并内脏外翻的确定,有利于提高胎儿的畸形情况的判断准确性;以及通过胎儿超声图像的切面的颅脑结构特征轮廓与臀部结构特征轮廓和/或颅脑结构特征的几何参数,能够实现胎儿超声图像对应的胎儿为无脑儿的异常确定;以及通过确定颅脑的左右丘脑融合、大脑镰缺如以及脉络膜丛与丘脑未分离中的至少一种情况,不仅能够实现胎儿超声图像的胎儿为无叶全前脑胎儿的确定,还能够丰富无叶全前脑胎儿的确定方式;以及通过提供至少两种无叶全前脑胎儿的确定方式,能够提高无叶全前脑胎儿的确定准确性以及可靠性。
在另一个可选的实施例中,如图3所示,第一判断模块302根据切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,确定该切面的颅脑结构特征的目标几何参数的方式具体为:
获取切面的颅脑结构特征的内轮廓的第一周长与颅脑结构特征的外轮廓的第二周长,并基于第一周长和第二周长,确定颅脑结构特征对应的头围参数;
确定颅脑结构特征的脑中线对应的中垂线与颅脑结构特征的外轮廓的第一交点以及中垂线与颅脑结构特征的内轮廓的第二交点,并基于第一交点与第二交点,确定颅脑结构特征对应的双顶径参数。
可见,实施图3所描述的确定装置能够通过获取颅脑结构特征的内、外轮廓的周长,以及颅脑结构特征的脑中线对应的中垂线与内、外轮廓的交点,能够实现颅脑结构特征的头围周长以及双顶径长度的获取,从而实现胎儿颅脑脑膜膨出的确定。
在又一个可选的实施例中,如图4所示,该装置还可以包括第二判断模块303以及校正模块304,其中:
第二判断模块303,用于当胎儿超声图像的切面包括颅脑切面时,在获取到胎儿超声图像的切面之后,以及在确定模块301确定胎儿超声图像的切面对应的目标信息之前,判断胎儿超声图像的切面与颅脑标准切面是否相匹配,当判断结果为是时,触发确定模块301执行上述的确定胎儿超声图像的切面对应的目标信息的操作;
校正模块304,用于当第二判断模块303判断出不匹配时,基于获取到的结构特征校正胎儿超声图像的切面以使该切面与颅脑标准切面相匹配,并触发确定模块301执行上述的确定胎儿超声图像的切面对应的目标信息的操作;
其中,当颅脑切面为颅脑水平切面时,颅脑标准切面包括侧脑室切面,当颅脑切面为颅脑矢状切面时,颅脑标准切面包括颅脑正中矢状切面。
可见,实施图4所描述的确定装置通过先判断获取到的胎儿超声图像的切面与颅脑标准切面是否相匹配,若匹配,则继续执行后续的获取切面的目标信息的操作,若否,则获取与其颅脑标准切面相匹配的切面,并重新获取切面的目标信息,有利于提高胎儿的畸形情况的确定准确性以及可靠性。
在又一个可选的实施例中,如图3或4所示,胎儿超声图像的切面对应的目标信息包括该切面的肢体结构特征的特征参数,其中,当胎儿超声图像的切面包括肢体切面,且肢体切面为双上肢切面时,切面的肢体结构特征包括该切面的手、前臂以及上臂中的至少一种;当胎儿超声图像的切面包括肢体切面,且肢体切面为双下肢切面时,切面的肢体结构特征包括该切面的足、大腿以及小腿中的至少一种;
以及,第一判断模块302根据胎儿超声图像的切面对应的目标信息,判断该胎儿超声图像对应的胎儿是否存在异常的方式具体为:
根据胎儿超声图像的切面的肢体结构特征的特征参数,判断肢体结构特征是否与胎儿超声图像的切面相匹配,当判断出不相匹配时,确定胎儿超声图像对应的胎儿存在畸形。
可见,实施图3或4所描述的确定装置能够通过胎儿超声图像的切面中的 肢体结构特征(例如:手、足)的特征参数,能够实现胎儿超声图像对应的胎儿是否为肢体缺失的胎儿的确定。
在又一个可选的实施例中,如图3或4所示,第一判断模块302根据胎儿超声图像的切面对应的目标信息,判断胎儿超声图像对应的胎儿是否存在畸形的方式具体为:
当胎儿超声图像的切面为颅脑切面或肱骨长径切面或股骨长径切面时,胎儿超声图像的切面对应的目标信息包括该切面的骨结构特征的特征参数,根据切面的骨结构特征的特征参数判断骨结构特征是否与该切面相匹配,当判断出不相匹配时,确定胎儿超声图像对应的胎儿存在畸形,骨结构特征的特征参数包括骨结构特征对应的轮廓、长度、面积、形状以及位置中的至少一种。
可见,实施图3或4所描述的确定装置能够通过将测量到的胎儿超声图像的切面的骨结构特征的特征参数(例如:长度)与对应切面相匹配,能够实现胎儿超声图像的胎儿为致死性骨发育不良的确定。
在又一个可选的实施例中,如图3或4所示,第一判断模块302根据胎儿超声图像的切面对应的目标信息,判断胎儿超声图像的对应的胎儿是否存在异常的方式具体为:
当胎儿超声图像的切面为心脏切面时,胎儿超声图像的切面对应的目标信息包括该切面的特征参数,判断切面特征参数与预设切面参数是否相匹配,当判断出不匹配时,确定胎儿超声图像的对应的胎儿存在畸形,切面的特征参数包括切面的多普勒血流参数和/或轮廓参数;
当胎儿超声图像的切面为心脏切面时,胎儿超声图像的切面对应的目标信息包括该切面的心脏结构特征的特征参数,根据切面的心脏结构特征的特征参数判断切面的心脏结构特征是否与该切面的标准心脏结构特征相匹配,当判断出不匹配时,确定胎儿超声图像的对应的胎儿存在畸形,切面的心脏结构特征的特征参数包括心脏结构特征的数量和/或心脏结构特征对应的面积;
其中,第一判断模块302根据切面的心脏结构特征的特征参数判断切面的心脏结构特征是否与该切面的标准心脏结构特征相匹配的方式具体为:
当切面的心脏结构特征的特征参数为心脏结构特征的数量时,判断切面的心脏结构特征的数量是否小于等于预设数量,当判断结果为是时,确定切面的心脏结构特征与该切面的标准心脏结构特征不相匹配;
当切面的心脏结构特征的特征参数为心脏结构特征对应的面积时,判断切面的心脏结构特征对应的面积是否大于等于预设面积阈值,当判断结果为是时,确定切面的心脏结构特征与该切面的标准心脏结构特征不相匹配。
可见,实施图3或4所描述的确定装置能够通过心脏切面的多普勒血流参数、轮廓参数、心脏结构特征数量以及每个心脏结构特征的面积,判断胎儿超声图像的胎儿的心脏是否存在异常,能够提高胎儿的心脏存在异常的确定准确性以及可靠性,有利于提高胎儿的心脏畸形情况的确定准确性以及可靠性。
在又一个可选的实施例中,如图4所示,该装置还包括分析模块305以及获取模块306,其中:
分析模块305,用于将获取到的连续多帧胎儿超声图像输入确定出的结构特征检测模型中进行分析。
获取模块306,用于获取结构特征检测模型输出的分析结果,作为每帧胎儿超声图像的结构特征信息,每帧胎儿超声图像的结构特征信息包括该胎儿超声图像的部位结构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结构特征信息至少包括该胎儿超声图像的部位结构特征的类别,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;
确定模块301,还用于根据每帧胎儿超声图像的部位结构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的切面。
可见,实施图4所描述的确定装置能够通过获取连续多帧胎儿超声图像的部位结构特征以及结构特征,并结合胎儿超声图像的部位结构特征以及结构特征,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 图像输入结构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,从而有利于提高胎儿的畸形情况的检测准确性以及可靠性。
在又一个可选的实施例中,如图4所示,该装置还包括设置模块307,其中:
设置模块307,用于当第一判断模块302判断出胎儿超声图像对应的胎儿存在畸形时,为胎儿超声图像的每个异常结构特征设置对应的结构特征标签,每个异常结构特征对应的结构特征标签用于表示该异常结构特征的畸形类型。
可见,实施图4所描述的确定装置能够通过在判断出胎儿存在畸形时,为异常结构特征设置对应的结构特征标签,便于医护人员根据结构特征标签清楚且快速知晓胎儿的畸形情况。
实施例四
请参阅图5,图5是本发明实施例公开的又一种基于胎儿超声图像的胎儿严重畸形检测装置。其中,图5所描述的基于胎儿超声图像的胎儿严重畸形检测装置可以应用于检测服务器(服务设备/服务***)中,其中,该检测服务器可以包括本地检测服务器或云检测服务器,本发明实施例不做限定。如图5所示,该基于胎儿超声图像的胎儿严重畸形检测装置可以包括:
存储有可执行程序代码的存储器501;
与存储器501耦合的处理器502;
进一步的,还可以包括与处理器502耦合的输入接口503以及输出接口504;
其中,处理器502调用存储器501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基于胎儿超声图像的胎儿严重畸形检测方法中部分或者全部的步骤。
实施例五
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基于胎儿超声图像的胎儿严重畸形检测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基于胎儿超声图像的胎儿严重畸形检测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基于胎儿超声图像的胎儿严重畸形检测方法基于胎儿超声图像的胎儿严重畸形检测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 术结构特征进行等同替换;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 (12)

  1. 一种基于胎儿超声图像的胎儿严重畸形检测方法,其特征在于,所述方法包括:
    在获取到胎儿超声图像的切面之后,确定所述胎儿超声图像的切面对应的目标信息,所述胎儿超声图像的切面对应的目标信息用于确定所述胎儿超声图像对应的胎儿的发育情况;
    根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形,当判断结果为是时,根据所述胎儿超声图像的切面对应的目标信息确定所述胎儿超声图像对应的胎儿的畸形情况,所述畸形情况包括所述胎儿超声图像对应的胎儿的畸形类型。
  2. 根据权利要求1所述的基于胎儿超声图像的胎儿严重畸形检测方法,其特征在于,所述胎儿超声图像的切面包括颅脑切面、肢体切面、腹部切面、脊髓切面、心脏切面、肱骨长径切面以及股骨长径切面中的其中一种,所述胎儿超声图像的颅脑切面包括颅脑水平切面和/或颅脑矢状切面,所述胎儿超声图像的肢体切面包括双上肢切面或双下肢切面,所述胎儿超声图像的腹部切面包括腹部水平切面和/或腹部矢状切面,所述胎儿超声图像的脊髓切面包括脊髓水平切面和/或脊髓矢状切面。
  3. 根据权利要求2所述的基于胎儿超声图像的胎儿严重畸形检测方法,其特征在于,所述根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形,包括:
    当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑矢状切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的颅脑结构特征轮廓以及臀部结构特征轮廓,根据所述胎儿超声图像的切面的颅脑结构特征轮廓以及臀部结构特征轮廓,测量所述胎儿超声图像的头臀长度,并判断所述头臀长度是否在预设头臀长度范围内,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形;
    当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑矢状切面时,所述胎儿超声图像的切面对应的目标信息包括所述胎儿超声图像的切面的颅脑结构特征的几何参数,判断所述切面的颅脑结构特征的几何参数与预设几何参数是否相匹配,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形,所述切面的颅脑结构特征包括该切面内的颅盖骨结构特征、大脑半球结构特征以及中脑结构特征中的至少一种,所述颅脑结构特征的几何参数包括该颅脑结构特征的形状、尺寸、位置以及面积中的至少一种,所述颅脑结构特征的位置为该颅脑结构特征在所述胎儿超声图像的切面中的位置;
    当所述胎儿超声图像的切面包括所述颅脑切面或所述脊髓切面或所述腹部切面时,所述胎儿超声图像的切面对应的目标信息包括所述胎儿超声图像的切面的形状,判断所述切面的形状是否与预设切面形状相匹配,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形;
    当所述胎儿超声图像的切面为所述颅脑切面或所述腹部切面或所述脊髓切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的目标结构特征对应的特征参数,根据所述切面的目标结构特征对应的特征参数判断所述目标结构特征是否与该切面相匹配,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形;
    当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑水平切面时,所述胎儿超声图像的切面对应的目标信息包括所述胎儿超声图像的切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,根据所述切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,确定所述切面的颅脑结构特征的目标几何参数,并判断所述颅脑结构特征的目标几何参数是否在预设几何参数范围内,当判断出不在所述预设几何参数范围内时,确定所述胎儿超 声图像对应的胎儿存在畸形,所述切面的颅脑结构特征的目标几何参数包括该颅脑结构特征的头围参数和/或双顶径参数;
    当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑水平切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的左丘脑轮廓以及右丘脑轮廓,获取所述左丘脑轮廓与所述右丘脑轮廓的第一拟合度,并判断所述第一拟合度是否大于等于第一预设拟合度阈值,当判断结果为是时,确定所述胎儿超声图像对应的胎儿存在畸形;
    当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑水平切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的大脑镰的位置,根据所述大脑镰的位置判断所述大脑镰在所述胎儿超声图像的脑中线位置的出现情况是否满足预设出现情况,当判断出不满足所述预设出现情况时,确定所述胎儿超声图像对应的胎儿存在畸形;
    当所述胎儿超声图像的切面包括所述颅脑切面,且所述颅脑切面为所述颅脑水平切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的脉络膜丛轮廓以及丘脑轮廓,获取所述脉络膜丛轮廓与所述丘脑轮廓之间的第二拟合度,并判断所述第二拟合度是否大于等于第二预设拟合度阈值,当判断结果为是时,确定所述胎儿超声图像对应的胎儿存在畸形。
  4. 根据权利要求3所述的胎儿结构特征的自动确定方法,其特征在于,所述根据所述切面的颅脑结构特征的内轮廓与该颅脑结构特征的外轮廓,确定所述切面的颅脑结构特征的目标几何参数,包括:
    获取所述切面的颅脑结构特征的内轮廓的第一周长与所述颅脑结构特征的外轮廓的第二周长,并基于所述第一周长和所述第二周长,确定所述颅脑结构特征对应的头围参数;
    确定所述颅脑结构特征的脑中线对应的中垂线与所述颅脑结构特征的外轮廓的第一交点以及所述中垂线与所述颅脑结构特征的内轮廓的第二交点,并基于所述第一交点与所述第二交点,确定所述颅脑结构特征对应的双顶径参数。
  5. 根据权利要求2-4任一项所述的胎儿结构特征的自动确定方法,其特征在于,在获取到胎儿超声图像的切面之后,以及所述确定所述胎儿超声图像的切面对应的目标信息之前,所述方法还包括:
    当所述胎儿超声图像的切面包括所述颅脑切面时,判断所述胎儿超声图像的切面与颅脑标准切面是否相匹配,当判断结果为是时,触发执行所述的确定所述胎儿超声图像的切面对应的目标信息的操作;
    当判断出不匹配时,基于获取到的结构特征校正所述胎儿超声图像的切面以使该切面与所述颅脑标准切面相匹配,并触发执行所述的确定所述胎儿超声图像的切面对应的目标信息的操作;
    其中,当所述颅脑切面为所述颅脑水平切面时,所述颅脑标准切面包括侧脑室切面,当所述颅脑切面为所述颅脑矢状切面时,所述颅脑标准切面包括颅脑正中矢状切面。
  6. 根据权利要求2所述的基于胎儿超声图像的胎儿严重畸形检测方法,其特征在于,所述胎儿超声图像的切面对应的目标信息包括该切面的肢体结构特征的特征参数,其中,当所述胎儿超声图像的切面包括所述肢体切面,且所述肢体切面为所述双上肢切面时,所述切面的肢体结构特征包括该切面的手、前臂以及上臂中的至少一种;当所述胎儿超声图像的切面包括所述肢体切面,且所述肢体切面为所述双下肢切面时,所述切面的肢体结构特征包括该切面的足、大腿以及小腿中的至少一种;
    以及,所述根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形,包括:
    根据所述胎儿超声图像的切面的肢体结构特征的特征参数,判断所述肢体结构特征是否与所述胎儿超声图像的切面相匹配,当判断出不相匹配时,确定 所述胎儿超声图像对应的胎儿存在畸形。
  7. 根据权利要求2所述的基于胎儿超声图像的胎儿严重畸形检测方法,其特征在于,所述根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形,包括:
    当所述胎儿超声图像的切面为所述颅脑切面或所述肱骨长径切面或所述股骨长径切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的骨结构特征的特征参数,根据所述切面的骨结构特征的特征参数判断所述骨结构特征是否与该切面相匹配,当判断出不相匹配时,确定所述胎儿超声图像对应的胎儿存在畸形,所述骨结构特征的特征参数包括所述骨结构特征对应的轮廓、长度、面积、形状以及位置中的至少一种。
  8. 根据权利要求2所述的基于胎儿超声图像的胎儿严重畸形检测方法,其特征在于,所述根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形,包括:
    当所述胎儿超声图像的切面为所述心脏切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的特征参数,判断所述切面特征参数与预设切面参数是否相匹配,当判断出不匹配时,确定所述胎儿超声图像的对应的胎儿存在畸形,所述切面的特征参数包括所述切面的多普勒血流参数和/或轮廓参数;
    当所述胎儿超声图像的切面为所述心脏切面时,所述胎儿超声图像的切面对应的目标信息包括该切面的心脏结构特征的特征参数,根据所述切面的心脏结构特征的特征参数判断所述切面的心脏结构特征是否与该切面的标准心脏结构特征相匹配,当判断出不匹配时,确定所述胎儿超声图像的对应的胎儿存在畸形,所述切面的心脏结构特征的特征参数包括所述心脏结构特征的数量和/或所述心脏结构特征对应的面积;
    其中,所述根据所述切面的心脏结构特征的特征参数判断所述切面的心脏结构特征是否与该切面的标准心脏结构特征相匹配,包括:
    当所述切面的心脏结构特征的特征参数为所述心脏结构特征的数量时,判断所述切面的心脏结构特征的数量是否小于等于预设数量,当判断结果为是时,确定所述切面的心脏结构特征与该切面的标准心脏结构特征不相匹配;
    当所述切面的心脏结构特征的特征参数为所述心脏结构特征对应的面积时,判断所述切面的心脏结构特征对应的面积是否大于等于预设面积阈值,当判断结果为是时,确定所述切面的心脏结构特征与该切面的标准心脏结构特征不相匹配。
  9. 根据权利要求1-8任一项所述的基于胎儿超声图像的胎儿严重畸形检测方法,其特征在于,所述方法还包括:
    将获取到的连续多帧胎儿超声图像输入确定出的结构特征检测模型中进行分析;
    获取所述结构特征检测模型输出的分析结果,作为每帧所述胎儿超声图像的结构特征信息,每帧所述胎儿超声图像的结构特征信息包括该胎儿超声图像的部位结构特征信息以及该胎儿超声图像的结构特征信息,每帧所述胎儿超声图像的部位结构特征信息至少包括该胎儿超声图像的部位结构特征的类别,每帧所述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;
    根据每帧所述胎儿超声图像的部位结构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的切面。
  10. 根据权利要求1-8任一项所述的胎儿结构特征的自动确定方法,其特征在于,所述方法还包括:
    当判断出所述胎儿超声图像对应的胎儿存在畸形时,为所述胎儿超声图像的每个异常结构特征设置对应的结构特征标签,每个所述异常结构特征对应的 结构特征标签用于表示该异常结构特征的畸形类型。
  11. 一种基于胎儿超声图像的胎儿严重畸形检测装置,其特征在于,所述装置包括:
    确定模块,用于在获取到胎儿超声图像的切面之后,确定所述胎儿超声图像的切面对应的目标信息,所述胎儿超声图像的切面对应的目标信息用于确定所述胎儿超声图像对应的胎儿的发育情况;
    第一判断模块,用于根据所述胎儿超声图像的切面对应的目标信息,判断所述胎儿超声图像对应的胎儿是否存在畸形;
    确定模块,还用于当所述第一判断模块判断出所述胎儿超声图像对应的胎儿存在畸形时,根据所述胎儿超声图像的切面对应的目标信息确定所述胎儿超声图像对应的胎儿的畸形情况,所述畸形情况包括所述胎儿超声图像对应的胎儿的畸形类型。
  12. 一种基于胎儿超声图像的胎儿严重畸形检测装置,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1-10任一项所述的基于胎儿超声图像的胎儿严重畸形检测方法。
PCT/CN2021/096813 2020-09-24 2021-05-28 基于胎儿超声图像的胎儿严重畸形检测方法及装置 WO2022062455A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011016343.9 2020-09-24
CN202011016343.9A CN112155604B (zh) 2020-09-24 2020-09-24 基于胎儿超声图像的胎儿严重畸形检测方法及装置

Publications (1)

Publication Number Publication Date
WO2022062455A1 true WO2022062455A1 (zh) 2022-03-31

Family

ID=73863925

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/096813 WO2022062455A1 (zh) 2020-09-24 2021-05-28 基于胎儿超声图像的胎儿严重畸形检测方法及装置

Country Status (2)

Country Link
CN (1) CN112155604B (zh)
WO (1) WO2022062455A1 (zh)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112155604B (zh) * 2020-09-24 2023-03-31 广州爱孕记信息科技有限公司 基于胎儿超声图像的胎儿严重畸形检测方法及装置
CN114073549A (zh) * 2021-11-23 2022-02-22 南方科技大学医院 一种智能获取胎儿颜面部骨性结构参数的方法及终端设备
CN114255867B (zh) * 2021-12-22 2024-01-16 华中科技大学协和深圳医院 一种超声图像检测方法、装置、设备及介质

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050245825A1 (en) * 2003-07-29 2005-11-03 Krantz David A System and method for assessing fetal abnormality based on landmarks
CN110613483A (zh) * 2019-09-09 2019-12-27 李胜利 一种基于机器学习检测胎儿颅脑异常的方法和***
CN110634125A (zh) * 2019-01-14 2019-12-31 广州爱孕记信息科技有限公司 基于深度学习的胎儿超声图像识别方法及***
CN110652317A (zh) * 2019-09-24 2020-01-07 深圳度影医疗科技有限公司 一种产前胎儿超声容积图像中标准切面的自动定位方法
CN110742654A (zh) * 2019-11-05 2020-02-04 深圳度影医疗科技有限公司 一种基于三维超声图像的标准切面的定位和测量方法
CN110991254A (zh) * 2019-11-08 2020-04-10 深圳大学 超声图像视频分类预测方法及***
CN112155604A (zh) * 2020-09-24 2021-01-01 广州爱孕记信息科技有限公司 基于胎儿超声图像的胎儿严重畸形检测方法及装置

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105405119B (zh) * 2015-10-21 2018-07-13 复旦大学 基于深度置信网络和三维模型的胎儿正中矢状面自动检测方法
CN108186051B (zh) * 2017-12-26 2021-11-30 珠海艾博罗生物技术股份有限公司 一种从超声图像中自动测量胎儿双顶径长度的图像处理方法及处理***
CN108378869B (zh) * 2017-12-26 2021-04-20 珠海艾博罗生物技术股份有限公司 一种从超声图像中自动测量胎儿头围长度的图像处理方法及处理***
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***
CN111326256B (zh) * 2020-02-28 2023-12-29 李胜利 胎儿超声标准切面图像智能识别自我训练学习***及考试方法
CN111462060A (zh) * 2020-03-24 2020-07-28 湖南大学 胎儿超声图像中标准切面图像的检测方法和装置

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050245825A1 (en) * 2003-07-29 2005-11-03 Krantz David A System and method for assessing fetal abnormality based on landmarks
CN110634125A (zh) * 2019-01-14 2019-12-31 广州爱孕记信息科技有限公司 基于深度学习的胎儿超声图像识别方法及***
CN110613483A (zh) * 2019-09-09 2019-12-27 李胜利 一种基于机器学习检测胎儿颅脑异常的方法和***
CN110652317A (zh) * 2019-09-24 2020-01-07 深圳度影医疗科技有限公司 一种产前胎儿超声容积图像中标准切面的自动定位方法
CN110742654A (zh) * 2019-11-05 2020-02-04 深圳度影医疗科技有限公司 一种基于三维超声图像的标准切面的定位和测量方法
CN110991254A (zh) * 2019-11-08 2020-04-10 深圳大学 超声图像视频分类预测方法及***
CN112155604A (zh) * 2020-09-24 2021-01-01 广州爱孕记信息科技有限公司 基于胎儿超声图像的胎儿严重畸形检测方法及装置

Also Published As

Publication number Publication date
CN112155604B (zh) 2023-03-31
CN112155604A (zh) 2021-01-01

Similar Documents

Publication Publication Date Title
WO2022062455A1 (zh) 基于胎儿超声图像的胎儿严重畸形检测方法及装置
Salomon et al. ISUOG Practice Guidelines (updated): performance of the routine mid-trimester fetal ultrasound scan
US8010175B2 (en) Patient-specific coronary territory mapping
Bethune et al. A pictorial guide for the second trimester ultrasound
US20230233177A1 (en) Method and Apparatus for Identification of Imaging Quality of Fetal Ultrasound Images
US20090012432A1 (en) State Based Birth Monitoring System
Hallowell et al. Reliability of quantitative echocardiography in adult sheep and goats
Abuhamad et al. First trimester ultrasound diagnosis of fetal abnormalities
JP2019526334A5 (zh)
CN111275755A (zh) 基于人工智能的二尖瓣瓣口面积检测方法、***和设备
Ruiz-Correa et al. New severity indices for quantifying single-suture metopic craniosynostosis
EP1773203A2 (en) State based birth monitoring system
Schmidt et al. Two-dimensional echocardiographic determination of ventricular volumes in the fetal heart. Validation studies in fetal lambs.
Ibrahim et al. Traumatic pericarditis in cattle: risk factors, clinical features and ultrasonographic findings
Muller et al. Prenatal ultrasonographic features of the Pena–Shokeir I syndrome and the trisomy 18 syndrome
Chervenak et al. Current status of fetal age and growth assessment
Zizzo et al. Growth‐restricted human fetuses have preserved respiratory sinus arrhythmia but reduced heart rate variability estimates of vagal activity during quiescence
Pittyanont et al. Cardio‐STIC Based Reference Ranges of Fetal Thymus Size in Singleton Pregnancies
Domínguez-Gallardo et al. Gestational Age-Adjusted Reference Ranges for Fetal Left Ventricle Longitudinal Strain by Automated Cardiac Motion Quantification between 24 and 37 Weeks’ Gestation
RU2529377C1 (ru) Способ дородовой оценки адаптации плода к повторной гипоксии по н.а. ураковой
Karadeniz et al. Results of fetal anomaly screening performed at 11–14 weeks of gestation at a tertiary center
TWI496557B (zh) 骨板製作方法
Zheng et al. Fetal Left Ventricular Mass Determination on 2‐Dimensional Echocardiography Using Area‐Length Calculation Methods
Robinson et al. Ultrasonic visualization of the placenta
Gembicki et al. P04. 09: The use of automated assessment of fractional limb volume for fetal birthweight estimation: a prospective study.

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: 21870842

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: 21870842

Country of ref document: EP

Kind code of ref document: A1