WO2024037019A1 - 一种胎心的超声成像方法和系统 - Google Patents

一种胎心的超声成像方法和系统 Download PDF

Info

Publication number
WO2024037019A1
WO2024037019A1 PCT/CN2023/089944 CN2023089944W WO2024037019A1 WO 2024037019 A1 WO2024037019 A1 WO 2024037019A1 CN 2023089944 W CN2023089944 W CN 2023089944W WO 2024037019 A1 WO2024037019 A1 WO 2024037019A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
dimensional
volume
ultrasonic
dimensional ultrasound
Prior art date
Application number
PCT/CN2023/089944
Other languages
English (en)
French (fr)
Inventor
陈子轩
邹耀贤
林穆清
杨俊英
何绪金
Original Assignee
武汉迈瑞医疗技术研究院有限公司
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 武汉迈瑞医疗技术研究院有限公司, 深圳迈瑞生物医疗电子股份有限公司 filed Critical 武汉迈瑞医疗技术研究院有限公司
Priority to PCT/CN2023/089944 priority Critical patent/WO2024037019A1/zh
Priority to CN202380009329.1A priority patent/CN116867440A/zh
Publication of WO2024037019A1 publication Critical patent/WO2024037019A1/zh

Links

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/02—Measuring pulse or heart rate

Definitions

  • the present invention relates to the field of ultrasound imaging of fetal heart, and specifically relates to an ultrasound imaging method and system of fetal heart.
  • congenital heart disease According to statistics from a clinical study, 8 out of 1,000 babies that survive successful birth will have congenital heart disease, and among cases of neonatal death due to malformations, congenital heart disease accounts for nearly 50% . It can be seen that the probability of congenital heart disease is high and the consequences are serious, which make fetal heart (fetal heart) examination a particularly important part of prenatal diagnosis.
  • the motion characteristics of the fetal heart make it difficult for traditional three-dimensional collection methods (free-arm 3D, static 3D and volume probe 4D collection) to meet the requirements for fast-moving 3D fetal heart data collection.
  • traditional two-dimensional ultrasound examination cannot directly provide the structural shape and positional relationship of the three-dimensional space of the lesion due to the complexity of the fetal heart structure, making it difficult to detect many potential lesions in two-dimensional ultrasound.
  • STIC spatio-temporal correlation imaging technology
  • the imaging principle of STIC is to first collect two-dimensional images of multiple cardiac cycles, calculate the heart rate, and reorganize the two-dimensional images of different cardiac phase positions into three-dimensional data based on the heart rate. Finally, the reorganized three-dimensional data will be played in a loop according to the cardiac phase sequence. This achieves the purpose of displaying a 3D beating heart.
  • STIC takes a long time to collect (for example, it generally takes about 25 cardiac cycles, about 12 seconds), and is easily affected by fetal movement during this process. , causing the collection process to fail.
  • the present invention mainly provides an ultrasound imaging method and system for fetal heartbeat, which will be described in detail below.
  • an ultrasound imaging method of fetal heartbeat including:
  • the data rearrangement includes at least placing part of the three-dimensional ultrasonic data with subsequent scanning timing. Forward adjustment, and backward adjustment for part of the three-dimensional ultrasound data whose scan timing is earlier;
  • the fetal heart rate is displayed based on the data-rearranged multi-volume three-dimensional ultrasound data.
  • an ultrasound imaging method of fetal heartbeat including:
  • the three-dimensional ultrasound data in m ⁇ n-1 volumes are rearranged according to the order of cardiac phases to obtain the three-dimensional ultrasound data in m ⁇ n-1 volumes after data rearrangement; the data rearrangement includes at least part of the scanning time sequence.
  • the latter three-dimensional ultrasound data is adjusted forward, and part of the three-dimensional ultrasound data with earlier scan timing is adjusted backward;
  • the fetal heart rate is displayed according to the three-dimensional ultrasound data of the m ⁇ n-1 volume after data rearrangement.
  • an ultrasound imaging method of fetal heartbeat including:
  • each of the three-dimensional ultrasonic scans includes: transmitting ultrasonic waves to the fetal heart and receiving corresponding ultrasonic echoes to obtain a volume of three-dimensional ultrasonic data containing the fetal heart;
  • the fetal heart rate is displayed based on the data-rearranged multi-volume three-dimensional ultrasound data.
  • an ultrasound imaging system including:
  • Ultrasound probe used to transmit ultrasonic waves to the fetal heart and receive corresponding ultrasonic echo signals
  • a transmitting and receiving control circuit for controlling the ultrasonic probe to transmit ultrasonic waves and receive ultrasonic echo signals
  • a processor configured to perform a method as described in any embodiment herein.
  • Figure 1 is a schematic diagram of an ultrasound image obtained through spatiotemporal correlation imaging technology
  • Figure 2 is a schematic structural diagram of an ultrasound imaging system according to an embodiment
  • Figure 3 is a schematic diagram of a three-dimensional ultrasound scanning sequence according to an embodiment
  • Figure 4 is a schematic diagram of an ultrasound imaging method according to an embodiment
  • Figure 5 is a schematic diagram of data rearrangement according to an embodiment
  • Figure 6 is a schematic diagram of data registration according to an embodiment
  • Figure 7 is a schematic diagram of data registration according to an embodiment
  • Figure 8 is a schematic diagram of boundary correction according to an embodiment
  • Figure 9 is a flow chart of an ultrasound imaging method according to an embodiment
  • Figure 10 is a flow chart of an ultrasound imaging method according to an embodiment
  • Figure 11 is a flow chart for setting time resolution in an embodiment
  • Figure 12 is a flow chart for setting the heart phase time interval in an embodiment
  • Figure 13 is a flow chart of an ultrasound imaging method according to an embodiment.
  • connection and “connection” mentioned in this application include direct and indirect connections (connections) unless otherwise specified.
  • an ultrasound imaging system includes an ultrasound probe.
  • the ultrasonic probe 10 is used to transmit ultrasonic waves to a region of interest, such as the fetal heart, and receive corresponding ultrasonic echo signals to obtain ultrasonic data, such as two-dimensional ultrasonic data or three-dimensional ultrasonic data.
  • the ultrasonic probe 10 includes multiple array elements, which are used to realize mutual conversion between electrical pulse signals and ultrasonic waves, thereby transmitting ultrasonic waves to the area of interest and receiving corresponding ultrasonic echo signals.
  • the array element can emit ultrasonic waves according to the excitation electrical signal, or convert the received ultrasonic waves into electrical signals. Therefore, each array element can be used to transmit ultrasonic waves to the biological tissue in the area of interest, and can also be used to receive ultrasonic echoes returned by the tissue.
  • the transmitting sequence and the receiving sequence can be used to control which array elements are used to transmit ultrasonic waves and which array elements are used to receive ultrasonic waves, or the array elements can be controlled to transmit ultrasonic waves or receive ultrasonic echoes in time slots. All array elements participating in ultrasonic emission can be excited by electrical signals at the same time, thereby emitting ultrasonic waves at the same time; or the array elements participating in ultrasonic emission can also be excited by several electrical signals with a certain time interval, thereby continuously emitting ultrasonic waves with a certain time interval.
  • the ultrasound probe 10 may be a volumetric probe. In some embodiments, the ultrasound probe 10 may be an area array probe.
  • the transmitting and receiving control circuit 20 is used to control the ultrasonic probe 10 to transmit ultrasonic waves and receive ultrasonic echo signals.
  • the transmitting and receiving control circuit 20 is used to control the ultrasonic probe 10 to transmit ultrasonic waves to the area of interest on the one hand, and to control the ultrasonic probe 10 to receive the ultrasonic echo signals reflected by the ultrasonic waves through the area of interest on the other hand.
  • the transmitting and receiving control circuit 20 is used to generate a transmitting sequence and a receiving sequence and output them to the ultrasonic probe 10 .
  • the transmitting sequence is used to control some or all of the multiple array elements in the ultrasonic probe 10 to transmit ultrasonic waves to the biological tissue 60.
  • the parameters of the transmitting sequence include the number of array elements for transmitting and ultrasonic wave transmitting parameters (such as amplitude, frequency, number of waves, emission interval, emission angle, wave pattern and/or focus position, etc.).
  • the receiving sequence is used to control some or all of the multiple array elements to receive the echo after the ultrasonic wave is transmitted through the tissue.
  • the parameters of the receiving sequence include the number of receiving array elements and the receiving parameters of the echo (such as receiving angle, depth, etc.).
  • the ultrasonic parameters in the transmitting sequence and the echo parameters in the receiving sequence are also different.
  • the echo processing module 30 is used to process the ultrasonic echo signal received by the ultrasonic probe 10, such as filtering, amplifying, and beamforming the ultrasonic echo signal to obtain ultrasonic echo data.
  • the echo processing module 30 can output the ultrasonic echo data to the processor 40, or can first store the ultrasonic echo data in a memory.
  • the processor 40 Read the ultrasonic echo data from the memory.
  • the echo processing module 30 can also be omitted.
  • the processor 40 is used to obtain ultrasonic echo data or signals, and use relevant algorithms to obtain the required parameters or images.
  • the processor 40 in some embodiments of the present invention includes, but is not limited to, a Central Processing Unit (CPU), a Micro Controller Unit (MCU), and a Field-Programmable Gate Array (FPGA). and digital signal processing (DSP) are devices used to interpret computer instructions and process data in computer software.
  • the processor 40 is used to execute each computer application program in the non-transitory computer-readable storage medium, thereby causing the sample analysis device to perform a corresponding detection process.
  • the display 50 may be used to display information, such as parameters and images calculated by the processor 40 .
  • the ultrasound imaging system itself may not integrate a display module, but may be connected to a computer device (such as a computer) to display information through the display module (such as a display screen) of the computer device.
  • the processor 40 acquires scanning rules; and the processor 40 controls multiple three-dimensional ultrasound scans of the fetal heart according to the scanning rules, and acquires multiple volumes of three-dimensional ultrasound data containing multiple cardiac cycles, so that the three-dimensional ultrasound data of different volumes are The ultrasound data corresponds to different cardiac phase positions of the fetal heart; wherein each three-dimensional ultrasound scan includes: transmitting ultrasound waves to the fetal heart and receiving corresponding ultrasound echoes to obtain a volume of three-dimensional ultrasound data containing the fetal heart; in some embodiments , each volume of three-dimensional ultrasound data contains a complete fetal heart rate.
  • the processor 40 performs data rearrangement on the multiple volumes of three-dimensional ultrasound data according to the sequence of the cardiac phases to obtain the multiple volumes of three-dimensional ultrasound data after data rearrangement; the processor 40 controls the display 50 based on the multiple volumes of three-dimensional ultrasound data after data rearrangement. Display the fetal heart rate.
  • performing multiple three-dimensional ultrasound scans on the fetal heart to obtain multi-volume three-dimensional ultrasound data containing multiple cardiac cycles refers to performing multiple three-dimensional ultrasound scans on the fetal heart to obtain multi-volume three-dimensional ultrasound data.
  • This multi-volume three-dimensional ultrasound data Ultrasound data contains data from multiple cardiac cycles.
  • the scanning rules include: in the process of performing multiple three-dimensional ultrasonic scans, each time the three-dimensional ultrasonic scan is completed, the next three-dimensional ultrasonic scan will be performed after the same delay time. For example, a total of N1 three-dimensional ultrasonic scans have been performed.
  • the next three-dimensional ultrasonic scan will be performed after the same delay time; in other words, After completing the first three-dimensional ultrasonic scan, a delay period elapses before performing a second three-dimensional ultrasonic scan; after completing the second three-dimensional ultrasonic scan, a same delay period elapses before performing a third three-dimensional ultrasonic scan, and so on.
  • a delay period elapses before performing a second three-dimensional ultrasonic scan
  • a same delay period elapses before performing a third three-dimensional ultrasonic scan
  • a volume of three-dimensional ultrasonic data is obtained through a three-dimensional ultrasonic scan.
  • a three-dimensional ultrasonic scan includes transmitting ultrasonic waves and receiving corresponding ultrasonic echoes. Therefore, after completing a three-dimensional ultrasonic scan in this article, the delay time Performing the next three-dimensional ultrasonic scan refers to: after the previous three-dimensional ultrasonic scan completes transmitting ultrasonic waves and receiving corresponding ultrasonic echoes, timing based on this time node, After the delay period, the next three-dimensional ultrasound scan is started.
  • the delay duration is calculated based on the fetal heart rate, the basic zero delay volume rate, and the first parameter.
  • the basic zero-delay volume rate in this article is the scan volume rate of the first 3D ultrasound scan among the multiple 3D ultrasound scans performed.
  • the first parameter includes temporal resolution or interphase interval.
  • the central phase time interval is the time interval between the central phase positions corresponding to two adjacent volumes of three-dimensional ultrasound data in the multi-volume three-dimensional ultrasound data after data rearrangement.
  • the processor 40 obtains the delay duration; the processor 40 controls multiple three-dimensional ultrasound scans of the fetal heart to acquire multiple volumes of three-dimensional ultrasound data containing multiple cardiac cycles; wherein, during these multiple three-dimensional ultrasound scans, During the scanning process: each time a three-dimensional ultrasonic scan is completed, the next three-dimensional ultrasonic scan will be performed after the delay time; and each three-dimensional ultrasonic scan includes: transmitting ultrasonic waves to the fetal heart and receiving corresponding ultrasonic echoes to obtain a
  • the volumes contain three-dimensional ultrasound data of the fetal heart; in some embodiments, each volume of the three-dimensional ultrasound data contains a complete fetal heart.
  • the processor 40 performs data rearrangement on the multiple volumes of three-dimensional ultrasound data according to the sequence of the cardiac phases to obtain the multiple volumes of three-dimensional ultrasound data after data rearrangement; the processor 40 controls the display 50 based on the multiple volumes of three-dimensional ultrasound data after data rearrangement. Display the fetal heart rate.
  • the delay length is such that the acquired multiple volumes of three-dimensional ultrasound data correspond to different cardiac phase positions of the fetal heart.
  • the delay length is such that the acquired multi-volume three-dimensional ultrasound data corresponds to different cardiac phase positions of the fetal heart, and the distance between adjacent cardiac phase positions is equal.
  • the delay length is such that the cardiac phase time interval and the time resolution are multiplied to 1, and the cardiac phase time interval is the cardiac phase corresponding to two adjacent volumes of three-dimensional ultrasound data among the multiple volumes of three-dimensional ultrasound data after data rearrangement. The time interval between locations.
  • the processor 40 obtains the heart rate of the fetal heart, and calculates the delay duration based on the heart rate, the basic zero-delay volume rate, and the first parameter.
  • the basic zero-delay volume rate in this article is the scan volume rate of the first 3D ultrasound scan among the multiple 3D ultrasound scans performed.
  • the first parameter includes temporal resolution or interphase time interval.
  • the central phase time interval is the time interval between the central phase positions corresponding to two adjacent volumes of three-dimensional ultrasound data in the multi-volume three-dimensional ultrasound data after data rearrangement.
  • the processor 40 calculates the delay duration based on the heart rate, the basic zero-delay volume rate and the first parameter, including:
  • DT represents the delay duration
  • BVR represents the basic zero-delay volume rate
  • TS represents the time resolution
  • FR represents the fetal heart rate
  • floor represents the downward rounding operation.
  • the processor 40 controls the display 50 to generate a setting interface, and the setting interface is used to set the time resolution; in response to the setting instruction, the processor 40 sets the time resolution. For example, the user sets the time resolution on the setting interface through an input tool such as a mouse or keyboard.
  • the processor 40 calculates the delay duration based on the heart rate, the basic zero-delay volume rate and the first parameter, including:
  • DT represents the delay duration
  • BVR represents the basic zero-delay volume rate
  • TI represents the heart phase time interval
  • FR represents the fetal heart rate
  • floor represents the downward rounding operation.
  • the processor 40 controls the display 50 to generate a setting interface, which is used to set the heart phase time interval; in response to the setting instruction, the processor 40 sets the heart phase time interval. For example, the user sets the heart phase time interval on the setting interface through an input tool such as a mouse or keyboard.
  • the fetal heart rate refers to the number of times the fetal heart beats per unit time (such as the number of times it runs per minute), which can be obtained through multiple methods, such as fetal heart signal monitoring, ultrasound Doppler Or M-ultrasound to measure fetal heart rate.
  • the method of fetal heart signal monitoring is usually to extract and separate the fetal ECG signal from the mixed ECG signal in the mother's abdomen, and calculate the fetal heart rate based on the separated fetal ECG signal.
  • the separation algorithms include adaptive filtering algorithm, template subtraction and blind signal. Analysis etc.
  • the method based on ultrasound Doppler mainly utilizes the Doppler effect; since fetal heart movement is cyclical, the echo frequency shift signal is also cyclical.
  • the fetal heart rate can be calculated.
  • M-ultrasound uses the periodicity of heart beating to move the echo signal from left to right and displays the time position curve image of the periodic heartbeat, so that the time position relationship can be further used to calculate the fetal heart rate.
  • the unit of heart rate is beats/minute
  • the unit of heart-phase time interval is seconds
  • the unit of volume rate (including basic zero-delay volume rate) is volume/second.
  • the unit of time resolution is volume/one cardiac cycle.
  • FIG 4 is a schematic diagram of some embodiments of the present application. It can be seen that by setting a reasonable delay length, the roll rate (that is, how many rolls can be scanned in 1 second) can be changed, so that each page obtained during the scanning process can be changed.
  • the volume exactly corresponds to the different cardiac phase positions of the fetal heart, and the time interval between the rearranged data and the preset time resolution are multiplied by 1, so that the corresponding positions of different cardiac phases of a complete cardiac cycle are exactly equidistantly filled. Therefore, the core of the algorithm is implemented by establishing the equation of time resolution. Assume that the preset time resolution is represented by TS, the heart rate is represented by FR, and the basic zero-delay volume rate is represented by BVR.
  • TS heart phase time interval
  • TS time resolution
  • TS 60/FR/TI
  • floor is the downward or tail-rounding operation (for example, 2.5 is 2 after rounding), from which the integer m1 can be obtained.
  • the actual volume rate AVR after adding delay can be obtained as:
  • Delay DT 1/AVR-1/BVR.
  • the processor 40 controls multiple three-dimensional ultrasound scans of the fetal heart to obtain multiple volumes of three-dimensional ultrasound data; in some embodiments, the number of these multiple three-dimensional ultrasound scans is m ⁇ n-1; in some embodiments, The number of volumes of these multi-volume three-dimensional ultrasound data is m ⁇ n-1.
  • m floor[(TS+1) ⁇ BVR ⁇ 60 ⁇ FR ⁇ TS];
  • BVR represents the basic zero-delay volume rate
  • FR represents the fetal heart rate
  • TS represents the time resolution
  • floor represents the downward rounding operation.
  • m floor[(60 ⁇ FR ⁇ TI+1) ⁇ BVR ⁇ TI]
  • BVR represents the basic zero-delay volume rate
  • TI represents the cardiac phase interval
  • FR represents the fetal heart rate
  • floor represents the downward rounding operation.
  • data rearrangement at least includes adjusting part of the three-dimensional ultrasound data with a later scan timing forward, and adjusting part of the three-dimensional ultrasound data with a previous scan timing backward.
  • data rearrangement at least includes adjusting part of the three-dimensional ultrasound data with a later scan timing forward, and adjusting part of the three-dimensional ultrasound data with a previous scan timing backward.
  • Figure 5 assume that there are three volumes of three-dimensional ultrasound data A, B and C in the scanning order. If the data are rearranged to obtain A, C and B, then it can be considered that part of the scan will be The three-dimensional ultrasound data with a later timing sequence are adjusted forward, and some of the three-dimensional ultrasound data with a previous scan timing are adjusted backward. For example, the three-dimensional ultrasound number C is adjusted forward, and the three-dimensional ultrasound number B is adjusted backward.
  • the data rearrangement by the processor 40 includes: the (i-1) ⁇ m+j volume of three-dimensional ultrasound data in the m ⁇ n-1 volumes of three-dimensional ultrasound data before data rearrangement is set as data
  • the rearranged (j-1) ⁇ n+i Volume three-dimensional ultrasound data; i and j are positive integers, and i is less than or equal to n, and j is less than or equal to m.
  • the three-dimensional ultrasound data before data rearrangement are arranged according to the scanning time sequence, that is, they are arranged according to the time sequence of the corresponding three-dimensional ultrasound scans.
  • the three-dimensional ultrasound data after data rearrangement are arranged according to the cardiac phase sequence.
  • the data rearrangement by the processor 40 includes: dividing the m ⁇ n-1 volume of three-dimensional ultrasound data before data rearrangement into n parts according to the time sequence of their respective three-dimensional ultrasound scans, from the first to the Each of the n-1 parts contains the three-dimensional ultrasound data in the m volume, and the n-th part contains the three-dimensional ultrasound data in the m-1 volume; any i-th volume of the m ⁇ n-1 three-dimensional ultrasound data before the data is rearranged
  • the jth volume of three-dimensional ultrasound data in each part is set to the (j-1) ⁇ n+ith volume of three-dimensional ultrasound data after data rearrangement; i and j are positive integers, and i is less than or equal to n, j Less than or equal to m.
  • the data rearrangement by the processor 40 includes: dividing the m ⁇ n-1 volume of three-dimensional ultrasound data before data rearrangement into n parts according to the time sequence of their respective three-dimensional ultrasound scans, from the first to the Each of the n-1 parts contains the three-dimensional ultrasound data described in volume m, and the n-th part contains the three-dimensional ultrasound data described in volume m-1; the j-th volume in the first to n-th parts before data rearrangement
  • the three-dimensional ultrasound data are respectively set to the 1st to nth volumes of the 3D ultrasound data in the jth part after data rearrangement, where j is a positive integer less than m; the 1st to nth volumes before data rearrangement are
  • the three-dimensional ultrasonic data of the m-th volume in each of the 1 parts are respectively set to the three-dimensional ultrasonic data of the 1st to n-1 volumes in the m-th part after data rearrangement.
  • the multiple volumes of three-dimensional ultrasound data after data rearrangement correspond to different cardiac phase positions of a complete cardiac cycle equidistantly.
  • the processor 40 also obtains the total scan time; after the total scan time is reached, the processor 40 controls to stop the three-dimensional ultrasound scan; or, the processor 40 generates a prompt to stop the three-dimensional ultrasound scan based on the total scan time.
  • T represents the total scanning time
  • FR represents the fetal heart rate
  • TI represents the cardiac phase interval
  • BVR represents the basic zero-delay volume rate
  • floor represents the downward rounding operation.
  • T represents the total scanning time
  • TS represents the time resolution
  • FR represents the fetal heart rate
  • BVR represents the basic zero-delay volume rate
  • floor represents the downward rounding operation.
  • the processor 40 also obtains the number of three-dimensional ultrasonic scans. After completing the three-dimensional ultrasonic scans, After the number of scanning times, the processor 40 controls to stop performing the three-dimensional ultrasonic scanning; or, the processor 40 generates a prompt to stop performing the three-dimensional ultrasonic scanning according to the number of times of the three-dimensional ultrasonic scanning. In some embodiments, the number of three-dimensional ultrasound scans is m ⁇ n-1.
  • the total number of scans involved in some embodiments is m ⁇ n-1, this refers to the number of scans related to the solution of this application; in the actual process, more scan times may be performed due to various reasons. , for example, in the above-mentioned m ⁇ n-1 scanning process, several scans can be inserted before it, or several scans can be inserted after it, but since these scans and the data obtained by the scans have nothing to do with this application, Therefore, these scans are not mentioned in this application.
  • the processor 40 may also perform data registration and/or boundary correction on the above-mentioned multi-volume three-dimensional ultrasound data after the data rearrangement.
  • the processor 40 can perform registration based on 3D data; in some embodiments, the processor 40 can perform registration based on 3D data of a two-dimensional plane, which will be described separately below.
  • registration based on 3D data is to directly search in three-dimensional space to achieve direct registration of two 3D data. It can usually be divided into traditional algorithms, deep learning-based registration algorithms and hybrid algorithms.
  • SIFT image grayscale or feature information
  • Harris extracts significant corner points in the image
  • HOG extracts features through the directional gradient histogram
  • SURF extracts features through the Hessian matrix and scale space.
  • Similarity measurement methods mainly include absolute squared error (SSD), mutual information (MI), normalized mutual information (NMI), normalized cross-correlation (NCC), structural similarity (SSIM), etc.
  • SSD absolute squared error
  • MI mutual information
  • NMI normalized mutual information
  • NCC normalized cross-correlation
  • SSIM structural similarity
  • the Demons algorithm is based on the optical flow theory and uses the gradient of the reference image and the grayscale difference between the reference image and the floating image to calculate the coordinate offset of each point, and iterates based on this offset to achieve 3D data registration;
  • the ICP algorithm first extracts feature points from 3D data and turns them into 3D point cloud data. Then, through the algorithm of nearest point search and iteration, the two 3D point clouds are aligned, thereby registering the 3D data.
  • 3D data registration algorithms based on deep learning can be divided into two categories.
  • the first category inputs two 3D data and outputs a three-dimensional rigid transformation matrix of the two 3D data; the second category inputs two 3D data (floating data). and reference data), the output is floating data aligned with the reference data.
  • the first type of registration algorithm as an example, by constructing a training sample library and using a regression strategy to estimate three-dimensional transformation parameters, two different 3D data can be aligned to achieve registration of 3D data.
  • the main steps are:
  • Step 1 Construct a training sample library:
  • 3D Given a 3D data and initializing the three-dimensional transformation parameters, 3D is realized by changing the three-dimensional rigid transformation parameters. Rigid transformation of data (translation and rotation in three directions, 6 parameters in total).
  • the training sample library can be composed of two 3D data before and after transformation and three-dimensional rigid transformation parameters, and there is a one-to-one correspondence between the two 3D data before and after transformation and the transformation parameters, which meets the needs of samples and labels in deep learning algorithms. constraint relationship.
  • the two 3D data before and after transformation are the input data
  • the three-dimensional rigid transformation parameters are the output label values.
  • Step 2 Network design and training steps:
  • the design of the network mainly includes convolutional layers, pooling layers, excitation layers, splicing layers, fully connected layers and set loss functions. By combining and stacking these layers, the features in the training samples are learned.
  • input the designed network model For any two 3D data in the training sample library, input the designed network model and output the estimated value of the three-dimensional rigid transformation parameters.
  • the loss function of this type of model generally consists of two parts. One is to use the difference between the estimated value and the true value to form the loss function. L1 norm, L2 norm, etc. are commonly used as the loss function; the other is to use the estimated value to calculate the third loss function.
  • the two 3D data are subjected to three-dimensional rigid transformation, and the output is compared with the first input 3D data.
  • Another loss function is formed through the difference between the two 3D data.
  • Commonly used are SSD, L1 norm, L2 norm, NC, SSIM etc. as loss function.
  • the model parameters are optimized to train the network.
  • Commonly used network models include AlexNet, VGGNet, GoogLeNet, MobileNet, SYMNet, STN, etc. This method includes but is not limited to the above network model structure and loss function.
  • Step 3 Reasoning steps:
  • the two 3D data to be registered are input into the network trained in Step 2 above, and the three-dimensional rigid transformation parameters can be obtained in real time.
  • the obtained three-dimensional rigid transformation parameters are applied to the first 3D data and performed on it. Transformation realizes the alignment of the first 3D data and the second 3D data, thus realizing the registration of the two 3D data.
  • the second type of deep learning-based 3D data registration algorithm is similar to the above steps.
  • a training sample library is constructed, and then a regression network model is designed.
  • the input of the model is two 3D data (floating data and fixed data), and the output is floating data aligned and registered with the fixed data.
  • the model is optimized through the loss function for training. network.
  • the inference stage by inputting floating 3D data and fixed 3D data, the registered floating 3D data can be obtained in real time, thereby realizing the registration of the two 3D data.
  • the 3D data hybrid registration algorithm is mainly different from traditional algorithms in that it combines deep learning algorithms with traditional algorithms.
  • the deep learning algorithm is used to reduce the dimensionality of the 3D data and extract features.
  • the extracted features are used in conjunction with the similarity evaluation criteria of traditional algorithms, such as MI, SSIM, NCC, etc., to achieve similarity comparison at the feature level, and then achieve registration.
  • the main algorithms that use deep learning to extract features include AlexNet, ZFNet, VGGNet, GoogLeNet, etc.
  • the algorithms for feature extraction based on deep learning include but are not limited to the above feature extraction algorithms.
  • the 3D data registration method based on the two-dimensional plane mainly performs registration on the two-dimensional level, and achieves three-dimensional registration by registering the two-dimensional plane at different angles.
  • the gray dotted line in the figure is the intersection line of surfaces A and B.
  • First align the A surface obtain the two-dimensional rigid transformation parameters in the x and z directions, and then align the B surface based on the alignment of the A surface.
  • the two-dimensional rigid transformation parameters in the y and z directions are obtained, and the registration of the two 3D data is achieved through multiple iterations of the above process.
  • This registration algorithm is mostly suitable for registration scenarios where only translation transformation exists.
  • the two-dimensional plane registration algorithm can also be divided into traditional algorithms, deep learning-based registration algorithms and hybrid algorithms. It is similar to the three-dimensional data registration algorithm. The difference is that the two-dimensional plane obtains two-dimensional rigid transformation parameters (two direction translation and a rotation, a total of 3 parameters). In addition to the registration methods already described, registration can also be achieved through other methods.
  • the block registration algorithm performs similarity matching on each image block to obtain the mapping relationship between the center positions of multiple image blocks, and then fits the two-dimensional rigid transformation parameters according to the least squares method to achieve image registration; the pyramid algorithm uses Construct two-dimensional images of different scales, start with low-resolution images for rough registration, gradually increase the image resolution and improve the registration accuracy step by step, to achieve the purpose of registering two two-dimensional images.
  • two offsets x bias , y bias , z bias , x bias ', y bias ' and z bias ' can be obtained in each of the x, y and z directions, and the non-overlapping areas in the 3D data are cropped, leaving only Image data within a small cubic area.
  • the processor 40 controls the display 50 to display the fetal heartbeat, including: sequentially displaying three-dimensional ultrasound data images of one cardiac cycle in sequence of cardiac phases.
  • the processor 40 controls m ⁇ n-1 three-dimensional ultrasound scans to the fetal heart to obtain m ⁇ n-1 volumes of three-dimensional ultrasound data; m and n are both greater than or equal to An integer of 2; wherein, in each cardiac cycle of the first n-1 fetal heart: m times of the three-dimensional ultrasound scans are performed to obtain m volumes of the three-dimensional ultrasound data; in the n-th fetal heart beat Within the period: m-1 times of the three-dimensional ultrasonic scans are performed to obtain m-1 volumes of the three-dimensional ultrasonic data.
  • each three-dimensional ultrasound scan includes: transmitting ultrasound waves to the fetal heart and receiving corresponding ultrasound echoes to obtain a volume of three-dimensional ultrasound data containing the fetal heart; in some embodiments, each volume of three-dimensional ultrasound data contains a complete of fetal heart rate.
  • the processor 40 rearranges the m ⁇ n-1 volume of three-dimensional ultrasound data according to the order of cardiac phase, and obtains the m ⁇ n-1 volume of three-dimensional ultrasound data after data rearrangement; the processor 40 performs data rearrangement according to the m ⁇ n volume after data rearrangement. -1 volume of three-dimensional ultrasound data, controlling the display 50 to display the fetal heart rate.
  • the total number of scans involved in some embodiments is m ⁇ n-1, this refers to the number of scans related to the solution of this application; in the actual process, more scan times may be performed due to various reasons. , for example, in the above m ⁇ n-1 scan process, several scans can be inserted before it, or it can be It then inserts several scans, but since these scans and the data obtained by the scans are irrelevant to this application, this application does not mention these scans.
  • data rearrangement at least includes adjusting part of the three-dimensional ultrasound data with a later scan timing forward, and adjusting part of the three-dimensional ultrasound data with a previous scan timing backward.
  • the m ⁇ n-1 volume of the three-dimensional ultrasound data after data rearrangement corresponds to different cardiac phase positions of a complete cardiac cycle equidistantly.
  • the processor 40 rearranges the m ⁇ n-1 volumes of three-dimensional ultrasound data in cardiac phase order to obtain the m ⁇ n-1 volumes of three-dimensional ultrasound data after data rearrangement, including: data When the rearranged m ⁇ n-1 volume of three-dimensional ultrasound data is divided into m parts according to the cardiac phase sequence, the first to m-1th parts respectively contain the three-dimensional ultrasound data of n volumes, and the The m parts include n-1 volumes of the three-dimensional ultrasound data; in any i-th cardiac cycle of the first n-1 fetal heart cycles, the m times of the three-dimensional ultrasound scans are obtained.
  • the m volumes of the three-dimensional ultrasound data are set to the i-th volume of the three-dimensional ultrasound data of each part in the m ⁇ n-1 volumes of the three-dimensional ultrasound data after data rearrangement; i is a positive integer, and i is less than or equal to n-1; in the cardiac cycle of the n-th fetal heart, m-1 volumes of the three-dimensional ultrasound data are obtained from the m-1 three-dimensional ultrasound scans performed , is set as the nth volume of the three-dimensional ultrasonic data of each of the 1st to m-1th parts in the m ⁇ n-1 volume of the three-dimensional ultrasonic data after data rearrangement.
  • the processor 40 rearranges the m ⁇ n-1 volumes of three-dimensional ultrasound data in cardiac phase order to obtain the m ⁇ n-1 volumes of three-dimensional ultrasound data after data rearrangement, including: data The three-dimensional ultrasound data in the (i-1) ⁇ m+jth volume of the m ⁇ n-1 volumes of three-dimensional ultrasound data before rearrangement is set to the (j-1) ⁇ n+th volume after data rearrangement.
  • Volume i is the three-dimensional ultrasound data; the i and j are positive integers, and i is less than or equal to n, and j is less than or equal to m.
  • the three-dimensional ultrasound data before data rearrangement are arranged according to the scanning time sequence, that is, they are arranged according to the time sequence of the corresponding three-dimensional ultrasound scans.
  • the three-dimensional ultrasound data after data rearrangement are arranged according to the cardiac phase sequence.
  • the processor 40 performs data rearrangement on the m ⁇ n-1 volume of three-dimensional ultrasound data in cardiac phase order to obtain the m ⁇ n-1 volume of three-dimensional ultrasound data after data rearrangement, including:
  • the three-dimensional ultrasound data described in the m ⁇ n-1 volume before data rearrangement is divided into n parts according to the time sequence of the corresponding three-dimensional ultrasound scans.
  • the first to n-1th parts each contain the three-dimensional ultrasound data described in the m volume.
  • the nth part contains the three-dimensional ultrasound data described in volume m-1; the j-th volume three-dimensional ultrasound data in any i-th part of the three-dimensional ultrasound data described in volume m ⁇ n-1 before data rearrangement, Set to the (j-1) ⁇ n+ith volume of three-dimensional ultrasound data after data rearrangement; the i and j are positive integers, and i is less than or equal to n, and j is less than or equal to m.
  • the processor 40 performs data rearrangement on the m ⁇ n-1 volume of three-dimensional ultrasound data in cardiac phase order to obtain the m ⁇ n-1 volume of three-dimensional ultrasound data after data rearrangement, including: m before data rearrangement
  • the three-dimensional ultrasound data described in ⁇ n-1 volumes are divided into n parts according to the time sequence of the corresponding three-dimensional ultrasound scans.
  • the first to n-1th parts each contain the three-dimensional ultrasound data described in m volumes, and the n-th part Part contains the three-dimensional ultrasound data described in volume m-1; the three-dimensional ultrasound data described in the j-th volume in the first to nth parts before data rearrangement are respectively set to the j-th part after data rearrangement.
  • m floor[(TS+1) ⁇ BVR ⁇ 60 ⁇ FR ⁇ TS];
  • BVR represents the basic zero-delay volume rate
  • FR represents the fetal heart rate
  • TS represents the time resolution
  • floor represents the downward rounding operation.
  • m floor[(60 ⁇ FR ⁇ TI+1) ⁇ BVR ⁇ TI]
  • BVR represents the basic zero-delay volume rate
  • TI represents the cardiac phase interval
  • FR represents the fetal heart rate
  • floor represents the downward rounding operation.
  • the processor 40 also obtains the delay duration; in the process of performing m ⁇ n-1 three-dimensional ultrasound scans: each time the three-dimensional ultrasound scan is completed, after the delay duration, the next scan is performed. The three-dimensional ultrasound scan.
  • the delay duration is such that the acquired m ⁇ n-1 volumes of three-dimensional ultrasound data correspond to different cardiac positions of the fetal heart.
  • the delay length is such that the acquired m ⁇ n-1 volumes of three-dimensional ultrasound data correspond to different cardiac phase positions of the fetal heart, and the distance between adjacent cardiac phase positions is equal.
  • the delay length is such that the heart phase interval and the time resolution are multiplied to one.
  • the delay duration is calculated based on the fetal heart rate, the basic zero delay volume rate, and the first parameter.
  • the basic zero-delay volume rate in this article is the scan volume rate of the first 3D ultrasound scan among the multiple 3D ultrasound scans performed.
  • the first parameter includes temporal resolution or interphase interval.
  • the central phase time interval is the time interval between the central phase positions corresponding to two adjacent volumes of three-dimensional ultrasound data in the multi-volume three-dimensional ultrasound data after data rearrangement.
  • the processor 40 obtains the heart rate of the fetal heart, and calculates the delay duration based on the heart rate, the basic zero-delay volume rate, and the first parameter.
  • the basic zero-delay volume rate of the multiple three-dimensional ultrasonic scans was Scan volume rate of the first 3D ultrasound scan in the scan.
  • the first parameter includes temporal resolution or interphase time interval.
  • the central phase time interval is the time interval between the central phase positions corresponding to two adjacent volumes of three-dimensional ultrasound data in the multi-volume three-dimensional ultrasound data after data rearrangement.
  • the processor 40 calculates the delay duration based on the heart rate, the basic zero-delay volume rate and the first parameter, including:
  • DT represents the delay duration
  • BVR represents the basic zero-delay volume rate
  • TS represents the time resolution
  • FR represents the fetal heart rate
  • floor represents the downward rounding operation.
  • the processor 40 controls the display 50 to generate a setting interface, and the setting interface is used to set the time resolution; in response to the setting instruction, the processor 40 sets the time resolution. For example, the user sets the time resolution on the setting interface through an input tool such as a mouse or keyboard.
  • the processor 40 calculates the delay duration based on the heart rate, the basic zero-delay volume rate and the first parameter, including:
  • DT represents the delay duration
  • BVR represents the basic zero-delay volume rate
  • TI represents the heart phase time interval
  • FR represents the fetal heart rate
  • floor represents the downward rounding operation.
  • the value of TI may be 0.01 to 0.03 seconds.
  • the processor 40 controls the display 50 to generate a setting interface, which is used to set the heart phase time interval; in response to the setting instruction, the processor 40 sets the heart phase time interval. For example, the user sets the heart phase time interval on the setting interface through an input tool such as a mouse or keyboard.
  • the processor 40 also obtains the total scan time; after the total scan time is reached, the processor 40 controls to stop the three-dimensional ultrasound scan; or, the processor 40 generates a prompt to stop the three-dimensional ultrasound scan based on the total scan time.
  • T represents the total scanning time
  • FR represents the fetal heart rate
  • TI represents the cardiac phase interval
  • BVR represents the basic zero-delay volume rate
  • floor represents the downward rounding operation.
  • T represents the total scanning time
  • TS represents the time resolution
  • FR represents the fetal heart rate
  • BVR represents the basic zero-delay volume rate
  • floor represents the downward rounding operation.
  • the processor 40 may also perform data registration and/or boundary correction on the above-mentioned multi-volume three-dimensional ultrasound data after the data rearrangement. For example, the processor 40 performs data registration and/or boundary correction on the m ⁇ n-1 volume of the three-dimensional ultrasound data after data rearrangement, and obtains the processed m ⁇ n-1 volume of the three-dimensional ultrasound data; the processor 40 According to the processed m ⁇ n-1 volume of three-dimensional ultrasound data, the display 50 is controlled to display the fetal heart rate.
  • the processor 40 controls the display 50 to display the fetal heartbeat, including: sequentially displaying three-dimensional ultrasound data images of one cardiac cycle in sequence of cardiac phases.
  • the solutions of some embodiments can significantly reduce the acquisition time (for example, only 3 to 4 seconds) and reduce the impact of fetal movement on the final imaging effect.
  • Some embodiments also disclose an ultrasound imaging method.
  • some embodiments of ultrasound imaging methods include the following steps:
  • Step 100 Obtain scanning rules.
  • Step 110 Control multiple three-dimensional ultrasound scans of the fetal heart according to the scanning rules, and obtain multiple volumes of three-dimensional ultrasound data containing multiple cardiac cycles, so that different volumes of the three-dimensional ultrasound data correspond to different patterns of the fetal heart.
  • Heart phase position wherein, each three-dimensional ultrasound scan includes: transmitting ultrasound waves to the fetal heart and receiving corresponding ultrasound echoes to obtain a volume of three-dimensional ultrasound data containing the fetal heart.
  • each volume of three-dimensional ultrasound data contains a complete fetal heartbeat.
  • Step 120 Perform data rearrangement on the multiple volumes of the three-dimensional ultrasound data according to the sequence of cardiac phases to obtain the multiple volumes of the three-dimensional ultrasound data after data rearrangement.
  • Step 130 Display the fetal heart rate based on the data-rearranged multi-volume three-dimensional ultrasound data.
  • the scanning rules include: in the process of performing multiple three-dimensional ultrasonic scans, each time the three-dimensional ultrasonic scan is completed, the next three-dimensional ultrasonic scan will be performed after the same delay time. For example, a total of N1 three-dimensional ultrasonic scans have been performed.
  • the next three-dimensional ultrasonic scan will be performed after the same delay time; in other words, After completing the first three-dimensional ultrasonic scan, perform a second three-dimensional ultrasonic scan after a delay period; after completing the second three-dimensional ultrasonic scan, perform a third three-dimensional ultrasonic scan after the same delay period, and so on.
  • the same delay elapses after completing the N1-1th three-dimensional ultrasound scan
  • the N1th three-dimensional ultrasound scan will be performed after a long period of time.
  • the delay duration is calculated based on the fetal heart rate, the basic zero delay volume rate, and the first parameter.
  • the basic zero-delay volume rate in this article is the scan volume rate of the first 3D ultrasound scan among the multiple 3D ultrasound scans performed.
  • the first parameter includes temporal resolution or interphase time interval.
  • the central phase time interval is the time interval between the central phase positions corresponding to two adjacent volumes of three-dimensional ultrasound data in the multi-volume three-dimensional ultrasound data after data rearrangement.
  • the ultrasound imaging method of some embodiments includes the following steps:
  • Step 101 Get the delay duration.
  • Step 111 Control multiple three-dimensional ultrasound scans on the fetal heart to obtain multi-volume three-dimensional ultrasound data containing multiple cardiac cycles; wherein, during the process of multiple three-dimensional ultrasound scans: each time the three-dimensional ultrasound scan is completed, Ultrasound scanning, after the delay time, performs the next three-dimensional ultrasonic scan; and each three-dimensional ultrasonic scan includes: transmitting ultrasonic waves to the fetal heart and receiving corresponding ultrasonic echoes to obtain a roll containing fetal heart.
  • Three-dimensional ultrasound data of the heart In some embodiments, each volume of three-dimensional ultrasound data contains a complete fetal heartbeat.
  • Step 120 Perform data rearrangement on the multiple volumes of the three-dimensional ultrasound data according to the sequence of cardiac phases to obtain the multiple volumes of the three-dimensional ultrasound data after data rearrangement.
  • Step 130 Display the fetal heart rate based on the data-rearranged multi-volume three-dimensional ultrasound data.
  • the delay length is such that the acquired multiple volumes of three-dimensional ultrasound data correspond to different cardiac phase positions of the fetal heart.
  • the delay length is such that the acquired multi-volume three-dimensional ultrasound data corresponds to different cardiac phase positions of the fetal heart, and the distance between adjacent cardiac phase positions is equal.
  • the delay length is such that the cardiac phase time interval and the time resolution are multiplied to 1, and the cardiac phase time interval is the cardiac phase corresponding to two adjacent volumes of three-dimensional ultrasound data among the multiple volumes of three-dimensional ultrasound data after data rearrangement. The time interval between locations.
  • step 101 obtains the heart rate of the fetal heart, and calculates the delay duration based on the heart rate, the basic zero-delay volume rate, and the first parameter.
  • the basic zero-delay volume rate in this article is the scan volume rate of the first 3D ultrasound scan among the multiple 3D ultrasound scans performed.
  • the first parameter includes temporal resolution or interphase time interval.
  • the central phase time interval is the time interval between the central phase positions corresponding to two adjacent volumes of three-dimensional ultrasound data in the multi-volume three-dimensional ultrasound data after data rearrangement.
  • step 101 calculates the delay duration based on the heart rate, basic zero-delay volume rate and the first parameter, including:
  • DT represents the delay duration
  • BVR represents the basic zero-delay volume rate
  • TS represents the time resolution
  • FR represents the fetal heart rate
  • floor represents the downward rounding operation.
  • the ultrasound imaging method further includes:
  • Step 102 Generate a setting interface, which is used to set the time resolution.
  • Step 103 In response to the setting instruction, set the time resolution. For example, the user sets the time resolution on the setting interface through an input tool such as a mouse or keyboard.
  • step 101 calculates the delay duration based on the heart rate, basic zero-delay volume rate and the first parameter, including:
  • DT represents the delay duration
  • BVR represents the basic zero-delay volume rate
  • TI represents the heart phase time interval
  • FR represents the fetal heart rate
  • floor represents the downward rounding operation.
  • the ultrasound imaging method further includes:
  • Step 104 Generate a setting interface, which is used to set the heart phase time interval.
  • Step 105 In response to the setting instruction, set the heart phase time interval. For example, the user sets the heart phase time interval on the setting interface through an input tool such as a mouse or keyboard.
  • the fetal heart rate refers to the number of times the fetal heart beats per unit time (such as the number of times it runs per minute), which can be obtained through multiple methods, such as fetal heart signal monitoring, ultrasound Doppler Or M-ultrasound to measure fetal heart rate.
  • the method of fetal heart signal monitoring is usually to extract and separate the fetal ECG signal from the mixed ECG signal in the mother's abdomen, and calculate the fetal heart rate based on the separated fetal ECG signal.
  • the separation algorithms include adaptive filtering algorithm, template subtraction and blind signal. Analysis etc.
  • the method based on ultrasound Doppler mainly utilizes the Doppler effect; since fetal heart movement is cyclical, the echo frequency shift signal is also cyclical.
  • the fetal heart rate can be calculated.
  • M-ultrasound uses the periodicity of heart beating to move the echo signal from left to right and displays the time position curve image of the periodic heartbeat, so that the time position relationship can be further used to calculate the fetal heart rate.
  • the number of the above-mentioned multiple three-dimensional ultrasound scans is m ⁇ n-1.
  • the number of volumes of the above-mentioned multi-volume three-dimensional ultrasound data is m ⁇ n-1.
  • m floor[(TS+1) ⁇ BVR ⁇ 60 ⁇ FR ⁇ TS];
  • BVR represents the basic zero-delay volume rate
  • FR represents the fetal heart rate
  • TS represents the time resolution
  • floor represents the downward rounding operation.
  • m floor[(60 ⁇ FR ⁇ TI+1) ⁇ BVR ⁇ TI]
  • BVR represents the basic zero-delay volume rate
  • TI represents the cardiac phase interval
  • FR represents the fetal heart rate
  • floor represents the downward rounding operation.
  • data rearrangement in step 120 at least includes adjusting part of the three-dimensional ultrasound data with a later scan timing forward, and adjusting part of the three-dimensional ultrasound data with a previous scan timing backward.
  • performing data rearrangement in step 120 includes: the (i-1) ⁇ m+j volume of three-dimensional ultrasound data in the m ⁇ n-1 volumes of three-dimensional ultrasound data before data rearrangement is set to data rearrangement.
  • the ranked (j-1) ⁇ n+i volume of three-dimensional ultrasound data; i and j are positive integers, and i is less than or equal to n, and j is less than or equal to m.
  • the three-dimensional ultrasound data before data rearrangement are arranged according to the scanning time sequence, that is, they are arranged according to the time sequence of the corresponding three-dimensional ultrasound scans.
  • the three-dimensional ultrasound data after data rearrangement are arranged according to the cardiac phase sequence.
  • data rearrangement in step 120 includes: dividing the m ⁇ n-1 volume of three-dimensional ultrasound data before data rearrangement into n parts according to the time sequence of their corresponding three-dimensional ultrasound scans, from the 1st to the nth part.
  • -1 parts each contain the three-dimensional ultrasound data in volume m, and the n-th part contains the three-dimensional ultrasound data in volume m-1; any i-th part of the m ⁇ n-1 volumes of three-dimensional ultrasound data before rearranging the data
  • the jth volume of three-dimensional ultrasound data in the part is set to the (j-1) ⁇ n+ith volume of three-dimensional ultrasound data after data rearrangement; i and j are positive integers, and i is less than or equal to n, and j is less than Or equal to m.
  • data rearrangement in step 120 includes: dividing the m ⁇ n-1 volume of three-dimensional ultrasound data before data rearrangement into n parts according to the time sequence of their corresponding three-dimensional ultrasound scans, from the 1st to the nth part.
  • the -1 parts each contain the three-dimensional ultrasound data described in volume m, and the n-th part contains the three-dimensional ultrasound data described in volume m-1; the j-th parts in the first to nth parts before the data are rearranged
  • the three-dimensional ultrasonic data are respectively set to the three-dimensional ultrasonic data of the 1st to nth volumes in the jth part after data rearrangement, where j is a positive integer less than m; the 1st to nth volumes before data rearrangement are The three-dimensional ultrasonic data of the respective m-th volume in one part are respectively set to the three-dimensional ultrasonic data of the first to n-1 volumes in the m-th part after data rearrangement.
  • the multiple volumes of three-dimensional ultrasound data after data rearrangement correspond to different cardiac phase positions of a complete cardiac cycle equidistantly.
  • step 130 performs data registration and/or boundary correction on the multiple volumes of three-dimensional ultrasound data after data rearrangement to obtain processed multiple volumes of three-dimensional ultrasound data; according to the processed multiple volumes of three-dimensional ultrasound data
  • the three-dimensional ultrasound data is used to display the fetal heart rate.
  • step 130 displays the fetal heartbeat, including: sequentially displaying three-dimensional ultrasound data images of one cardiac cycle in sequence of cardiac phases.
  • step 110 or step 111 also obtains the total scan time; after the total scan time is reached, Step 110 or step 111 controls to stop the three-dimensional ultrasonic scanning; or, step 110 or step 111 generates a prompt to stop the three-dimensional ultrasonic scanning based on the total scanning time.
  • T represents the total scanning time
  • FR represents the fetal heart rate
  • TI represents the cardiac phase interval
  • BVR represents the basic zero-delay volume rate
  • floor represents the downward rounding operation.
  • T represents the total scanning time
  • TS represents the time resolution
  • FR represents the fetal heart rate
  • BVR represents the basic zero-delay volume rate
  • floor represents the downward rounding operation.
  • step 110 or step 111 also obtains the number of three-dimensional ultrasound scans. After completing the number of three-dimensional ultrasound scans, step 110 or step 111 controls to stop performing the three-dimensional ultrasound scan; or, step 110 or step 111 is based on The number of three-dimensional ultrasonic scans is used to generate a prompt to stop performing the three-dimensional ultrasonic scans. In some embodiments, the number of three-dimensional ultrasound scans is m ⁇ n-1.
  • some embodiments of ultrasound imaging methods include the following steps:
  • Step 200 Control to perform m ⁇ n-1 three-dimensional ultrasound scans on the fetal heart to obtain m ⁇ n-1 volumes of three-dimensional ultrasound data; m and n are integers greater than or equal to 2; among them, the first n-1 In each cardiac cycle of the fetal heart: perform m times of the three-dimensional ultrasound scans to obtain m volumes of the three-dimensional ultrasound data; in the nth cardiac cycle of the fetal heart: perform m-1 times of the three-dimensional ultrasound scans Obtain the three-dimensional ultrasound data of volume m-1.
  • each three-dimensional ultrasound scan includes: transmitting ultrasound waves to the fetal heart and receiving corresponding ultrasound echoes to obtain a volume of three-dimensional ultrasound data containing the fetal heart; in some embodiments, each volume of three-dimensional ultrasound data contains a complete of fetal heart rate.
  • m floor[(TS+1) ⁇ BVR ⁇ 60 ⁇ FR ⁇ TS];
  • BVR represents the basic zero-delay volume rate
  • FR represents the fetal heart rate
  • TS represents the time resolution
  • floor represents the downward rounding operation.
  • m floor[(60 ⁇ FR ⁇ TI+1) ⁇ BVR ⁇ TI]
  • BVR represents the basic zero-delay volume rate
  • TI represents the cardiac phase interval
  • FR represents the fetal heart rate
  • floor represents the downward rounding operation.
  • step 200 also obtains the delay time; in the process of performing m ⁇ n-1 three-dimensional ultrasonic scans: each time the three-dimensional ultrasonic scan is completed, after the delay time, the next time is performed. Three-dimensional ultrasound scan.
  • the delay duration is such that the acquired m ⁇ n-1 volumes of three-dimensional ultrasound data correspond to different cardiac positions of the fetal heart.
  • the delay duration is such that the acquired m ⁇ n-1 volumes of three-dimensional ultrasound data correspond to different cardiac phase positions of the fetal heart, and the distance between adjacent cardiac phase positions is equal.
  • the delay length is such that the heart phase interval and the time resolution are multiplied to one.
  • the delay duration is calculated based on the fetal heart rate, the basic zero delay volume rate, and the first parameter.
  • the basic zero-delay volume rate in this article is the scan volume rate of the first 3D ultrasound scan among the multiple 3D ultrasound scans performed.
  • the first parameter includes temporal resolution or interphase time interval.
  • the central phase time interval is the time interval between the central phase positions corresponding to two adjacent volumes of three-dimensional ultrasound data in the multi-volume three-dimensional ultrasound data after data rearrangement.
  • step 200 obtains the heart rate of the fetal heart, and calculates the delay duration based on the heart rate, the basic zero-delay volume rate, and the first parameter.
  • the basic zero-delay volume rate in this article is the scan volume rate of the first 3D ultrasound scan among the multiple 3D ultrasound scans performed.
  • the first parameter includes temporal resolution or interphase time interval.
  • the central phase time interval is the time interval between the central phase positions corresponding to two adjacent volumes of three-dimensional ultrasound data in the multi-volume three-dimensional ultrasound data after data rearrangement.
  • step 200 calculates the delay duration based on the heart rate, basic zero-delay volume rate and the first parameter, including:
  • DT represents the delay duration
  • BVR represents the basic zero-delay volume rate
  • TS represents the time resolution
  • FR represents the fetal heart rate
  • floor represents the downward rounding operation.
  • step 200 calculates the delay duration based on the heart rate, basic zero-delay volume rate and the first parameter, including:
  • DT represents the delay duration
  • BVR represents the basic zero-delay volume rate
  • TI represents the heart phase time interval
  • FR represents the fetal heart rate
  • floor represents the downward rounding operation.
  • the value of TI may be 0.01 to 0.03 seconds.
  • step 200 also obtains the total scanning time; after the total scanning time is reached, step 200 controls to stop the three-dimensional ultrasonic scanning; or, step 200 generates a prompt to stop the three-dimensional ultrasonic scanning based on the total scanning time.
  • T represents the total scanning time
  • FR represents the fetal heart rate
  • TI represents the cardiac phase interval
  • BVR represents the basic zero-delay volume rate
  • floor represents the downward rounding operation.
  • T represents the total scanning time
  • TS represents the time resolution
  • FR represents the fetal heart rate
  • BVR represents the basic zero-delay volume rate
  • floor represents the downward rounding operation.
  • Step 210 Rearrange the m ⁇ n-1 volume of three-dimensional ultrasound data according to the sequence of cardiac phases to obtain the m ⁇ n-1 volume of three-dimensional ultrasound data after data rearrangement.
  • data rearrangement at least includes adjusting part of the three-dimensional ultrasound data with a later scan timing forward, and adjusting part of the three-dimensional ultrasound data with a previous scan timing backward.
  • the m ⁇ n-1 volume of the three-dimensional ultrasound data after data rearrangement corresponds to different cardiac phase positions of a complete cardiac cycle equidistantly.
  • step 210 performs data rearrangement on the m ⁇ n-1 volume of three-dimensional ultrasound data according to the order of cardiac phases to obtain the m ⁇ n-1 volume of three-dimensional ultrasound data after data rearrangement, including: data rearrangement.
  • the 1st to m-1th parts respectively contain the three-dimensional ultrasound data described in the n volumes
  • the m-th part Parts include n-1 volumes of the three-dimensional ultrasound data; m obtained by m times of the three-dimensional ultrasound scans in any i-th cardiac cycle of the first n-1 fetal heart cycles
  • the volume of the three-dimensional ultrasonic data is set to the i-th volume of the three-dimensional ultrasonic data of each part in the m ⁇ n-1 volume of the three-dimensional ultrasonic data after data rearrangement; the i is a positive integer, and i is less than or equal to n-1; in
  • step 210 performs data rearrangement on the m ⁇ n-1 volume of three-dimensional ultrasound data according to the order of cardiac phases to obtain the m ⁇ n-1 volume of three-dimensional ultrasound data after data rearrangement, including: data rearrangement.
  • the three-dimensional ultrasound data described in the (i-1) ⁇ m+jth volume among the m ⁇ n-1 volumes of the three-dimensional ultrasound data ranked first is set to the (j-1) ⁇ n+ith volume after data rearrangement.
  • Volume the three-dimensional ultrasound data; the i and j are positive integers, and i is less than or equal to n, and j is less than or equal to m.
  • the three-dimensional ultrasound data before data rearrangement are arranged according to the scanning time sequence, that is, they are arranged according to the time sequence of the corresponding three-dimensional ultrasound scans.
  • the three-dimensional ultrasound data after data rearrangement are arranged according to the cardiac phase sequence.
  • step 210 performs data rearrangement on the m ⁇ n-1 volume of three-dimensional ultrasound data according to the order of cardiac phases to obtain the m ⁇ n-1 volume of three-dimensional ultrasound data after data rearrangement, including:
  • the three-dimensional ultrasound data in the m ⁇ n-1 volume before rearrangement is divided into n parts according to the time sequence of the corresponding three-dimensional ultrasound scans.
  • the first to n-1th parts each contain the three-dimensional ultrasound data in the m volume.
  • the nth part contains the three-dimensional ultrasound data described in volume m-1;
  • the j-th volume three-dimensional ultrasound data in any i-th part of the three-dimensional ultrasound data described in volume m ⁇ n-1 before data rearrangement, set is the three-dimensional ultrasound data of the (j-1) ⁇ n+ith volume after data rearrangement;
  • the i and j are positive integers, and i is less than or equal to n, and j is less than or equal to m.
  • step 210 performs data rearrangement on the m ⁇ n-1 volume of three-dimensional ultrasound data according to the order of cardiac phases to obtain the m ⁇ n-1 volume of three-dimensional ultrasound data after data rearrangement, including:
  • the three-dimensional ultrasound data in the m ⁇ n-1 volume before rearrangement is divided into n parts according to the time sequence of the corresponding three-dimensional ultrasound scans.
  • the first to n-1th parts each contain the three-dimensional ultrasound data in the m volume.
  • Data, the nth part contains the three-dimensional ultrasound data described in volume m-1; the three-dimensional ultrasound data described in the jth volume in the first to nth parts before data rearrangement are respectively set to the third volume after data rearrangement.
  • the data are respectively set to the three-dimensional ultrasound data described in volumes 1 to n-1 in the m-th part after data rearrangement.
  • Step 220 Display the fetal heart rate based on the m ⁇ n-1 volume of three-dimensional ultrasound data after data rearrangement.
  • step 220 may also perform data registration and/or boundary correction on the multi-volume three-dimensional ultrasound data after the data rearrangement.
  • step 220 performs data registration and/or boundary correction on the m ⁇ n-1 volume of three-dimensional ultrasonic data after data rearrangement to obtain the processed m ⁇ n-1 volume of three-dimensional ultrasonic data; step 220 is based on the processing After the m ⁇ n-1 volume of the three-dimensional ultrasound data is displayed, the fetal heart rate is displayed.
  • step 220 displays the fetal heartbeat, including: sequentially displaying three-dimensional ultrasound data images of one cardiac cycle in order of heart phases.
  • These computer program instructions may be loaded onto a general-purpose computer, special-purpose computer, or other programmable data processing apparatus to form a machine, such that the instructions executed on the computer or other programmable data processing apparatus may generate a device that implements the specified functions.
  • These computer program instructions may also be stored in a computer-readable memory, which may instruct a computer or other programmable data processing device to operate in a specific manner, such that the instructions stored in the computer-readable memory may form a Manufactured articles include devices that perform specified functions.
  • Computer program instructions may also be loaded onto a computer or other programmable data processing device to perform a series of operating steps on the computer or other programmable device to produce a computer-implemented process such that the execution on the computer or other programmable device Instructions can provide steps for implementing a specified function.
  • the term “comprises” and any other variations thereof are intended to be non-exclusively inclusive such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also those not expressly listed or otherwise not part of the process , methods, systems, articles or other elements of equipment.
  • the term “coupled” and any other variations thereof as used herein refers to physical connection, electrical connection, magnetic connection, optical connection, communication connection, functional connection and/or any other connection.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Medical Informatics (AREA)
  • Biophysics (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Pathology (AREA)
  • Radiology & Medical Imaging (AREA)
  • Cardiology (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Physics & Mathematics (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

一种胎心的超声成像方法和系统，包括：取延迟时长（101）；控制向胎心进行多次三维超声扫描，获取包含多个心动周期的多卷三维超声数据；其中，每完成一次三维超声扫描，经过延迟时长，再进行下一次三维超声扫描（111）；对多卷三维超声数据按照心相顺序进行数据重排，得到数据重排后的多卷三维超声数据（120）；根据数据重排后的多卷三维超声数据，进行胎心的显示（130）。

Description

一种胎心的超声成像方法和系统 技术领域
本发明涉及对胎儿心脏的超声成像领域，具体涉及一种胎心的超声成像方法和系统。
背景技术
根据一份临床研究的统计，在1000个顺利生产存活的婴儿里面会有8个带有先天性心脏病，而在由于畸形导致新生儿死亡的病例里面，先天性心脏病就占到了近五成。由此可见先心病发生概率之大，后果影响之严重，这些都让胎心(胎儿心脏)检查成为产前诊断尤为重要的环节。胎儿心脏的运动特性，使得传统的三维采集方式(自由臂3D、静态3D和容积探头4D采集)难以满足快速运动的3D胎心数据采集要求。而传统的二维超声检查，又由于胎儿心脏结构的复杂性，不能直接提供病变部位三维空间的结构形态及其位置关系，导致很多潜在的病变在二维超声中难以检查出来。
时空关联成像技术(STIC)的提出，可以获取高的空间分辨率剖面图像，且具有很高的时间分辨率，不仅降低了胎心扫查技术的门槛，减少操作者技能依赖性，也使得胎心检查更加方便、快捷，对于胎心复杂结构的观察、心脏血流动力学的检查诊断都具有至关重要的意义。
STIC的成像原理是首先采集多个心动周期的二维图像，计算心率，并根据心率将不同心相位置的二维图像重组为三维数据，最终将重组后的三维数据根据心相顺序循环播放，从而达到显示一个3D跳动心脏的目的。为了获得足够多的心动周期来提高重建后3D数据的分辨率，STIC的采集时间较长(例如一般需25个心动周期左右，约12秒左右)，在此过程中极易受到胎儿运动的影响，导致采集过程失败。此外，受成像原理的影响，STIC重建后的数据B、C面分辨率较低(如图1所示，边缘存在锯齿状)，低质量的数据可能会导致诊断过程忽略一些微小的隔膜病变，造成误诊。
如何对胎心进行三维超声成像实现运动胎心的跳动显示，是本领域技术人员一直在研究和改进的问题。
发明内容
基于上述考虑，本发明主要提供一种胎心的超声成像方法和系统，下面具体说明。
根据第一方面，一种实施例中提供一种胎心的超声成像方法，包括：
获取延迟时长；
控制向所述胎心进行多次三维超声扫描，获取包含多个心动周期的多卷三维超声数据；其中，在进行多次所述三维超声扫描的过程中：每完成一次所述三维超声扫描，经过所述延迟时长，再进行下一次所述三维超声扫描；并且，每次所述三维超声扫描包括：向所述胎心发射超声波以及接收相应的超声波回波，得到一卷包含胎心的三维超声数据；
对多卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的多卷所述三维超声数据；所述数据重排至少包括将部分扫描时序在后的所述三维超声数据向前调整，以及部分扫描时序在前的所述三维超声数据调整向后调整；
根据所述数据重排后的多卷三维超声数据，进行所述胎心的显示。
根据第二方面，一种实施例中提供一种胎心的超声成像方法，包括：
控制向所述胎心进行m×n-1次三维超声扫描，获取m×n-1卷三维超声数据；m和n都为大于或等于2的整数；其中，在前n-1个所述胎心的每个心动周期内：进行m次所述三维超声扫描得到m卷所述三维超声数据；在第n个所述胎心的心动周期内：进行m-1次所述三维超声扫描得到m-1卷所述三维超声数据；
对m×n-1卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷所述三维超声数据；所述数据重排至少包括将部分扫描时序在后的所述三维超声数据向前调整，以及部分扫描时序在前的所述三维超声数据调整向后调整；
根据数据重排后的m×n-1卷所述三维超声数据，进行所述胎心的显示。
根据第三方面，一种实施例中提供一种胎心的超声成像方法，包括：
获取扫描规则；
根据所述扫描规则控制对所述胎心进行多次三维超声扫描，获取包含多个心动周期的多卷三维超声数据，使得不同卷的所述三维超声数据对应所述胎心的不同心相位置；其中，每次所述三维超声扫描包括：向所述胎心发射超声波以及接收相应的超声波回波，得到一卷包含胎心的三维超声数据；
对多卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的多卷所述三维超声数据；
根据所述数据重排后的多卷三维超声数据，进行所述胎心的显示。
根据第四方面，一种实施例中提供一种超声成像系统，包括：
超声探头，用于向胎心发射超声波，以及接收相应的超声波回波信号；
发射和接收控制电路，用于控制所述超声探头执行发射超声波和接收超声波回波信号；
处理器，用于执行如本文任一实施例所述的方法。
附图说明
图1为通过时空关联成像技术所得到的超声图像的示意图；
图2为一种实施例的超声成像系统的结构示意图；
图3为一种实施例的三维超声扫描时序的示意图；
图4为一种实施例的超声成像方法的原理图；
图5为一种实施例的数据重排的示意图；
图6为一种实施例的数据配准的示意图；
图7为一种实施例的数据配准的示意图；
图8为一种实施例的边界修正的示意图；
图9为一种实施例的超声成像方法的流程图；
图10为一种实施例的超声成像方法的流程图；
图11为一种实施例中设置时间分辨率的流程图；
图12为一种实施例中设置心相时间间隔的流程图；
图13为一种实施例的超声成像方法的流程图。
具体实施方式
下面通过具体实施方式结合附图对本发明作进一步详细说明。其中不同实施方式中类似元件采用了相关联的类似的元件标号。在以下的实施方式中，很多细节描述是为了使得本申请能被更好的理解。然而，本领域技术人员可以毫不费力的认识到，其中部分特征在不同情况下是可以省略的，或者可以由其他元件、材料、方法所替代。在某些情况下，本申请相关的一些操作并没有在说明书中显示或者描述，这是为了避免本申请的核心部分被过多的描述所淹没，而对于本领域技术人员而言，详细描述这些相关操作并不是必要的，他们根据说明书中的描述以及本领域的一般技术知识即可完整了解相关操作。
另外，说明书中所描述的特点、操作或者特征可以以任意适当的方式结合形成各种实施方式。同时，方法描述中的各步骤或者动作也可以按照本领域技术人员所能显而易见的方式进行顺序调换或调整。因此，说明书和附图中的各种顺序只是为了清楚描述某一个实施例，并不意味着是必须的顺序，除非另有说明其中某个顺序是必须遵循的。
本文中为部件所编序号本身，例如“第一”、“第二”等，仅用于区分所描述的对象，不具有任何顺序或技术含义。而本申请所说“连接”、“联接”，如无特别说明，均包括直接和间接连接(联接)。
请参照图2，一些实施例中提供一种超声成像系统，超声成像系统包括超声探 头10、发射和接收控制电路20和处理器40；一些实施例中还可以包括回波处理模块30和/或显示器50，下面对各部件进行说明。
超声探头10用于向感兴趣区域例如胎心发射超声波，以及接收相应的超声波回波信号，得到超声数据，例如二维的超声数据或三维的超声数据。一些具体实施例中，超声探头10包括多个阵元，用于实现电脉冲信号和超声波的相互转换，从而实现向感兴趣区域发射超声波并接收相应的超声波回波信号。阵元可根据激励电信号发射超声波，或将接收的超声波变换为电信号。因此每个阵元可用于向感兴趣区域的生物组织发射超声波，也可用于接收经组织返回的超声波回波。在进行超声检测时，可通过发射序列和接收序列控制哪些阵元用于发射超声波，哪些阵元用于接收超声波，或者控制阵元分时隙用于发射超声波或接收超声回波。参与超声波发射的所有阵元可以被电信号同时激励，从而同时发射超声波；或者参与超声波发射的阵元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。
一些实施例中，超声探头10可以为容积探头。一些实施例中，超声探头10可以为面阵探头。
发射和接收控制电路20用于控制超声探头10执行发射超声波和接收超声波回波信号。例如发射和接收控制电路20一方面用于控制超声探头10向感兴趣区域发射超声波，另一方面用于控制超声探头10接收超声波经感兴趣区域反射的超声波回波信号。一些具体实施例中，发射和接收控制电路20用于产生发射序列和接收序列，并输出至超声探头10。发射序列用于控制超声探头10中多个阵元中的部分或者全部向生物组织60发射超声波，发射序列的参数包括发射用的阵元数和超声波发射参数(例如幅度、频率、发波次数、发射间隔、发射角度、波型和/或聚焦位置等)。接收序列用于控制多个阵元中的部分或者全部接收超声波经组织后的回波，接收序列的参数包括接收用的阵元数以及回波的接收参数(例如接收角度、深度等)。对超声回波的用途不同或根据超声回波生成的图像不同，发射序列中的超声波参数和接收序列中的回波参数也有所不同。
回波处理模块30用于对超声探头10接收到的超声回波信号进行处理，例如对超声回波信号进行滤波、放大、波束合成等处理，得到超声回波数据。在具体实施例中，回波处理模块30可以将超声回波数据输出给处理器40，也可以将超声回波数据先存储在一存储器中，在需要基于超声回波数据进行运算时，处理器40从存储器中读取超声回波数据。本领域技术人员应当理解，在有的实施例中，当不需要对超声回波信号进行滤波、放大、波束合成等处理时，回波处理模块30也可以省略。
处理器40用于获取超声回波数据或者说信号，并采用相关算法得到所需要的 参数或图像。本发明一些实施例中的处理器40包括但不限于中央处理器(Central Processing Unit，CPU)、微控制单元(Micro Controller Unit，MCU)、现场可编程门阵列(Field-Programmable Gate Array，FPGA)和数字信号处理(DSP)等用于解释计算机指令以及处理计算机软件中的数据的装置。一些实施例中，处理器40用于执行该非暂时性计算机可读存储介质中的各计算机应用程序，从而使样本分析装置执行相应的检测流程。
显示器50可以用于显示信息，例如显示由处理器40计算得到的参数和图像等。本领域技术人员应当理解，在有的实施例中，超声成像系统本身可以不集成显示模块，而是连接一个计算机设备(例如电脑)，通过计算机设备的显示模块(例如显示屏)来显示信息。
以上为超声成像系统的一些说明。
一些实施例中，处理器40获取扫描规则；以及处理器40根据扫描规则控制对胎心进行多次三维超声扫描，获取包含多个心动周期的多卷三维超声数据，使得不同卷的所述三维超声数据对应所述胎心的不同心相位置；其中，每次三维超声扫描包括：向胎心发射超声波以及接收相应的超声波回波，得到一卷包含胎心的三维超声数据；一些实施例中，每一卷三维超声数据都包含完整的胎心。处理器40对这多卷三维超声数据按照心相顺序进行数据重排，得到数据重排后的多卷三维超声数据；处理器40根据数据重排后的这多卷三维超声数据，控制显示器50进行胎心的显示。
可以理解地，对胎心进行多次三维超声扫描获取包含多个心动周期的多卷三维超声数据，指的是对胎心进行了多次三维超声扫描得到多卷三维超声数据，这多卷三维超声数据包含了多个心动周期的数据。
一些实施例中，扫描规则包括：在进行多次上述三维超声扫描的过程中，每完成一次所述三维超声扫描，经过相同的延迟时长，再进行下一次三维超声扫描。例如总共进行了N1次三维超声扫描，在进行这N1次三维超声扫描的过程中，每完成一次所述三维超声扫描，经过相同的延迟时长，再进行下一次三维超声扫描；换句话说，在完成第1次三维超声扫描后，经过一延迟时长后再进行第2次三维超声扫描；在完成第2次三维超声扫描后，经过一相同的延迟时长后再进行第3次三维超声扫描，如此进行，直到在完成第N1-1次三维超声扫描后，经过一相同的延迟时长后再进行第N1次三维超声扫描。
请参照图3，如上所述，通过一次三维超声扫描得到一卷三维超声数据，一次三维超声扫描包括发射超声波以及接收相应的超声波回波，因此，本文中完成一次三维超声扫描后，经过延迟时长再进行下一次三维超声扫描，指的是：在前一次三维超声扫描完成发射超声波以及接收相应的超声波回波后，以这个时间节点计时， 经过延迟时长后，再开始进行下一次三维超声扫描。
一些实施例中，延迟时长基于胎心的心率、基础零延时卷率和第一参数计算得到。本文中基础零延时卷率为所进行的这多次三维超声扫描中第一次三维超声扫描的扫描卷率。一些实施例中，第一参数包括时间分辨率或心相时间间隔。本文中心相时间间隔为数据重排后的这多卷三维超声数据中相邻两卷三维超声数据所对应的心相位置之间的时间间隔。
因此，一些实施例中，处理器40获取延迟时长；处理器40控制向胎心进行多次三维超声扫描，获取包含多个心动周期的多卷三维超声数据；其中，在进行这多次三维超声扫描的过程中：每完成一次三维超声扫描，经过所述延迟时长，再进行下一次三维超声扫描；并且，每次三维超声扫描包括：向胎心发射超声波以及接收相应的超声波回波，得到一卷包含胎心的三维超声数据；一些实施例中，每一卷三维超声数据都包含完整的胎心。处理器40对这多卷三维超声数据按照心相顺序进行数据重排，得到数据重排后的多卷三维超声数据；处理器40根据数据重排后的这多卷三维超声数据，控制显示器50进行胎心的显示。
一些实施例中，延迟时长使得所获取的多卷三维超声数据对应胎心的不同心相位置。
一些实施例中，延迟时长使得所获取的多卷三维超声数据对应胎心的不同心相位置，且各相邻的心相位置之间的距离相等。
一些实施例中，延迟时长使得心相时间间隔和时间分辨率相乘为1，心相时间间隔为数据重排后的这多卷三维超声数据中相邻两卷三维超声数据所对应的心相位置之间的时间间隔。
一些实施例中，处理器40获取胎心的心率，并根据心率、基础零延时卷率和第一参数计算得到延迟时长。本文中基础零延时卷率为所进行的这多次三维超声扫描中第一次三维超声扫描的扫描卷率。一些实施例中，第一参数包括时间分辨率或心相时间间隔。本文中心相时间间隔为数据重排后的这多卷三维超声数据中相邻两卷三维超声数据所对应的心相位置之间的时间间隔。
一些实施例中，处理器40根据心率、基础零延时卷率和第一参数计算得到延迟时长，包括：

其中，DT表示延迟时长，BVR表示基础零延时卷率，TS表示时间分辨率，FR表示胎心的心率，floor表示向下取整操作。
一些实施例中，处理器40控制显示器50生成设置界面，设置界面用于设置时间分辨率；响应于设置指令，处理器40设置时间分辨率。例如，用户通过输入工具例如鼠标或键盘等在设置界面上设置时间分辨率。
一些实施例中，处理器40根据心率、基础零延时卷率和第一参数计算得到延迟时长，包括：

其中，DT表示延迟时长，BVR表示基础零延时卷率，TI表示心相时间间隔，FR表示胎心的心率，floor表示向下取整操作。
一些实施例中，处理器40控制显示器50生成设置界面，设置界面用于设置心相时间间隔；响应于设置指令，处理器40设置心相时间间隔。例如，用户通过输入工具例如鼠标或键盘等在设置界面上设置心相时间间隔。
需要说明的是，胎心的心率是指胎儿心脏在单位时间内跳动的次数(例如每分钟跑去的次数)，这可以通过多次方法获取，例如胎心信号监测的方式、超声多普勒或M超测量胎心心率图。胎心信号监测的方式通常是从母亲腹部混合的心电信号中提取分离出胎儿的心电信号，根据分离的胎儿心电信号计算胎儿心率；分离算法有自适应滤波算法、模板减法和盲信号分析等。基于超声多普勒的方法主要是利用多普勒效应；由于胎心运动具有周期性，因此回波频移信号也具有周期性，通过对此频移信号进行处理，可以计算出胎儿的心率。M超则是利用心脏跳动的周期性，使回波信号从左向右自行移动扫描，显示出周期性心跳的时间位置曲线像，从而可以进一步利用时间位置关系计算出胎儿心率。
需要说明的是，本文中公式所涉及到的各物理量，其中心率的单位是次/分钟，心相时间间隔的单位是秒，卷率(包括基础零延时卷率)的单位是卷/秒，时间分辨率的单位是卷/一个心动周期。
下面对如何计算延时时长进行原理上的说明和推导。
请参照图4，为本申请一些实施例的原理图，可以看到，通过设定合理的延时时长，从而能够改变卷率(即1秒能够扫多少卷)，使得扫描过程得到的每一卷正好对应胎心不同的心相位置，且重排后的数据间时间间隔与预设的时间分辨率相乘为1，从而恰好等距的填充一个完整心动周期不同心相对应位置。因此，算法的核心是通过建立时间分辨率的等式实现，假设预设的时间分辨率用TS表示，心率用FR表示，基础零延时卷率用BVR表示，则心相时间间隔TI与时间分辨率TS关系为TS＝60/FR/TI；同时，时间分辨率通过基础零延时扫描卷率及心相时间间隔计算可 以表示为TS＝1/(BVR×TI)×m-1；其中，m表示扫描过程将一个完整心动周期分割为m部分，对应根据卷率在单个心动周期中扫描的次数，因此，m必定为整数，由此可以得到m的计算公式：
m1＝floor((TS+1)×BVR×TI)
其中，floor为向下或者说去尾取整操作(例如2.5取整后就是2)，由此可以得到整数m1。基于上述，可以求得增加延时后的实际卷率AVR为：
AVR＝m1/(TS+1)/TI＝floor((TS+1)×BVR×TI)/(TS+1)/TI；其中TS＝60/FR/TI；
根据基础零延时卷率以及实际卷率可以得到：
延时DT＝1/AVR-1/BVR。
以上是计算延时时长进行原理上的说明和推导。
如上所述，处理器40控制对胎心进行多次三维超声扫描，获取多卷三维超声数据；一些实施例中，这多次三维超声扫描的次数为m×n-1；一些实施例中，这多卷三维超声数据的卷数为m×n-1。
一些实施例中：

m＝floor[(TS+1)×BVR×60÷FR÷TS]；
BVR表示基础零延时卷率；FR表示胎心的心率；TS表示时间分辨率；floor表示向下取整操作。
一些实施例中：

m＝floor[(60÷FR÷TI+1)×BVR×TI]；
BVR表示基础零延时卷率；TI表示心相时间间隔；FR表示胎心的心率，floor表示向下取整操作。
在获取上述多卷三维超声数据后，需要进行数据重排。一些实施例中，数据重排至少包括将部分扫描时序在后的三维超声数据向前调整，以及部分扫描时序在前的三维超声数据调整向后调整。需要说明的是，请参照图5，假设有扫描顺序先后顺序上的三卷三维超声数据A、B和C，若将其数据重排后得到A、C和B，那么可以是认为将部分扫描时序在后的三维超声数据向前调整，以及部分扫描时序在前的三维超声数据调整向后调整，例如将三维超声数C向前调整了，以及将三维超声数B向后调整了。
一些实施例中，处理器40进行数据重排包括：数据重排前的m×n-1卷所述三维超声数据中第(i-1)×m+j卷三维超声数据，被设置为数据重排后的第(j-1)×n+i 卷三维超声数据；i和j为正整数，且i小于或等于n，j小于或等于m。数据重排前的三维超声数据是按照扫描时序所排列的，即按照各自对应的三维超声扫描的时间先后顺序所排列的，数据重排后的三维超声数据则是按照心相顺序所排列的。
一些实施例中，处理器40进行数据重排包括：将数据重排前的m×n-1卷三维超声数据按照各自对应的三维超声扫描的时间先后顺序划分为n个部分，第1至第n-1个部分分别都包含m卷所述三维超声数据，第n个部分包含m-1卷所述三维超声数据；将数据重排前的m×n-1卷三维超声数据中任意第i个部分中的第j卷三维超声数据，设置为数据重排后的第(j-1)×n+i卷所述三维超声数据；i和j为正整数，且i小于或等于n，j小于或等于m。
一些实施例中，处理器40进行数据重排包括：将数据重排前的m×n-1卷三维超声数据按照各自对应的三维超声扫描的时间先后顺序划分为n个部分，第1至第n-1个部分分别都包含m卷所述三维超声数据，第n个部分包含m-1卷所述三维超声数据；将数据重排前的第1至第n个部分中各自的第j卷所述三维超声数据分别设置为数据重排后的第j个部分中的第1至n卷所述三维超声数据，其中j为小于m的正整数；将数据重排前的第1至第n-1个部分中各自的第m卷所述三维超声数据分别设置为数据重排后的第m个部分中的第1至n-1卷所述三维超声数据。
一些实施例中，数据重排后的这多卷三维超声数据等距地对应一个完整心动周期的不同心相位置。
一些实施例中，处理器40还获取总扫描时间；总扫描时间到达后，处理器40控制停止进行所述三维超声扫描；或者，处理器40根据总扫描时间生成停止进行三维超声扫描的提示。
一些实施例中，总扫描时间为：
T＝60÷FR÷TI÷AVR；
其中，T表示总扫描时间；FR表示胎心的心率；TI表示心相时间间隔；BVR表示基础零延时卷率；floor表示向下取整操作。
一些实施例中，总扫描时间为：
T＝TS÷AVR；
其中，T表示总扫描时间；TS表示时间分辨率；FR表示胎心的心率；BVR表示基础零延时卷率；floor表示向下取整操作。
一些实施例中，处理器40还获取三维超声扫描的次数，在进行完三维超声扫 描的次数后，处理器40控制停止进行所述三维超声扫描；或者，处理器40根据所述三维超声扫描的次数，生成停止进行所述三维超声扫描的提示。一些实施例中，三维超声扫描的次数m×n-1。
可以理解地，一些实施例中所涉及的总扫描次数虽然为m×n-1，但这是指与本申请方案相关的扫描次数；在实际过程中，可以由于各种原因而扫描更多次数，例如在上述的扫描的m×n-1次过程中，可以在它之前插入若干次扫描，也可以在它之后再插入若干次扫描，但由于这些扫描及扫描得到的数据与本申请无关，因此，本申请并没有提及这些扫描。
一些实施例中，在对数据重排后，处理器40还可以对数据重排后的上述多卷三维超声数据进行数据配准和/或边界修正。
由于在数据采集过程中胎儿运动是不可避免的，而数据又经过了重排，所以重排后的数据间有可能会缺失连续性，导致重排后的数据循环播放会存在跳动感，从而使得胎心的跳动感反而无法体现。一些实施例中，处理器40可以基于3D数据进行配准；一些实施例中，处理器40可以基于二维平面的3D数据进行配准，下面分别进行说明。
如图6所示，基于3D数据配准是直接在三维空间中搜索，实现两个3D数据的直接配准，通常可以分为传统算法、基于深度学习的配准算法和混合算法。
传统3D配准算法主要利用图像灰度或特征信息，通过一定的相似性度量进行配准。特征提取方法如下：SIFT提取了一种具有尺度不变性的局部特征描述子；Harris提取图像中显著的角点；HOG通过方向梯度直方图实现特征的提取；SURF通过Hessian矩阵与尺度空间，实现特征点的提取。相似性度量方法主要有绝对平方误差(SSD)、互信息(MI)、归一化互信息(NMI)、归一化互相关(NCC)、结构相似性(SSIM)等。此外，还有一些其他传统配准算法。Demons算法基于光流理论，利用参考图像的梯度及以及参考图像和浮动图像的灰度差值来计算每一个点的坐标偏移量，根据此偏移量进行迭代，从而实现3D数据配准；ICP算法先将3D数据提取特征点，变为三维点云数据，之后通过最近点搜索并迭代的算法，实现两个三维点云对齐，从而将3D数据配准。
基于深度学习的3D数据配准算法可以分为两类，第一类输入为两个3D数据，输出为两个3D数据的三维刚性变换矩阵；第二类是输入为两个3D数据(浮动数据和参考数据)，输出为与参考数据对齐配准后的浮动数据。以第一类配准算法为例，通过构建训练样本库，利用回归的策略实现三维变换参数的估计，从而对齐两个不同3D数据，实现3D数据的配准。主要步骤为：
Step 1：构建训练样本库步骤：
给定一个3D数据并初始化三维变换参数，通过改变三维刚性变换参数实现3D 数据的刚性变换(三个方向的平移和旋转，共6个参数)。训练样本库可以由变换前、后的两个3D数据以及三维刚性变换参数构成，且变换前、后的两个3D数据与变换参数之间具有一一对应关系，满足深度学习算法中样本与标签的约束关系。其中，变换前、后的两个3D数据为输入数据，三维刚性变换参数为输出的标签值。
Step 2：网络设计与训练步骤：
构建好训练样本库后，设计一种网络模型，回归拟合三维刚性变换参数。网络的设计主要包括卷积层、池化层、激励层、拼接层、全连接层以及所设定的损失函数，通过对这些层进行组合、堆叠来学习训练样本中的特征。对于训练样本库中的任意两个3D数据，输入所设计的网络模型，输出三维刚性变换参数的估计值。此类模型的损失函数一般由两部分组成，其一是利用估计值、真值间的差异组成部分损失函数，常用L1范数、L2范数等作为损失函数；其二是利用估计值对第二个3D数据进行三维刚性变换，并将输出与第一个输入的3D数据进行对比，通过两个3D数据间的差异组成另外的损失函数，常用SSD、L1范数、L2范数、NC、SSIM等作为损失函数。通过给定上述两种损失函数不同的权重关系，对模型参数进行优化，从而训练网络。常用的网络模型有AlexNet、VGGNet、GoogLeNet、MobileNet、SYMNet、STN等，该方法包含但不仅限于上述网络模型结构及损失函数。
Step 3：推理步骤：
在推理阶段，将待配准的两个3D数据输入至上述Step 2训练好的网络中，可实时得到三维刚性变换参数，将得到的三维刚性变换参数作用于第一个3D数据并对其进行变换，实现了第一个3D数据和第二个3D数据的对此，从而实现了两个3D数据的配准。
第二类基于深度学习的3D数据配准算法与上述步骤类似。首先构建训练样本库，之后设计一种回归网络模型，该模型输入是两个3D数据(浮动数据和固定数据)，输出是与固定数据对齐配准后的浮动数据，通过损失函数优化模型从而训练网络。在推理阶段，输入浮动3D数据和固定3D数据可实时得到配准后的浮动3D数据，从而实现两个3D数据的配准。
3D数据混合配准算法主要区别于传统算法，将深度学习算法与传统算法相结合。先通过深度学习算法对3D数据降维并提取特征，利用提取到的特征并结合传统算法的相似度评价准则，如MI、SSIM、NCC等，实现特征层面的相似性比较，进而实现配准。利用深度学习提取特征的主要算法有AlexNet、ZFNet、VGGNet、GoogLeNet等，基于深度学习提取特征的算法包含但不仅限于上述特征提取算法。
基于二维平面的3D数据配准方法主要基于二维层面进行配准，通过对不同角度的二维平面配准从而实现三维配准。以图7所示为例，通过对3D胎心数据中心A面和中心B面配准，实现整个3D数据的配准。图中灰色的虚线为A、B面的交线， 以交线作为中线，对A、B两个面分别提取特征，先将A面对齐，求得x和z方向的二维刚性变换参数，再在A面对齐的基础上对齐B面，求得y和z方向的二维刚性变换参数，通过多次迭代上述过程，实现两个3D数据的配准，这种配准算法多适用于仅存在平移变换的配准场景。
二维平面的配准算法也可分为传统算法、基于深度学习的配准算法和混合算法，与三维数据配准算法类似，区别在于二维平面求得的是二维刚性变换参数(两个方向的平移和一个旋转，共3个参数)。除已叙述过的配准方法外，也可以通过其他方法实现配准。分块配准算法通过对各个图像块进行相似度匹配，从而求得多个图像块中心位置的映射关系，再根据最小二乘法拟合二维刚性变换参数，从而实现图像配准；金字塔算法通过构建不同尺度的二维图像，从低分辨率图像着手粗配准，逐渐提高图像分辨率并一步步提高配准精度，达到配准两幅二维图像的目的。
数据配准后的多个3D数据间有可能存在明显的平移或旋转关系，从而导致多个3D数据的边界不一致，使得重建后的各个体数据间边界不同。为了使得平移后体数据间一致，需要去除部分边界。如图8所示，以两个3D数据配准后为例，大立方体表示原始3D数据大小，小立方体表示配准后两个3D数据的重叠区域，根据配准过程求得的三维变换参数矩阵，可以求得x、y和z方向上各两个偏移量xbias、ybias、zbias、xbias’、ybias’和zbias’，将3D数据中非重叠区域进行裁剪，仅保留小立方体区域内的图像数据。将上述原理应用于多个3D数据，保留多个3D数据的重叠区域，将其他区域进行裁剪，实现多3D数据的边界修正。
一些实施例中，处理器40控制显示器50进行胎心的显示，包括：按照心相顺序依次显示一个心动周期的三维超声数据图像。
从心动周期的角度来看，一些实施例中，处理器40控制向胎心进行m×n-1次三维超声扫描，获取m×n-1卷三维超声数据；m和n都为大于或等于2的整数；其中，在前n-1个所述胎心的每个心动周期内：进行m次所述三维超声扫描得到m卷所述三维超声数据；在第n个所述胎心的心动周期内：进行m-1次所述三维超声扫描得到m-1卷所述三维超声数据。一些实施例中，每次三维超声扫描包括：向胎心发射超声波以及接收相应的超声波回波，得到一卷包含胎心的三维超声数据；一些实施例中，每一卷三维超声数据都包含完整的胎心。处理器40对m×n-1卷三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷三维超声数据；处理器40根据数据重排后的m×n-1卷三维超声数据，控制显示器50进行胎心的显示。
可以理解地，一些实施例中所涉及的总扫描次数虽然为m×n-1，但这是指与本申请方案相关的扫描次数；在实际过程中，可以由于各种原因而扫描更多次数，例如在上述的扫描的m×n-1次过程中，可以在它之前插入若干次扫描，也可以在 它之后再插入若干次扫描，但由于这些扫描及扫描得到的数据与本申请无关，因此，本申请并没有提及这些扫描。
一些实施例中，数据重排至少包括将部分扫描时序在后的三维超声数据向前调整，以及部分扫描时序在前的三维超声数据调整向后调整。
一些实施例中，数据重排后的m×n-1卷所述三维超声数据等距地对应一个完整心动周期的不同心相位置。
一些实施例中，处理器40对m×n-1卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷所述三维超声数据，包括：数据重排后的m×n-1卷所述三维超声数据，按照心相顺序被划分为m个部分的情况下，第1至第m-1个部分分别包含n卷所述三维超声数据，第m个部分包含n-1卷所述三维超声数据；在前n-1个所述胎心的心动周期中的任意第i个心动周期内，其进行的m次所述三维超声扫描所得到的m卷所述三维超声数据，被设置为数据重排后的m×n-1卷所述三维超声数据中第1至第m个部分中各个部分的第i卷所述三维超声数据；所述i为正整数，且i小于或等于n-1；在第n个所述胎心的心动周期内，其进行的m-1次所述三维超声扫描所得到m-1卷所述三维超声数据，被设置为数据重排后的m×n-1卷所述三维超声数据中第1至第m-1个部分中各个部分的第n卷所述三维超声数据。
一些实施例中，处理器40对m×n-1卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷所述三维超声数据，包括：数据重排前的m×n-1卷所述三维超声数据中第(i-1)×m+j卷所述三维超声数据，被设置为数据重排后的第(j-1)×n+i卷所述三维超声数据；所述i和j为正整数，且i小于或等于n，j小于或等于m。
数据重排前的三维超声数据是按照扫描时序所排列的，即按照各自对应的三维超声扫描的时间先后顺序所排列的，数据重排后的三维超声数据则是按照心相顺序所排列的。
一些实施例中，处理器40对m×n-1卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷所述三维超声数据，包括：将数据重排前的m×n-1卷所述三维超声数据按照各自对应的三维超声扫描的时间先后顺序划分为n个部分，第1至第n-1个部分分别都包含m卷所述三维超声数据，第n个部分包含m-1卷所述三维超声数据；将数据重排前的m×n-1卷所述三维超声数据中任意第i个部分中的第j卷三维超声数据，设置为数据重排后的第(j-1)×n+i卷所述三维超声数据；所述i和j为正整数，且i小于或等于n，j小于或等于m。
一些实施例中，处理器40对m×n-1卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷所述三维超声数据，包括：将数据重排前的m ×n-1卷所述三维超声数据按照各自对应的三维超声扫描的时间先后顺序划分为n个部分，第1至第n-1个部分分别都包含m卷所述三维超声数据，第n个部分包含m-1卷所述三维超声数据；将数据重排前的第1至第n个部分中各自的第j卷所述三维超声数据分别设置为数据重排后的第j个部分中的第1至n卷所述三维超声数据，其中j为小于m的正整数；将数据重排前的第1至第n-1个部分中各自的第m卷所述三维超声数据分别设置为数据重排后的第m个部分中的第1至n-1卷所述三维超声数据。
一些实施例中：

m＝floor[(TS+1)×BVR×60÷FR÷TS]；
BVR表示基础零延时卷率；FR表示胎心的心率；TS表示时间分辨率；floor表示向下取整操作。
一些实施例中：

m＝floor[(60÷FR÷TI+1)×BVR×TI]；
BVR表示基础零延时卷率；TI表示心相时间间隔；FR表示胎心的心率，floor表示向下取整操作。
一些实施例中，处理器40还获取延时时长；在进行m×n-1次所述三维超声扫描的过程中：每完成一次所述三维超声扫描，经过所述延迟时长，再进行下一次所述三维超声扫描。
一些实施例中，延迟时长使得所获取的所述m×n-1卷三维超声数据对应所述胎心的不同心目位置。
一些实施例中，延迟时长使得所获取的所述m×n-1卷三维超声数据对应所述胎心的不同心相位置，且各相邻的心相位置之间的距离相等。
一些实施例中，延迟时长使得心相时间间隔和时间分辨率相乘为1。
一些实施例中，延迟时长基于胎心的心率、基础零延时卷率和第一参数计算得到。本文中基础零延时卷率为所进行的这多次三维超声扫描中第一次三维超声扫描的扫描卷率。一些实施例中，第一参数包括时间分辨率或心相时间间隔。本文中心相时间间隔为数据重排后的这多卷三维超声数据中相邻两卷三维超声数据所对应的心相位置之间的时间间隔。
一些实施例中，处理器40获取胎心的心率，并根据心率、基础零延时卷率和第一参数计算得到延迟时长。本文中基础零延时卷率为所进行的这多次三维超声扫 描中第一次三维超声扫描的扫描卷率。一些实施例中，第一参数包括时间分辨率或心相时间间隔。本文中心相时间间隔为数据重排后的这多卷三维超声数据中相邻两卷三维超声数据所对应的心相位置之间的时间间隔。
一些实施例中，处理器40根据心率、基础零延时卷率和第一参数计算得到延迟时长，包括：

其中，DT表示延迟时长，BVR表示基础零延时卷率，TS表示时间分辨率，FR表示胎心的心率，floor表示向下取整操作。
一些实施例中，处理器40控制显示器50生成设置界面，设置界面用于设置时间分辨率；响应于设置指令，处理器40设置时间分辨率。例如，用户通过输入工具例如鼠标或键盘等在设置界面上设置时间分辨率。
一些实施例中，处理器40根据心率、基础零延时卷率和第一参数计算得到延迟时长，包括：

其中，DT表示延迟时长，BVR表示基础零延时卷率，TI表示心相时间间隔，FR表示胎心的心率，floor表示向下取整操作。
一些实施例中，TI的值可以为0.01到0.03秒。
一些实施例中，处理器40控制显示器50生成设置界面，设置界面用于设置心相时间间隔；响应于设置指令，处理器40设置心相时间间隔。例如，用户通过输入工具例如鼠标或键盘等在设置界面上设置心相时间间隔。
一些实施例中，处理器40还获取总扫描时间；总扫描时间到达后，处理器40控制停止进行所述三维超声扫描；或者，处理器40根据总扫描时间生成停止进行三维超声扫描的提示。
一些实施例中，总扫描时间为：
T＝60÷FR÷TI÷AVR；
其中，T表示总扫描时间；FR表示胎心的心率；TI表示心相时间间隔；BVR表示基础零延时卷率；floor表示向下取整操作。
一些实施例中，总扫描时间为：
T＝TS÷AVR；
其中，T表示总扫描时间；TS表示时间分辨率；FR表示胎心的心率；BVR表示基础零延时卷率；floor表示向下取整操作。
一些实施例中，在对数据重排后，处理器40还可以对数据重排后的上述多卷三维超声数据进行数据配准和/或边界修正。例如处理器40对数据重排后的m×n一1卷所述三维超声数据进行数据配准和/或边界修正，得到处理后的m×n一1卷所述三维超声数据；处理器40根据处理后的m×n一1卷所述三维超声数据，控制显示器50进行胎心的显示。
具体如何进行数据配准以及边界修正见上文的记载，在此不再赘述。
一些实施例中，处理器40控制显示器50进行胎心的显示，包括：按照心相顺序依次显示一个心动周期的三维超声数据图像。
一些实施例的方案，能够大幅减少采集时间(例如仅需3到4s)，降低了胎儿运动对最终成像效果的影响。
一些实施例还公开了一种超声成像方法。
请参照图9，一些实施例的超声成像方法，包括以下步骤：
步骤100：获取扫描规则。
步骤110：根据所述扫描规则控制对所述胎心进行多次三维超声扫描，获取包含多个心动周期的多卷三维超声数据，使得不同卷的所述三维超声数据对应所述胎心的不同心相位置；其中，每次所述三维超声扫描包括：向所述胎心发射超声波以及接收相应的超声波回波，得到一卷包含胎心的三维超声数据。一些实施例中，每一卷三维超声数据都包含完整的胎心。
步骤120：对多卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的多卷所述三维超声数据。
步骤130：根据所述数据重排后的多卷三维超声数据，进行所述胎心的显示。
一些实施例中，扫描规则包括：在进行多次上述三维超声扫描的过程中，每完成一次所述三维超声扫描，经过相同的延迟时长，再进行下一次三维超声扫描。例如总共进行了N1次三维超声扫描，在进行这N1次三维超声扫描的过程中，每完成一次所述三维超声扫描，经过相同的延迟时长，再进行下一次三维超声扫描；换句话说，在完成第1次三维超声扫描后，经过一延迟时长后再进行第2次三维超声扫描；在完成第2次三维超声扫描后，经过一相同的延迟时长后再进行第3次三维超声扫描，如此进行，直到在完成第N1-1次三维超声扫描后，经过一相同的延迟时 长后再进行第N1次三维超声扫描。
一些实施例中，延迟时长基于胎心的心率、基础零延时卷率和第一参数计算得到。本文中基础零延时卷率为所进行的这多次三维超声扫描中第一次三维超声扫描的扫描卷率。一些实施例中，第一参数包括时间分辨率或心相时间间隔。本文中心相时间间隔为数据重排后的这多卷三维超声数据中相邻两卷三维超声数据所对应的心相位置之间的时间间隔。
请参照图10，一些实施例的超声成像方法，包括以下步骤：
步骤101：获取延迟时长。
步骤111：控制向所述胎心进行多次三维超声扫描，获取包含多个心动周期的多卷三维超声数据；其中，在进行多次所述三维超声扫描的过程中：每完成一次所述三维超声扫描，经过所述延迟时长，再进行下一次所述三维超声扫描；并且，每次所述三维超声扫描包括：向所述胎心发射超声波以及接收相应的超声波回波，得到一卷包含胎心的三维超声数据。一些实施例中，每一卷三维超声数据都包含完整的胎心。
步骤120：对多卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的多卷所述三维超声数据。
步骤130：根据所述数据重排后的多卷三维超声数据，进行所述胎心的显示。
一些实施例中，延迟时长使得所获取的多卷三维超声数据对应胎心的不同心相位置。
一些实施例中，延迟时长使得所获取的多卷三维超声数据对应胎心的不同心相位置，且各相邻的心相位置之间的距离相等。
一些实施例中，延迟时长使得心相时间间隔和时间分辨率相乘为1，心相时间间隔为数据重排后的这多卷三维超声数据中相邻两卷三维超声数据所对应的心相位置之间的时间间隔。
一些实施例中，步骤101获取胎心的心率，并根据心率、基础零延时卷率和第一参数计算得到延迟时长。本文中基础零延时卷率为所进行的这多次三维超声扫描中第一次三维超声扫描的扫描卷率。一些实施例中，第一参数包括时间分辨率或心相时间间隔。本文中心相时间间隔为数据重排后的这多卷三维超声数据中相邻两卷三维超声数据所对应的心相位置之间的时间间隔。
一些实施例中，步骤101根据心率、基础零延时卷率和第一参数计算得到延迟时长，包括：

其中，DT表示延迟时长，BVR表示基础零延时卷率，TS表示时间分辨率，FR表示胎心的心率，floor表示向下取整操作。
一些实施例中，请参照图11，超声成像方法还包括：
步骤102：生成设置界面，设置界面用于设置时间分辨率。
步骤103：响应于设置指令，设置时间分辨率。例如，用户通过输入工具例如鼠标或键盘等在设置界面上设置时间分辨率。
一些实施例中，步骤101根据心率、基础零延时卷率和第一参数计算得到延迟时长，包括：

其中，DT表示延迟时长，BVR表示基础零延时卷率，TI表示心相时间间隔，FR表示胎心的心率，floor表示向下取整操作。
一些实施例中，请参照图12，超声成像方法还包括：
步骤104：生成设置界面，设置界面用于设置心相时间间隔。
步骤105：响应于设置指令，设置心相时间间隔。例如，用户通过输入工具例如鼠标或键盘等在设置界面上设置心相时间间隔。
需要说明的是，胎心的心率是指胎儿心脏在单位时间内跳动的次数(例如每分钟跑去的次数)，这可以通过多次方法获取，例如胎心信号监测的方式、超声多普勒或M超测量胎心心率图。胎心信号监测的方式通常是从母亲腹部混合的心电信号中提取分离出胎儿的心电信号，根据分离的胎儿心电信号计算胎儿心率；分离算法有自适应滤波算法、模板减法和盲信号分析等。基于超声多普勒的方法主要是利用多普勒效应；由于胎心运动具有周期性，因此回波频移信号也具有周期性，通过对此频移信号进行处理，可以计算出胎儿的心率。M超则是利用心脏跳动的周期性，使回波信号从左向右自行移动扫描，显示出周期性心跳的时间位置曲线像，从而可以进一步利用时间位置关系计算出胎儿心率。
一些实施例中，上述多次三维超声扫描的次数为m×n-1。
一些实施例中，上述多卷三维超声数据的卷数为m×n-1。
一些实施例中：

m＝floor[(TS+1)×BVR×60÷FR÷TS]；
BVR表示基础零延时卷率；FR表示胎心的心率；TS表示时间分辨率；floor表示向下取整操作。
一些实施例中：

m＝floor[(60÷FR÷TI+1)×BVR×TI]；
BVR表示基础零延时卷率；TI表示心相时间间隔；FR表示胎心的心率，floor表示向下取整操作。
在获取上述多卷三维超声数据后，需要进行数据重排。一些实施例中，步骤120进行数据重排至少包括将部分扫描时序在后的三维超声数据向前调整，以及部分扫描时序在前的三维超声数据调整向后调整。
一些实施例中，步骤120进行数据重排包括：数据重排前的m×n-1卷所述三维超声数据中第(i-1)×m+j卷三维超声数据，被设置为数据重排后的第(j-1)×n+i卷三维超声数据；i和j为正整数，且i小于或等于n，j小于或等于m。数据重排前的三维超声数据是按照扫描时序所排列的，即按照各自对应的三维超声扫描的时间先后顺序所排列的，数据重排后的三维超声数据则是按照心相顺序所排列的。
一些实施例中，步骤120进行数据重排包括：将数据重排前的m×n-1卷三维超声数据按照各自对应的三维超声扫描的时间先后顺序划分为n个部分，第1至第n-1个部分分别都包含m卷所述三维超声数据，第n个部分包含m-1卷所述三维超声数据；将数据重排前的m×n-1卷三维超声数据中任意第i个部分中的第j卷三维超声数据，设置为数据重排后的第(j-1)×n+i卷所述三维超声数据；i和j为正整数，且i小于或等于n，j小于或等于m。
一些实施例中，步骤120进行数据重排包括：将数据重排前的m×n-1卷三维超声数据按照各自对应的三维超声扫描的时间先后顺序划分为n个部分，第1至第n-1个部分分别都包含m卷所述三维超声数据，第n个部分包含m-1卷所述三维超声数据；将数据重排前的第1至第n个部分中各自的第j卷所述三维超声数据分别设置为数据重排后的第j个部分中的第1至n卷所述三维超声数据，其中j为小于m的正整数；将数据重排前的第1至第n-1个部分中各自的第m卷所述三维超声数据分别设置为数据重排后的第m个部分中的第1至n-1卷所述三维超声数据。
一些实施例中，数据重排后的这多卷三维超声数据等距地对应一个完整心动周期的不同心相位置。
在对数据重排后，还可以对数据重排后的上述多卷三维超声数据进行数据配准和/或边界修正。因此，一些实施例中，步骤130对数据重排后的多卷所述三维超声数据进行数据配准和/或边界修正，得到处理后的多卷所述三维超声数据；根据处理后的多卷所述三维超声数据，进行所述胎心的显示。一些实施例中，步骤130进行胎心的显示，包括：按照心相顺序依次显示一个心动周期的三维超声数据图像。
一些实施例中，步骤110或步骤111还获取总扫描时间；总扫描时间到达后， 步骤110或步骤111控制停止进行所述三维超声扫描；或者，步骤110或步骤111根据总扫描时间生成停止进行三维超声扫描的提示。
一些实施例中，总扫描时间为：
T＝60÷FR÷TI÷AVR；
其中，T表示总扫描时间；FR表示胎心的心率；TI表示心相时间间隔；BVR表示基础零延时卷率；floor表示向下取整操作。
一些实施例中，总扫描时间为：
T＝TS÷AVR；
其中，T表示总扫描时间；TS表示时间分辨率；FR表示胎心的心率；BVR表示基础零延时卷率；floor表示向下取整操作。
一些实施例中，步骤110或步骤111还获取三维超声扫描的次数，在进行完三维超声扫描的次数后，步骤110或步骤111控制停止进行所述三维超声扫描；或者，步骤110或步骤111根据所述三维超声扫描的次数，生成停止进行所述三维超声扫描的提示。一些实施例中，三维超声扫描的次数m×n-1。
请参照图13，一些实施例的超声成像方法，包括以下步骤：
步骤200：控制向胎心进行m×n一1次三维超声扫描，获取m×n-1卷三维超声数据；m和n都为大于或等于2的整数；其中，在前n-1个所述胎心的每个心动周期内：进行m次所述三维超声扫描得到m卷所述三维超声数据；在第n个所述胎心的心动周期内：进行m-1次所述三维超声扫描得到m-1卷所述三维超声数据。一些实施例中，每次三维超声扫描包括：向胎心发射超声波以及接收相应的超声波回波，得到一卷包含胎心的三维超声数据；一些实施例中，每一卷三维超声数据都包含完整的胎心。
一些实施例中：

m＝floor[(TS+1)×BVR×60÷FR÷TS]；
BVR表示基础零延时卷率；FR表示胎心的心率；TS表示时间分辨率；floor表示向下取整操作。
一些实施例中：

m＝floor[(60÷FR÷TI+1)×BVR×TI]；
BVR表示基础零延时卷率；TI表示心相时间间隔；FR表示胎心的心率，floor表示向下取整操作。
一些实施例中，步骤200还获取延时时长；在进行m×n-1次所述三维超声扫描的过程中：每完成一次所述三维超声扫描，经过所述延迟时长，再进行下一次所述三维超声扫描。
一些实施例中，延迟时长使得所获取的所述m×n-1卷三维超声数据对应所述胎心的不同心目位置。
一些实施例中，延迟时长使得所获取的所述m×n-1卷三维超声数据对应所述胎心的不同心相位置，且各相邻的心相位置之间的距离相等。
一些实施例中，延迟时长使得心相时间间隔和时间分辨率相乘为1。
一些实施例中，延迟时长基于胎心的心率、基础零延时卷率和第一参数计算得到。本文中基础零延时卷率为所进行的这多次三维超声扫描中第一次三维超声扫描的扫描卷率。一些实施例中，第一参数包括时间分辨率或心相时间间隔。本文中心相时间间隔为数据重排后的这多卷三维超声数据中相邻两卷三维超声数据所对应的心相位置之间的时间间隔。
一些实施例中，步骤200获取胎心的心率，并根据心率、基础零延时卷率和第一参数计算得到延迟时长。本文中基础零延时卷率为所进行的这多次三维超声扫描中第一次三维超声扫描的扫描卷率。一些实施例中，第一参数包括时间分辨率或心相时间间隔。本文中心相时间间隔为数据重排后的这多卷三维超声数据中相邻两卷三维超声数据所对应的心相位置之间的时间间隔。
一些实施例中，步骤200根据心率、基础零延时卷率和第一参数计算得到延迟时长，包括：

其中，DT表示延迟时长，BVR表示基础零延时卷率，TS表示时间分辨率，FR表示胎心的心率，floor表示向下取整操作。
一些实施例中，步骤200根据心率、基础零延时卷率和第一参数计算得到延迟时长，包括：

其中，DT表示延迟时长，BVR表示基础零延时卷率，TI表示心相时间间隔，FR表示胎心的心率，floor表示向下取整操作。
一些实施例中，TI的值可以为0.01到0.03秒。
一些实施例中，步骤200还获取总扫描时间；总扫描时间到达后，步骤200控制停止进行所述三维超声扫描；或者，步骤200根据总扫描时间生成停止进行三维超声扫描的提示。
一些实施例中，总扫描时间为：
T＝60÷FR÷TI÷AVR；
其中，T表示总扫描时间；FR表示胎心的心率；TI表示心相时间间隔；BVR表示基础零延时卷率；floor表示向下取整操作。
一些实施例中，总扫描时间为：
T＝TS÷AVR；
其中，T表示总扫描时间；TS表示时间分辨率；FR表示胎心的心率；BVR表示基础零延时卷率；floor表示向下取整操作。
步骤210：对m×n-1卷三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷三维超声数据。一些实施例中，数据重排至少包括将部分扫描时序在后的三维超声数据向前调整，以及部分扫描时序在前的三维超声数据调整向后调整。
一些实施例中，数据重排后的m×n-1卷所述三维超声数据等距地对应一个完整心动周期的不同心相位置。
一些实施例中，步骤210对m×n-1卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷所述三维超声数据，包括：数据重排后的m×n-1卷所述三维超声数据，按照心相顺序被划分为m个部分的情况下，第1至第m-1个部分分别包含n卷所述三维超声数据，第m个部分包含n-1卷所述三维超声数据；在前n-1个所述胎心的心动周期中的任意第i个心动周期内，其进行的m次所述三维超声扫描所得到的m卷所述三维超声数据，被设置为数据重排后的m×n-1卷所述三维超声数据中第1至第m个部分中各个部分的第i卷所述三维超声数据；所述i为正整数，且i小于或等于n-1；在第n个所述胎心的心动周期内，其进行的m-1次所述三维超声扫描所得到m-1卷所述三维超声数据，被设置为数据重排后的m×n-1卷所述三维超声数据中第1至第m-1个部分中各个部分的第n卷所述三 维超声数据。
一些实施例中，步骤210对m×n-1卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷所述三维超声数据，包括：数据重排前的m×n-1卷所述三维超声数据中第(i-1)×m+j卷所述三维超声数据，被设置为数据重排后的第(j-1)×n+i卷所述三维超声数据；所述i和j为正整数，且i小于或等于n，j小于或等于m。
数据重排前的三维超声数据是按照扫描时序所排列的，即按照各自对应的三维超声扫描的时间先后顺序所排列的，数据重排后的三维超声数据则是按照心相顺序所排列的。
一些实施例中，步骤210对m×n-1卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷所述三维超声数据，包括：将数据重排前的m×n-1卷所述三维超声数据按照各自对应的三维超声扫描的时间先后顺序划分为n个部分，第1至第n-1个部分分别都包含m卷所述三维超声数据，第n个部分包含m-1卷所述三维超声数据；将数据重排前的m×n-1卷所述三维超声数据中任意第i个部分中的第j卷三维超声数据，设置为数据重排后的第(j-1)×n+i卷所述三维超声数据；所述i和j为正整数，且i小于或等于n，j小于或等于m。
一些实施例中，步骤210对m×n-1卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷所述三维超声数据，包括：将数据重排前的m×n-1卷所述三维超声数据按照各自对应的三维超声扫描的时间先后顺序划分为n个部分，第1至第n-1个部分分别都包含m卷所述三维超声数据，第n个部分包含m-1卷所述三维超声数据；将数据重排前的第1至第n个部分中各自的第j卷所述三维超声数据分别设置为数据重排后的第j个部分中的第1至n卷所述三维超声数据，其中j为小于m的正整数；将数据重排前的第1至第n-1个部分中各自的第m卷所述三维超声数据分别设置为数据重排后的第m个部分中的第1至n-1卷所述三维超声数据。
步骤220：根据数据重排后的m×n-1卷三维超声数据，进行胎心的显示。一些实施例中，在对数据重排后，步骤220还可以对数据重排后的上述多卷三维超声数据进行数据配准和/或边界修正。例如步骤220对数据重排后的m×n-1卷所述三维超声数据进行数据配准和/或边界修正，得到处理后的m×n-1卷所述三维超声数据；步骤220根据处理后的m×n-1卷所述三维超声数据，进行胎心的显示。
具体如何进行数据配准以及边界修正见上文的记载，在此不再赘述。
一些实施例中，步骤220进行胎心的显示，包括：按照心相顺序依次显示一个心动周期的三维超声数据图像。
本文参照了各种示范实施例进行说明。然而，本领域的技术人员将认识到，在 不脱离本文范围的情况下，可以对示范性实施例做出改变和修正。例如，各种操作步骤以及用于执行操作步骤的组件，可以根据特定的应用或考虑与系统的操作相关联的任何数量的成本函数以不同的方式实现(例如一个或多个步骤可以被删除、修改或结合到其他步骤中)。
在上述实施例中，可以全部或部分地通过软件、硬件、固件或者其任意组合来实现。另外，如本领域技术人员所理解的，本文的原理可以反映在计算机可读存储介质上的计算机程序产品中，该可读存储介质预装有计算机可读程序代码。任何有形的、非暂时性的计算机可读存储介质皆可被使用，包括磁存储设备(硬盘、软盘等)、光学存储设备(CD至ROM、DVD、Blu Ray盘等)、闪存和/或诸如此类。这些计算机程序指令可被加载到通用计算机、专用计算机或其他可编程数据处理设备上以形成机器，使得这些在计算机上或其他可编程数据处理装置上执行的指令可以生成实现指定的功能的装置。这些计算机程序指令也可以存储在计算机可读存储器中，该计算机可读存储器可以指示计算机或其他可编程数据处理设备以特定的方式运行，这样存储在计算机可读存储器中的指令就可以形成一件制造品，包括实现指定功能的实现装置。计算机程序指令也可以加载到计算机或其他可编程数据处理设备上，从而在计算机或其他可编程设备上执行一系列操作步骤以产生一个计算机实现的进程，使得在计算机或其他可编程设备上执行的指令可以提供用于实现指定功能的步骤。
虽然在各种实施例中已经示出了本文的原理，但是许多特别适用于特定环境和操作要求的结构、布置、比例、元件、材料和部件的修改可以在不脱离本披露的原则和范围内使用。以上修改和其他改变或修正将被包含在本文的范围之内。
前述具体说明已参照各种实施例进行了描述。然而，本领域技术人员将认识到，可以在不脱离本披露的范围的情况下进行各种修正和改变。因此，对于本披露的考虑将是说明性的而非限制性的意义上的，并且所有这些修改都将被包含在其范围内。同样，有关于各种实施例的优点、其他优点和问题的解决方案已如上所述。然而，益处、优点、问题的解决方案以及任何能产生这些的要素，或使其变得更明确的解决方案都不应被解释为关键的、必需的或必要的。本文中所用的术语“包括”和其任何其他变体，皆属于非排他性包含，这样包括要素列表的过程、方法、文章或设备不仅包括这些要素，还包括未明确列出的或不属于该过程、方法、系统、文章或设备的其他要素。此外，本文中所使用的术语“耦合”和其任何其他变体都是指物理连接、电连接、磁连接、光连接、通信连接、功能连接和/或任何其他连接。
具有本领域技术的人将认识到，在不脱离本发明的基本原理的情况下，可以对上述实施例的细节进行许多改变。因此，本发明的范围应仅由权利要求确定。

Claims (52)

  1. 一种胎心的超声成像方法，其特征在于，包括：
    获取延迟时长；
    控制向所述胎心进行多次三维超声扫描，获取包含多个心动周期的多卷三维超声数据；其中，在进行多次所述三维超声扫描的过程中：每完成一次所述三维超声扫描，经过所述延迟时长，再进行下一次所述三维超声扫描；并且，每次所述三维超声扫描包括：向所述胎心发射超声波以及接收相应的超声波回波，得到一卷包含胎心的三维超声数据；
    对多卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的多卷所述三维超声数据；所述数据重排至少包括将部分扫描时序在后的所述三维超声数据向前调整，以及部分扫描时序在前的所述三维超声数据调整向后调整；
    根据所述数据重排后的多卷三维超声数据，进行所述胎心的显示。
  2. 如权利要求1所述的超声成像方法，其特征在于，所述获取延迟时长包括：
    获取所述胎心的心率；
    根据所述心率、基础零延时卷率和第一参数计算得到所述延迟时长；其中，所述基础零延时卷率为所述多次三维超声扫描中第一次三维超声扫描的扫描卷率，所述第一参数包括时间分辨率或心相时间间隔，所述心相时间间隔为数据重排后的多卷所述三维超声数据中相邻两卷所述三维超声数据所对应的心相位置之间的时间间隔。
  3. 如权利要求2所述的超声成像方法，其特征在于，所述根据所述心率、基础零延时卷率和第一参数计算得到所述延迟时长，包括：

    其中，DT表示所述延迟时长，BVR表示所述基础零延时卷率，TS表示所述时间分辨率，FR表示所述胎心的心率，floor表示向下取整操作。
  4. 如权利要求2所述的超声成像方法，其特征在于，所述根据所述心率、基础零延时卷率和第一参数计算得到所述延迟时长，包括：

    其中，DT表示所述延迟时长，BVR表示所述基础零延时卷率，TI表示所述心相 时间间隔，FR表示所述胎心的心率，floor表示向下取整操作。
  5. 如权利要求2或3所述的超声成像方法，其特征在于，还包括：
    生成设置界面，所述设置界面用于设置所述时间分辨率；
    响应于设置指令，设置所述时间分辨率。
  6. 如权利要求2或4所述的超声成像方法，其特征在于，还包括：
    生成设置界面，所述设置界面用于设置所述心相时间间隔；
    响应于设置指令，设置所述心相时间间隔。
  7. 如权利要求1所述的超声成像方法，其特征在于，所述多次三维超声扫描的次数为m×n-1，和/或，所述多卷三维超声数据的卷数为m×n-1；其中：

    m＝floor[(TS+1)×BVR×60÷FR÷TS]；
    BVR表示基础零延时卷率，所述基础零延时卷率为所述多次三维超声扫描中第一次三维超声扫描的扫描卷率；FR表示所述胎心的心率；TS表示时间分辨率；floor表示向下取整操作。
  8. 如权利要求1所述的超声成像方法，其特征在于，所述多次三维超声扫描的次数为m×n-1，和/或，所述多卷三维超声数据的卷数为m×n-1；其中：

    m＝floor[(60÷FR÷TI+1)×BVR×TI]；
    BVR表示基础零延时卷率，所述基础零延时卷率为所述多次三维超声扫描中第一次三维超声扫描的扫描卷率；TI表示心相时间间隔，所述心相时间间隔为数据重排后的多卷所述三维超声数据中相邻两卷所述三维超声数据所对应的心相位置之间的时间间隔；FR表示所述胎心的心率，floor表示向下取整操作。
  9. 如权利要求7或8所述的超声成像方法，其特征在于，所述对多卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的多卷所述三维超声数据，包括：数据重排前的m×n-1卷所述三维超声数据中第(i-1)×m+j卷所述三维超声数据，被设置为数据重排后的第(j-1)×n+i卷所述三维超声数据；所述i和j为正整数，且i小于或等于n，j小于或等于m。
  10. 如权利要求7或8所述的超声成像方法，其特征在于，所述对多卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的多卷所述三维超声数据，包括：
    将数据重排前的m×n-1卷所述三维超声数据按照各自对应的三维超声扫描的时间先后顺序划分为n个部分，第1至第n-1个部分分别都包含m卷所述三维超声 数据，第n个部分包含m-1卷所述三维超声数据；
    将数据重排前的m×n-1卷所述三维超声数据中任意第i个部分中的第j卷三维超声数据，设置为数据重排后的第(j-1)×n+i卷所述三维超声数据；所述i和j为正整数，且i小于或等于n，j小于或等于m。
  11. 如权利要求7或8所述的超声成像方法，其特征在于，所述对多卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的多卷所述三维超声数据，包括：
    将数据重排前的m×n-1卷所述三维超声数据按照各自对应的三维超声扫描的时间先后顺序划分为n个部分，第1至第n-1个部分分别都包含m卷所述三维超声数据，第n个部分包含m-1卷所述三维超声数据；
    将数据重排前的第1至第n个部分中各自的第j卷所述三维超声数据分别设置为数据重排后的第j个部分中的第1至n卷所述三维超声数据，其中j为小于m的正整数；
    将数据重排前的第1至第n-1个部分中各自的第m卷所述三维超声数据分别设置为数据重排后的第m个部分中的第1至n-1卷所述三维超声数据。
  12. 如权利要求1所述的超声成像方法，其特征在于，数据重排后的多卷所述三维超声数据等距地对应一个完整心动周期的不同心相位置。
  13. 如权利要求1或12所述的超声成像方法，其特征在于，所述延迟时长使得所获取的所述多卷三维超声数据对应所述胎心的不同心相位置。
  14. 如权利要求1或12所述的超声成像方法，其特征在于，所述延迟时长使得所获取的所述多卷三维超声数据对应所述胎心的不同心相位置，且各相邻的心相位置之间的距离相等。
  15. 如权利要求1、12、13或14所述的超声成像方法，其特征在于，所述延迟时长使得心相时间间隔和时间分辨率相乘为1，所述心相时间间隔为数据重排后的多卷所述三维超声数据中相邻两卷所述三维超声数据所对应的心相位置之间的时间间隔。
  16. 如权利要求1所述的超声成像方法，其特征在于，还包括：获取总扫描时间；
    在所述总扫描时间到达后，控制停止进行所述三维超声扫描；或者，根据所述总扫描时间生成停止进行所述三维超声扫描的提示。
  17. 如权利要求16所述的超声成像方法，其特征在于，所述总扫描时间为：
    T＝60÷FR÷TI÷AVR；
    其中，T表示所述总扫描时间；FR表示所述胎心的心率；TI表示心相时间间隔，所述心相时间间隔为数据重排后的多卷所述三维超声数据中相邻两卷所述三维超声数据所对应的心相位置之间的时间间隔；BVR表示基础零延时卷率，所述基础零延时卷率为所述多次三维超声扫描中第一次三维超声扫描的扫描卷率；floor表示向下取整操作。
  18. 如权利要求16所述的超声成像方法，其特征在于，所述总扫描时间为：
    T＝TS÷AVR；
    其中，T表示所述总扫描时间；TS表示时间分辨率；FR表示所述胎心的心率；BVR表示基础零延时卷率，所述基础零延时卷率为所述多次三维超声扫描中第一次三维超声扫描的扫描卷率；floor表示向下取整操作。
  19. 如权利要求7或8所述的超声成像方法，其特征在于，还包括：获取所述三维超声扫描的次数；
    在进行完所述三维超声扫描的次数后，控制停止进行所述三维超声扫描；或者，根据所述三维超声扫描的次数，生成停止进行所述三维超声扫描的提。
  20. 如权利要求1所述的超声成像方法，其特征在于，所述根据所述数据重排后的多卷三维超声数据，进行所述胎心的显示，包括：
    对数据重排后的多卷所述三维超声数据进行数据配准和/或边界修正，得到处理后的多卷所述三维超声数据；
    根据处理后的多卷所述三维超声数据，进行所述胎心的显示。
  21. 如权利要求1或20所述的超声成像方法，其特征在于，所述进行所述胎心的显示，包括：按照心相顺序依次显示一个心动周期的三维超声数据图像。
  22. 如权利要求1所述的超声成像方法，其特征在于，每一卷所述三维超声数据都包含完整的胎心。
  23. 一种胎心的超声成像方法，其特征在于，包括：
    控制向所述胎心进行m×n-1次三维超声扫描，获取m×n-1卷三维超声数据；m和n都为大于或等于2的整数；其中，在前n-1个所述胎心的每个心动周期内：进行m次所述三维超声扫描得到m卷所述三维超声数据；在第n个所述胎心的心动周期内：进行m-1次所述三维超声扫描得到m-1卷所述三维超声数据；
    对m×n-1卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷所述三维超声数据；所述数据重排至少包括将部分扫描时序在后的所述三维超声数据向前调整，以及部分扫描时序在前的所述三维超声数据调整向后调整；
    根据数据重排后的m×n-1卷所述三维超声数据，进行所述胎心的显示。
  24. 如权利要求23所述的超声成像方法，其特征在于，所述对m×n-1卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷所述三维超声数据，包括：
    数据重排后的m×n-1卷所述三维超声数据，按照心相顺序被划分为m个部分的情况下，第1至第m-1个部分分别包含n卷所述三维超声数据，第m个部分包含n-1卷所述三维超声数据；
    在前n-1个所述胎心的心动周期中的任意第i个心动周期内，其进行的m次所述三维超声扫描所得到的m卷所述三维超声数据，被设置为数据重排后的m×n-1卷所述三维超声数据中第1至第m个部分中各个部分的第i卷所述三维超声数据；所述i为正整数，且i小于或等于n-1；
    在第n个所述胎心的心动周期内，其进行的m-1次所述三维超声扫描所得到m-1卷所述三维超声数据，被设置为数据重排后的m×n-1卷所述三维超声数据中第1至第m-1个部分中各个部分的第n卷所述三维超声数据。
  25. 如权利要求23所述的超声成像方法，其特征在于，所述对m×n-1卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷所述三维超声数据，包括：
    数据重排前的m×n-1卷所述三维超声数据中第(i-1)×m+j卷所述三维超声数据，被设置为数据重排后的第(j-1)×n+i卷所述三维超声数据；所述i和j为正整数，且i小于或等于n，j小于或等于m。
  26. 如权利要求23所述的超声成像方法，其特征在于，所述对m×n-1卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷所述三维超声数据，包括：
    将数据重排前的m×n-1卷所述三维超声数据按照各自对应的三维超声扫描的时间先后顺序划分为n个部分，第1至第n-1个部分分别都包含m卷所述三维超声数据，第n个部分包含m-1卷所述三维超声数据；
    将数据重排前的m×n-1卷所述三维超声数据中任意第i个部分中的第j卷三维超声数据，设置为数据重排后的第(j-1)×n+i卷所述三维超声数据；所述i和j为正整数，且i小于或等于n，j小于或等于m。
  27. 如权利要求23所述的超声成像方法，其特征在于，所述对m×n-1卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的m×n-1卷所述三维超声数据，包括：
    将数据重排前的m×n-1卷所述三维超声数据按照各自对应的三维超声扫描的时间先后顺序划分为n个部分，第1至第n-1个部分分别都包含m卷所述三维超声 数据，第n个部分包含m-1卷所述三维超声数据；
    将数据重排前的第1至第n个部分中各自的第j卷所述三维超声数据分别设置为数据重排后的第j个部分中的第1至n卷所述三维超声数据，其中j为小于m的正整数；
    将数据重排前的第1至第n-1个部分中各自的第m卷所述三维超声数据分别设置为数据重排后的第m个部分中的第1至n-1卷所述三维超声数据。
  28. 如权利要求23所述的超声成像方法，其特征在于：

    m＝floor[(TS+1)×BVR×60÷FR÷TS]；
    BVR表示基础零延时卷率，所述基础零延时卷率为所述m×n-1次三维超声扫描中第一次三维超声扫描的扫描卷率；FR表示所述胎心的心率；TS表示时间分辨率；floor表示向下取整操作。
  29. 如权利要求22所述的超声成像方法，其特征在于：

    m＝floor[(60÷FR÷TI+1)×BVR×TI]；
    BVR表示基础零延时卷率，所述基础零延时卷率为所述m×n-1次三维超声扫描中第一次三维超声扫描的扫描卷率；TI表示心相时间间隔，所述心相时间间隔为数据重排后的m×n-1卷所述三维超声数据中相邻两卷所述三维超声数据所对应的心相位置之间的时间间隔；FR表示所述胎心的心率，floor表示向下取整操作。
  30. 如权利要求23所述的超声成像方法，其特征在于，还包括：获取延迟时长；
    在进行m×n-1次所述三维超声扫描的过程中：每完成一次所述三维超声扫描，经过所述延迟时长，再进行下一次所述三维超声扫描。
  31. 如权利要求30所述的超声成像方法，其特征在于，所述获取延迟时长包括：
    获取所述胎心的心率；
    根据所述心率、基础零延时卷率和第一参数计算得到所述延迟时长；其中，所述基础零延时卷率为所述m×n-1次三维超声扫描中第一次三维超声扫描的扫描卷率，所述第一参数包括时间分辨率或心相时间间隔，所述心相时间间隔为数据重排后的m×n-1卷所述三维超声数据中相邻两卷所述三维超声数据所对应的心相位置之间的时间间隔。
  32. 如权利要求31所述的超声成像方法，其特征在于，所述根据所述心率、 基础零延时卷率和第一参数计算得到所述延迟时长，包括：

    其中，DT表示所述延迟时长，BVR表示所述基础零延时卷率，TS表示所述时间分辨率，FR表示所述胎心的心率，floor表示向下取整操作。
  33. 如权利要求31所述的超声成像方法，其特征在于，所述根据所述心率、基础零延时卷率和第一参数计算得到所述延迟时长，包括：

    其中，DT表示所述延迟时长，BVR表示所述基础零延时卷率，TI表示所述心相时间间隔，FR表示所述胎心的心率，floor表示向下取整操作。
  34. 如权利要求30所述的超声成像方法，其特征在于，所述延迟时长使得所获取的所述m×n-1卷三维超声数据对应所述胎心的不同心相位置。
  35. 如权利要求30所述的超声成像方法，其特征在于，所述延迟时长使得所获取的所述m×n-1卷三维超声数据对应所述胎心的不同心相位置，且各相邻的心相位置之间的距离相等。
  36. 如权利要求30、34或35所述的超声成像方法，其特征在于，所述延迟时长使得心相时间间隔和时间分辨率相乘为1。
  37. 如权利要求23所述的超声成像方法，其特征在于，数据重排后的m×n-1卷所述三维超声数据等距地对应一个完整心动周期的不同心相位置。
  38. 如权利要求23所述的超声成像方法，其特征在于，还包括：获取总扫描时间；
    在所述总扫描时间到达后，控制停止进行所述三维超声扫描；或者，根据所述总扫描时间生成停止进行所述三维超声扫描的提示。
  39. 如权利要求38所述的超声成像方法，其特征在于，所述总扫描时间为：
    T＝60÷FR÷TI÷AVR；
    其中，T表示所述总扫描时间；FR表示所述胎心的心率；TI表示心相时间间隔，所述心相时间间隔为数据重排后的m×n-1卷所述三维超声数据中相邻两卷所述三维超声数据所对应的心相位置之间的时间间隔；BVR表示基础零延时卷率，所述基 础零延时卷率为所述m×n-1次三维超声扫描中第一次三维超声扫描的扫描卷率；floor表示向下取整操作。
  40. 如权利要求29所述的超声成像方法，其特征在于，所述总扫描时间为：
    T＝TS÷AVR；
    其中，T表示所述总扫描时间；TS表示时间分辨率；FR表示所述胎心的心率；BVR表示基础零延时卷率，所述基础零延时卷率为所述m×n-1次三维超声扫描中第一次三维超声扫描的扫描卷率；floor表示向下取整操作。
  41. 如权利要求23所述的超声成像方法，其特征在于，所述根据数据重排后的m×n-1卷所述三维超声数据，进行所述胎心的显示，包括：
    对所述数据重排后的m×n-1卷所述三维超声数据进行数据配准和/或边界修正，得到处理后的m×n-1卷所述三维超声数据；
    根据处理后的m×n-1卷所述三维超声数据，进行所述胎心的显示。
  42. 如权利要求23或41所述的超声成像方法，其特征在于，所述进行所述胎心的显示，包括：按照心相顺序依次显示一个心动周期的三维超声数据图像。
  43. 如权利要求23所述的超声成像方法，其特征在于，每次所述三维超声扫描包括：向所述胎心发射超声波以及接收相应的超声波回波，得到一卷包含完整胎心的三维超声数据。
  44. 一种胎心的超声成像方法，其特征在于，包括：
    获取扫描规则；
    根据所述扫描规则控制对所述胎心进行多次三维超声扫描，获取包含多个心动周期的多卷三维超声数据，使得不同卷的所述三维超声数据对应所述胎心的不同心相位置；其中，每次所述三维超声扫描包括：向所述胎心发射超声波以及接收相应的超声波回波，得到一卷包含胎心的三维超声数据；
    对多卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的多卷所述三维超声数据；
    根据所述数据重排后的多卷三维超声数据，进行所述胎心的显示。
  45. 如权利要求44所述的超声成像方法，其特征在于，所述扫描规则包括：在进行多次所述三维超声扫描的过程中，每完成一次所述三维超声扫描，经过相同的延迟时长，再进行下一次所述三维超声扫描。
  46. 如权利要求45所述的超声成像方法，其特征在于，所述延迟时长基于所述胎心的心率、基础零延时卷率和第一参数计算得到；所述基础零延时卷率为所述多次三维超声扫描中第一次三维超声扫描的扫描卷率，所述第一参数包括时间分 辨率或心相时间间隔，所述心相时间间隔为数据重排后的多卷所述三维超声数据中相邻两卷所述三维超声数据所对应的心相位置之间的时间间隔。
  47. 如权利要求44所述的超声成像方法，其特征在于，所述数据重排至少包括将部分扫描时序在后的所述三维超声数据向前调整，以及部分扫描时序在前的所述三维超声数据调整向后调整。
  48. 如权利要求44所述的超声成像方法，其特征在于，所述多次三维超声扫描的次数为m×n-1，和/或，所述多卷三维超声数据的卷数为m×n-1；其中：

    m＝floor[(TS+1)×BVR×60÷FR÷TS]；
    BVR表示基础零延时卷率，所述基础零延时卷率为所述多次三维超声扫描中第一次三维超声扫描的扫描卷率；FR表示所述胎心的心率；TS表示时间分辨率；floor表示向下取整操作。
  49. 如权利要求44所述的超声成像方法，其特征在于，所述多次三维超声扫描的次数为m×n-1，和/或，所述多卷三维超声数据的卷数为m×n-1；其中：

    m＝floor[(60÷FR÷TI+1)×BVR×TI]；
    BVR表示基础零延时卷率，所述基础零延时卷率为所述多次三维超声扫描中第一次三维超声扫描的扫描卷率；TI表示心相时间间隔，所述心相时间间隔为数据重排后的多卷所述三维超声数据中相邻两卷所述三维超声数据所对应的心相位置之间的时间间隔；FR表示所述胎心的心率，floor表示向下取整操作。
  50. 如权利要求48或49所述的超声成像方法，其特征在于，所述对多卷所述三维超声数据按照心相顺序进行数据重排，得到数据重排后的多卷所述三维超声数据，包括：数据重排前的m×n-1卷所述三维超声数据中第(i-1)×m+j卷所述三维超声数据，被设置为数据重排后的第(j-1)×n+i卷所述三维超声数据；所述i和j为正整数，且i小于或等于n，j小于或等于m。
  51. 一种超声成像系统，其特征在于，包括：
    超声探头，用于向胎心发射超声波，以及接收相应的超声波回波信号；
    发射和接收控制电路，用于控制所述超声探头执行发射超声波和接收超声波回波信号；
    处理器，用于执行如权利要求1至50中任一项所述的方法。
  52. 一种计算机可读存储介质，其特征在于，包括程序，所述程序能够被处理器执行以实现如权利要求1至50中任一项所述的方法。
PCT/CN2023/089944 2023-04-21 2023-04-21 一种胎心的超声成像方法和系统 WO2024037019A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2023/089944 WO2024037019A1 (zh) 2023-04-21 2023-04-21 一种胎心的超声成像方法和系统
CN202380009329.1A CN116867440A (zh) 2023-04-21 2023-04-21 一种胎心的超声成像方法和系统

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2023/089944 WO2024037019A1 (zh) 2023-04-21 2023-04-21 一种胎心的超声成像方法和系统

Publications (1)

Publication Number Publication Date
WO2024037019A1 true WO2024037019A1 (zh) 2024-02-22

Family

ID=88225432

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/089944 WO2024037019A1 (zh) 2023-04-21 2023-04-21 一种胎心的超声成像方法和系统

Country Status (2)

Country Link
CN (1) CN116867440A (zh)
WO (1) WO2024037019A1 (zh)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6414487B1 (en) * 2000-11-22 2002-07-02 Philips Medical Systems (Cleveland), Inc. Time and memory optimized method of acquiring and reconstructing multi-shot 3D MRI data
US20090148021A1 (en) * 2007-12-10 2009-06-11 Masao Yui Magnetic resonance imaging apparatus and magnetic resonance imaging method
CN101455565A (zh) * 2007-12-10 2009-06-17 株式会社东芝 磁共振成像装置和磁共振成像方法
CN102469984A (zh) * 2009-06-30 2012-05-23 皇家飞利浦电子股份有限公司 通过非心电图生理门控采集进行的三维胎心成像
CN105266813A (zh) * 2015-11-30 2016-01-27 中国科学院深圳先进技术研究院 一种自门控三维心脏成像的运动信号提取方法及装置
US20160171716A1 (en) * 2014-12-12 2016-06-16 Siemens Medical Solutions Usa, Inc. Reconstruction of Aneurysm Wall Motion
CN107886554A (zh) * 2016-09-30 2018-04-06 西门子保健有限责任公司 流数据的重构

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6414487B1 (en) * 2000-11-22 2002-07-02 Philips Medical Systems (Cleveland), Inc. Time and memory optimized method of acquiring and reconstructing multi-shot 3D MRI data
US20090148021A1 (en) * 2007-12-10 2009-06-11 Masao Yui Magnetic resonance imaging apparatus and magnetic resonance imaging method
CN101455565A (zh) * 2007-12-10 2009-06-17 株式会社东芝 磁共振成像装置和磁共振成像方法
CN102469984A (zh) * 2009-06-30 2012-05-23 皇家飞利浦电子股份有限公司 通过非心电图生理门控采集进行的三维胎心成像
US20160171716A1 (en) * 2014-12-12 2016-06-16 Siemens Medical Solutions Usa, Inc. Reconstruction of Aneurysm Wall Motion
CN105266813A (zh) * 2015-11-30 2016-01-27 中国科学院深圳先进技术研究院 一种自门控三维心脏成像的运动信号提取方法及装置
CN107886554A (zh) * 2016-09-30 2018-04-06 西门子保健有限责任公司 流数据的重构

Also Published As

Publication number Publication date
CN116867440A (zh) 2023-10-10

Similar Documents

Publication Publication Date Title
US11354791B2 (en) Methods and system for transforming medical images into different styled images with deep neural networks
US10192032B2 (en) System and method for saving medical imaging data
US9179890B2 (en) Model-based positioning for intracardiac echocardiography volume stitching
US11931201B2 (en) Device and method for obtaining anatomical measurements from an ultrasound image
JPWO2010134512A1 (ja) 医用画像診断装置とその関心領域設定方法
US11712224B2 (en) Method and systems for context awareness enabled ultrasound scanning
US20220361848A1 (en) Method and system for generating a synthetic elastrography image
EP4061231B1 (en) Intelligent measurement assistance for ultrasound imaging and associated devices, systems, and methods
CN106030657B (zh) 医学4d成像中的运动自适应可视化
US20220233171A1 (en) Systems and methods for controlling volume rate
CN110956076A (zh) 基于容积渲染在三维超声数据中进行结构识别的方法和系统
WO2020089416A1 (en) Identifying an interventional device in medical images
JP7346586B2 (ja) 合成３ｄ超音波画像を取得するための方法及びシステム
US9033883B2 (en) Flow quantification in ultrasound using conditional random fields with global consistency
CN112351742A (zh) 用于执行跨瓣压量化的方法和系统
US11717268B2 (en) Ultrasound imaging system and method for compounding 3D images via stitching based on point distances
WO2024037019A1 (zh) 一种胎心的超声成像方法和系统
EP2189812B1 (en) Providing volume information on a periodically moving target object in an ultrasound system
Wang et al. Optical flow networks for heartbeat estimation in 4d ultrasound images
US20220207743A1 (en) System and method for two dimensional acoustic image compounding via deep learning
US20230360225A1 (en) Systems and methods for medical imaging
WO2024013114A1 (en) Systems and methods for imaging screening
CN116258736A (zh) 用于分割图像的系统和方法
CN117557591A (zh) 一种基于超声图像的轮廓编辑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: 23853928

Country of ref document: EP

Kind code of ref document: A1