WO2022099705A1 - 早孕期胎儿的超声成像方法和超声成像系统 - Google Patents

早孕期胎儿的超声成像方法和超声成像系统 Download PDF

Info

Publication number
WO2022099705A1
WO2022099705A1 PCT/CN2020/129117 CN2020129117W WO2022099705A1 WO 2022099705 A1 WO2022099705 A1 WO 2022099705A1 CN 2020129117 W CN2020129117 W CN 2020129117W WO 2022099705 A1 WO2022099705 A1 WO 2022099705A1
Authority
WO
WIPO (PCT)
Prior art keywords
standard
target
trimester
fetus
section
Prior art date
Application number
PCT/CN2020/129117
Other languages
English (en)
French (fr)
Inventor
梁天柱
罗瑚
邹耀贤
林穆清
陈志杰
Original Assignee
深圳迈瑞生物医疗电子股份有限公司
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 深圳迈瑞生物医疗电子股份有限公司 filed Critical 深圳迈瑞生物医疗电子股份有限公司
Priority to PCT/CN2020/129117 priority Critical patent/WO2022099705A1/zh
Priority to CN202080105099.5A priority patent/CN116171131A/zh
Publication of WO2022099705A1 publication Critical patent/WO2022099705A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/13—Tomography
    • A61B8/14—Echo-tomography

Definitions

  • the present application relates to the technical field of ultrasound imaging, and more particularly, to an ultrasound imaging method and an ultrasound imaging system for a fetus in early pregnancy.
  • ultrasonography Due to its advantages of safety, convenience, non-radiation and low cost, ultrasonography is widely used in clinical examinations and has become one of the main auxiliary means for doctors to diagnose diseases. As the most important imaging examination in prenatal examination, prenatal ultrasonography provides the most important imaging evidence for fetal growth and development measurement and structural abnormal screening. Prenatal ultrasonography is already one of the tests that must be performed in the first, second and third trimesters.
  • fetal structural examination and malformation screening in early pregnancy are the current clinical development trends and research hotspots.
  • Fetal structural screening in early pregnancy can detect fatal malformations as early as possible, provide pregnant women with the opportunity to terminate pregnancy as soon as possible, and minimize physical and mental harm, which has very important clinical significance and value.
  • the screening of fetal structural abnormalities in early pregnancy has a large number of sections, and the requirements for each section are different. The acquisition of the section is cumbersome, and the quality of the section is greatly affected by the doctor's experience and techniques.
  • a first aspect of the embodiments of the present application provides an ultrasound imaging method for a fetus during early pregnancy, the method comprising:
  • the at least one standard slice is displayed.
  • a second aspect of the embodiments of the present application provides an ultrasound imaging method for a fetus in early pregnancy, the method comprising:
  • the at least one standard slice is displayed.
  • a third aspect of the embodiments of the present application provides an ultrasound imaging method for a fetus in early pregnancy, the method comprising:
  • the first-trimester standard section includes at least one of the following: Ventricular horizontal transverse section, biparietal diameter section, head and neck midsagittal section, breast diameter section, abdominal girth section, abdominal wall umbilical cord insertion section, bladder section, longitudinal axis section of the spine, trunk long axis section and top-rump section;
  • the target standard slice is displayed.
  • a fourth aspect of the embodiments of the present application provides an ultrasound imaging method for a fetus in early pregnancy, the method comprising:
  • the target standard slice is displayed.
  • a fifth aspect of the embodiments of the present application provides an ultrasound imaging method for a fetus in early pregnancy, the method comprising:
  • the orientation of the target area of the first trimester fetus is displayed.
  • a sixth aspect of the embodiments of the present application provides an ultrasound imaging system, where the ultrasound imaging system includes:
  • a transmitting/receiving circuit used to excite the ultrasonic probe to transmit ultrasonic waves to the fetus in early pregnancy, and receive the echoes of the ultrasonic waves to obtain ultrasonic echo signals;
  • processor for:
  • a display for displaying the at least one standard slice.
  • a seventh aspect of the embodiments of the present application provides an ultrasound imaging system, the ultrasound imaging system comprising:
  • a transmitting/receiving circuit used to excite the ultrasonic probe to transmit ultrasonic waves to the fetus in early pregnancy, and receive the echoes of the ultrasonic waves to obtain ultrasonic echo signals;
  • processor for:
  • the at least one standard slice obtains three-dimensional ultrasound data of the fetus in the first trimester according to the ultrasound echo signal
  • Regions of at least two distinct early pregnancy target feature structures from the three-dimensional ultrasound data are Regions of at least two distinct early pregnancy target feature structures from the three-dimensional ultrasound data
  • a display for displaying the at least one standard slice.
  • An eighth aspect of the embodiments of the present application provides an ultrasound imaging system, where the ultrasound imaging system includes:
  • a transmitting/receiving circuit used to excite the ultrasonic probe to transmit ultrasonic waves to the fetus in early pregnancy, and receive the echoes of the ultrasonic waves to obtain ultrasonic echo signals;
  • processor for:
  • the first-trimester standard section includes at least one of the following: Ventricular horizontal transverse section, biparietal diameter section, head and neck midsagittal section, breast diameter section, abdominal girth section, abdominal wall umbilical cord insertion section, bladder section, longitudinal axis section of the spine, trunk long axis section and top-rump section;
  • a ninth aspect of the embodiments of the present application provides an ultrasound imaging system, the ultrasound imaging system comprising:
  • a transmitting/receiving circuit used to excite the ultrasonic probe to transmit ultrasonic waves to the fetus in early pregnancy, and receive the echoes of the ultrasonic waves to obtain ultrasonic echo signals;
  • processor for:
  • a tenth aspect of the embodiments of the present application provides an ultrasound imaging system, where the ultrasound imaging system includes:
  • a transmitting/receiving circuit used to excite the ultrasonic probe to transmit ultrasonic waves to the fetus in early pregnancy, and receive the echoes of the ultrasonic waves to obtain ultrasonic echo signals;
  • processor for:
  • a display for displaying the direction of the target area of the first trimester fetus.
  • the standard section of the fetus in the early pregnancy can be automatically extracted according to the three-dimensional ultrasound data collected in a single time, without the need for doctors to manually extract the standard sections one by one, which greatly optimizes the prenatal period.
  • the inspection workflow can effectively improve work efficiency, and can improve the stability of the quality of the obtained standard slices, and promote the promotion and application of early pregnancy structural screening.
  • FIG. 1 shows a schematic block diagram of an ultrasound imaging system according to an embodiment of the present application
  • FIG. 2 shows a schematic flowchart of a method for ultrasound imaging of a fetus in early pregnancy according to an embodiment of the present invention
  • FIG. 3 shows a schematic flow chart of a method for ultrasound imaging of a fetus in the first trimester according to another embodiment of the present invention
  • FIG. 4 shows a schematic flowchart of an ultrasound imaging method for a fetus in early pregnancy according to another embodiment of the present invention
  • FIG. 5 shows a schematic flow chart of a method for ultrasound imaging of a fetus in early pregnancy according to yet another embodiment of the present invention
  • FIG. 6 shows a schematic flowchart of a method for ultrasound imaging of a fetus in the first trimester according to yet another embodiment of the present invention.
  • FIG. 1 shows a schematic structural block diagram of an ultrasound imaging system 100 according to an embodiment of the present application.
  • the ultrasound imaging system 100 includes an ultrasound probe 110 , a transmit/receive circuit 112 , a processor 114 , a display 116 , and a memory 118 . Further, the ultrasound imaging system 100 may further include a beam forming circuit, a transmit/receive selection switch, and the like.
  • the ultrasonic probe 110 includes a plurality of transducer array elements, and the plurality of transducer array elements can be arranged in a row to form a linear array, or arranged in a two-dimensional matrix to form an area array, and the plurality of transducer array elements can also be arranged form a convex array.
  • the transducer is used to transmit ultrasonic waves according to the excitation electrical signal, or convert the received ultrasonic waves into electrical signals, so each array element can be used to realize the mutual conversion of electrical pulse signals and ultrasonic waves, so as to realize the tissue emission to the target area of the measured object.
  • Ultrasound can also be used to receive ultrasound echoes reflected back by tissue.
  • transducers are used for transmitting ultrasonic waves and which transducers are used for receiving ultrasonic waves, or which transducers are used for transmitting ultrasonic waves or receiving ultrasonic waves in time slots through the transmitting sequence and receiving sequence.
  • the transducers participating in ultrasonic emission can be excited by electrical signals at the same time, so as to emit ultrasonic waves at the same time; or, the transducers participating in ultrasonic beam emission can also be excited by several electrical signals with a certain time interval, so as to continuously emit a certain time interval. Ultrasound.
  • the transmit/receive circuit 112 may be connected to the ultrasound probe 110 through a transmit/receive selection switch.
  • the transmit/receive selection switch may also be called a transmit/receive controller, which may include a transmit controller and a receive controller.
  • the transmit controller is used to excite the ultrasound probe 110 to transmit ultrasound to the area where the fetus is located in the first trimester via the transmit circuit; the receive controller uses The ultrasound probe 110 is used to receive ultrasound echoes returned from the region where the fetus is located in the first trimester through the receiving circuit, so as to obtain ultrasound echo data.
  • the transmitting/receiving circuit 112 sends the electrical signal of the ultrasonic echo into the beam forming circuit, and the beam forming circuit performs processing such as focusing delay, weighting and channel summation on the electrical signal, and then sends the processed ultrasonic echo data to the beam forming circuit. into the processor 114.
  • the processor 114 may be implemented by software, hardware, firmware or any combination thereof, and may use circuits, single or multiple application specific integrated circuits (ASICs), single or multiple general-purpose integrated circuits, single or multiple microprocessors, single or multiple programmable logic devices, or any combination of the foregoing circuits and/or devices, or other suitable circuits or devices, thereby enabling the processor 114 to perform the various corresponding steps of the method. Also, the processor 114 may control other components in the ultrasound imaging system 100 to perform desired functions.
  • ASICs application specific integrated circuits
  • microprocessors single or multiple programmable logic devices
  • the processor 114 may control other components in the ultrasound imaging system 100 to perform desired functions.
  • the processor 114 processes the received ultrasound echo data to obtain three-dimensional ultrasound data of the fetus in the first trimester.
  • the ultrasound probe 110 transmits/receives ultrasound in a series of scanning planes, and is integrated by the processor 114 according to its three-dimensional spatial relationship, so as to realize the scanning of the fetus in the three-dimensional space in the first trimester and the reconstruction of the three-dimensional image.
  • the image post-processing steps such as denoising, smoothing, and enhancement are performed by the processor 114, the three-dimensional ultrasound data of the fetus in the first trimester is acquired.
  • the processor 114 may acquire three-dimensional ultrasound data of the whole body of the fetus in the first trimester, or may acquire only the three-dimensional ultrasound data of the head or body of the fetus in the first trimester.
  • the processor 114 is also used to extract standard slices of the first trimester fetus from the three-dimensional ultrasound data. Standard slices obtained by processor 114 may be stored in memory or displayed on display 116 . Also, the processor 114 may render and display the three-dimensional ultrasound data on the display 116 .
  • the display 116 is connected to the processor 114, and the display 116 may be a touch display screen, a liquid crystal display screen, etc.; or the display 116 may be an independent display device such as a liquid crystal display, a television set, etc. independent of the ultrasound imaging system 100; or the display 116 may be Displays of electronic devices such as smartphones, tablets, etc.
  • the number of displays 116 may be one or more.
  • the display 116 may include a main screen and a touch screen, where the main screen is mainly used for displaying ultrasound images, and the touch screen is mainly used for human-computer interaction.
  • Display 116 may display ultrasound images obtained by processor 114 .
  • the display 116 can also provide a graphical interface for the user to perform human-computer interaction while displaying the ultrasonic image, set one or more controlled objects on the graphical interface, and provide the user with a human-computer interaction device to input operating instructions to control these objects.
  • the controlled object so as to perform the corresponding control operation.
  • an icon is displayed on a graphical interface, and the icon can be operated by using a human-computer interaction device to perform a specific function.
  • the ultrasound imaging system 100 may further include other human-computer interaction devices other than the display 116, which are connected to the processor 114.
  • the processor 114 may be connected to the human-computer interaction device through an external input/output port, and the external input/output port may be connected to the human-computer interaction device.
  • the output port can be a wireless communication module, a wired communication module, or a combination of the two.
  • External input/output ports may also be implemented based on USB, bus protocols such as CAN, and/or wired network protocols, and the like.
  • the human-computer interaction device may include an input device for detecting the user's input information, for example, the input information may be a control instruction for the ultrasonic transmission/reception sequence, or a point, line or frame drawn on the ultrasonic image. Manipulate input instructions, or may also include other instruction types.
  • the input device may include one or a combination of a keyboard, a mouse, a scroll wheel, a trackball, a mobile input device (eg, a mobile device with a touch display screen, a cell phone, etc.), a multi-function knob, and the like.
  • the human-computer interaction apparatus may also include an output device such as a printer.
  • the ultrasound imaging system 100 may also include memory for storing instructions executed by the processor, storing received ultrasound echoes, storing ultrasound images, and the like.
  • the memory may be a flash memory card, solid state memory, hard disk, or the like. It may be volatile memory and/or non-volatile memory, removable memory and/or non-removable memory, and the like.
  • the components included in the ultrasound imaging system 100 shown in FIG. 1 are only illustrative, and may include more or less components. This application is not limited to this.
  • FIG. 2 is a schematic flowchart of a method 200 for ultrasound imaging of a fetus in early pregnancy according to an embodiment of the present application.
  • a method 200 for ultrasound imaging of a fetus in early pregnancy includes the following steps:
  • step S210 ultrasonic waves are transmitted to the fetus in the first trimester, and echoes of the ultrasonic waves are received to obtain ultrasonic echo signals.
  • the first trimester fetus generally refers to the fetus under 14 weeks of gestation.
  • the fetus in the first trimester has grown to a certain size, most of the organs have been differentiated and formed, and a considerable part of the characteristic structure has been identified by ultrasound imaging.
  • the ultrasound imaging method provided in this application can automatically extract the standard section that reflects the characteristic structure information, so that in early pregnancy Compared with the ultrasound examination in the second and third trimesters, it can provide pregnant women with relevant pregnancy information as early as possible.
  • ultrasound image acquisition may be performed based on the ultrasound imaging system 100 shown in FIG. 1 .
  • the user moves the ultrasound probe 110 to select an appropriate position and angle, and the transmit circuit in the transmit/receive circuit 120 sends a set of delayed focused pulses to the ultrasound probe 110, and the ultrasound probe 110 transmits ultrasound waves along the 2D scan plane to the fetus in the first trimester.
  • the ultrasonic probe 110 receives the reflected ultrasonic echoes, it converts them into electrical signals, and the beamforming circuit performs corresponding delay and weighted summation processing on the signals obtained from multiple transmissions/receptions to realize beamformation, and then send the echoes back. into the processor 114 for subsequent signal processing.
  • step S210 the function of automatically extracting standard slices may be automatically enabled or manually enabled by the user.
  • a user interface may also be provided to allow the user to manually select standard slices desired to be extracted.
  • this step is optional, and in other embodiments, all standard slices involved in the following may be extracted by default.
  • step S220 three-dimensional ultrasound data of the fetus in the first trimester is obtained based on the ultrasound echo signal.
  • the three-dimensional spatial relationship of the ultrasound echoes obtained by the ultrasound probe 110 transmitted/received in a series of scanning planes can be integrated, so as to realize the scanning of the fetus in the three-dimensional space in the first trimester and the reconstruction of the 3D image.
  • image post-processing steps such as denoising, smoothing, and enhancement, the 3D ultrasound data of the fetus in the first trimester is obtained.
  • the three-dimensional ultrasound data of the whole body of the fetus in the first trimester may be acquired, or only the three-dimensional data of the fetal head region or the fetal body region in the first trimester may be acquired, depending on the standard slice to be extracted.
  • the standard section to be extracted includes a standard section corresponding to the fetal head region in the first trimester
  • the three-dimensional ultrasound data includes at least the three-dimensional ultrasound data of the fetal head region in the first trimester.
  • the standard section to be extracted includes the standard section corresponding to the fetal body region in the first trimester
  • the three-dimensional ultrasound data includes at least the three-dimensional ultrasound data of the fetal body region in the first trimester.
  • the standard slice is a two-dimensional slice containing key information in the three-dimensional ultrasound image, and a physiological characteristic structure with clinical value can be observed through the two-dimensional slice.
  • the subsequently extracted standard slices include at least one of the following standard slices: apical-rump diameter slice (ie, the whole body midsagittal slice), lateral ventricle horizontal slice, biparietal diameter slice (ie thalamus horizontal slice), NT Standard view (i.e.
  • lateral ventricle horizontal transverse view, biparietal diameter view and NT standard view are the standard views corresponding to the fetal head region in the first trimester, and the 3D ultrasound data is required to include at least the 3D ultrasound data of the fetal head region in the first trimester;
  • breast diameter view The abdominal girth section, the abdominal wall umbilical cord insertion section, the bladder section, the longitudinal axis section of the spine, and the long axis section of the trunk are the standard sections corresponding to the fetal body area in the first trimester, and the 3D ultrasound data is required to include at least the 3D ultrasound data of the fetal body area in the first trimester;
  • the top-rump diameter section is the standard section corresponding to the whole body region of the fetus in the first trimester, so the 3D ultrasound data is required to include the 3D ultrasound data of the whole body region of the fetus in the first trimester.
  • a visualization algorithm may be used to render the three-dimensional ultrasound data, so as to obtain a three-dimensional ultrasound image and display it by using a display device.
  • the rendering includes, for example, a surface rendering method or a volume rendering method, which is not limited in this embodiment of the present invention.
  • step S230 the direction of the target region of the fetus in the first trimester is determined according to the three-dimensional ultrasound data.
  • determining the direction of the target region of the fetus in the first trimester may include determining the direction of the head region of the fetus in the first trimester, determining the direction of the body region of the fetus in the first trimester, or determining the direction of the head region and the body region of the fetus in the first trimester .
  • the standard section to be extracted includes the standard section corresponding to the fetal head region in the first trimester
  • the corresponding direction of the fetal head region in the first trimester is determined according to the three-dimensional ultrasound data of the fetal head region in the first trimester.
  • the orientation of the head region can also be used to extract standard views corresponding to the whole body region of the fetus in the first trimester, such as the top-rump diameter view.
  • the standard section to be extracted includes the standard section corresponding to the fetal body region in the early pregnancy, the direction of the fetal body region in the early pregnancy is determined according to the three-dimensional ultrasound data of the fetal body region in the early pregnancy.
  • the direction of the target area includes the first direction of the head area of the fetus in early pregnancy, wherein the first direction is the left and right direction of the head area. Since the standard section corresponding to the head region is generally parallel or perpendicular to the left and right directions of the head, and the determination of the left and right directions of the head is relatively more accurate, determining the direction of the head region of the fetus in the first trimester may at least include determining the fetal head in the first trimester. the left and right directions of the top area. However, in other embodiments, the determined direction of the head region of the fetus in the first trimester may also be other directions, such as the up and down direction of the head region.
  • the orientation of the head region of the fetus in the first trimester may be determined by a trained machine learning model.
  • the database includes a large amount of ultrasound data of the first-trimester fetus.
  • Each ultrasound data marks the direction of the fetal head region in the first-trimester, or the position and size of the head region. direction.
  • the position of the head area can be the position of the region of interest frame (ROI) of the head, such as the position of each vertex of the ROI frame or the position of the center point of the ROI frame and the size of the frame, or the position of the head area can also be the position of the fetus
  • ROI region of interest frame
  • the specific area range of the head area; the direction of the head area can be the left and right directions of the head area.
  • the ultrasound data in the database is used as a training sample to train a traditional machine learning model or a deep learning model to predict the position and orientation of the fetal head region in the first trimester.
  • an optimal mapping function from the ultrasound data to the position and direction of the head region can be learned according to the ultrasound data in the fetal ultrasound database in the early pregnancy, so that the position and direction of the head region obtained by mapping the early pregnancy data in the database are the same as the actual calibrated ones.
  • the error between the position and orientation of the head region is minimal. Executing this optimal mapping function on the three-dimensional ultrasound data obtained in step S220 can obtain the prediction result of the position and direction of the head region.
  • Traditional machine learning methods include SVM support vector machine, logistic regression method, least square method, etc.
  • image features such as Sift features, gradient features, texture features such as LBP, PCA, LDA, Harr features, HOG and LOG features
  • image features such as Sift features, gradient features, texture features such as LBP, PCA, LDA, Harr features, HOG and LOG features
  • Optimal mapping function between features and head region position and orientation For the deep learning model, an end-to-end neural network can be trained as the optimal mapping function to directly construct the mapping relationship between the ultrasound data of the fetus in the first trimester and the direction of the head region.
  • the first direction of the head region ie, the left-right direction
  • the first direction of the head region may be determined according to the symmetry of the head. Specifically, firstly, the head region of the fetus in the early pregnancy is detected in the 3D ultrasound data obtained in step S220, and then the first direction of the fetal head in the early pregnancy, that is, the left and right directions, is determined according to the symmetry of the head region of the fetus in the early pregnancy.
  • a machine learning method can be used to detect the head of a fetus in the first trimester in the three-dimensional ultrasound data.
  • a first-trimester fetal ultrasound database is first established.
  • the 3D ultrasound data of each first-trimester fetus in the database is marked with the position of the head region, such as the position of the head region of interest (ROI) or the head.
  • traditional machine learning methods or deep learning methods are used to learn an optimal mapping function, which is used to obtain the position or head region of the region of interest box (ROI) in the head region from the 3D ultrasound data of the fetus in the first trimester.
  • the specific area of the head can be used to detect or segment the fetal head in the first trimester.
  • Hough transform In the skull halo detection method, Hough transform, RANSAC and other methods can be used to detect the circle or ellipse, sphere or ellipse with the brightest brightness, the strongest gradient or the greatest brightness and gradient weighting in the three-dimensional ultrasound data of the fetus in the first trimester A spherical surface as a result of detection of the fetal head region in the first trimester.
  • the left and right directions can be determined according to the symmetry of the head region.
  • the left hemisphere area and the right hemisphere area of the head area can be determined according to the symmetry of the head of the fetus in the first trimester, and the normal direction of the interface between the left hemisphere area and the right hemisphere area of the head area is determined as: Left and right directions of the head area.
  • the left hemisphere region and the right hemisphere region with optimal symmetry can be obtained by preset search, gradient update, reinforcement learning and other methods.
  • the preset search method a plurality of candidate interfaces of the left hemisphere region and the right hemisphere region of the head are first generated according to the established rules, and the two sides of the interface are the candidate left hemisphere regions and right hemisphere regions; The left and right hemisphere regions with optimal symmetry are found in it.
  • a fixed section is firstly taken as the initial interface between the left hemisphere region and the right hemisphere region, and then the left hemisphere of the head is continuously iterated according to the gradient direction of the symmetry function or the iteration direction obtained by the reinforcement learning method. Region and right hemisphere region interface until the symmetry of the left and right hemisphere regions is optimal.
  • the left hemisphere can be classified according to various suitable indicators such as the absolute error, square error, correlation function, correlation coefficient, and the output results of deep learning networks such as twin network and CNN, etc.
  • the symmetry of the region and the right hemisphere region is evaluated in order to find the left and right hemisphere regions with the best symmetry.
  • the direction of the head region may also be determined by means of mid-sagittal plane detection. Specifically, the mid-sagittal plane of the fetal head region in the first trimester is extracted from the three-dimensional ultrasound data acquired in step S220. , and the normal direction of the midsagittal plane is determined as the left-right direction of the head region.
  • methods such as preset search, gradient update, and reinforcement learning can be used to extract the midsagittal plane from the three-dimensional ultrasound data of the fetus in the first trimester.
  • a database of fetal slice images in the first trimester is first established.
  • Each slice image in the database is a slice in the three-dimensional data of the fetus in the first trimester, and whether the slice is a midsagittal plane is marked.
  • the traditional machine learning method or deep learning method can be used to learn an optimal mapping function from the fetal section image in the first trimester to the midsagittal plane, which is used to determine whether each section is the midsagittal plane. , or the probability that each slice belongs to the midsagittal plane.
  • multiple candidate midsagittal planes are first generated according to a predetermined rule, and then one candidate midsagittal plane with the highest probability of belonging to the midsagittal plane is used as the midsagittal plane detection result.
  • the gradient update method a fixed section is first taken as the starting position of the midsagittal plane, and then the section is iteratively updated according to the gradient direction of the probability function of the midsagittal plane until the probability value belonging to the midsagittal plane reaches the maximum , or the probability value exceeds a predetermined threshold, that is, the detection result of the midsagittal plane is obtained.
  • a reward function is determined for each slice and the transformation (translation, rotation, etc.) of the fetal 3D ultrasound data in the first trimester, and the reward function can be related to the difference between the current slice and the actual midsagittal plane.
  • Image similarity, image error, position and orientation deviation, etc., and the reward function of slice transformation is the difference between the slice reward functions before and after transformation.
  • the direction of the target area includes at least one of the following: a second direction of the body area, a third direction of the body area, and a fourth direction of the body area, wherein the second direction of the body area, the third direction and the The fourth direction is the up-down direction, the front-rear direction, and the left-right direction of the body region, respectively, and any two of the second direction, the third direction, and the fourth direction are perpendicular to each other.
  • Determining the body orientation of the fetus in the first trimester includes but is not limited to the following implementations:
  • the body orientation may be determined by a trained machine learning model.
  • This method is similar to the above-mentioned determination of the head direction through the trained machine learning model. Specifically, an ultrasound database of the fetus in the first trimester is first established, and each ultrasound data in the database marks the body region of the fetus in the first trimester. Orientation, or the location and orientation of areas of the fetal body during early pregnancy. After the database is constructed, a traditional machine learning model or a deep learning model is trained to predict the orientation of the body region, or the position and orientation of the body region, from the three-dimensional ultrasound data of the fetus in the first trimester to be identified.
  • an optimal mapping function from the three-dimensional data of the fetus in the early pregnancy to the direction of the body region or the position and direction of the body is learned, so that the optimal mapping function is used to map the ultrasound data in the database to the body region direction or body region position obtained by mapping
  • the error between the direction and the true value of the actual calibration is minimal.
  • the above-mentioned optimal mapping function can be executed to obtain the direction of the body region or the prediction result of the direction and position of the body region.
  • the direction of the body region can be determined by means of spine detection. Specifically, the spine region of the fetus in the early pregnancy is detected in the three-dimensional ultrasound data obtained in step S220, and the direction of the spine region is determined in the early pregnancy. Orientation of the fetal body region. Since the spine region is more obvious in the three-dimensional ultrasound image, the orientation of the body region of the fetus in the first trimester can be accurately determined according to the position of the spine region.
  • Methods to detect or segment fetal spine regions in the first trimester include machine learning methods or traditional image processing methods.
  • the machine learning method is similar to the above machine learning method, that is, a fetal ultrasound database in early pregnancy is established in advance, and each 3D ultrasound data in the database is marked with the position corresponding to the fetal spine region in the early pregnancy, such as the region of interest box of the spine region ( ROI) or the specific area of the spine region, and then use traditional machine learning methods or deep learning methods to learn an optimal mapping function to obtain the spine region Region of Interest Box (ROI) or The specific area range of the spine area is used to detect or segment the fetal spine area in the first trimester.
  • ROI region of interest box of the spine region
  • ROI region of interest box of the spine region
  • ROI region of interest box of the spine region
  • specific area range of the spine area is used to detect or segment the fetal spine area in the first trimester.
  • Hough transform In traditional image processing methods, Hough transform, RANSAC and other methods can be used to detect the brightest straight line or arc in the 3D ultrasound data of the fetus in the first trimester, as the detection result of the fetal spine region in the first trimester.
  • the up-down direction and front-rear direction of the fetal body region in the first trimester can be directly determined through the detection results of the spine region.
  • determining the up and down direction of the fetal body region in early pregnancy according to the direction of the spinal region includes: determining a straight line close to the spinal region, and determining the direction of the straight line as the second direction of the fetal body region in early pregnancy, that is, the up and down direction.
  • a method such as least squares can be used to fit the straight line closest to the spine region, or, the brightest straight line obtained by using Hough transform, RANSAC and other methods in the above-mentioned traditional image processing methods can be directly used as the most bright line.
  • a straight line close to the spine area is determining a straight line close to the spinal region, and determining the direction of the straight line as the second direction of the fetal body region in early pregnancy, that is, the up and down direction.
  • a method such as least squares can be used to fit the straight line closest to the
  • the third direction of the fetal body region in the early pregnancy can be further determined according to the up-down direction.
  • one or more cross-sections of the body that are perpendicular to the upper and lower directions determined above can be extracted from the three-dimensional ultrasound data of the fetus in the first trimester, and machine learning or traditional image processing methods are used to detect the center point of the body and the spine in the cross-section of the body.
  • the position of the center of the body and the position of the spine is determined as the anterior-posterior direction of the fetal body region in the first trimester. On this line, the position of the spine is close to the rear of the cross-section of the body, and the position of the center of the body is close to the front of the cross-section of the body.
  • the front-rear direction of the body region can also be determined directly according to the detection result of the spine region, without the need to determine the front-rear direction based on the up-down direction of the body region.
  • a curve close to the spine can be determined, and the convex direction of the curve can be determined as the front-rear direction of the fetal body region in the early pregnancy.
  • a method such as least squares can be used to fit an arc closest to the spine as the above curve, or the arc with the brightest brightness obtained in the above traditional image processing method can be directly obtained as the above curve.
  • the direction of the body region can also be determined according to the shape characteristics of the fetal body region in early pregnancy, that is, the body region of the fetus in early pregnancy can be detected from the three-dimensional ultrasound data, and the body direction can be determined according to the shape of the body region. Since the shape features of the fetal body region in the up and down direction in the early pregnancy are more obvious, and the shape difference between the front and rear directions and the left and right directions is small, this method is mainly used to determine the up and down direction of the fetal body region in the early pregnancy.
  • a machine learning or image processing method may be used to determine the body region of the fetus in the first trimester from the three-dimensional ultrasound data acquired in step S220, where the determined body region may be a region of interest frame (ROI frame) that surrounds the body of the fetus in the first trimester. ), or to segment specific regions of the fetal body during the first trimester. Then, the long axis of the body region is determined according to the shape of the body region according to the detection or segmentation result of the fetal body region in the early pregnancy, and the long axis direction is determined as the up-down direction of the fetal body region in the early pregnancy.
  • a principal component analysis (PCA) method can be used to determine the long axis of the body region, or to detect the two most distant points in the above body region, and the line connecting the two points is the long axis of the body region.
  • PCA principal component analysis
  • the up-down direction and the up-down direction and the front-to-back direction of the fetal body region in the first trimester can be determined.
  • the direction perpendicular to the anterior-posterior direction is determined as the left-right direction of the fetal body in the first trimester.
  • the left and right directions of the fetal body region in the early pregnancy can also be determined according to some specific feature structures of the fetus in the early pregnancy.
  • the direction of the line connecting the regions of the symmetrical features is determined as the left-right direction of the fetal body region in the first trimester.
  • the symmetrical characteristic structure is, for example, a symmetrical characteristic structure such as two kidneys, two lungs, and left and right ribs.
  • the left and right atriums of the fetus in early pregnancy can also be detected.
  • the direction of the connection between the left and right atrium is generally at a 45° angle with the left and right direction of the fetal body area in the early pregnancy. According to this characteristic, the left and right direction of the fetal body area in the early pregnancy can also be determined.
  • step S230 may be performed automatically after acquiring the three-dimensional ultrasound data, or may be performed according to a received user instruction.
  • the user can activate the function of automatically extracting standard slices by triggering a key for extracting standard slices, and the key can be a virtual key set on the user interface of the display, or a physical key.
  • step S230 is started.
  • the user may also activate the function of automatically extracting standard slices before starting to acquire the ultrasound data, and after obtaining the three-dimensional ultrasound data, step S230 is automatically performed.
  • the standard slice to be extracted may also be determined according to the received user input.
  • the names of ten standard cut planes such as apical-rump diameter cut surface and double-apex diameter cut surface can be displayed on the user interface, and the standard cut surface to be extracted is determined according to the user's selection.
  • the options for the standard section of the head region, the standard section of the body region, and the standard section of the whole body region may be displayed on the user interface, and the standard section of the head region, the standard section of the body region or the standard section of the whole body region may be extracted according to the user's selection.
  • the direction of each target area can be more accurately determined according to the three-dimensional ultrasound data of the fetus in the first trimester, the direction of the target area is first determined, and then the standard slice is further extracted according to the direction of the target area, which can improve the quality of the extracted standard slice.
  • step S240 at least one standard slice corresponding to the target area is extracted from the three-dimensional ultrasound data according to the target direction.
  • the direction of the target area determined in step S230 mainly includes the direction of the head area and the direction of the body area
  • the three-dimensional ultrasound including the fetal head area in the first trimester can be A standard slice of the head region is determined from the data, or, based on the orientation of the body region, in the three-dimensional ultrasound data including the fetal body region of the first trimester.
  • the standard view of the head region includes at least one of the lateral ventricle horizontal transverse view, the biparietal diameter view and the midsagittal view of the head and neck
  • the standard view of the body region includes the breast diameter view, the abdominal circumference view, and the abdominal wall umbilical cord insertion view.
  • the standard view can also include standard views of the whole body region of the fetus in the first trimester, such as the top-rump plane (whole body mid-sagittal plane). direction to determine.
  • extracting at least one standard slice according to the direction of the target area includes: determining a plurality of candidate slices in the three-dimensional ultrasound data of the fetus in the first trimester according to the direction of the target area determined in step S230, and selecting the slices in the plurality of candidate slices Pick out at least one standard cut.
  • the optimal head area is determined from the multiple candidate sections of the head area.
  • Standard cut For example, in the detection of the NT section and the top-rump diameter section, the interface or midsagittal plane between the left hemisphere region and the right hemisphere region of the head obtained in step S230 can be directly used as the NT section or the top-rump section.
  • a group of parallel multiple candidate slice planes perpendicular to the direction may be generated, or a group of multiple candidate cut planes approximately perpendicular to the direction may be generated within a certain angle range Section, and then select the final NT section or top-rump diameter section from multiple candidate sections.
  • fetal head cross-sections such as lateral ventricle cross-sections, biparietal cross-sections, cerebellum cross-sections, etc.
  • a set of parallel Or multiple candidate slices approximately parallel to this direction, and then select the final lateral ventricle transverse slice, biparietal diameter slice or cerebellum transverse slice among the multiple candidate slices.
  • the standard section of the body area similar to the standard section of the head area, first, according to the direction of the body area of the fetus in the first trimester, a set of candidate sections of the body area are generated, and then the most candidate section of the body area is determined. Excellent body area standard cut.
  • the cross-section of the fetal body region in the first trimester such as the four-chamber heart view, the gastric bubble view, the abdominal wall umbilical cord insertion view, the bladder view, etc.
  • the upper and lower directions of the fetal body region in the first trimester determined in step S230 Generate a set of multiple parallel tangent planes perpendicular to the up-down direction as candidate tangent planes, or generate a set of multiple tangent planes that are approximately perpendicular to the up-down direction within a certain angle range as candidate tangent planes, and then select criteria from the multiple candidate tangent planes cross section.
  • a set of parallel to the direction can be generated according to the left-right direction or the up-down direction of the body region determined in step S230.
  • a set of candidate cut planes parallel to the direction may be generated according to the up-down direction or the anterior-posterior direction of the body region determined in step S230, or a set of candidate sections may be generated within a certain angle range.
  • the candidate slices approximately parallel to the direction are grouped, and then the standard midsagittal plane is selected from the multiple candidate slices.
  • the trained machine learning model can be used to determine the probability of each candidate slice corresponding to the target area as a standard slice, and the candidate slice whose probability satisfies the first threshold is determined as the target area. Corresponds to at least one standard slice. Exemplarily, if there are at least two candidate slices whose probability satisfies the first threshold, then the final selected standard slice may be determined according to the received selection operation, that is, at least two candidate slices whose probability satisfies the first threshold is determined by the user. Choose the best standard cut.
  • the machine learning model may be a traditional machine learning model or a deep learning model.
  • a trained machine learning model can also be used to determine the probability of each candidate slice corresponding to the target area as a standard slice, and the candidate slice with the highest probability can be determined as a standard slice corresponding to the target area. , that is, the final standard section is directly determined by the system, thereby simplifying the operation process.
  • selecting a standard slice from the multiple candidate slices includes: detecting an early pregnancy target feature structure corresponding to at least one standard slice corresponding to the target area on the multiple candidate slices corresponding to the target area, The candidate slice whose probability of existence of the target eigenstructure satisfies the second threshold or the candidate slice with the highest probability of existence of the target eigenstructure of early pregnancy is determined as the standard slice.
  • the characteristic structures of the early pregnancy target corresponding to the top-rump diameter section include nasal bone, genital ridge, etc.; the characteristic structures of the early pregnancy target corresponding to the horizontal transverse section of the lateral ventricle include falx cerebellum, lateral ventricle, choroid plexus, etc.; the first pregnancy target characteristics corresponding to the biparietal diameter section
  • the structures include thalamus, skull halo, etc.; the early pregnancy target characteristic structures corresponding to the midsagittal plane of the head and neck include the posterior cervical translucency, nasal bone, etc.; the early pregnancy target characteristic structures corresponding to the breast diameter section include the four-chambered heart, etc.;
  • the characteristic structures of the early pregnancy target include gastric bubbles, etc.; the characteristic structures of the early pregnancy target corresponding to the umbilical cord insertion section of the abdominal wall include the umbilical cord insertion section, etc.; the corresponding early pregnancy target characteristic structures of the bladder section include legs or bladder, etc.; the early pregnancy target corresponding to the longitudinal
  • the standard slice of the head area after generating a set of multiple candidate slices parallel or approximately parallel to the direction according to the left and right directions of the head area, you can detect and detect the difference between the head area and the head area on the multiple candidate slices respectively.
  • the target characteristic structure of early pregnancy corresponding to the standard section such as the lateral ventricle or choroid plexus corresponding to the transverse section of the lateral ventricle, the thalamus corresponding to the biparietal diameter section, and the cerebellum corresponding to the transverse section of the cerebellum, etc.
  • the candidate slices are used as the detection results of the corresponding standard slices of the head.
  • the characteristic structure of the early pregnancy target corresponding to the standard slice to be determined may be detected in the three-dimensional ultrasound data acquired in step S220, and then In combination with the direction of the target area determined above, a cut plane that at least partially coincides with the target feature structure of early pregnancy and whose included angle meets the preset requirements with the direction of the target area is determined as at least one criterion corresponding to the target area section. Further, a cut plane that basically coincides with the feature structure of the early pregnancy target and has an included angle of 0 degrees with the direction of the determined target area can be determined as a standard cut plane corresponding to the target area. Alternatively, a cut plane that substantially coincides with the feature structure of the early pregnancy target and has an included angle of 90 degrees with the direction of the determined target area may be determined as a standard cut plane corresponding to the target area.
  • the characteristic structure of the early pregnancy target corresponding to the standard section of the head to be determined can be detected, and then combined with the direction of the head area of the fetus in the first trimester, the corresponding standard section of the head area can be determined .
  • the first-pregnancy target feature structures corresponding to the horizontal cross section of the lateral ventricle include falx cerebral, lateral ventricle, and choroid plexus
  • the first-pregnancy target feature structures corresponding to the biparietal radial section include the thalamus and skull halo
  • the corresponding early pregnancy target feature structures include the translucent layer at the back of the neck and the nasal bone.
  • the standard section of the whole body area can also be determined according to the direction of the head area and the early pregnancy target characteristic structure of the head area.
  • the standard section is the top-rump diameter section
  • the corresponding early pregnancy target characteristic structure includes nasal bone, reproductive crest.
  • the nasal bone, NT (rear cervical translucency), IT (intracranial translucency), and posterior fossa cisterns can be detected in 3D ultrasound data.
  • One or more feature structures and then determine a cut plane that at least partially coincides with the above feature structure and is perpendicular to the left-right direction of the head region, as the detection result of the NT standard cut plane.
  • one or more feature structures in the nasal bone, NT (transparent layer of the back of the neck), bladder, and genital ridge can be detected, and then it is determined in the three-dimensional ultrasound data that at least part of the above feature structures coincide with the head. Sections perpendicular to the left and right directions of the upper region are used as the detection results of the top-rump diameter section.
  • the lateral ventricle and/or choroid plexus corresponding to the lateral ventricle cross section, the thalamus and/or cerebral peduncles corresponding to the biparietal section, and the cerebellum can be detected in the 3D ultrasound data.
  • the target feature structure of early pregnancy corresponding to the standard section of the body area to be determined can be detected, and then combined with the direction of the body area of the fetus in the first trimester, the standard section of the corresponding body area can be determined. .
  • the position of the spine region or its fitted arc can be used to determine the one closest to the spine or
  • the fitted arc is parallel or approximately parallel to the cut plane in the up-down direction or the anterior-posterior direction of the body area (that is, the angle with the up-down direction or the anterior-posterior direction is 0°), and this section is taken as the mid-sagittal of the fetal body area in the first trimester surface detection results.
  • the heart corresponding to the four-chamber view, the gastric vesicle corresponding to the gastric bubble section, the umbilical cord and its abdominal wall insertion section corresponding to the abdominal wall umbilical cord insertion section, and the abdominal wall corresponding to the section can be detected.
  • the bladder or umbilical artery corresponding to the bladder section determine one or more sections so that they respectively at least partially overlap with the above first-pregnancy target feature structures, and are perpendicular or approximately perpendicular to the up-down direction of the fetal body region in the first-trimester determined in step S230 , and use this section as the detection result of the four-chamber heart section, the gastric bubble section, the abdominal wall umbilical cord insertion section or the bladder section.
  • the kidneys, gastric vesicles or spine corresponding to the coronal section of the kidneys, the spine or ribs corresponding to the coronal section of the spine can be detected, and then one or more sections can be determined so that They are respectively coincident with the above feature structures of the early pregnancy target and are parallel to the left and right or up and down directions of the fetal body region in the early pregnancy determined in step S230, and the section is used as the detection result of the coronal section of the kidneys or the coronal section of the spine.
  • other standard slices may be determined according to the determined one or more standard slices and the location of the early pregnancy target feature in the determined standard slices.
  • the interface between the left and right hemisphere regions of the head obtained above, the midsagittal plane, the NT standard plane, or the top-rump diameter plane can be detected.
  • the position of the gastric vesicle and the spine can be determined in the detected gastric vesicle section, and then a section can be determined to pass through the stomach.
  • the vesicle or spine is perpendicular to the section of the gastric vesicle, and this section is used as the detection result of the coronal section of both kidneys.
  • the position of the spine and ribs can be determined in one or more transverse sections of the body area that have been detected, and then a section can be determined so that it passes through the spine or ribs and is perpendicular to the transverse plane of the body, and This section was taken as the detection result of the coronal section of the spine.
  • Step S250 displaying the at least one standard slice.
  • the displayed standard cut planes may be all standard cut planes extracted in step S240, or may be part of the standard cut planes. Exemplarily, when the standard slices extracted in step S240 are displayed, some of the standard slices may be displayed according to the received user instruction. The corresponding standard slice is displayed.
  • step S230 or step S240 the feature structure of the early pregnancy target is detected in the three-dimensional ultrasound data
  • the feature structure of the early pregnancy target detected from the three-dimensional ultrasound data may also be displayed for the user to perform Control analysis.
  • the target direction determined in step S230 may also be displayed, so that the user can view the associated standard slice in combination with the target direction.
  • the target direction may be displayed at the same time as the three-dimensional ultrasound image is displayed.
  • the left and right directions of the fetal head region in the first trimester, or the up-down, front-rear, and left-right directions of the fetal body region in the first trimester can be displayed by graphical markers such as arrows.
  • the name of each determined standard slice may also be displayed, so that the user can intuitively determine the type of the standard slice, or select the standard slice to be viewed according to the name of the standard slice.
  • the names of the standard slices can be displayed synchronously with the standard slices, or, the names of each standard slice can be displayed on the display interface first, and when the user selects the name of the standard slice, the corresponding standard slice is displayed.
  • the ultrasound imaging method 200 of the fetus in the first trimester of the present application can automatically determine at least one standard slice corresponding to the target area according to the direction of the target area of the fetus in the first trimester, without requiring the doctor to manually extract the standard slices one by one, greatly reducing the need for It optimizes the workflow of prenatal examination, effectively improves work efficiency, and can improve the stability of the quality of the obtained standard slices, and promote the popularization and application of early pregnancy structural screening.
  • the ultrasound imaging system 100 provided by the embodiments of the present application may be used to implement the above-mentioned ultrasound imaging method 200 of a fetus in the first trimester.
  • the ultrasound imaging system 100 may include the ultrasound probe 110 , the transmit/receive circuit 112 , the processor 114 , and the display 116 , and the relevant descriptions of the various components can be referred to above.
  • the transmitting/receiving circuit 112 is used to excite the ultrasonic probe 110 to transmit ultrasonic waves to the fetus in early pregnancy, and receive the echoes of the ultrasonic waves to obtain ultrasonic echo signals; the processor 114, for: obtaining three-dimensional ultrasound data of the fetus in the first trimester based on the ultrasonic echo signal; determining the direction of the target area of the fetus in the first trimester according to the three-dimensional ultrasound data; according to the direction of the target area, from the At least one standard slice corresponding to the target area is extracted from the three-dimensional ultrasound data; the display 116 is used to display the at least one standard slice.
  • the target area includes at least one of: a head area, a body area, and a full body area.
  • the target area includes a head area
  • the direction of the target area includes a first direction of the head area, wherein the first direction is a left-right direction of the head area.
  • the target area includes a body area
  • the direction of the target area includes at least one of: a second direction, a third direction, and a fourth direction of the body area, wherein the second, third, and fourth directions
  • the directions are the up-down direction, the front-rear direction, and the left-right direction of the body region in the three-dimensional ultrasound data, respectively, and any two of the second, third, and fourth directions are perpendicular to each other.
  • extracting at least one standard slice corresponding to the target area from the three-dimensional ultrasound data according to the direction of the target area includes: determining a plurality of candidate slices corresponding to the target area in the three-dimensional ultrasound data according to the direction of the target area; At least one standard slice corresponding to the target area is selected from the multiple candidate slices corresponding to the target area.
  • extracting at least one standard slice corresponding to the target area from the three-dimensional ultrasound data according to the direction of the target area including: detecting the early pregnancy target feature structure corresponding to the standard slice corresponding to the target area in the three-dimensional ultrasound data; The cut plane whose characteristic structure of the early pregnancy target is at least partially coincident and whose included angle with the direction of the target area satisfies a preset requirement is determined as at least one standard cut plane corresponding to the target area.
  • the display 116 is also used to display the first pregnancy target feature.
  • Display 116 may also be used to display the direction of the determined target area.
  • Display 116 may also be used to display the name of the determined standard slice.
  • the ultrasound imaging system 100 of the embodiment of the present application can automatically determine the standard section of the fetus in the first trimester, which improves the work efficiency and the quality of the standard section.
  • FIG. 3 is a schematic flowchart of a method 300 for ultrasound imaging of a fetus in early pregnancy according to an embodiment of the present application.
  • the ultrasound imaging method 300 of the fetus in the first trimester includes the following steps:
  • step S310 transmitting ultrasonic waves to the fetus in early pregnancy, and receiving echoes of the ultrasonic waves to obtain ultrasonic echo signals;
  • step S320 obtain three-dimensional ultrasound data of the fetus in the first trimester according to the ultrasound echo signal
  • step S330 detecting at least two different regions of early pregnancy target feature structures from the three-dimensional ultrasound data
  • step S340 at least one section that at least partially overlaps with each of the regions of the at least two different early pregnancy target feature structures is determined as at least one criterion for the first trimester fetus section;
  • step S350 the at least one standard slice is displayed.
  • Steps S310 , S320 and S350 in the ultrasound imaging method 300 of the first trimester fetus according to the embodiment of the present application are substantially similar to steps S210 , S220 and S250 in the ultrasound imaging method 200 described with reference to FIG. 2 , for the sake of brevity , the same details will not be repeated here, and the following mainly describes in detail the manner in which the standard slice is determined according to the three-dimensional ultrasound data in steps S330 and S340.
  • the extracted standard slices include at least one of the standard slices of the head region, the standard slices of the body region, or the standard slices of the whole body region of the first trimester fetus.
  • the standard section includes the standard section of the head region
  • the three-dimensional ultrasound data acquired in step S320 at least includes the three-dimensional ultrasound data of the fetal head region in the first trimester.
  • the standard section includes the standard section of the body region
  • the three-dimensional ultrasound data acquired in step S320 includes at least the three-dimensional ultrasound data of the fetal body region in the first trimester.
  • the standard slice includes the standard slice of the whole body area
  • the three-dimensional ultrasound data acquired in step S320 includes the three-dimensional ultrasound data of the whole body area of the fetus in the first trimester.
  • the standard view may include at least one of the following standard views: parietal rump diameter view, lateral ventricle horizontal transverse view, biparietal diameter view, NT standard view, breast diameter view, abdominal circumference view, abdominal wall umbilical cord insertion view, bladder view , longitudinal section of the spine, and long-axis section of the trunk.
  • the standard sections of the fetal head area in the first trimester include lateral ventricle horizontal cross section, biparietal diameter section and NT standard section, and the 3D ultrasound data is required to include at least the 3D ultrasound data of the fetal head area in the first trimester;
  • Standard views include breast diameter view, abdominal circumference view, abdominal wall umbilical cord insertion view, bladder view, spine longitudinal axis view and trunk long axis view.
  • the standard view of the region includes the top-rump diameter view, and the three-dimensional ultrasound data is required to include the three-dimensional ultrasound data of the whole body region of the fetus during the first trimester.
  • any suitable image detection or segmentation method can be used to determine the position of the early pregnancy target feature structure.
  • a traditional machine learning method or a deep learning method can be used to train a machine learning model for the feature structure corresponding to each standard slice, so as to determine the position of the feature structure.
  • a first-trimester fetal ultrasound database is established in advance, and each 3D ultrasound data in the database marks the position of the first-trimester fetal target feature structure, such as its region of interest (ROI) or specific area range, and then Use traditional machine learning methods or deep learning methods to learn an optimal mapping function, which is used to obtain the region of interest (ROI) or specific area range of the feature structure of the early pregnancy target from the 3D ultrasound data of the fetus in early pregnancy, so as to achieve the early pregnancy target. Detection or segmentation of feature structures.
  • step S340 for the standard slice to be extracted, the type of the standard slice is first obtained, the characteristic structure of the early pregnancy target corresponding to the type of the standard slice is determined, and at least two types of the standard slice are detected from the three-dimensional ultrasound data. Regions of different early pregnancy target characteristic structures. Afterwards, a cut plane that at least partially coincides with the at least two different early pregnancy target feature structures is determined as the detection result of the standard cut plane. Specifically, a cut plane substantially coincident with each region of the early pregnancy target feature structure in the at least two different early pregnancy target feature structure regions may be determined as a standard cut plane. In order to make the obtained standard slices more accurate, slices at least partially coincident with three or more different early pregnancy target feature structures may be determined as the detection results of the standard slices.
  • the corresponding early pregnancy target feature structures include the nasal bone and the genital ridge; when the standard section is the lateral ventricle horizontal transverse section, the corresponding early pregnancy target feature structures include the falx cerebellum, the lateral ventricle and the genital ridge.
  • the corresponding early pregnancy target feature structures include thalamus and skull halo; when the first pregnancy target feature structures include the midsagittal plane of the head and neck, the corresponding early pregnancy target feature structures include neck transparency layer and nasal bone; when the standard section is the DBH section, the corresponding early pregnancy target feature structures include the heart, ribs and spinal bones; when the standard section is the abdominal girth section, the corresponding early pregnancy target feature structures include gastric vesicles, spine bones and liver; When the standard section is the umbilical cord insertion section of the abdominal wall, the corresponding early pregnancy target feature structures include the umbilical cord insertion section, umbilical cord, spine and anterior abdominal wall; when the standard section is the bladder section, the corresponding early pregnancy target feature structures include legs and bladder; When the standard view is the longitudinal axis of the spine, the corresponding early pregnancy target feature structures include spine and skin margin; when the standard view is the trunk long axis view, the corresponding early pregnancy target feature structures
  • the displayed standard slices may be part or all of the standard slices extracted in step S340.
  • the names of the standard slices can also be displayed.
  • the first-trimester target feature structures detected from the 3D ultrasound data can also be displayed.
  • the first-trimester target feature structures can be displayed in the 3D ultrasound image of the fetus in the first trimester. , for example, to display a ROI box surrounding the early pregnancy target feature structure or to display the outline of the early pregnancy target feature structure.
  • the ultrasound imaging system 100 provided by the embodiment of the present application can be used to implement the above-mentioned ultrasound imaging method 300 of a fetus in the first trimester.
  • the ultrasound imaging system 100 may include the ultrasound probe 110 , the transmit/receive circuit 112 , the processor 114 , and the display 116 , and the relevant descriptions of the various components can be referred to above.
  • the transmitting/receiving circuit 112 is used to excite the ultrasonic probe 110 to transmit ultrasonic waves to the fetus in early pregnancy, and receive the echoes of the ultrasonic waves to obtain ultrasonic echo signals;
  • the processor 114 uses In: obtaining three-dimensional ultrasound data of the fetus in the first trimester based on the ultrasound echo signal; detecting regions of at least two early-pregnancy target feature structures from the three-dimensional ultrasound data; determining early-pregnancy target features different from the at least two early-pregnancy target features Each of the first-pregnancy target feature structures in the structure at least partially overlaps at least one slice plane to serve as at least one standard slice plane of the first-trimester fetus; the display 116 is used to display the at least one standard slice plane.
  • the ultrasound imaging method 300 and the ultrasound imaging system of the fetus in the first trimester of the present application can automatically determine the standard section of the fetus in the first trimester according to the target feature structure of the fetus in the first trimester, without the need for doctors to manually extract the standard sections one by one, which greatly optimizes the prenatal period.
  • the workflow of inspection can effectively improve work efficiency, and can improve the stability of the obtained standard section quality, and promote the promotion and application of early pregnancy structural screening.
  • FIG. 4 is a schematic flowchart of a method 400 for ultrasound imaging of a fetus in early pregnancy according to an embodiment of the present application.
  • a method 400 for ultrasound imaging of a fetus in early pregnancy includes the following steps:
  • step S410 transmitting ultrasonic waves to the fetus in the first trimester, and receiving echoes of the ultrasonic waves to obtain ultrasonic echo signals;
  • step S420 obtain three-dimensional ultrasound data of the fetus in the first trimester according to the ultrasound echo signal
  • step S430 match the three-dimensional ultrasound data with an ultrasound data template pre-configured with standard slices in early pregnancy, and determine a target standard slice in the three-dimensional ultrasonic data according to the matching result, wherein the standard slices in early pregnancy include the following At least one: lateral ventricle horizontal cross section, biparietal diameter section, midsagittal head and neck section, breast diameter section, abdominal circumference section, abdominal wall umbilical cord insertion section, bladder section, longitudinal axis of the spine, long axis of the trunk and top buttocks radial section;
  • step S440 the target standard slice is displayed.
  • Steps S410 , S420 and S440 in the ultrasound imaging method 400 of the first trimester fetus according to the embodiment of the present application are substantially similar to steps S210 , S220 and S250 in the ultrasound imaging method 200 described with reference to FIG. 2 , for the sake of brevity , the same details will not be repeated here, and the following mainly describes in detail the manner of determining the standard slice according to the three-dimensional ultrasound data in step S430.
  • the standard slice is determined by matching the three-dimensional ultrasound data of the fetus in the first trimester with the preconfigured ultrasound data template of the standard slice in the first trimester.
  • the preconfigured ultrasound data template of the standard slice in the first trimester is Including matching including pre-configured three-dimensional data template of standard section in early pregnancy (hereinafter referred to as standard three-dimensional data template), pre-configured two-dimensional data template of standard section in early pregnancy (hereinafter referred to as standard two-dimensional section template) and pre-configured
  • the first-pregnancy target feature structure template data template (hereinafter referred to as the standard first-pregnancy target feature structure template) of the first-trimester standard section is configured, and the key point template (hereinafter referred to as the standard key-point template) of the first-trimester standard section is preconfigured.
  • the matching with the template includes at least one of the following: three-dimensional data matching with a standard three-dimensional data template, two-dimensional section matching with a standard two-dimensional section template, early pregnancy target feature structure matching with a standard early pregnancy target feature structure template, and Keypoint matching with standard keypoint templates.
  • matching the three-dimensional data of the fetus in the early pregnancy with the standard three-dimensional data template includes finding an optimal three-dimensional space transformation relationship so that the three-dimensional ultrasound data obtained in step S420 has the highest similarity or the smallest difference between the standard three-dimensional data template.
  • image feature extraction such as gradient features, texture features such as LBP, Harr features, HOG/LOG features, etc. Then find an optimal three-dimensional space transformation relationship, so that the similarity between the image features extracted from the three-dimensional ultrasound data of the fetus in the first trimester and the standard three-dimensional data template is the highest or the difference is the smallest.
  • the position corresponding to the target standard slice in the 3D ultrasound data of the fetus in the first trimester can be determined according to the position of the standard slice in the first trimester pre-configured in the standard 3D data template and the 3D space transformation relationship obtained by matching.
  • matching the 3D ultrasound data of the fetus in the first trimester with the standard 2D slice template includes finding an optimal 2D slice in the 3D ultrasound data of the fetus in the first trimester acquired in step S420, so that the 2D slice matches the standard 2D slice.
  • the similarity between the two-dimensional slice templates is the highest or the difference is the smallest, or the similarity between the image features extracted from the optimal two-dimensional slice and the standard two-dimensional slice template is the highest or the difference is the smallest.
  • the image features include but Not limited to gradient features, texture features such as LBP, Harr features, HOG/LOG features.
  • matching the three-dimensional ultrasound data of the fetus in the first trimester with the standard first-pregnancy target feature structure template includes finding an optimal image block in the three-dimensional ultrasound data acquired in step S420, so that the image block matches the standard first-pregnancy target feature structure
  • the similarity between the templates is the highest or the difference is the smallest, or the similarity between the image features extracted from the image block and the standard early pregnancy target feature structure template is the highest or the difference is the smallest.
  • the feature structure matching of the early pregnancy target may also include using target detection methods such as Faster RCNN, Mask RCNN, SSD, YOLO, Retinanet, Efficientnet, Cornernet, Centernet, FCOS, etc., to detect in the three-dimensional ultrasound data acquired in step S420
  • the candidate early pregnancy target feature structure region is then matched with the standard early pregnancy target feature structure template.
  • matching with the standard early pregnancy target feature structure template may include finding an optimal candidate early pregnancy target feature structure in the 3D ultrasound data, so that the similarity between the standard early pregnancy target feature structure template and the standard early pregnancy target feature structure template is the highest or the difference is the smallest; or It includes extracting the image features of the candidate early pregnancy target feature structure and the standard early pregnancy target feature structure template, and then finding an optimal candidate early pregnancy target feature structure, so that the image features extracted from it and the image features extracted from the standard early pregnancy target feature structure template are consistent. Alternatively, it may also include finding the optimal candidate early pregnancy target feature structure and an optimal spatial transformation to minimize the spatial position difference between the candidate early pregnancy target feature structure and the standard early pregnancy target feature structure template.
  • the position of the target standard slice can be determined according to the position of the early pregnancy target characteristic structure in the three ultrasound data and the early pregnancy target characteristic structure corresponding to the standard slice.
  • the corresponding target feature structures in early pregnancy include nasal bone and genital ridge
  • the corresponding target feature structures in early pregnancy include falx cerebrum, Lateral ventricle and choroid plexus
  • the corresponding target feature structures in early pregnancy include thalamus and skull halo
  • the corresponding target feature structures in early pregnancy include the translucent layer at the back of the neck and the nasal bone
  • the standard section in the first trimester is the breast diameter section, the corresponding target feature structure in early
  • matching the three-dimensional ultrasound data with the standard keypoint template includes finding one or more optimal points in the three-dimensional ultrasound data acquired in step S420, and matching the image features near the optimum point with the image features of the standard keypoint template The similarity is the highest or the difference is the smallest.
  • Keypoint matching can also include using injection feature point extraction methods (such as SIFT method), corner detection methods (such as Harris method), or predicting candidate keypoint coordinates or the region where candidate keypoints are based on neural networks, so as to determine in 3D ultrasound data.
  • injection feature point extraction methods such as SIFT method
  • corner detection methods such as Harris method
  • predicting candidate keypoint coordinates or the region where candidate keypoints are based on neural networks so as to determine in 3D ultrasound data.
  • At least one candidate keypoint and then matching the candidate keypoint with the standard keypoint template.
  • matching the candidate keypoints with the standard keypoint template may include finding an optimal candidate keypoint to make its nearby image features have the highest similarity or the smallest difference between the image features of the standard keypoint template; or, Matching the candidate keypoint with the standard keypoint template may also include finding the optimal candidate keypoint and an optimal spatial transformation relationship to minimize the spatial position difference between the optimal candidate keypoint and the standard keypoint template.
  • a cut plane including these key points can be determined as the target standard cut plane according to the positions of the key points corresponding to the standard cut planes; or according to the obtained optimal spatial transformation relationship, according to the standard key point template according to the first trimester standard cut plane. position, to determine the position corresponding to the target standard slice in the 3D ultrasound data.
  • the ultrasound imaging system 100 provided by the embodiment of the present application can be used to implement the above-mentioned ultrasound imaging method 400 of a fetus in the first trimester.
  • the ultrasound imaging system 100 may include the ultrasound probe 110 , the transmit/receive circuit 112 , the processor 114 , and the display 116 , and the relevant descriptions of the various components can be referred to above.
  • the transmitting/receiving circuit 112 is used to excite the ultrasonic probe 110 to transmit ultrasonic waves to the fetus in early pregnancy, and receive the echoes of the ultrasonic waves to obtain ultrasonic echo signals; the processor 114, used for: obtaining the three-dimensional ultrasound data of the fetus in the first trimester based on the ultrasound echo signal; matching the three-dimensional ultrasound data with the ultrasound data template of the pre-configured standard slice in the first trimester, and determining the three-dimensional ultrasound according to the matching result
  • the target standard view in the ultrasound data wherein the first trimester standard view includes at least one of the following: lateral ventricle horizontal transverse view, biparietal diameter view, head and neck midsagittal view, breast diameter view, abdominal circumference view, abdominal wall umbilical cord insertion site slices, bladder slices, longitudinal spine slices, trunk long-axis slices and apical rump slices; the display 116 is used to display the target standard slices
  • the ultrasound imaging method 400 and the ultrasound imaging system of the first-trimester fetus in the embodiment of the present application can automatically determine the standard section of the first-trimester fetus by matching with the pre-configured ultrasound data template of the standard section of the first-trimester, without the need for a doctor to manually extract the standard sections one by one. It greatly optimizes the workflow of prenatal examination, effectively improves work efficiency, and can improve the stability of the obtained standard section quality, and promote the promotion and application of early pregnancy structural screening.
  • FIG. 5 is a schematic flowchart of a method 500 for ultrasound imaging of a fetus in early pregnancy according to an embodiment of the present application.
  • a method 500 for ultrasound imaging of a fetus in early pregnancy includes the following steps:
  • step S510 transmitting ultrasonic waves to the fetus in early pregnancy, and receiving echoes of the ultrasonic waves to obtain ultrasonic echo signals;
  • step S520 obtain three-dimensional ultrasound data of the fetus in the first trimester according to the ultrasound echo signal
  • step S530 the image features of the target area of the three-dimensional ultrasound data are extracted according to the three-dimensional ultrasound data
  • step S540 the normal direction of the target standard slice and the position information of the preset point on the target standard slice are determined according to the image feature of the target area;
  • step S550 the target standard tangent plane is determined according to the normal direction of the target standard tangent plane and the position information of the preset point on the target standard tangent plane;
  • step S560 the target standard slice is displayed.
  • Steps S510 and S520 in the ultrasound imaging method 500 of the first trimester fetus according to the embodiment of the present application are substantially similar to steps S210 and S220 in the ultrasound imaging method 200 described with reference to FIG.
  • the following mainly describes in detail the manner in which the standard slice is determined according to the three-dimensional ultrasound data in the method 500 .
  • the image features of the target area in the three-dimensional ultrasound data of the fetus in the first trimester are first extracted.
  • the extraction method of the image feature of the target area includes a traditional image processing method or a deep learning method.
  • traditional image processing methods include extraction of Sift features, gradient features, texture features such as LBP, PCA, LDA, Harr features, HOG and LOG features and other image features, as well as image edge extraction, such as using canny operator to achieve edge extraction, etc.
  • Deep learning methods include training the neural network model for one or more specific tasks, such as regressing the position and orientation of standard slices, estimating the position, size and/or orientation of the fetus in the first trimester, identifying the target feature structure of the fetus in the first trimester and/or marker points, etc., and then take the output of one or more network nodes in the middle of the neural network model as the image feature of the target area, or take the front-end feature extraction network in the trained neural network to extract the image of the target area. feature.
  • the normal direction of the target standard section can be determined according to the mapping relationship between the image features of the target area and the normal direction of the standard section corresponding to the image features of the target area, and the normal direction of the target standard section can be determined according to the image features of the target area.
  • the mapping relationship between the feature and the position information of the preset point on the standard tangent plane corresponding to the image feature of the target area determines the position information of the preset point on the target standard tangent plane.
  • the preset point may be the point where the standard tangent plane intersects the X axis, the Y axis, or the Z axis, or may also be the position information of the standard tangent plane in the preset direction, for example, it may be the intersection of the standard tangent plane and the normal direction. .
  • a machine learning model or a deep learning model can be used to regress the mapping relationship between the image feature of the target area and the normal direction of the standard section corresponding to the target image feature; a machine learning model or a deep learning model can also be used to regress the The mapping relationship between the image features of the target area and the preset points on the standard slice corresponding to the image features of the target area.
  • a first-trimester fetal ultrasound database is pre-established, wherein each ultrasound data includes first-trimester three-dimensional data and/or image features of its target area, as well as preset point positions and normal directions of one or more standard slices.
  • mapping function When training the model, find an optimal mapping function from the image features of the target area to the position and normal direction of the preset point, so that the image features of the target area in the database can be obtained through the mapping function to the preset point position and normal line.
  • the error between the direction and the preset point position and normal direction of the actual standard tangent plane is minimal.
  • the mapping function the position of the preset point and the direction of the normal line can be predicted according to the image features of the target area of the three-dimensional ultrasound image.
  • standard slices can be extracted from the 3D ultrasound data according to the preset point position and the normal direction, for example, slices passing through the preset point are extracted from all slices perpendicular to the normal direction.
  • Standard views include one or more of the following views: lateral ventricle horizontal cross, biparietal diameter, head and neck midsagittal, breast diameter, abdominal circumference, abdominal wall umbilical cord insertion, bladder, longitudinal spine Slices, trunk long-axis slices, and top-rump diameter slices.
  • mapping functions include linear functions, polynomial functions, logistic functions, etc.
  • deep learning methods use deep neural networks as mapping functions, Including but not limited to CNN convolutional neural network, MLP multilayer perceptron, RNN recurrent neural network, etc.
  • the ultrasound imaging system 100 provided by the embodiment of the present application can be used to implement the above-mentioned ultrasound imaging method 500 of a fetus in the first trimester.
  • the ultrasound imaging system 100 may include the ultrasound probe 110, the transmit/receive circuit 112, the processor 114, and the display 116, and the relevant descriptions of the various components can be referred to above.
  • the transmitting/receiving circuit 112 is used to excite the ultrasonic probe 110 to transmit ultrasonic waves to the fetus in early pregnancy, and receive the echoes of the ultrasonic waves to obtain ultrasonic echo signals;
  • the processor 114 uses In: obtaining the three-dimensional ultrasound data of the fetus in the first trimester based on the ultrasound echo signal; extracting the image features of the target area of the three-dimensional ultrasound data according to the three-dimensional ultrasound data; determining the target standard according to the image features of the target area the normal direction of the tangent plane and the position information of the preset point on the target standard tangent plane; the target standard tangent plane is determined according to the normal direction of the target standard tangent plane and the position information of the preset point on the target standard tangent plane; the display 116 is used to display the standard slice.
  • the ultrasound imaging method 500 and the ultrasound imaging system of the first-trimester fetus can automatically determine the standard section of the first-trimester fetus according to the image features of the target area of the three-dimensional ultrasound data of the first-trimester fetus, without the need for doctors to manually extract the standard sections one by one, thereby improving the The work efficiency and the quality of standard slices are improved.
  • FIG. 6 is a schematic flowchart of a method 600 for ultrasound imaging of a fetus in early pregnancy according to an embodiment of the present application.
  • a method 600 for ultrasound imaging of a fetus in early pregnancy includes the following steps:
  • step S610 transmitting ultrasonic waves to the fetus in early pregnancy, and receiving echoes of the ultrasonic waves to obtain ultrasonic echo signals;
  • step S620 obtain three-dimensional ultrasound data of the fetus in the first trimester based on the ultrasound echo signal
  • step S630 the direction of the target area of the fetus in the first trimester is determined according to the three-dimensional ultrasound data
  • step S640 the direction of the target area of the first trimester fetus is displayed.
  • the method of determining the direction of the target region in the ultrasound imaging method 600 of the first trimester fetus according to the embodiment of the present application is substantially the same as the method of determining the direction of the target region in the ultrasound imaging method 200 described with reference to FIG. 2 .
  • the target area includes at least one of the following: a head area, a body area, and a whole body area.
  • the target area includes a head area
  • the direction of the target area includes a first direction of the head area, wherein the first direction is a left-right direction of the head area.
  • a machine learning method or a skull halo detection method can be used to detect the head region of the fetus in the first trimester in the three-dimensional ultrasound data, and the left hemisphere region and the right hemisphere region of the head region can be determined according to the symmetry of the head region of the fetus in the first trimester. hemisphere area, and the normal direction of the interface between the left hemisphere area and the right hemisphere area of the head area is determined as the first direction of the head area.
  • the midsagittal plane of the head region of the fetus in the first trimester can be extracted from the three-dimensional ultrasound data; the normal direction of the midsagittal plane of the head region is determined as the first direction of the head region.
  • the target area may also include a body area, and the direction of the target area includes at least one of the following: the second direction of the body area, the third direction of the body area, and the fourth direction of the body area, wherein the second direction, the third direction
  • the direction and the fourth direction are the up-down direction, the front-rear direction, and the left-right direction of the body area, respectively, and any two of the second, third, and fourth directions are perpendicular to each other.
  • the spine region of the fetus in the first trimester may be detected in the three-dimensional ultrasound data; the orientation of the body region is determined according to the orientation of the spine region. For example, a straight line close to the spine region can be determined, and the direction of the straight line can be determined as the second direction of the body region, ie, the up-down direction.
  • one or more cross-section planes of the body perpendicular to the second direction can be extracted from the three-dimensional ultrasound data of the fetal body in the first trimester;
  • the line connecting the position of the point and the spine is determined as the third direction of the fetal body region in the first trimester, that is, the anterior-posterior direction.
  • the method for determining the anterior-posterior direction may also be: determining a curve close to the spine region, and determining the convex direction of the curve as the third direction of the fetal body region in the first trimester.
  • the method of determining a body region may include: detecting a body region of a fetus in early pregnancy from three-dimensional ultrasound data; and determining an orientation of the body region according to the shape of the body region.
  • the method for determining the fourth direction of the body region may include: determining the second direction and the third direction of the fetal body region in the first trimester, and determining the direction perpendicular to the second direction and the third direction as the fetal body region in the first trimester. the fourth direction.
  • the positions of the kidney regions in the fetal body in the first trimester can also be detected in the three-dimensional ultrasound data, and the direction of the line connecting the regions of the symmetrical feature structures is determined as the fourth direction of the fetal body region in the first trimester.
  • At least one standard slice corresponding to the target area may be extracted from the 3D ultrasound data according to the direction of the target area.
  • the corresponding multiple candidate slices; at least one standard slice corresponding to the target area is selected from the multiple candidate slices corresponding to the target area.
  • the early pregnancy target feature structure corresponding to the standard section corresponding to the target area can be detected in the three-dimensional ultrasound data, and the angle between the early pregnancy target feature structure and the early pregnancy target feature structure at least partially overlaps, and the angle between the direction of the target area and the target area meets the preset requirements.
  • the slice is determined as at least one standard slice corresponding to the target area.
  • the ultrasound imaging system 100 provided by the embodiment of the present application can be used to implement the above-mentioned ultrasound imaging method 600 of a fetus in the first trimester.
  • the ultrasound imaging system 100 may include the ultrasound probe 110 , the transmit/receive circuit 112 , the processor 114 , and the display 116 , and the relevant descriptions of the various components can be referred to above.
  • the transmitting/receiving circuit 112 is used to excite the ultrasonic probe 110 to transmit ultrasonic waves to the fetus in early pregnancy, and receive the echoes of the ultrasonic waves to obtain ultrasonic echo signals; the processor 114 uses In: obtaining three-dimensional ultrasound data of the fetus in the first trimester based on the ultrasound echo signal; determining the direction of the target area of the fetus in the first trimester according to the three-dimensional ultrasound data; the display 116 is used to display the direction of the target area.
  • a computer storage medium is also provided, where program instructions are stored on the computer storage medium, and the program instructions are used to execute the methods of the embodiments of the present application when the program instructions are run by a computer or a processor 200 , method 300 , method 400 , method 500 or corresponding steps of method 600 .
  • the storage medium may include, for example, a memory card of a smartphone, a storage component of a tablet computer, a hard disk of a personal computer, read only memory (ROM), erasable programmable read only memory (EPROM), portable compact disk read only memory (CD-ROM), USB memory, or any combination of the above storage media.
  • the computer-readable storage medium can be any combination of one or more computer-readable storage media.
  • a computer program is also provided, and the computer program can be stored in the cloud or on a local storage medium.
  • the computer program is run by a computer or a processor, it is used to execute the corresponding steps of the ultrasound imaging method of the fetus in the first trimester of the embodiments of the present application.
  • the standard section of the fetus in the first trimester can be automatically determined according to the three-dimensional ultrasound data collected in a single time, without the need for a doctor to manually extract the standard section one by one.
  • Dadi optimizes the workflow of prenatal examination, effectively improves work efficiency, and improves the stability of the quality of the obtained standard slices, and promotes the popularization and application of early pregnancy structural screening.
  • the disclosed apparatus and method may be implemented in other manners.
  • the device embodiments described above are only illustrative.
  • the division of the units is only a logical function division. In actual implementation, there may be other division methods.
  • multiple units or components may be combined or May be integrated into another device, or some features may be omitted, or not implemented.
  • Various component embodiments of the present application may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof.
  • a microprocessor or a digital signal processor (DSP) may be used in practice to implement some or all functions of some modules according to the embodiments of the present application.
  • DSP digital signal processor
  • the present application can also be implemented as a program of apparatus (eg, computer programs and computer program products) for performing part or all of the methods described herein.
  • Such a program implementing the present application may be stored on a computer-readable medium, or may be in the form of one or more signals. Such signals may be downloaded from Internet sites, or provided on carrier signals, or in any other form.

Landscapes

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

Abstract

一种早孕期胎儿的超声成像方法(200)和超声成像系统(100)，早孕期胎儿的超声成像方法(200)包括：向早孕期胎儿发射超声波，接收超声波的回波，以获得超声回波信号(S210)；基于超声回波信号获得早孕期胎儿的三维超声数据(S220)；根据三维超声数据确定早孕期胎儿的目标区域的方向(S230)；根据目标区域的方向，从三维超声数据中提取与目标区域对应的至少一个标准切面(S240)；显示至少一个标准切面(S250)。早孕期胎儿的超声成像方法(200)和超声成像系统(100)根据单次采集的三维超声数据自动提取早孕期胎儿的标准切面，无需医生逐一手动提取标准切面，提高了提取标准切面的效率和提取到的标准切面的质量。

Description

早孕期胎儿的超声成像方法和超声成像系统
说明书
技术领域
本申请涉及超声成像技术领域，更具体地涉及一种早孕期胎儿的超声成像方法和超声成像系统。
背景技术
超声检查由于其安全、方便、无辐射、廉价等优势，在临床检查上具有广泛的应用，成为医生进行疾病诊断的主要辅助手段之一。产前超声检查作为产前检查中的最主要的影像学检查，为胎儿的生长发育测量及结构异常筛查提供了最重要的影像学证据。产前超声检查已经是早孕、中孕及晚孕期必须要进行的检查之一。
在实际临床应用中，在早孕期开展胎儿结构检查及畸形筛查是当前临床的发展趋势与研究热点。在早孕期开展胎儿结构筛查可尽早发现致死性畸形，为孕妇尽可能早地提供终止妊娠的机会，最大程度减少身体健康及精神上的伤害，具有非常重要的临床意义与价值。然而，早孕期胎儿结构异常筛查的切面数众多，各个切面要求不一，切面的获取比较繁琐，并且切面质量受医生的经验与手法影响比较大。
发明内容
在发明内容部分中引入了一系列简化形式的概念，这将在具体实施方式部分中进一步详细说明。本发明的发明内容部分并不意味着要试图限定出所要求保护的技术方案的关键特征和必要技术特征，更不意味着试图确定所要求保护的技术方案的保护范围。
本申请实施例第一方面提供一种早孕期胎儿的超声成像方法，所述方法包括：
向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号；
基于所述超声回波信号获得所述早孕期胎儿的三维超声数据；
根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向；
根据所述目标区域的方向，从所述三维超声数据中提取与所述目标区域对应的至少一个标准切面；
显示所述至少一个标准切面。
本申请实施例第二方面提供一种早孕期胎儿的超声成像方法，所述方法包括：
向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号；
根据所述超声回波信号获得早孕期胎儿的三维超声数据；
从所述三维超声数据中检测至少两个不同的早孕目标特征结构的区域；
确定与所述至少两个不同的早孕目标特征结构的区域中每个所述早孕目标特征结构的区域均至少部分重合的至少一个切面，以作为所述早孕期胎儿的至少一个标准切面；
显示所述至少一个标准切面。
本申请实施例第三方面提供一种早孕期胎儿的超声成像方法，所述方法包括：
向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号；
根据所述超声回波信号获得早孕期胎儿的三维超声数据；
将所述三维超声数据与预先配置好早孕期标准切面的超声数据模板进行匹配，并根据匹配结果确定所述三维超声数据中的目标标准切面，其中所述早孕期标准切面包括以下至少一个：侧脑室水平横切面、双顶径切面、头颈部正中矢状切面、胸径切面、腹围切面、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面、躯干长轴切面和顶臀径切面；
显示所述目标标准切面。
本申请实施例第四方面提供一种早孕期胎儿的超声成像方法，所述方法 包括：
向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号；
根据所述超声回波信号获得所述早孕期胎儿的三维超声数据；
根据所述三维超声数据提取所述三维超声数据的目标区域的图像特征；
根据所述目标区域的图像特征确定目标标准切面的法线方向和所述目标标准切面上预设点的位置信息；
根据所述目标标准切面的法线方向和所述目标标准切面上预设点的位置信息确定所述目标标准切面；
显示所述目标标准切面。
本申请实施例第五方面提供一种早孕期胎儿的超声成像方法，所述方法包括：
向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号；
基于所述超声回波信号获得所述早孕期胎儿的三维超声数据；
根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向；
显示所述早孕期胎儿的目标区域的方向。
本申请实施例第六方面提供一种超声成像系统，所述超声成像系统包括：
超声探头；
发射/接收电路，用于激励所述超声探头向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；
处理器，用于：
基于所述超声回波信号获得所述早孕期胎儿的三维超声数据；
根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向；
根据所述目标区域的方向，从所述三维超声数据中提取与所述目标区域对应的至少一个标准切面；
显示器，用于显示所述至少一个标准切面。
本申请实施例第七方面提供一种超声成像系统，所述超声成像系统包括：
超声探头；
发射/接收电路，用于激励所述超声探头向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；
处理器，用于：
显示所述至少一个标准切面根据所述超声回波信号获得所述早孕期胎儿的三维超声数据；
从所述三维超声数据中至少两个不同的早孕目标特征结构的区域；
确定与所述至少两个不同的早孕目标特征结构的区域中每个所述早孕目标特征结构的区域均至少部分重合的至少一个切面，以作为所述早孕期胎儿的至少一个标准切面；
显示器，用于显示所述至少一个标准切面。
本申请实施例第八方面提供一种超声成像统，所述超声成像系统包括：
超声探头；
发射/接收电路，用于激励所述超声探头向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；
处理器，用于：
根据所述超声回波信号获得所述早孕期胎儿的三维超声数据；
将所述三维超声数据与预先配置好早孕期标准切面的超声数据模板进行匹配，并根据匹配结果确定所述三维超声数据中的目标标准切面，其中所述早孕期标准切面包括以下至少一个：侧脑室水平横切面、双顶径切面、头颈部正中矢状切面、胸径切面、腹围切面、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面、躯干长轴切面和顶臀径切面；
显示器，用于显示所述目标标准切面。
本申请实施例第九方面提供一种超声成像系统，所述超声成像系统包括：
超声探头；
发射/接收电路，用于激励所述超声探头向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；
处理器，用于：
根据所述三维超声数据提取所述三维超声数据的目标区域的图像特征；
根据所述目标区域的图像特征确定目标标准切面的法线方向和所述目标标准切面上预设点的位置信息；
根据所述超声回波信号获得所述早孕期胎儿的三维超声数据；
根据所述三维超声数据提取所述三维超声数据的目标区域的图像特征；
根据所述目标区域的图像特征确定目标标准切面的法线方向和所述目标标准切面上预设点的位置信息；
根据所述目标标准切面的法线方向和所述目标标准切面上预设点的位置信息确定所述目标标准切面；
显示器，用于显示所述目标标准切面。
本申请实施例第十方面提供一种超声成像系统，所述超声成像系统包括：
超声探头；
发射/接收电路，用于激励所述超声探头向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；
处理器，用于：
基于所述超声回波信号获得所述早孕期胎儿的三维超声数据；
根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向；
显示器，用于显示所述早孕期胎儿的目标区域的方向。
根据本申请实施例的早孕期胎儿的超声成像方法和超声成像系统，能够根据单次采集的三维超声数据自动提取早孕期胎儿的标准切面，无需医生逐一手动提取标准切面，极大地优化了产前检查的工作流，有效地提升工作效率，并且能提升所获取的标准切面质量的稳定性，促进早孕结构筛查的推广与应用。
附图说明
为了更清楚地说明本申请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性的前提下，还可以根据这些附图获得其他的附图。
在附图中：
图1示出根据本申请实施例的超声成像系统的示意性框图；
图2示出根据本发明一实施例的早孕期胎儿的超声成像方法的示意性流程图；
图3示出根据本发明另一实施例的早孕期胎儿的超声成像方法的示意性流程图；
图4示出根据本发明另一实施例的早孕期胎儿的超声成像方法的示意性流程图；
图5示出根据本发明又一实施例的早孕期胎儿的超声成像方法的示意性流程图；
图6示出根据本发明再一实施例的早孕期胎儿的超声成像方法的示意性流程图。
具体实施方式
为了使得本申请的目的、技术方案和优点更为明显，下面将参照附图详细描述根据本申请的示例实施例。显然，所描述的实施例仅仅是本申请的一部分实施例，而不是本申请的全部实施例，应理解，本申请不受这里描述的示例实施例的限制。基于本申请中描述的本申请实施例，本领域技术人员在没有付出创造性劳动的情况下所得到的所有其它实施例都应落入本申请的保护范围之内。
在下文的描述中，给出了大量具体的细节以便提供对本申请更为彻底的理解。然而，对于本领域技术人员而言显而易见的是，本申请可以无需一个或多个这些细节而得以实施。在其他的例子中，为了避免与本申请发生混淆，对于本领域公知的一些技术特征未进行描述。
应当理解的是，本申请能够以不同形式实施，而不应当解释为局限于这里提出的实施例。相反地，提供这些实施例将使公开彻底和完全，并且将本申请的范围完全地传递给本领域技术人员。
在此使用的术语的目的仅在于描述具体实施例并且不作为本申请的限制。 在此使用时，单数形式的“一”、“一个”和“所述/该”也意图包括复数形式，除非上下文清楚指出另外的方式。还应明白术语“组成”和/或“包括”，当在该说明书中使用时，确定所述特征、整数、步骤、操作、元件和/或部件的存在，但不排除一个或更多其它的特征、整数、步骤、操作、元件、部件和/或组的存在或添加。在此使用时，术语“和/或”包括相关所列项目的任何及所有组合。
为了彻底理解本申请，将在下列的描述中提出详细的结构，以便阐释本申请提出的技术方案。本申请的可选实施例详细描述如下，然而除了这些详细描述外，本申请还可以具有其他实施方式。
下面，首先参考图1描述根据本申请一个实施例的超声成像系统，图1示出了根据本申请实施例的超声成像系统100的示意性结构框图。
如图1所示，超声成像系统100包括超声探头110、发射/接收电路112、处理器114、显示器116以及存储器118。进一步地，超声成像系统100还可以包括波束合成电路和发射/接收选择开关等。
具体地，超声探头110包括多个换能器阵元，多个换能器阵元可以排列成一排构成线阵，或排布成二维矩阵构成面阵，多个换能器阵元也可以构成凸阵列。换能器用于根据激励电信号发射超声波，或将接收的超声波转换为电信号，因此每个阵元可用于实现电脉冲信号和超声波的相互转换，从而实现向被测对象的目标区域的组织发射超声波、也可用于接收经组织反射回的超声波回波。在进行超声成像时，可通过发射序列和接收序列控制哪些换能器用于发射超声波，哪些换能器用于接收超声波，或者控制换能器分时隙用于发射超声波或接收超声波的回波。参与超声波发射的换能器可以同时被电信号激励，从而同时发射超声波；或者，参与超声波束发射的换能器也可以被具有一定时间间隔的若干电信号激励，从而持续发射具有一定时间间隔的超声波。
发射/接收电路112可以通过发射/接收选择开关与超声探头110连接。发射/接收选择开关也可以被称为发送/接收控制器，其可以包括发送控制器和接收控制器，发送控制器用于激励超声探头110经由发射电路向早孕期胎儿所在区域发射超声波；接收控制器用于通过超声探头110经由接收电路接收从早孕期胎儿所在区域返回的超声回波，从而获得超声回波数据。之后，发射/接收电路112将超声回波的电信号送入波束合成电路，波束合成电路对该电 信号进行聚焦延时、加权和通道求和等处理，然后将处理后的超声回波数据送入处理器114。
可选地，处理器114可以通过软件、硬件、固件或其任意组合来实现，可以使用电路、单个或多个专用集成电路(Application Specific Integrated Circuit，ASIC)、单个或多个通用集成电路、单个或多个微处理器、单个或多个可编程逻辑器件、或者前述电路和/或器件的任意组合、或者其他适合的电路或器件，从而使得处理器114可以执行本说明书中的各个实施例中的方法的相应步骤。并且，处理器114可以控制所述超声成像系统100中的其它组件以执行期望的功能。
处理器114对其接收到的超声回波数据进行处理，得到早孕期胎儿的三维超声数据。作为示例，超声探头110在一系列扫描平面内发射/接收超声波，由处理器114根据其三维空间关系进行整合，实现早孕期胎儿在三维空间的扫描以及三维图像的重建。最后，由处理器114对其进行去噪、平滑、增强等部分或全部图像后处理步骤后，获取早孕期胎儿的三维超声数据。处理器114可以获取早孕期胎儿全身的三维超声数据，也可以仅获取早孕期胎儿头部或者身体的三维超声数据。处理器114还用于从三维超声数据中提取早孕期胎儿的标准切面。处理器114得到的标准切面可以存储于存储器中或在显示器116上显示。并且，处理器114还可以对三维超声数据进行绘制并在显示器116上显示。
显示器116与处理器114连接，显示器116可以为触摸显示屏、液晶显示屏等；或者显示器116可以为独立于超声成像系统100之外的液晶显示器、电视机等独立显示设备；或者显示器116可以是智能手机、平板电脑等电子设备的显示屏，等等。其中，显示器116的数量可以为一个或多个。例如，显示器116可以包括主屏和触摸屏，主屏主要用于显示超声图像，触摸屏主要用于人机交互。
显示器116可以显示处理器114得到的超声图像。此外，显示器116在显示超声图像的同时还可以提供给用户进行人机交互的图形界面，在图形界面上设置一个或多个被控对象，提供给用户利用人机交互装置输入操作指令来控制这些被控对象，从而执行相应的控制操作。例如，在图形界面上显示图标，利用人机交互装置可以对该图标进行操作，用来执行特定的功能。
可选地，超声成像系统100还可以包括显示器116之外的其他人机交互 装置，其与处理器114连接，例如，处理器114可以通过外部输入/输出端口与人机交互装置连接，外部输入/输出端口可以是无线通信模块，也可以是有线通信模块，或者两者的组合。外部输入/输出端口也可基于USB、如CAN等总线协议、和/或有线网络协议等来实现。
其中，人机交互装置可以包括输入设备，用于检测用户的输入信息，该输入信息例如可以是对超声波发射/接收时序的控制指令，可以是在超声图像上绘制出点、线或框等的操作输入指令，或者还可以包括其他指令类型。输入设备可以包括键盘、鼠标、滚轮、轨迹球、移动式输入设备(比如带触摸显示屏的移动设备、手机等等)、多功能旋钮等等其中之一或者多个的结合。人机交互装置还可以包括诸如打印机之类的输出设备。
超声成像系统100还可以包括存储器，用于存储处理器执行的指令、存储接收到的超声回波、存储超声图像，等等。存储器可以为闪存卡、固态存储器、硬盘等。其可以为易失性存储器和/或非易失性存储器，为可移除存储器和/或不可移除存储器等。
应理解，图1所示的超声成像系统100所包括的部件只是示意性的，其可以包括更多或更少的部件。本申请对此不限定。
下面，将参考图2描述根据本申请实施例的早孕期胎儿的超声成像方法。图2是本申请实施例的早孕期胎儿的超声成像方法200的一个示意性流程图。
如图2所示，本申请一个实施例的早孕期胎儿的超声成像方法200包括如下步骤：
首先，在步骤S210，向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号。
其中，早孕期胎儿一般指在孕14周以下的胎儿。早孕期胎儿已经生长到了一定大小，大部分器官已经分化形成，相当一部分特征结构已经能够通过超声成像所辨别，通过本申请提供的超声成像方法能够自动提取反应特征结构信息的标准切面，从而在早孕期实现胎儿的结构检查及畸形筛查，相对于中晚孕期的超声检查来说，能够尽可能早地为孕妇提供相关的妊娠信息。
示例性地，可以基于图1所示的超声成像系统100进行超声图像采集。用户移动超声探头110选择合适的位置和角度，发射/接收电路120中的发射电路将一组经过延迟聚焦的脉冲发送到超声探头110，超声探头110沿2D扫 描平面向早孕期胎儿发射超声波形。超声探头110接收到反射回的超声回波后，将其转化为电信号，由波束合成电路对多次发射/接收得到的信号进行相应的延时与加权求和的处理，实现波束合成，再送入处理器114进行后续的信号处理。
示例性地，在执行步骤S210之前，可以自动开启或由用户手动开启自动提取标准切面的功能。在一些实施例中，在执行步骤S210之前或之后，还可以提供用户界面，以允许用户手动选择期望提取的标准切面。然而，该步骤是可选的，在其他实施例中，可以默认提取下文所涉及的全部标准切面。
在步骤S220，基于所述超声回波信号获得所述早孕期胎儿的三维超声数据。
具体地，可以对超声探头110在一系列扫描平面内发射/接收获得的超声回波的三维空间关系进行整合，从而实现早孕期胎儿在三维空间的扫描以及3D图像的重建。最后，经过去噪、平滑、增强等部分或全部图像后处理步骤后，获得早孕期胎儿的三维超声数据。
其中，可以获取早孕期胎儿全身的三维超声数据，也可以仅获取早孕期胎儿头部区域或者早孕期胎儿身体区域的三维数据，具体取决于待提取的标准切面。当待提取的标准切面包括早孕期胎儿头部区域对应的标准切面时，三维超声数据至少包括早孕期胎儿头部区域的三维超声数据。当待提取的标准切面包括早孕期胎儿身体区域对应的标准切面时，三维超声数据至少包括早孕期胎儿身体区域的三维超声数据。当待提取的标准切面包括早孕期胎儿全身区域对应的标准切面时，则需要获取早孕期胎儿全身区域的三维超声数据。
在本申请实施例中，标准切面为三维超声图像中包含关键信息的二维切面，通过该二维切面可以观察到具有临床价值的生理特征结构。示例性地，后续提取的标准切面包括以下标准切面中的至少一个：顶臀径切面(即全身正中矢状切面)、侧脑室水平横切面、双顶径切面(即丘脑水平横切面)、NT标准切面(即头颈部正中矢状切面)、胸径切面(即四腔心切面)、腹围切面(即胃泡切面)、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面(即身体正中矢状切面)、躯干长轴切面(即身体冠状切面)。后续可以提取以上标准切面中的多个标准切面，以对早孕期胎儿进行全面的筛查。
以上标准切面中，侧脑室水平横切面、双顶径切面及NT标准切面为早 孕期胎儿头部区域对应的标准切面，要求三维超声数据至少包括早孕期胎儿头部区域的三维超声数据；胸径切面、腹围切面、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面、躯干长轴切面为早孕期胎儿身体区域对应的标准切面，要求三维超声数据至少包括早孕期胎儿身体区域的三维超声数据；而顶臀径切面为早孕期胎儿全身区域对应的标准切面，因而要求三维超声数据需包括早孕期胎儿全身区域的三维超声数据。
在一些实施例中，获取三维超声数据之后，可以对三维超声数据使用可视化算法进行绘制，从而获得三维超声图像，并利用显示设备进行显示。所述绘制例如包括表面绘制方法或体绘制方法等，本发明实施例对此不做限制。
在步骤S230，根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向。
其中，确定早孕期胎儿的目标区域的方向可以包括确定早孕期胎儿的头部区域的方向、确定早孕期胎儿的身体区域的方向，或确定早孕期胎儿的头部区域的方向及身体区域的方向。当待提取的标准切面包括早孕期胎儿头部区域对应的标准切面时，则根据早孕期胎儿头部区域的三维超声数据确定早孕期胎儿的头部区域对应方向。头部区域的方向还可以用于提取早孕期胎儿全身区域对应的标准切面，例如顶臀径切面。当待提取的标准切面包括早孕期胎儿身体区域对应的标准切面时，则根据早孕期胎儿身体区域的三维超声数据确定早孕期胎儿的身体区域的方向。
下面，首先对确定早孕期胎儿头部区域的方向的几种实现方式进行示例性的描述。其中，目标区域的方向包括早孕期胎儿的头部区域的第一方向，其中该第一方向为头部区域的左右方向。由于头部区域对应的标准切面一般与头部的左右方向平行或垂直，并且头部的左右方向的确定相对更加准确，因而确定早孕期胎儿的头部区域的方向可以至少包括确定早孕期胎儿头部区域的左右方向。但在其他实施例中，所确定的早孕期胎儿的头部区域的方向也可以是其他方向，例如头部区域的上下方向。
在一个实施例中，可以通过训练好的机器学习模型确定早孕期胎儿的头部区域的方向。
在模型训练之前，首先需要建立一个早孕期胎儿超声数据库，数据库中包括大量的早孕期胎儿的超声数据，每个超声数据均标记了早孕期胎儿头部区域的方向，或者头部区域的位置和方向。其中头部区域的位置可以是头部 的感兴趣区域框(ROI)的位置，如ROI框各顶点的位置或ROI框中心点位置及框的大小，或者，头部区域的位置也可以是胎儿头部区域的具体区域范围；头部区域的方向可以是头部区域的左右方向。
构建好数据库后，以数据库中的超声数据作为训练样本训练传统机器学习模型或深度学习模型，用于预测早孕期胎儿头部区域的位置与方向。具体地，可以根据早孕期胎儿超声数据库中的超声数据，学习一个超声数据到头部区域位置与方向的最优映射函数，使得数据库中早孕数据映射得到的头部区域位置与方向与实际标定的头部区域位置与方向之间的误差最小。对于步骤S220中获取的三维超声数据执行这个最优映射函数，即可获得头部区域位置与方向的预测结果。
传统的机器学习方法包括SVM支持向量机、逻辑回归方法、最小二乘法等。对于传统机器学习方法，首先在早孕期胎儿超声数据库中的超声图像中提取图像特征，如Sift特征、梯度特征、LBP等纹理特征、PCA、LDA、Harr特征、HOG与LOG特征等，再学习图像特征与头部区域位置和方向之间的最优映射函数。而对于深度学习模型来说，则可以训练一个端到端的神经网络作为最优映射函数，直接构建早孕期胎儿的超声数据与其头部区域方向之间的映射关系。
在另一个实施例中，可以根据头部对称性确定头部区域的第一方向，即左右方向。具体地，首先在步骤S220获取的三维超声数据中检测出早孕期胎儿的头部区域，之后根据早孕期胎儿的头部区域的对称性确定早孕期胎儿头部的第一方向，即左右方向。
示例性地，可以采用机器学习方法、颅骨光环检测方法或其他任何合适的三维图像分割方法在三维超声数据中检测出早孕期胎儿的头部。
在机器学习方法中，首先建立一个早孕期胎儿超声数据库，数据库中每一个早孕期胎儿的三维超声数据均标记了头部区域的位置，如头部感兴趣区域框(ROI)的位置或头部的具体区域范围，之后，采用传统机器学习方法或深度学习方法，学习一个最优映射函数，用于从早孕期胎儿的三维超声数据中获取头部区域感兴趣区域框(ROI)的位置或头部的具体区域范围，从而实现早孕期胎儿头部的检测或分割。
在颅骨光环检测方法中，可以采用霍夫变换、RANSAC等方法，在早孕期胎儿的三维超声数据中检测亮度最亮、梯度最强或亮度和梯度加权最大的 圆形或椭圆形、球面或椭球面，以作为早孕期胎儿头部区域的检测结果。
确定早孕期胎儿头部区域以后，则可以根据头部区域的对称性确定其左右方向。示例性地，可以根据早孕期胎儿的头部的对称性确定头部区域的左半球区域和右半球区域，并将头部区域的左半球区域和右半球区域的分界面的法线方向确定为头部区域的左右方向。
其中，可以通过预设搜索、梯度更新、强化学习等方法获得对称性最优的左半球区域和右半球区域。在预设搜索方法中，首先按既定规则生成多个候选的头部左半球区域和右半球区域的分界面，分界面两侧为候选的左半球区域和右半球区域；之后，再逐一判断而在其中寻找到对称性最优的左半球区域和右半球区域。在梯度更新和强化学习方法中，首先取一个固定切面作为左半球区域和右半球区域的初始分界面，再根据对称性函数的梯度方向或强化学习方法得到的迭代方向，不断迭代头部左半球区域和右半球区域分界面，直到左半球区域和右半球区域的对称性达到最优。
以上方法中，可以根据左半球区域和右半球区域灰度值的绝对误差、平方误差、相关函数、相关系数、以及孪生网络、CNN等深度学习网络的输出结果等各种合适的指标对左半球区域和右半球区域的对称性进行评价，以便于寻找到对称性最优的左半球区域和右半球区域。
在另一个实施例中，还可以采用正中矢状面检测的方式确定头部区域的方向，具体地，从步骤S220中获取的三维超声数据中提取出早孕期胎儿头部区域的正中矢状面，并将正中矢状面的法线方向确定为头部区域的左右方向。
其中，示例性地，可以采用预设搜索、梯度更新、强化学习等方法，从早孕期胎儿的三维超声数据中提取正中矢状面。
在预设搜索及梯度更新方法中，首先建立早孕期胎儿切面图像的数据库，数据库中每一张切面图像均为早孕期胎儿三维数据中的一个切面，并标记了该切面是否为正中矢状面。建立数据库后，可以采用传统机器学习方法或深度学习方法，学习一个从早孕期胎儿切面图像到是否为正中矢状面的类别的最优映射函数，用于判断每个切面是否为正中矢状面，或每个切面属于正中矢状面的概率。
在预设搜索方法中，首先按既定规则生成多个候选正中矢状面，然后再将属于正中矢状面的概率最高的一个候选正中矢状面作为正中矢状面检测结果。在梯度更新的方法中，首先取一个固定切面作为正中矢状面的起始位置， 再根据正中矢状面的概率函数的梯度方向，迭代更新切面，直到属于正中矢状面的概率值达到最大，或概率值超过预定阈值，即获得了正中矢状面的检测结果。
在强化学习方法中，首先为早孕期胎儿三维超声数据中的每个切面以及切面的变换(平移、旋转等)确定一个奖励函数，该奖励函数可以与当前切面与实际正中矢状面之间的图像相似度、图像误差、位置与朝向的偏差等有关，而切面变换的奖励函数则是变换前后切面奖励函数的差。实际寻找正中矢状面时，首先取一个固定切面作为正中矢状面的起始位置，再不断迭代更新切面，使最终切面的奖励函数达到最大，即获得了正中矢状面的检测结果。
当目标区域包括身体区域时，目标区域的方向包括以下至少一个：身体区域的第二方向、身体区域的第三方向和身体区域的第四方向，其中身体区域的第二方向、第三方向和第四方向分别为身体区域的上下方向、前后方向和左右方向，且第二方向、第三方向和第四方向之间任意两个相互垂直。确定早孕期胎儿的身体方向包括但不限于以下几种实现方式：
在一种实现方式中，可以通过训练好的机器学习模型确定身体方向。
该方式与上文所述的通过训练好的机器学习模型确定头部方向类似，具体地，首先建立一个早孕期胎儿的超声数据库，数据库中每一个超声数据均标记了早孕期胎儿的身体区域的方向，或早孕期胎儿身体区域的位置与方向。构建好数据库后，再训练传统的机器学习模型或深度学习模型，用于对待识别的早孕期胎儿三维超声数据预测其身体区域的方向，或身体区域的位置与方向。在训练过程中，学习一个早孕期胎儿三维数据到身体区域方向或身体位置与方向的最优映射函数，使得利用该最优映射函数对数据库中的超声数据映射得到的身体区域方向或身体区域位置与方向与实际标定的真值之间的误差最小。对待识别的早孕期胎儿三维超声数据，执行上述的最优映射函数，即可获得其身体区域的方向或身体区域的方向与位置的预测结果。
作为另一种实现方式，可以采用脊柱检测的方式确定身体区域的方向，具体地，在步骤S220中获取的三维超声数据中检测出早孕期胎儿的脊柱区域，并根据脊柱区域的方向确定早孕期胎儿的身体区域的方向。由于脊柱区域在三维超声图像中较为明显，因而根据脊柱区域的位置能够准确地确定早孕期胎儿的身体区域的方向。
检测或分割早孕期胎儿脊柱区域的方法包括机器学习方法或传统的图像 处理方法。机器学习方法与上文中的机器学习方法类似，即预先建立一个早孕期胎儿超声数据库，数据库中每一个三维超声数据均标记了早孕期胎儿脊柱区域对应的位置，如脊柱区域的感兴趣区域框(ROI)或脊柱区域的具体区域范围，之后采用传统机器学习方法或深度学习方法，学习一个最优映射函数，用于从早孕期胎儿的三维超声数据中获取脊柱区域感兴趣区域框(ROI)或脊柱区域的具体区域范围，实现早孕期胎儿脊柱区域的检测或分割。在传统图像处理方法中，可以采用霍夫变换、RANSAC等方法，在早孕期胎儿的三维超声数据中检测亮度最亮的直线或弧线，作为早孕期胎儿脊柱区域的检测结果。
通过脊柱区域的检测结果可以直接确定早孕期胎儿身体区域的上下方向和前后方向。其中，根据脊柱区域的方向确定早孕期胎儿身体区域的上下方向包括：确定贴近脊柱区域的直线，并将该直线的方向确定为早孕期胎儿身体区域的第二方向，即上下方向。示例性地，可以采用最小二乘等方法拟合出最贴近脊柱区域的直线，或者，也可以直接采用上述传统图像处理方法中采用霍夫变换、RANSAC等方法获得的亮度最亮的直线作为最贴近脊柱区域的直线。
根据脊柱区域的检测结果确定早孕期胎儿身体区域的上下方向以后，可以根据该上下方向进一步确定早孕期胎儿身体区域的第三方向，即前后方向。具体地，可以从早孕期胎儿的三维超声数据中提取与上文确定的上下方向相垂直的一个或多个身体横切面，使用机器学习或传统图像处理方法检测身体横切面中身体中心点与脊柱的位置，并将身体中心点与脊柱的位置连线方向确定为早孕期胎儿身体区域的前后方向。在该连线上，脊柱的位置靠近身体横切面的后方，身体中心点的位置靠近身体横切面的前方。
在另一实施例中，也可以直接根据脊柱区域的检测结果确定身体区域的前后方向，而无需基于身体区域的上下方向来确定前后方向。具体地，由于脊柱向身体的后方凸起，因而可以确定贴近脊柱的曲线，并将该曲线的凸起方向确定为早孕期胎儿身体区域的前后方向。其中，可以采用最小二乘等方法拟合出最贴近脊柱的一条弧线作为上述曲线，或者直接获取上述传统图像处理方法中获得的亮度最亮的弧线作为上述曲线。
除此之外，还可以根据早孕期胎儿身体区域的形状特征确定身体区域的方向，即从三维超声数据中检测出早孕期胎儿的身体区域，并根据身体区域 的形状确定所述身体方向。由于早孕期胎儿身体区域的上下方向的形状特征较为明显，而前后方向和左右方向的形状差别较小，因而该方式主要用于确定早孕期胎儿身体区域的上下方向。
具体地，可以采用机器学习或图像处理方法在步骤S220中获取的三维超声数据中确定出早孕期胎儿的身体区域，其中确定的身体区域可以是确定包围早孕期胎儿身体的感兴趣区域框(ROI)，或分割出早孕期胎儿身体的具体区域范围。之后，再根据早孕期胎儿身体区域的检测或分割结果，根据身体区域的形状确定身体区域的长轴，并将该长轴方向确定为早孕期胎儿身体区域的上下方向。示例性地，可以采用主成分分析(PCA)方法确定身体区域的长轴，或检测上述身体区域中距离最远的两个点，两点间的连线即为身体区域的长轴。
至于早孕期胎儿身体区域的第四方向(即左右方向)的确定，可以在采用上述任意方式或其他任何可行的方式确定出早孕期胎儿身体区域的上下方向和前后方向之后，将与上下方向和前后方向相垂直的方向确定为早孕期胎儿身体的左右方向。除此之外，也可以根据早孕期胎儿的一些特定特征结构确定早孕期胎儿身体区域的左右方向，例如，可以在三维超声数据中检测早孕期胎儿身体区域中对称特征结构的区域的位置，并将对称特征结构的区域连线的方向确定为早孕期胎儿身体区域的左右方向。其中，对称特征结构例如为双肾、双肺、左右肋骨等具有对称性的特征结构。或者，也可以检测早孕期胎儿的左右心房，左右心房连线方向与早孕期胎儿身体区域的左右方向一般成45°角，根据该特性也可以确定早孕期胎儿身体区域的左右方向。
示例性地，步骤S230可以在获取到三维超声数据后自动执行，也可以据接收到的用户指令而执行。例如，用户可以通过触发提取标准切面的按键而启动自动提取标准切面的功能，该按键可以是设置在显示器用户交互界面上的虚拟按键，也可以是实体按键。在获取三维超声数据后，当接收到提取标准切面的用户指令时，开始执行步骤S230。在一些实施例中，用户也可以在开始采集超声数据之前启动自动提取标准切面的功能，则在获得三维超声数据之后，自动执行步骤S230。
在一些实施例中，在执行步骤S230之前，还可以根据接收到的用户输入确定待提取的标准切面。例如，可以在用户交互界面上显示顶臀径切面、双顶径切面等十个标准切面的名称，并根据用户的选择确定待提取的标准切面。 或者，可以在用户交互界面上显示头部区域标准切面、身体区域标准切面以及全身区域标准切面的选项，根据用户的选择确定提取头部区域标准切面、身体区域标准切面或全身区域标准切面。
由于根据早孕期胎儿的三维超声数据能够较为准确地确定各个目标区域的方向，因而首先确定目标区域的方向、再根据目标区域的方向进一步提取标准切面，能够提高所提取到的标准切面的质量。
在步骤S240，根据所述目标方向，从所述三维超声数据中提取与所述目标区域对应的至少一个标准切面。
如上所述，步骤S230中确定的目标区域的方向主要包括头部区域的方向和身体区域的方向，则在步骤S240中，可以根据头部区域的方向在包括早孕期胎儿头部区域的三维超声数据中确定头部区域的标准切面，或者，根据身体区域的方向在包括早孕期胎儿身体区域的三维超声数据中确定身体区域的标准切面。其中，头部区域的标准切面包括侧脑室水平横切面、双顶径切面和头颈部正中矢状切面中的至少一个，身体区域的标准切面包括胸径切面、腹围切面、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面和躯干长轴切面中的以少一个。除此之外，标准切面还可以包括早孕期胎儿全身区域的标准切面，例如顶臀径切面(全身正中矢状切面)，顶臀径切面可以根据早孕期胎儿的头部区域的方向或身体区域的方向来确定。
在一个实施例中，根据目标区域的方向提取至少一个标准切面包括：根据步骤S230中确定的目标区域的方向在早孕期胎儿的三维超声数据中确定多个候选切面，并在多个候选切面中挑选出至少一个标准切面。
对于头部区域的标准切面来说，首先根据早孕期胎儿的头部区域的方向，生成一组头部区域的候选切面，再在头部区域的多个候选切面中确定最优的头部区域标准切面。例如，在NT切面及顶臀径切面的检测中，可以将步骤S230中获取的头部左半球区域和头部右半球区域的分界面或正中矢状面直接作为NT切面或顶臀径切面的检测结果；也可以根据步骤S230中确定的头部区域左右方向，生成一组垂直于该方向的平行的多个候选切面，或在一定角度范围内生成一组近似垂直于该方向的多个候选切面，再在多个候选切面中选择最终的NT切面或顶臀径切面。
又例如，在早孕期胎儿头部横切面(例如侧脑室横切面、双顶径切面、小脑横切面等)的检测中，可以根据步骤S230中确定的头部区域左右方向， 生成一组平行于或近似平行于该方向的多个候选切面，再在多个候选切面中选择最终的侧脑室横切面、双顶径切面或小脑横切面。
对于身体区域的标准切面来说，与头部区域的标准切面类似，首先根据早孕期胎儿的身体区域的方向，生成一组身体区域的候选切面，再在身体区域的多个候选切面中确定最优的身体区域标准切面。
例如，在确定早孕期胎儿身体区域的横切面，如四腔心切面、胃泡切面、腹壁脐带插入处切面、膀胱切面等时，可以根据步骤S230中确定的早孕期胎儿身体区域的上下方向，生成一组垂直于该上下方向的多个平行切面作为候选切面，或在一定角度范围内生成一组近似垂直于该上下方向的多个切面作为候选切面，再在多个候选切面中挑选出标准的横切面。
又例如，在确定早孕期胎儿身体区域的冠状切面，如双肾冠状切面、脊柱冠状切面等时，可以根据步骤S230中确定的身体区域的左右方向或上下方向，生成一组平行于该方向的候选切面，或在一定角度范围内生成一组近似平行于该方向的候选切面，再在多个候选切面中挑选出标准的冠状切面。
此外，在确定早孕期胎儿身体的正中矢状面时，可以根据步骤S230中确定的身体区域的上下方向或前后方向，生成一组平行于该方向的候选切面，或在一定角度范围内生成一组近似平行于该方向的候选切面，再在多个候选切面中挑选出标准的正中矢状面。
生成候选切面以后，在一个实施例中，可以采用训练好的机器学习模型判断与目标区域对应的每个候选切面作为标准切面的概率，并将概率满足第一阈值的候选切面确定为与目标区域对应的至少一个标准切面。示例性地，若存在至少两个概率满足第一阈值的候选切面，则可以根据接收到的选择操作确定其中最终选定的标准切面，即由用户在概率满足第一阈值的至少两个候选切面中选择最佳的标准切面。其中，所述机器学习模型可以为传统的机器学习模型或深度学习模型。
可选地，也可以采用训练好的机器学习模型判断与目标区域对应的每个所述候选切面作为标准切面的概率，并将概率最高的候选切面确定为与所述目标区域对应的一个标准切面，即由系统直接确定最终的标准切面，从而简化操作流程。在另一个实施例中，在多个候选切面中挑选出标准切面包括：在与目标区域对应的多个候选切面上检测与目标区域对应的至少一个标准切面对应的早孕目标特征结构，并将早孕目标特征结构存在的概率满足第二阈 值的候选切面或早孕目标特征结构存在的概率最高的候选切面确定为标准切面。其中，顶臀径切面对应的早孕目标特征结构包括鼻骨、生殖嵴等；侧脑室水平横切面对应的早孕目标特征结构包括大脑镰、侧脑室、脉络丛等；双顶径切面对应的早孕目标特征结构包括丘脑、颅骨光环等；头颈部正中矢状面对应的早孕目标特征结构包括颈后透明层、鼻骨等；胸径切面对应的早孕目标特征结构包括四腔心等；腹围切面对应的早孕目标特征结构包括胃泡等；腹壁脐带插入处切面对应的早孕目标特征结构包括脐带插入处等；膀胱切面时对应的早孕目标特征结构包括双腿或膀胱等；脊柱纵轴切面对应的早孕目标特征结构包括脊柱、皮肤缘等；躯干长轴切面对应的早孕目标特征结构包括双肾、胃泡、脊柱等。
例如，对于头部区域的标准切面来说，在根据头部区域左右方向生成一组平行于或近似平行于该方向的多个候选切面以后，可以在多个候选切面上分别检测与头部区域的标准切面对应的早孕目标特征结构，例如与侧脑室横切面对应的侧脑室或脉络丛、与双顶径切面对应的丘脑、与小脑横切面的小脑等，将早孕目标特征结构存在概率最高的候选切面作为相应的头部标准切面的检测结果。
作为根据目标区域的方向提取与目标区域对应的标准切面的另外一种实现方式，可以首先在步骤S220中获取的三维超声数据中检测出与待确定的标准切面相对应的早孕目标特征结构，再结合上文中所确定的目标区域的方向，将与早孕目标特征结构至少部分重合、并且与所述目标区域的方向之间的夹角满足预设要求的切面确定为与目标区域对应的至少一个标准切面。进一步地，可以将与早孕目标特征结构基本重合、并且与已确定的目标区域的方向之间的夹角为0度的切面确定为与目标区域对应的一个标准切面。或者，可以将与早孕目标特征结构基本重合、并且与已确定的目标区域的方向之间的夹角为90度的切面确定为与目标区域对应的一个标准切面。
其中，对于头部区域的标准切面来说，可以检测与待确定的头部标准切面相对应的早孕目标特征结构，再结合早孕期胎儿的头部区域的方向，确定相应的头部区域标准切面。作为示例，与侧脑室水平横切面对应的早孕目标特征结构包括大脑镰、侧脑室、脉络丛；与双顶径切面对应的早孕目标特征结构包括丘脑、颅骨光环；与头颈部正中矢状面对应的早孕目标特征结构包括颈后透明层、鼻骨。除此之外，全身区域的标准切面也可以根据头部区域 的方向和头部区域的早孕目标特征结构来确定，当标准切面为顶臀径切面时，对应的早孕目标特征结构包括鼻骨、生殖嵴。
例如，在NT标准切面(即头颈部正中矢状切面)的检测中，可以在三维超声数据中检测鼻骨、NT(颈后透明层)、IT(颅内透明层)、颅后窝池中一个或多个特征结构，再确定与以上特征结构至少部分重合且与头部区域的左右方向相垂直的切面，作为NT标准切面的检测结果。在顶臀径切面的检测中，可以检测鼻骨、NT(颈后透明层)、膀胱、生殖嵴中一个或多个特征结构，再在三维超声数据中确定与以上特征结构至少部分重合且与头部区域的左右方向相垂直的切面，作为顶臀径切面的检测结果。
在早孕期胎儿头部横切面的检测中，可以在三维超声数据中检测与侧脑室横切面对应的侧脑室和/或脉络丛、与双顶径切面对应的丘脑和/或大脑脚、与小脑横切面对应的丘脑、小脑和/或IT(颅内透明层)，再确定一个或多个切面，使其分别经过以上特征结构且与头部区域的左右方向相平行，以分别作为侧脑室横切面、双顶径切面或小脑横切面的检测结果。
类似地，对于身体区域的标准切面来说，可以检测与待确定的身体区域的标准切面相对应的早孕目标特征结构，再结合早孕期胎儿的身体区域的方向，确定相应的身体区域的标准切面。
例如，在早孕期胎儿身体区域的正中矢状面的检测中，若步骤S230中根据脊柱区域位置确定身体区域的方向，可以根据脊柱区域位置或其拟合的弧线，确定一个最贴近脊柱或其拟合的弧线、且平行或近似平行于身体区域上下方向或前后方向(即与上下方向或前后方向的夹角为0°)的切面，并将该切面作为早孕期胎儿身体区域正中矢状面的检测结果。
在早孕期胎儿身体区域的横切面的检测中，可以检测与四腔心切面对应的心脏、与胃泡切面对应的胃泡、与腹壁脐带插入处切面对应的脐带及其腹壁插入处、与膀胱切面对应的膀胱或脐动脉，再确定一个或多个切面，使其分别与以上早孕目标特征结构至少部分重合、且与步骤S230中确定的早孕期胎儿身体区域的上下方向相垂直或近似垂直，并以该切面作为四腔心切面、胃泡切面、腹壁脐带插入处切面或膀胱切面的检测结果。
在早孕期胎儿身体区域的冠状面的检测中，可以检测与双肾冠状切面对应的双肾、胃泡或脊柱、与脊柱冠状切面对应的脊柱或肋骨，再确定一个或多个切面，使其分别与以上早孕目标特征结构重合、且与步骤S230中确定的 早孕期胎儿身体区域的左右方向或上下方向相平行，并以该切面作为双肾冠状切面或脊柱冠状切面的检测结果。此外，在一些实施例中，可以根据已确定的一个或多个标准切面，以及早孕目标特征结构在已确定的标准切面中所在的位置，确定其他的标准切面。
例如，在早孕期胎儿头部横切面的检测中，可以在上文中获取的头部左半球区域和右半球区域的分界面、正中矢状面、NT标准切面或顶臀径切面上，检测与以上早孕期胎儿头部横切面相关的早孕目标特征结构的位置，例如脉络丛、丘脑、第三脑室、小脑蚓部、IT(颅内透明层)等，再确定一个或多个切面，使其分别经过以上特征结构且与头部左半球区域和右半球区域的分界面、正中矢状面、NT切面或顶臀径切面相垂直，以作为头部横切面的检测结果。
类似地，对于早孕期胎儿身体区域的标准切面，例如在双肾冠状切面的检测中，可以在已检测到的胃泡切面中确定胃泡及脊柱的位置，再确定一个切面，使其经过胃泡或脊柱且与胃泡切面相垂直，并将该切面作为双肾冠状切面的检测结果。在脊柱冠状切面的检测中，可以在已检测到的一个或多个身体区域横切面中确定脊柱及肋骨的位置，再确定一个切面，使其经过脊柱或肋骨且与身体横切面相垂直，并将该切面作为脊柱冠状切面的检测结果。
步骤S250，显示所述至少一个标准切面。
其中，所显示的标准切面可以是步骤S240中提取的全部的标准切面，也可以是其中的部分标准切面。示例性地，在显示步骤S240中提取的标准切面时，可以根据接收到的用户指令显示其中的部分标准切面，例如，可以在显示界面上显示标准切面的名称或缩略图，并根据用户的选择显示相应的标准切面。
在一些实施例中，若在步骤S230或步骤S240中，在三维超声数据中对早孕目标特征结构进行了检测，则还可以显示从三维超声数据中检测出的早孕目标特征结构，以供用户进行对照分析。
除此之外，还可以将步骤S230中所确定的目标方向进行显示，以便于用户结合目标方向查看与之相关联的标准切面。作为示例，可以在显示三维超声图像的同时显示目标方向。例如，可以通过箭头等图形标记显示早孕期胎儿头部区域的左右方向，或早孕期胎儿身体区域的上下方向、前后方向和左右方向。
在一些实施例中，还可以显示所确定的各个标准切面的名称，以便于用户直观地确定标准切面的类型，或根据标准切面的名称选择所要查看的标准切面。标准切面的名称可以与标准切面同步显示，或者，可以首先在显示界面上显示各个标准切面的名称，当用户选中标准切面的名称时，显示相应的标准切面。
综上所述，本申请实施例的早孕期胎儿的超声成像方法200能够根据早孕期胎儿的目标区域的方向自动确定与目标区域对应的至少一个标准切面，无需医生逐一手动提取标准切面，极大地优化了产前检查的工作流，有效地提升工作效率，并且能够提高所获取的标准切面质量的稳定性，促进早孕结构筛查的推广与应用。
现在重新参照图1，本申请实施例所提供的超声成像系统100可以用于实现上述早孕期胎儿的超声成像方法200。如上所述，超声成像系统100可以包括超声探头110、发射/接收电路112、处理器114以及显示器116，各个部件的相关描述可以参照上文。
当用于实现超声成像方法200时，发射/接收电路112用于激励所述超声探头110向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；处理器114，用于：基于所述超声回波信号获得所述早孕期胎儿的三维超声数据；根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向；根据所述目标区域的方向，从所述三维超声数据中提取与所述目标区域对应的至少一个标准切面；显示器116用于显示该至少一个标准切面。
作为示例，目标区域包括以下至少一个：头部区域、身体区域和全身区域。
在一个实施例中，目标区域包括头部区域，目标区域的方向包括所述头部区域的第一方向，其中所述第一方向为头部区域的左右方向。
在另一个实施例中，目标区域包括身体区域，目标区域的方向包括以下至少一个：身体区域的第二方向、第三方向和第四方向，其中所述第二方向、第三方向和第四方向分别为身体区域在所述三维超声数据中的上下方向、前后方向和左右方向，且所述第二方向、第三方向和第四方向之间任意两个相互垂直。
作为示例，根据目标区域的方向，从三维超声数据中提取与目标区域对应的至少一个标准切面，包括：根据目标区域的方向在三维超声数据中确定 与目标区域对应的多个候选切面；在与目标区域对应的多个候选切面中挑选出与目标区域对应的至少一个标准切面。
作为示例，根据目标区域的方向，从三维超声数据中提取与目标区域对应的至少一个标准切面，包括：在三维超声数据中检测与目标区域对应的标准切面相对应的早孕目标特征结构；将与早孕目标特征结构至少部分重合、并且与目标区域的方向之间的夹角满足预设要求的切面确定为与目标区域对应的至少一个标准切面。
在一个实施例中，显示器116还用于显示所述早孕目标特征结构。显示器116还可以用于显示所确定的目标区域的方向。显示器116还可以用于显示所确定的标准切面的名称。
以上仅描述了超声成像系统100各部件的主要功能，更多细节参见对早孕期胎儿的超声成像方法200进行的相关描述。本申请实施例的超声成像系统100能够自动确定早孕期胎儿的标准切面，提高了工作效率和标准切面的质量。
下面，将参考图3描述根据本申请另一实施例的早孕期胎儿的超声成像方法。图3是本申请实施例的早孕期胎儿的超声成像方法300的一个示意性流程图。
如图3所示，所述早孕期胎儿的超声成像方法300包括如下步骤：
在步骤S310，向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号；
在步骤S320，根据所述超声回波信号获得早孕期胎儿的三维超声数据；
在步骤S330，从所述三维超声数据中检测至少两个不同的早孕目标特征结构的区域；
在步骤S340，确定与所述至少两个不同的早孕目标特征结构的区域中每个所述早孕目标特征结构的区域均至少部分重合的至少一个切面，以作为所述早孕期胎儿的至少一个标准切面；
在步骤S350，显示所述至少一个标准切面。
根据本申请实施例的早孕期胎儿的超声成像方法300中的步骤S310、步骤S320以及步骤S350与参考图2描述的超声成像方法200中的步骤S210、步骤S220以及步骤S250大体上类似，为了简洁，此处不再赘述相同的细节 内容，以下主要对步骤S330和步骤S340中根据三维超声数据确定标准切面的方式进行详细描述。
与超声成像方法200类似，早孕期胎儿的超声成像方法300中，所提取的标准切面包括早孕期胎儿的头部区域的标准切面、身体区域的标准切面或全身区域的标准切面中的至少一个。当标准切面包括头部区域的标准切面时，步骤S320中获取的三维超声数据至少包括早孕期胎儿头部区域的三维超声数据。当标准切面包括身体区域的标准切面时，步骤S320中获取的三维超声数据至少包括早孕期胎儿身体区域的三维超声数据。当标准切面包括全身区域的标准切面时，步骤S320中获取的三维超声数据包括早孕期胎儿全身区域的三维超声数据。
具体地，标准切面可以包括以下标准切面中的至少一个：顶臀径切面、侧脑室水平横切面、双顶径切面、NT标准切面、胸径切面、腹围切面、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面、躯干长轴切面。其中，早孕期胎儿头部区域的标准切面包括侧脑室水平横切面、双顶径切面及NT标准切面，要求三维超声数据至少包括早孕期胎儿头部区域的三维超声数据；早孕期胎儿身体区域的标准切面包括胸径切面、腹围切面、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面以及躯干长轴切面，要求三维超声数据至少包括早孕期胎儿身体区域的三维超声数据；而早孕期胎儿全身区域的标准切面包括顶臀径切面，要求三维超声数据需包括早孕期胎儿全身区域的三维超声数据。
在步骤S330中，可以采用任何合适的图像检测或分割方法确定早孕目标特征结构的位置。例如，可以采用传统的机器学习方法或深度学习方法，为与各个标准切面对应的特征结构训练一个机器学习模型，以用于确定该特征结构的位置。在模型训练之前，预先建立一个早孕期胎儿超声数据库，数据库中每一个三维超声数据均标记了早孕期胎儿的早孕目标特征结构的位置，例如其感兴趣区域框(ROI)或具体区域范围，之后采用传统机器学习方法或深度学习方法，学习一个最优映射函数，用于从早孕期胎儿的三维超声数据中获取早孕目标特征结构的感兴趣区域框(ROI)或具体区域范围，实现该早孕目标特征结构的检测或分割。
在步骤S340中，对于待提取的标准切面，首先获取标准切面的类型，确定与该标准切面的类型对应的早孕目标特征结构，并从三维超声数据中检测与标准切面的类型相对应的至少两个不同的早孕目标特征结构的区域。之后， 再确定与所述至少两个不同的早孕目标特征结构均至少部分重合的切面，以作为该标准切面的检测结果。具体地，可以确定与至少两个不同的早孕目标特征结构区域中每个早孕目标特征结构的区域均基本重合的一个切面，以作为标准切面。为了使所得到的标准切面更加准确，可以确定与三个或三个以上不同的早孕目标特征结构至少部分重合的切面，以作为标准切面的检测结果。
示例性地，当标准切面为顶臀径切面时，对应的早孕目标特征结构包括鼻骨和生殖嵴；当标准切面为侧脑室水平横切面时，对应的早孕目标特征结构包括大脑镰、侧脑室和脉络丛；当标准切面为双顶径切面时，对应的早孕目标特征结构包括丘脑和颅骨光环；当早孕目标特征结构包括头颈部正中矢状面时，对应的早孕目标特征结构包括颈后透明层和鼻骨；当标准切面为胸径切面时，对应的早孕目标特征结构包括心脏、肋骨和脊柱骨；当标准切面为腹围切面时，对应的早孕目标特征结构包括胃泡、脊柱骨和肝脏；当标准切面为腹壁脐带插入处切面时，对应的早孕目标特征结构包括脐带插入处、脐带、脊柱骨和前腹壁；当标准切面为膀胱切面时，对应的早孕目标特征结构包括双腿和膀胱；当标准切面为脊柱纵轴切面时，对应的早孕目标特征结构包括脊柱和皮肤缘；当标准切面为躯干长轴切面时，对应的早孕目标特征结构包括双肾、胃泡和脊柱。
在步骤S350中，所显示的标准切面可以是步骤S340中提取的标准切面中的部分或全部标准切面。除了显示所提取的标准切面以外，还可以显示标准切面的名称，此外，还可以显示从三维超声数据中检测出的早孕目标特征结构，早孕目标特征结构可以显示在早孕期胎儿的三维超声图像中，例如，显示包围早孕目标特征结构的ROI框或显示早孕目标特征结构的轮廓。
现在重新参照图1，本申请实施例所提供的超声成像系统100可以用于实现上述早孕期胎儿的超声成像方法300。如上所述，超声成像系统100可以包括超声探头110、发射/接收电路112、处理器114以及显示器116，各个部件的相关描述可以参照上文。
当用于实现超声成像方法300时，发射/接收电路112用于激励所述超声探头110向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；处理器114用于：基于所述超声回波信号获得所述早孕期胎儿的三维超声数据；从所述三维超声数据中检测至少两个早孕目标特征结构的区 域；确定与所述至少两个不同的早孕目标特征结构中每个所述早孕目标特征结构均至少部分重合的至少一个切面，以作为所述早孕期胎儿的至少一个标准切面；显示器116用于显示所述至少一个标准切面。
以上仅描述了超声成像系统100各部件的主要功能，更多细节参见对早孕期胎儿的超声成像方法300进行的相关描述。
本申请实施例的早孕期胎儿的超声成像方法300以及超声成像系统能够根据早孕期胎儿的早孕目标特征结构自动确定早孕期胎儿的标准切面，无需医生逐一手动提取标准切面，极大地优化了产前检查的工作流，有效地提升工作效率，并且能够提高所获取的标准切面质量的稳定性，促进早孕结构筛查的推广与应用。
下面，将参考图4描述根据本申请另一实施例的早孕期胎儿的超声成像方法。图4是本申请实施例的早孕期胎儿的超声成像方法400的一个示意性流程图。
如图4所示，本申请一个实施例的早孕期胎儿的超声成像方法400包括如下步骤：
在步骤S410，向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号；
在步骤S420，根据所述超声回波信号获得早孕期胎儿的三维超声数据；
在步骤S430，将所述三维超声数据与预先配置好早孕期标准切面的超声数据模板进行匹配，并根据匹配结果确定所述三维超声数据中的目标标准切面，其中所述早孕期标准切面包括以下至少一个：侧脑室水平横切面、双顶径切面、头颈部正中矢状切面、胸径切面、腹围切面、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面、躯干长轴切面和顶臀径切面；
在步骤S440，显示所述目标标准切面。
根据本申请实施例的早孕期胎儿的超声成像方法400中的步骤S410、步骤S420以及步骤S440与参考图2描述的超声成像方法200中的步骤S210、步骤S220以及步骤S250大体上类似，为了简洁，此处不再赘述相同的细节内容，以下主要对步骤S430中根据三维超声数据确定标准切面的方式进行详细描述。
在步骤S430中，通过将早孕期胎儿的三维超声数据与预先配置好早孕期 标准切面的超声数据模板进行匹配来确定标准切面，示例性地，所述预先配置好早孕期标准切面的超声数据模板包括匹配包括预先配置好早孕期标准切面的三维数据模板(以下称之为标准三维数据模板)、预先配置好早孕期标准切面的二维数据模板(以下称之为标准二维切面模板)以及预先配置好早孕期标准切面的早孕目标特征结构模板数据模板(以下称之为标准早孕目标特征结构模板)以及预先配置好早孕期标准切面的关键点模板(以下称之为标准关键点模板)。与模板进行的匹配包括以下至少一种：与标准三维数据模板进行的三维数据匹配、与标准二维切面模板进行的二维切面匹配、与标准早孕目标特征结构模板进行的早孕目标特征结构匹配以及与标准关键点模板进行的关键点匹配。
作为示例，将早孕期胎儿的三维数据与标准三维数据模板相匹配，包括寻找一个最优的三维空间变换关系，使步骤S420中获取的三维超声数据与标准三维数据模板相似度最高或相差最小。在另一个示例中，也可以先对步骤S420中获取的早孕期胎儿的三维超声数据与标准三维数据模板进行图像特征提取，例如梯度特征、LBP等纹理特征、Harr特征、HOG/LOG特征等，再寻找一个最优的三维空间变换关系，使从早孕期胎儿的三维超声数据与标准三维数据模板中所提取的图像特征之间的相似度最高或相差最小。完成匹配后，可以根据预先在标准三维数据模板中配置好的早孕期标准切面的位置以及匹配获得的三维空间变换关系，确定早孕期胎儿的三维超声数据中目标标准切面对应的位置。
作为示例，将早孕期胎儿的三维超声数据与标准二维切面模板相匹配包括在步骤S420中获取的早孕期胎儿的三维超声数据中寻找一个最优的二维切面，使得该二维切面与标准二维切面模板之间相似度最高或相差最小，或是使从该最优的二维切面与标准二维切面模板中所提取的图像特征之间的相似度最高或相差最小，图像特征包括但不限于梯度特征、LBP等纹理特征、Harr特征、HOG/LOG特征。完成匹配后，可以将从三维数据中寻找到的最优的二维切面作为目标标准切面。
作为示例，将早孕期胎儿的三维超声数据与标准早孕目标特征结构模板相匹配，包括在步骤S420中获取的三维超声数据中寻找一个最优的图像块，使该图像块与标准早孕目标特征结构模板之间的相似度最高或相差最小，或从该图像块和标准早孕目标特征结构模板中提取的图像特征之间的相似度最 高或相差最小。
在另一示例中，早孕目标特征结构匹配还可以包括使用Faster RCNN、Mask RCNN、SSD、YOLO、Retinanet、Efficientnet、Cornernet、Centernet、FCOS等目标检测方法，在步骤S420中获取的三维超声数据中检测候选的早孕目标特征结构区域，再与标准早孕目标特征结构模板相匹配。
其中，与标准早孕目标特征结构模板相匹配可以包括在三维超声数据中寻找一个最优的候选早孕目标特征结构，使其与标准早孕目标特征结构模板之间的相似度最高或相差最小；也可以包括提取候选早孕目标特征结构与标准早孕目标特征结构模板的图像特征，再寻找一个最优的候选早孕目标特征结构，使从中提取的图像特征与从标准早孕目标特征结构模板中提取的图像特征之间的相似度最高或相差最小；或者，还可以包括寻找最优的候选早孕目标特征结构和一个最优的空间变换，使候选早孕目标特征结构与标准早孕目标特征结构模板的空间位置相差最小。
完成匹配后，可以根据三位超声数据中早孕目标特征结构的位置以及标准切面所对应的早孕目标特征结构确定目标标准切面的位置。具体地，当早孕期标准切面为顶臀径切面时，对应的早孕目标特征结构包括鼻骨、生殖嵴；当早孕期标准切面为侧脑室水平横切面时，对应的早孕目标特征结构包括大脑镰、侧脑室、脉络丛；当早孕期标准切面为双顶径切面时，对应的早孕目标特征结构包括丘脑、颅骨光环；当早孕期标准切面包括头颈部正中矢状面时，对应的早孕目标特征结构包括颈后透明层、鼻骨；当早孕期标准切面为胸径切面时，对应的早孕目标特征结构包括四腔心；当早孕期标准切面为腹围切面时，对应的早孕目标特征结构包括胃泡；当早孕期标准切面为腹壁脐带插入处切面时，对应的早孕目标特征结构包括脐带插入处；当早孕期标准切面为膀胱切面时，对应的早孕目标特征结构包括双腿或膀胱；当早孕期标准切面为脊柱纵轴切面时，对应的早孕目标特征结构包括脊柱、皮肤缘；当早孕期标准切面为躯干长轴切面时，对应的早孕目标特征结构包括双肾、胃泡、脊柱。或者，也可以根据所获得的最优的空间变换，根据标准早孕目标特征结构模板中早孕期标准切面的位置，确定三维超声数据中目标标准切面所对应的位置。
作为示例，将三维超声数据与标准关键点模板相匹配，包括在步骤S420中获取的三维超声数据中寻找一个或多个最优点，使该最优点附近的图像特 征与标准关键点模板的图像特征之间的相似度最高或相差最小。
关键点匹配也可以包括使用注入特征点提取方法(例如SIFT方法)、角点检测方法(例如Harris方法)或基于神经网络预测候选关键点坐标或候选关键点所在区域，从而在三维超声数据中确定至少一个候选关键点，再将候选关键点与标准关键点模板相匹配。其中，将候选关键点与标准关键点模板相匹配可以包括寻找一个最优的候选关键点，使其附近的图像特征与标准关键点模板的图像特征之间的相似度最高或相差最小；或者，将候选关键点与标准关键点模板相匹配也可以包括寻找最优的候选关键点和一个最优的空间变换关系，使该最优的候选关键点与标准关键点模板的空间位置相差最小。
完成匹配后，可以根据标准切面对应的关键点位置，确定一个包括这些关键点的切面以作为目标标准切面；也可根据获得的最优空间变换关系，根据标准关键点模板中早孕期标准切面的位置，确定三维超声数据中目标标准切面所对应的位置。
现在重新参照图1，本申请实施例所提供的超声成像系统100可以用于实现上述早孕期胎儿的超声成像方法400。如上所述，超声成像系统100可以包括超声探头110、发射/接收电路112、处理器114以及显示器116，各个部件的相关描述可以参照上文。
当用于实现超声成像方法400时，发射/接收电路112用于激励所述超声探头110向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；处理器114，用于：基于所述超声回波信号获得所述早孕期胎儿的三维超声数据；将所述三维超声数据与预先配置好早孕期标准切面的超声数据模板进行匹配，并根据匹配结果确定所述三维超声数据中的目标标准切面，其中所述早孕期标准切面包括以下至少一个：侧脑室水平横切面、双顶径切面、头颈部正中矢状切面、胸径切面、腹围切面、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面、躯干长轴切面和顶臀径切面；显示器116用于显示所述目标标准切面。
以上仅描述了超声成像系统100各部件的主要功能，更多细节参见对早孕期胎儿的超声成像方法400进行的相关描述。
本申请实施例的早孕期胎儿的超声成像方法400以及超声成像系统能够通过与预先配置后早孕标准切面的超声数据模板进行匹配而自动确定早孕期胎儿的标准切面，无需医生逐一手动提取标准切面，极大地优化了产前检查 的工作流，有效地提升工作效率，并且能够提高所获取的标准切面质量的稳定性，促进早孕结构筛查的推广与应用。
下面，将参考图5描述根据本申请另一实施例的早孕期胎儿的超声成像方法。图5是本申请实施例的早孕期胎儿的超声成像方法500的一个示意性流程图。
如图5所示，本申请一个实施例的早孕期胎儿的超声成像方法500包括如下步骤：
在步骤S510，向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号；
在步骤S520，根据所述超声回波信号获得早孕期胎儿的三维超声数据；
在步骤S530，根据所述三维超声数据提取所述三维超声数据的目标区域的图像特征；
在步骤S540，根据所述目标区域的图像特征确定目标标准切面的法线方向和所述目标标准切面上预设点的位置信息；
在步骤S550，根据所述目标标准切面的法线方向和所述目标标准切面上预设点的位置信息确定所述目标标准切面；
在步骤S560，显示所述目标标准切面。
根据本申请实施例的早孕期胎儿的超声成像方法500中的步骤S510和步骤S520与参考图2描述的超声成像方法200中的步骤S210和步骤S220大体上类似，为了简洁，此处不再赘述相同的细节内容，以下主要对方法500中根据三维超声数据确定标准切面的方式进行详细描述。
在步骤S530中，首先提取早孕期胎儿的三维超声数据中目标区域的图像特征。作为示例，目标区域的图像特征的提取方法包括传统图像处理方法或深度学习方法。其中，传统图像处理方法包括提取Sift特征、梯度特征、LBP等纹理特征、PCA、LDA、Harr特征、HOG与LOG特征等图像特征，也包括图像边缘的提取，如利用canny算子实现边缘提取等；深度学习方法则包括对神经网络模型进行一个或多个特定任务的训练，如回归标准切面的位置与方向、估计早孕期胎儿的位置、大小和/或方向、识别早孕期胎儿早孕目标特征结构和/或标志点等，然后取出神经网络模型中间一个或多个网络节点的输出作为目标区域的图像特征，或取训练好的神经网络中的前端特征提取网 络，以用于提取目标区域的图像特征。
提取目标区域的图像特征之后，可以根据目标区域的图像特征与目标区域的图像特征所对应的标准切面的法线方向之间的映射关系确定目标标准切面的法线方向，以及根据目标区域的图像特征与目标区域的图像特征所对应的标准切面上预设点的位置信息之间的映射关系确定目标标准切面上预设点的位置信息。示例性地，预设点可以是标准切面与X轴、Y轴或Z轴相交的点，或者还可以为标准切面在预设方向上的位置信息，例如可以是标准切面与法线方向的交点。
示例性地，可以采用机器学习模型或深度学习模型回归出目标区域的图像特征与目标图像特征对应的标准切面的法线方向之间的映射关系；也可以采用机器学习模型或深度学习模型回归出目标区域的图像特征与目标区域的图像特征对应的标准切面上预设点之间的映射关系。具体地，预先建立一个早孕期胎儿超声数据库，其中每一个超声数据包括了早孕三维数据和/或其目标区域的图像特征，以及一个或多个标准切面的预设点位置与法线方向。进行模型的训练时，寻找一个从目标区域的图像特征到预设点位置和法线方向的最优映射函数，使得数据库的目标区域的图像特征经过映射函数所得到的预设点位置和法线方向与实际标准切面的预设点位置和法线方向之间的误差最小。使用该映射函数即可根据三维超声图像的目标区域的图像特征预测预设点位置和法线方向。
之后，在步骤S550中即可根据预设点位置和法线方向从三维超声数据中提取标准切面，例如在垂直于法线方向上的全部切面中提取经过预设点的切面。标准切面包括以下切面中的一个或多个：侧脑室水平横切面、双顶径切面、头颈部正中矢状切面、胸径切面、腹围切面、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面、躯干长轴切面和、顶臀径切面。
示例性地，传统的机器学习方法包括SVM支持向量机、最小二乘法、逻辑回归法等，其映射函数包括线性函数、多项式函数、逻辑函数等；深度学习方法则使用深度神经网络作为映射函数，包括但不限于CNN卷积神经网络、MLP多层感知机、RNN循环神经网络等。
现在重新参照图1，本申请实施例所提供的超声成像系统100可以用于实现上述早孕期胎儿的超声成像方法500。如上所述，超声成像系统100可以包括超声探头110、发射/接收电路112、处理器114以及显示器116，各个 部件的相关描述可以参照上文。
当用于实现超声成像方法500时，发射/接收电路112用于激励所述超声探头110向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；处理器114用于：基于所述超声回波信号获得所述早孕期胎儿的三维超声数据；根据所述三维超声数据提取所述三维超声数据的目标区域的图像特征；根据所述目标区域的图像特征确定目标标准切面的法线方向和所述目标标准切面上预设点的位置信息；根据所述目标标准切面的法线方向和所述目标标准切面上预设点的位置信息确定所述目标标准切面；显示器116用于显示所述标准切面。
以上仅描述了超声成像系统100各部件的主要功能，更多细节参见对早孕期胎儿的超声成像方法500进行的相关描述。
本申请实施例的早孕期胎儿的超声成像方法500以及超声成像系统能够根据早孕期胎儿的三维超声数据的目标区域的图像特征自动确定早孕期胎儿的标准切面，无需医生逐一手动提取标准切面，提高了工作效率和标准切面的质量。
下面，将参考图6描述根据本申请另一实施例的早孕期胎儿的超声成像方法。图6是本申请实施例的早孕期胎儿的超声成像方法600的一个示意性流程图。
如图6所示，本申请一个实施例的早孕期胎儿的超声成像方法600包括如下步骤：
在步骤S610，向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号；
在步骤S620，基于所述超声回波信号获得所述早孕期胎儿的三维超声数据；
在步骤S630，根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向；
在步骤S640，显示所述早孕期胎儿的目标区域的方向。
根据本申请实施例的早孕期胎儿的超声成像方法600中确定目标区域的方向的方法与参考图2描述的超声成像方法200中的确定目标区域的方向的方法大体上一致。具体地，目标区域包括以下至少一个：头部区域、身体区 域和全身区域。
在一个实施例中，目标区域包括头部区域，目标区域的方向包括所述头部区域的第一方向，其中所述第一方向为头部区域的左右方向。
示例性地，可以采用机器学习方法或颅骨光环检测方法在三维超声数据中检测出早孕期胎儿的头部区域，根据早孕期胎儿的头部区域的对称性确定头部区域的左半球区域和右半球区域，并将所述头部区域的左半球区域和右半球区域的分界面的法线方向确定为头部区域的第一方向。
或者，可以从三维超声数据中提取出早孕期胎儿的头部区域的正中矢状面；将头部区域的正中矢状面的法线方向确定为头部区域的第一方向。
所述目标区域也可以包括身体区域，则目标区域的方向包括以下至少一个：身体区域的第二方向、身体区域的第三方向和身体区域的第四方向，其中所述第二方向、第三方向和第四方向分别为身体区域的上下方向、前后方向和左右方向，且所述第二方向、第三方向和第四方向之间任意两个相互垂直。
示例性地，可以在三维超声数据中检测出所述早孕期胎儿的脊柱区域；根据脊柱区域的方向确定身体区域的方向。例如，可以确定贴近所述脊柱区域的直线，并将该直线的方向确定为身体区域的第二方向，即上下方向。
确定第二方向之后，可以从早孕期胎儿身体的三维超声数据中提取与第二方向相垂直的一个或多个身体横切面；检测身体横切面中身体中心点与脊柱的位置，并将身体中心点与脊柱的位置连线确定为所述早孕期胎儿身体区域的第三方向，即前后方向。确定前后方向的方法也可以为：确定贴近脊柱区域的曲线，并将曲线的凸起方向确定为早孕期胎儿身体区域的第三方向。
在一个实施例中，确定身体区域的方法可以包括：从三维超声数据中检测出早孕期胎儿的身体区域；根据身体区域的形状确定身体区域的方向。
示例性地，确定身体区域第四方向的方法可以包括：确定早孕期胎儿身体区域的第二方向和第三方向，将与第二方向和第三方向相垂直的方向确定为早孕期胎儿身体区域的第四方向。或者，也可以在三维超声数据中检测早孕期胎儿身体中双肾区域的位置，并将对称特征结构的区域连线的方向确定为早孕期胎儿身体区域的第四方向。
作为示例，确定目标区域的方向以后，还可以根据目标区域的方向，从三维超声数据中提取与目标区域对应的至少一个标准切面，例如，根据目标 区域的方向在三维超声数据中确定与目标区域对应的多个候选切面；在与目标区域对应的多个候选切面中挑选出与目标区域对应的至少一个标准切面。其中，可以在三维超声数据中检测与目标区域对应的标准切面相对应的早孕目标特征结构，并将与早孕目标特征结构至少部分重合、并且与目标区域的方向之间的夹角满足预设要求的切面确定为与目标区域对应的至少一个标准切面。关于根据目标区域的方向确定标准切面的其他具体细节可以参照对超声成像方法200进行的相关描述。
现在重新参照图1，本申请实施例所提供的超声成像系统100可以用于实现上述早孕期胎儿的超声成像方法600。如上所述，超声成像系统100可以包括超声探头110、发射/接收电路112、处理器114以及显示器116，各个部件的相关描述可以参照上文。
当用于实现超声成像方法600时，发射/接收电路112用于激励所述超声探头110向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；处理器114用于：基于所述超声回波信号获得所述早孕期胎儿的三维超声数据；根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向；显示器116用于显示目标区域的方向。
以上仅描述了超声成像系统100各部件的主要功能，更多细节参见对早孕期胎儿的超声成像方法600进行的相关描述。此外，根据本申请实施例，还提供了一种计算机存储介质，在所述计算机存储介质上存储了程序指令，在所述程序指令被计算机或处理器运行时用于执行本申请实施例的方法200、方法300、方法400、方法500或方法600的相应步骤。所述存储介质例如可以包括智能电话的存储卡、平板电脑的存储部件、个人计算机的硬盘、只读存储器(ROM)、可擦除可编程只读存储器(EPROM)、便携式紧致盘只读存储器(CD-ROM)、USB存储器、或者上述存储介质的任意组合。所述计算机可读存储介质可以是一个或多个计算机可读存储介质的任意组合。
此外，根据本申请实施例，还提供了一种计算机程序，该计算机程序可以存储在云端或本地的存储介质上。在该计算机程序被计算机或处理器运行时用于执行本申请实施例的早孕期胎儿的超声成像方法的相应步骤。
基于上面的描述，根据本申请实施例的早孕期胎儿的超声成像方法和超声成像系统，能够根据单次采集的三维超声数据自动确定早孕期胎儿的标准切面，无需医生逐一手动提取标准切面，极大地优化了产前检查的工作流， 有效地提升工作效率，并且能提升所获取的标准切面质量的稳定性，促进早孕结构筛查的推广与应用。
尽管这里已经参考附图描述了示例实施例，应理解上述示例实施例仅仅是示例性的，并且不意图将本申请的范围限制于此。本领域普通技术人员可以在其中进行各种改变和修改，而不偏离本申请的范围和精神。所有这些改变和修改意在被包括在所附权利要求所要求的本申请的范围之内。
本领域普通技术人员可以意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本申请的范围。
在本申请所提供的几个实施例中，应该理解到，所揭露的设备和方法，可以通过其它的方式实现。例如，以上所描述的设备实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个设备，或一些特征可以忽略，或不执行。
在此处所提供的说明书中，说明了大量具体细节。然而，能够理解，本申请的实施例可以在没有这些具体细节的情况下实践。在一些实例中，并未详细示出公知的方法、结构和技术，以便不模糊对本说明书的理解。
类似地，应当理解，为了精简本申请并帮助理解各个发明方面中的一个或多个，在对本申请的示例性实施例的描述中，本申请的各个特征有时被一起分组到单个实施例、图、或者对其的描述中。然而，并不应将该本申请的方法解释成反映如下意图：即所要求保护的本申请要求比在每个权利要求中所明确记载的特征更多的特征。更确切地说，如相应的权利要求书所反映的那样，其发明点在于可以用少于某个公开的单个实施例的所有特征的特征来解决相应的技术问题。因此，遵循具体实施方式的权利要求书由此明确地并入该具体实施方式，其中每个权利要求本身都作为本申请的单独实施例。
本领域的技术人员可以理解，除了特征之间相互排斥之外，可以采用任何组合对本说明书(包括伴随的权利要求、摘要和附图)中公开的所有特征以及如此公开的任何方法或者设备的所有过程或单元进行组合。除非另外明 确陈述，本说明书(包括伴随的权利要求、摘要和附图)中公开的每个特征可以由提供相同、等同或相似目的的替代特征来代替。
此外，本领域的技术人员能够理解，尽管在此所述的一些实施例包括其它实施例中所包括的某些特征而不是其它特征，但是不同实施例的特征的组合意味着处于本申请的范围之内并且形成不同的实施例。例如，在权利要求书中，所要求保护的实施例的任意之一都可以以任意的组合方式来使用。
本申请的各个部件实施例可以以硬件实现，或者以在一个或者多个处理器上运行的软件模块实现，或者以它们的组合实现。本领域的技术人员应当理解，可以在实践中使用微处理器或者数字信号处理器(DSP)来实现根据本申请实施例的一些模块的一些或者全部功能。本申请还可以实现为用于执行这里所描述的方法的一部分或者全部的装置程序(例如，计算机程序和计算机程序产品)。这样的实现本申请的程序可以存储在计算机可读介质上，或者可以具有一个或者多个信号的形式。这样的信号可以从因特网网站上下载得到，或者在载体信号上提供，或者以任何其他形式提供。
应该注意的是上述实施例对本申请进行说明而不是对本申请进行限制，并且本领域技术人员在不脱离所附权利要求的范围的情况下可设计出替换实施例。在权利要求中，不应将位于括号之间的任何参考符号构造成对权利要求的限制。本申请可以借助于包括有若干不同元件的硬件以及借助于适当编程的计算机来实现。在列举了若干装置的单元权利要求中，这些装置中的若干个可以是通过同一个硬件项来具体体现。单词第一、第二、以及第三等的使用不表示任何顺序。可将这些单词解释为名称。
以上所述，仅为本申请的具体实施方式或对具体实施方式的说明，本申请的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本申请揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本申请的保护范围之内。本申请的保护范围应以权利要求的保护范围为准。

Claims (49)

  1. 一种早孕期胎儿的超声成像方法，其特征在于，所述方法包括：
    向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号；
    基于所述超声回波信号获得所述早孕期胎儿的三维超声数据；
    根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向；
    根据所述目标区域的方向，从所述三维超声数据中提取与所述目标区域对应的至少一个标准切面；
    显示所述至少一个标准切面。
  2. 根据权利要求1所述的方法，其特征在于，所述目标区域包括以下至少一个：头部区域、身体区域和全身区域。
  3. 根据权利要求2所述的方法，其特征在于，所述头部区域对应的标准切面包括以下至少一个：侧脑室水平横切面、双顶径切面和头颈部正中矢状切面，所述身体区域对应的标准切面包括以下至少一个：胸径切面、腹围切面、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面、躯干长轴切面，所述全身区域对应的标准切面包括顶臀径切面。
  4. 根据权利要求2所述的方法，其特征在于，所述目标区域包括头部区域，所述目标区域的方向包括所述头部区域的第一方向，其中所述第一方向为头部区域的左右方向。
  5. 根据权利要求4所述的方法，其特征在于，所述根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向包括：
    在所述三维超声数据中检测出早孕期胎儿的头部区域；
    根据所述早孕期胎儿的头部区域的对称性确定所述头部区域的第一方向。
  6. 根据权利要求5所述的方法，其特征在于，所述在所述三维超声数据中检测出早孕期胎儿的头部区域，包括：
    采用机器学习方法或颅骨光环检测方法在所述三维超声数据中检测出早孕期胎儿的头部区域。
  7. 根据权利要求5所述的方法，其特征在于，所述根据所述早孕期胎儿的头部区域的对称性确定所述头部区域的第一方向，包括：
    根据所述早孕期胎儿的头部区域的对称性确定所述头部区域的左半球区域和右半球区域，并将所述头部区域的左半球区域和右半球区域的分界面的法线方向确定为头部区域的第一方向。
  8. 根据权利要求4所述的方法，其特征在于，确定所述头部区域的方向 包括：
    从所述三维超声数据中提取出所述早孕期胎儿的头部区域的正中矢状面；
    将所述头部区域的正中矢状面的法线方向确定为所述头部区域的第一方向。
  9. 根据权利要求2所述的方法，其特征在于，所述目标区域包括身体区域，所述目标区域的方向包括以下至少一个：身体区域的第二方向、身体区域的第三方向和身体区域的第四方向，其中所述第二方向、第三方向和第四方向分别为身体区域的上下方向、前后方向和左右方向，且所述第二方向、第三方向和第四方向之间任意两个相互垂直。
  10. 根据权利要求9所述的方法，其特征在于，根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向包括：
    在所述三维超声数据中检测出所述早孕期胎儿的脊柱区域；
    根据所述脊柱区域的方向确定所述身体区域的方向。
  11. 根据权利要求10所述的方法，其特征在于，所述根据所述脊柱区域的方向确定所述身体区域的方向包括：
    确定贴近所述脊柱区域的直线，并将所述直线的方向确定为所述身体区域的第二方向。
  12. 根据权利要求11所述的方法，其特征在于，所述根据所述脊柱区域的方向确定所述身体区域方向包括：
    从所述早孕期胎儿的三维超声数据中提取与所述第二方向相垂直的一个或多个身体横切面；
    检测所述身体横切面中身体中心点与脊柱的位置，并将所述身体中心点与所述脊柱的位置连线确定为所述早孕期胎儿身体区域的第三方向。
  13. 根据权利要求10所述的方法，其特征在于，所述根据所述脊柱区域的方向确定所述身体区域方向包括：
    确定贴近所述脊柱区域的曲线，并将所述曲线的凸起方向确定为所述早孕期胎儿身体区域的第三方向。
  14. 根据权利要求9所述的方法，其特征在于，根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向包括：
    从所述三维超声数据中检测出所述早孕期胎儿的身体区域；
    根据所述身体区域的形状确定所述身体区域的方向。
  15. 根据权利要求14所述的方法，其特征在于，所述根据所述身体区域的形状确定所述身体区域的方向，包括：
    根据所述身体区域的形状确定所述身体区域的长轴；
    将所述长轴的方向确定为所述早孕期胎儿身体区域的第二方向。
  16. 根据权利要求9-15中任一项所述的方法，其特征在于，根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向包括：
    确定所述早孕期胎儿身体区域的第二方向和第三方向，将与所述第二方向和所述第三方向相垂直的方向确定为所述早孕期胎儿身体区域的第四方向。
  17. 根据权利要求9-15中任一项所述的方法，其特征在于，根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向包括：
    在所述三维超声数据中检测所述早孕期胎儿身体中对称特征结构的区域的位置，并将对称特征结构的区域连线的方向确定为所述早孕期胎儿身体区域的第四方向。
  18. 根据权利要求1-17中任一项所述的方法，其特征在于，所述根据所述目标区域的方向，从所述三维超声数据中提取与所述目标区域对应的至少一个标准切面，包括：
    根据所述目标区域的方向在所述三维超声数据中确定与所述目标区域对应的多个候选切面；
    在所述与所述目标区域对应的多个候选切面中挑选出与所述目标区域对应的至少一个标准切面。
  19. 根据权利要求18所述的方法，其特征在于，所述在所述与所述目标区域对应的多个候选切面中挑选出与所述目标区域对应的至少一个标准切面，包括：
    采用训练好的机器学习模型判断与所述目标区域对应的每个所述候选切面作为标准切面的概率，并将概率满足第一阈值的所述候选切面确定为与所述目标区域对应的至少一个标准切面。
  20. 根据权利要求18所述的方法，其特征在于，所述在所述与所述目标区域对应的多个候选切面中挑选出与所述目标区域对应的至少一个标准切面，包括：
    采用训练好的机器学习模型判断与所述目标区域对应的每个所述候选切面作为标准切面的概率，并将概率最高的所述候选切面确定为与所述目标区域对应的一个标准切面。
  21. 根据权利要求18所述的方法，其特征在于，所述在所述与所述目标区域对应的多个候选切面中挑选出与所述目标区域对应的至少一个标准切面，包括：
    在所述与所述目标区域对应的多个候选切面上检测与所述目标区域对应的至少一个标准切面对应的早孕目标特征结构，并将早孕目标特征结构存在的概率满足第二阈值的所述候选切面确定为与所述目标区域对应的至少一个标准切面。
  22. 根据权利要求1-17中任一项所述的方法，其特征在于，所述根据所述目标区域的方向，从所述三维超声数据中提取与所述目标区域对应的至少一个标准切面，包括：
    在所述三维超声数据中检测与所述目标区域对应的至少一个标准切面相对应的早孕目标特征结构；
    将与所述早孕目标特征结构至少部分重合、并且与所述目标区域的方向之间的夹角满足预设要求的切面确定为与所述目标区域对应的至少一个标准切面。
  23. 根据权利要求22中所述的方法，其特征在于，所述将与所述早孕目标特征结构至少部分重合、并且与所述目标区域的方向之间的夹角满足预设要求的切面确定为与所述目标区域对应的至少一个标准切面，包括：
    将与所述早孕目标特征结构基本重合、并且与所述目标区域的方向之间的夹角为0度的切面确定为与所述目标区域对应的一个标准切面。
  24. 一种早孕期胎儿的超声成像方法，其特征在于，所述方法包括：
    向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号；
    根据所述超声回波信号获得早孕期胎儿的三维超声数据；
    从所述三维超声数据中检测至少两个不同的早孕目标特征结构的区域；
    确定与所述至少两个不同的早孕目标特征结构的区域中每个所述早孕目标特征结构的区域均至少部分重合的至少一个切面，以作为所述早孕期胎儿的至少一个标准切面；
    显示所述至少一个标准切面。
  25. 根据权利要求24所述的方法，其特征在于，所述从所述三维超声数据中检测至少两个不同的早孕目标特征结构的区域包括：
    获取标准切面的类型；
    从所述三维超声数据中检测与所述标准切面的类型相对应的至少两个不同的早孕目标特征结构的区域。
  26. 根据权利要求24所述的方法，其特征在于，所述方法还包括：
    显示所述至少两个不同的早孕目标特征结构的区域。
  27. 根据权利要求24所述的方法，其特征在于，所述确定与所述至少两个不同的早孕目标特征结构区域中每个所述早孕目标特征结构的区域均至少部分重合的至少一个切面，以作为所述早孕期胎儿的至少一个标准切面，包括：
    确定与所述至少两个不同的早孕目标特征结构区域中每个所述早孕目标特征结构的区域均基本重合的一个切面，以作为所述早孕期胎儿的一个标准切面。
  28. 根据权利要求24或25所述的方法，其特征在于，所述标准切面包括以下至少一种：头部区域的标准切面、身体区域的标准切面和全身区域的标准切面；所述头部区域的标准切面包括以下至少一个：侧脑室水平横切面、双顶径切面和头颈部正中矢状切面，所述身体区域的标准切面包括以下至少一个：胸径切面、腹围切面、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面和躯干长轴切面，所述全身区域的标准切面包括顶臀径切面。
  29. 根据权利要求28所述的方法，其特征在于，当所述标准切面为所述顶臀径切面时，对应的所述早孕目标特征结构包括鼻骨和生殖嵴；当所述标准切面为侧脑室水平横切面时，对应的所述早孕目标特征结构包括大脑镰、侧脑室和脉络丛；当所述标准切面为双顶径切面时，对应的所述早孕目标特征结构包括丘脑和颅骨光环；当所述标准切面为头颈部正中矢状面时，对应的所述早孕目标特征结构包括颈后透明层和鼻骨；当所述标准切面为胸径切面时，对应的所述早孕目标特征结构包括心脏、肋骨和脊柱骨；当所述标准切面为腹围切面时，对应的所述早孕目标特征结构包括胃泡、脊柱骨和肝脏；当所述标准切面为腹壁脐带插入处切面时，对应的所述早孕目标特征结构包括脐带插入处、脐带、脊柱骨和前腹壁；当所述标准切面为所述膀胱切面时，对应的所述早孕目标特征结构包括双腿和膀胱；当所述标准切面为脊柱纵轴切面时，对应的所述早孕目标特征结构包括脊柱和皮肤缘；当所述标准切面为躯干长轴切面时，对应的所述早孕目标特征结构包括双肾、胃泡和脊柱。
  30. 一种早孕期胎儿的超声成像方法，其特征在于，所述方法包括：
    向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号；
    根据所述超声回波信号获得早孕期胎儿的三维超声数据；
    将所述三维超声数据与预先配置好早孕期标准切面的超声数据模板进行匹配，并根据匹配结果确定所述三维超声数据中的目标标准切面，其中所述早孕期标准切面包括以下至少一个：侧脑室水平横切面、双顶径切面、头颈 部正中矢状切面、胸径切面、腹围切面、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面、躯干长轴切面和顶臀径切面；
    显示所述目标标准切面。
  31. 根据权利要求30所述的方法，其特征在于，所述匹配包括以下至少一种：三维数据匹配、二维切面匹配、早孕目标特征结构匹配和关键点匹配。
  32. 一种早孕期胎儿的超声成像方法，其特征在于，所述方法包括：
    向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号；
    根据所述超声回波信号获得所述早孕期胎儿的三维超声数据；
    根据所述三维超声数据提取所述三维超声数据的目标区域的图像特征；
    根据所述目标区域的图像特征确定目标标准切面的法线方向和所述目标标准切面上预设点的位置信息；
    根据所述目标标准切面的法线方向和所述目标标准切面上预设点的位置信息确定所述目标标准切面；
    显示所述目标标准切面。
  33. 根据权利要求32所述的方法，其特征在于，所述根据所述目标区域的图像特征确定目标标准切面的法线方向和所述目标标准切面上预设点的位置信息包括：
    根据所述目标区域的图像特征与所述目标区域的图像特征对应的标准切面的法线方向之间的映射关系确定目标标准切面的法线方向；
    根据所述目标区域的图像特征与所述目标区域的图像特征对应的标准切面上预设点的位置信息之间的映射关系确定所述目标标准切面上预设点的位置信息。
  34. 根据权利要求33所述的方法，其特征在于，所述方法还包括：
    采用机器学习模型或深度学习模型回归出所述目标区域的图像特征与所述目标区域的图像特征对应的标准切面的法线方向之间的映射关系；
    采用机器学习模型或深度学习模型回归出所述目标区域的图像特征与所述目标区域的图像特征对应的标准切面上预设点之间的映射关系。
  35. 根据权利要求32-34中任一项所述的方法，其特征在于，所述标准切面包括以下切面中的一个或多个：侧脑室水平横切面、双顶径切面、头颈部正中矢状切面、胸径切面、腹围切面、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面、躯干长轴切面和顶臀径切面。
  36. 一种早孕期胎儿的超声成像方法，其特征在于，所述方法包括：
    向早孕期胎儿发射超声波，接收所述超声波的回波，以获得超声回波信号；
    基于所述超声回波信号获得所述早孕期胎儿的三维超声数据；
    根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向；
    显示所述早孕期胎儿的目标区域的方向。
  37. 根据权利要求36所述的方法，其特征在于，所述方法还包括：
    根据所述目标区域的方向，从所述三维超声数据中提取与所述目标区域对应的至少一个标准切面；
    显示所述至少一个标准切面。
  38. 根据权利要求36或37所述的方法，其特征在于，所述目标区域包括头部区域，所述目标区域的方向包括所述头部区域的第一方向，其中所述第一方向为头部区域的左右方向。
  39. 根据权利要求36或37所述的方法，其特征在于，所述目标区域包括身体区域，所述目标区域的方向包括以下至少一个：身体区域的第二方向、身体区域的第三方向和身体区域的第四方向，其中所述第二方向、第三方向和第四方向分别为身体区域的上下方向、前后方向和左右方向，且所述第二方向、第三方向和第四方向之间任意两个相互垂直。
  40. 一种超声成像系统，其特征在于，所述超声成像系统包括：
    超声探头；
    发射/接收电路，用于激励所述超声探头向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；
    处理器，用于：
    基于所述超声回波信号获得所述早孕期胎儿的三维超声数据；
    根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向；
    根据所述目标区域的方向，从所述三维超声数据中提取与所述目标区域对应的至少一个标准切面；
    显示器，用于显示所述至少一个标准切面。
  41. 根据权利要求40所述的超声成像系统，其特征在于，所述目标区域包括以下至少一个：头部区域、身体区域和全身区域。
  42. 根据权利要求41所述的超声成像系统，其特征在于，所述目标区域包括头部区域，所述目标区域的方向包括所述头部区域的第一方向，其中所述第一方向为头部区域的左右方向。
  43. 根据权利要求41所述的超声成像系统，其特征在于，所述目标区域 包括身体区域，所述目标区域的方向包括以下至少一个：身体区域的第二方向、第三方向和第四方向，其中所述第二方向、第三方向和第四方向分别为身体区域在所述三维超声数据中的上下方向、前后方向和左右方向，且所述第二方向、第三方向和第四方向之间任意两个相互垂直。
  44. 根据权利要求40-43中任一项所述的超声成像系统，其特征在于，所述根据所述目标区域的方向，从所述三维超声数据中提取与所述目标区域对应的至少一个标准切面，包括：
    根据所述目标区域的方向在所述三维超声数据中确定与所述目标区域对应的多个候选切面；
    在与所述目标区域对应的所述多个候选切面中挑选出与所述目标区域对应的至少一个标准切面。
  45. 根据权利要求40-43中任一项所述的超声成像系统，其特征在于，所述根据所述目标区域的方向，从所述三维超声数据中提取与所述目标区域对应的至少一个标准切面，包括：
    在所述三维超声数据中检测与所述目标区域对应的标准切面相对应的早孕目标特征结构；
    将与所述早孕目标特征结构至少部分重合、并且与所述目标区域的方向之间的夹角满足预设要求的切面确定为与所述目标区域对应的至少一个所述标准切面。
  46. 一种超声成像系统，其特征在于，所述超声成像系统包括：
    超声探头；
    发射/接收电路，用于激励所述超声探头向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；
    处理器，用于：
    显示所述至少一个标准切面根据所述超声回波信号获得所述早孕期胎儿的三维超声数据；
    从所述三维超声数据中至少两个不同的早孕目标特征结构的区域；
    确定与所述至少两个不同的早孕目标特征结构的区域中每个所述早孕目标特征结构的区域均至少部分重合的至少一个切面，以作为所述早孕期胎儿的至少一个标准切面；
    显示器，用于显示所述至少一个标准切面。
  47. 一种超声成像统，其特征在于，所述超声成像系统包括：
    超声探头；
    发射/接收电路，用于激励所述超声探头向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；
    处理器，用于：
    根据所述超声回波信号获得所述早孕期胎儿的三维超声数据；
    将所述三维超声数据与预先配置好早孕期标准切面的超声数据模板进行匹配，并根据匹配结果确定所述三维超声数据中的目标标准切面，其中所述早孕期标准切面包括以下至少一个：侧脑室水平横切面、双顶径切面、头颈部正中矢状切面、胸径切面、腹围切面、腹壁脐带插入处切面、膀胱切面、脊柱纵轴切面、躯干长轴切面和顶臀径切面；
    显示器，用于显示所述目标标准切面。
  48. 一种超声成像系统，其特征在于，所述超声成像系统包括：
    超声探头；
    发射/接收电路，用于激励所述超声探头向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；
    处理器，用于：
    根据所述三维超声数据提取所述三维超声数据的目标区域的图像特征；
    根据所述目标区域的图像特征确定目标标准切面的法线方向和所述目标标准切面上预设点的位置信息；
    根据所述超声回波信号获得所述早孕期胎儿的三维超声数据；
    根据所述三维超声数据提取所述三维超声数据的目标区域的图像特征；
    根据所述目标区域的图像特征确定目标标准切面的法线方向和所述目标标准切面上预设点的位置信息；
    根据所述目标标准切面的法线方向和所述目标标准切面上预设点的位置信息确定所述目标标准切面；
    显示器，用于显示所述目标标准切面。
  49. 一种超声成像系统，其特征在于，所述超声成像系统包括：
    超声探头；
    发射/接收电路，用于激励所述超声探头向早孕期胎儿发射超声波，并接收所述超声波的回波，以获得超声回波信号；
    处理器，用于：
    基于所述超声回波信号获得所述早孕期胎儿的三维超声数据；
    根据所述三维超声数据确定所述早孕期胎儿的目标区域的方向；
    显示器，用于显示所述早孕期胎儿的目标区域的方向。
PCT/CN2020/129117 2020-11-16 2020-11-16 早孕期胎儿的超声成像方法和超声成像系统 WO2022099705A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2020/129117 WO2022099705A1 (zh) 2020-11-16 2020-11-16 早孕期胎儿的超声成像方法和超声成像系统
CN202080105099.5A CN116171131A (zh) 2020-11-16 2020-11-16 早孕期胎儿的超声成像方法和超声成像系统

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2020/129117 WO2022099705A1 (zh) 2020-11-16 2020-11-16 早孕期胎儿的超声成像方法和超声成像系统

Publications (1)

Publication Number Publication Date
WO2022099705A1 true WO2022099705A1 (zh) 2022-05-19

Family

ID=81602059

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/129117 WO2022099705A1 (zh) 2020-11-16 2020-11-16 早孕期胎儿的超声成像方法和超声成像系统

Country Status (2)

Country Link
CN (1) CN116171131A (zh)
WO (1) WO2022099705A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114882019A (zh) * 2022-07-01 2022-08-09 首都医科大学附属北京妇产医院 孕早期或孕中期超声图像标准化质量控制方法、系统及设备
CN116452899A (zh) * 2023-06-20 2023-07-18 四川省医学科学院·四川省人民医院 一种基于深度学习的超声心动图标准切面识别及评分方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110313291A1 (en) * 2009-02-10 2011-12-22 Hitachi Medical Corporation Medical image processing device, medical image processing method, medical image diagnostic apparatus, operation method of medical image diagnostic apparatus, and medical image display method
CN104414680A (zh) * 2013-08-21 2015-03-18 深圳迈瑞生物医疗电子股份有限公司 一种三维超声成像方法及系统
CN108186051A (zh) * 2017-12-26 2018-06-22 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系统
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
CN111368586A (zh) * 2018-12-25 2020-07-03 深圳迈瑞生物医疗电子股份有限公司 超声成像方法及系统
CN107106143B (zh) * 2015-05-07 2020-10-20 深圳迈瑞生物医疗电子股份有限公司 三维超声成像方法和装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110313291A1 (en) * 2009-02-10 2011-12-22 Hitachi Medical Corporation Medical image processing device, medical image processing method, medical image diagnostic apparatus, operation method of medical image diagnostic apparatus, and medical image display method
CN104414680A (zh) * 2013-08-21 2015-03-18 深圳迈瑞生物医疗电子股份有限公司 一种三维超声成像方法及系统
CN107106143B (zh) * 2015-05-07 2020-10-20 深圳迈瑞生物医疗电子股份有限公司 三维超声成像方法和装置
CN108186051A (zh) * 2017-12-26 2018-06-22 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系统
CN111368586A (zh) * 2018-12-25 2020-07-03 深圳迈瑞生物医疗电子股份有限公司 超声成像方法及系统
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

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114882019A (zh) * 2022-07-01 2022-08-09 首都医科大学附属北京妇产医院 孕早期或孕中期超声图像标准化质量控制方法、系统及设备
CN116452899A (zh) * 2023-06-20 2023-07-18 四川省医学科学院·四川省人民医院 一种基于深度学习的超声心动图标准切面识别及评分方法

Also Published As

Publication number Publication date
CN116171131A (zh) 2023-05-26

Similar Documents

Publication Publication Date Title
US20240074675A1 (en) Adaptive ultrasound scanning
US20110255762A1 (en) Method and system for determining a region of interest in ultrasound data
US11931201B2 (en) Device and method for obtaining anatomical measurements from an ultrasound image
WO2018195946A1 (zh) 一种超声图像显示方法、设备及存储介质
WO2022099705A1 (zh) 早孕期胎儿的超声成像方法和超声成像系统
KR102063374B1 (ko) 초음파 볼륨의 자동 정렬
CN111281430A (zh) 超声成像方法、设备及可读存储介质
US8636662B2 (en) Method and system for displaying system parameter information
WO2024093911A1 (zh) 一种超声成像方法及超声设备
CN116138806A (zh) 分析心脏射血分数的方法及系统和超声成像系统
JP7427002B2 (ja) フレームのインデックス付け及び画像レビューのためのシステム及び方法
CN114521914A (zh) 超声参数测量方法和超声参数测量系统
WO2022099704A1 (zh) 中晚孕期胎儿的超声成像方法和超声成像系统
CN115670517A (zh) 胎儿脑室比率的测量方法及超声成像系统
WO2022133806A1 (zh) 胎儿颜面部容积图像修复方法和超声成像系统
CN114947939A (zh) 用于多平面成像的超声成像系统和方法
JP2024501181A (ja) 超音波画像取得、追跡、及びレビュー
CN113229850A (zh) 超声盆底成像方法和超声成像系统
WO2022134049A1 (zh) 胎儿颅骨的超声成像方法和超声成像系统
CN117814840A (zh) 早孕胎儿的超声成像方法和超声成像系统
EP4226863A1 (en) Fetal heart rate monitoring
US20240173007A1 (en) Method and apparatus with user guidance and automated image setting selection for mitral regurgitation evaluation
CN107007302A (zh) 超声设备、超声图像处理方法及装置
CN116327237A (zh) 超声成像系统及方法、超声图像处理系统及方法
CN114680942A (zh) 基于输卵管造影成像的评估方法和超声成像系统

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: 20961259

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: 20961259

Country of ref document: EP

Kind code of ref document: A1