Ultrasonic fetal nuchal translucency image segmentation, detection method and device

Abstract

The present invention relates to the field of ultrasonic image detection, and specifically provides an ultrasonic fetal nuchal translucency image segmentation, detection method and device. The ultrasonic fetal nuchal translucency image segmentation method comprises: acquiring an ultrasonic image containing a fetal nuchal translucency; performing double straight line detection on the ultrasonic image to obtain a fetal nuchal translucency reference area; and obtaining a target contour of the fetal nuchal translucency based on the reference area. This method can relatively accurately locate the position of the NT by performing straight line detection on the NT area, thereby obtaining the NT reference area. On the one hand, it can eliminate the interference of the area outside the baseline on subsequent calculations, thereby saving subsequent calculation time. On the other hand, it provides a relatively accurate basis for the subsequent construction of the NT target contour, thereby increasing the accuracy of the calculation results.

Description

Technical Field

[0001] The present invention relates to the field of ultrasonic image detection, and in particular to a method and a device for segmenting and detecting an ultrasonic fetal nuchal transparent layer image. Background Art

[0002] The thickness of the fetal nuchal translucency (NT) in ultrasound images plays a very important role in the screening of fetal chromosomal abnormalities. The increase in NT thickness means an increased possibility of fetal abnormalities. Therefore, the measurement of fetal NT thickness in early pregnancy has important clinical diagnostic value. At present, the measurement of NT thickness in clinical diagnosis is mainly performed manually by ultrasound physicians operating a trackball. However, there are a lot of speckle noise, blurred edges of the NT area, and artifact interference on ultrasound images, which leads to random errors caused by manual measurement and visual errors of clinicians, which will affect the accuracy of the measurement results and lead to inaccurate measurement results. Summary of the invention

[0003] In order to solve the technical problem that the edge of the fetal nuchal translucency layer on the ultrasound image is blurred, resulting in inaccurate measurement results, the present invention provides a method and device for segmenting the ultrasound fetal nuchal translucency layer image.

[0004] At the same time, in order to solve the technical problems of inaccurate results of existing manual measurement of fetal nuchal translucency and increased time cost due to repetitive operations, the present invention provides a method and device for automatically measuring the thickness of the fetal nuchal translucency layer by ultrasound.

[0005] In a first aspect, the present invention provides a method for segmenting an ultrasound fetal nuchal transparent layer image, comprising:

[0006] Obtain ultrasound images that include the fetal nuchal translucency;

[0007] Based on the fetal nuchal translucency feature, performing double straight line detection on the ultrasound image to obtain a fetal nuchal translucency reference area;

[0008] A target contour of the fetal nuchal translucency layer is obtained based on the reference area.

[0009] In some embodiments, performing double straight line detection on the ultrasound image includes:

[0010] Detecting two high echo band characteristic areas of the fetal nuchal transparent layer from the ultrasound image to obtain two parallel baselines;

[0011] The reference area is obtained based on the two parallel baselines.

[0012] In some embodiments, detecting two high echogenic band characteristic areas of the fetal nuchal translucency from the ultrasound image comprises:

[0013] Performing feature enhancement on two high echo band characteristic areas of the fetal nuchal transparent layer on the ultrasound image to obtain a high echo band characteristic image;

[0014] Based on the characteristic image, the two high echo band characteristic regions are detected.

[0015] In some embodiments, the double straight line detection method is Radon transform detection.

[0016] In some embodiments, performing double straight line detection using Radon transform includes:

[0017] Mapping the ultrasonic image to Radon space through Radon transformation to obtain a spatial image with two highlight areas;

[0018] Calculate the angle and position of two maximum cumulative characteristic peaks based on the aerial image;

[0019] Through inverse Radon transformation, the points corresponding to the peak values ​​are mapped onto the image to obtain the parameters of the parallel baseline.

[0020] In some embodiments, before the feature enhancement of the two high echo band feature areas of the fetal nuchal transparent layer on the ultrasound image, the method further includes:

[0021] The ultrasonic image is subjected to denoising processing.

[0022] In some embodiments, obtaining the target contour of the fetal nuchal translucency layer based on the reference area includes:

[0023] constructing an initial profile based on the two parallel baselines;

[0024] Based on the initial contour, a target contour of the fetal nuchal translucency layer is obtained by a region-based level set segmentation method.

[0025] In some embodiments, constructing an initial profile based on the two parallel baselines comprises:

[0026] The closed area formed by the two parallel baselines and the boundary of the ultrasound image constitutes the initial contour; or

[0027] The initial contour is formed by drawing a perpendicular line from the intersection of one of the parallel baselines and the boundary of the ultrasound image to the other parallel baseline, and the boundary of the rectangular area contained in the image is formed; or

[0028] A perpendicular line is drawn from the intersection point of one of the parallel baselines and the boundary of the ultrasonic image to the other parallel baseline to form the boundary of the region of the rectangular inscribed ellipse to constitute the initial contour.

[0029] In a second aspect, the present invention further provides an ultrasonic fetal nuchal transparent layer image detection method, comprising:

[0030] Acquire a target contour of a fetal nuchal translucency layer, wherein the target contour is obtained by using the ultrasonic fetal nuchal translucency layer image segmentation method described in any embodiment of the first aspect;

[0031] The target profile parameters are measured.

[0032] In some embodiments, measuring the target profile parameter comprises:

[0033] Based on the target contour, determining the upper boundary and the lower boundary of the fetal nuchal translucency layer on the ultrasound image, and calculating the midline of the upper boundary and the lower boundary;

[0034] Performing straight line fitting on the upper boundary, the lower boundary and the midline,

[0035] Calculate the distance between the normal line of the midline and the two intersection points of the upper and lower boundaries using the fitted straight line;

[0036] Find the position coordinates of the two intersection points at the maximum distance;

[0037] Convert the position coordinates to physical unit size.

[0038] In a third aspect, the present invention provides an ultrasonic fetal nuchal transparent layer image segmentation device, comprising:

[0039] An image acquisition module, used for acquiring an ultrasound image including a fetal nuchal transparent layer;

[0040] A straight line detection module, used for performing double straight line detection on the ultrasound image to obtain a fetal nuchal transparent layer reference area; and

[0041] An image segmentation module is used to obtain a target contour of the fetal nuchal translucency layer based on the reference area.

[0042] In a fourth aspect, the present invention provides an ultrasonic fetal nuchal transparent layer image detection device, comprising:

[0043] An acquisition module, configured to acquire a target contour of a fetal nuchal translucency layer, wherein the target contour is obtained by using the ultrasonic fetal nuchal translucency layer image segmentation method described in any embodiment of the first aspect;

[0044] The measuring module is used to measure the target profile parameters.

[0045] In a fifth aspect, the present invention provides an ultrasonic device, comprising:

[0046] Processor; and

[0047] A memory is communicatively connected to the processor and stores computer-readable instructions that can be executed by the processor. When the computer-readable instructions are executed, the processor executes the ultrasonic fetal nuchal transparent layer image segmentation method described in any embodiment of the first aspect.

[0048] The ultrasonic fetal nuchal transparent layer image segmentation method provided by the present invention comprises acquiring an ultrasonic image containing a fetal nuchal transparent layer, performing double straight line detection on the ultrasonic image, obtaining a fetal nuchal transparent layer reference area, and obtaining a target contour of the fetal nuchal transparent layer based on the reference area. The method automatically segments the NT contour, improves the degree of automation, and at the same time, by performing double straight line detection on the NT area, the position of the NT can be relatively accurately located, thereby obtaining the NT reference area. On the one hand, the interference of the area outside the baseline on the subsequent calculation can be eliminated, thereby saving the subsequent calculation time. On the other hand, a relatively accurate basis is provided for the subsequent construction of the NT target contour, thereby increasing the accuracy of the calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1 is a schematic diagram of a method for segmenting an ultrasound fetal nuchal transparent layer image according to some embodiments of the present invention;

[0051] Figure 2 is a schematic diagram of obtaining ultrasonic NT images according to some embodiments of the present invention;

[0052] Figure 3 is a schematic flow chart of a method for segmenting an ultrasound fetal nuchal transparent layer image according to some embodiments of the present invention;

[0053] Figure 4 is a schematic diagram of straight line detection and traditional edge detection results according to some embodiments of the present invention;

[0054] Figure 5 is a schematic diagram of a method for detecting an ultrasonic fetal nuchal translucency image according to some embodiments of the present invention;

[0055] Figure 6is a schematic diagram of the structure of a system for segmenting an ultrasound fetal nuchal transparent layer image according to some embodiments of the present invention;

[0056] Figure 7 is a schematic structural diagram of an ultrasonic fetal nuchal translucency image detection system according to some embodiments of the present invention;

[0057] Figure 8 It is a schematic diagram of a structure suitable for implementing the method or ultrasonic device in an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described implementation is a partial embodiment of the present invention, rather than all implementations. Based on the implementations in the present invention, all other implementations obtained by ordinary technicians in the field without creative work are within the scope of protection of the present invention. In addition, the technical features involved in the different implementations of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0059] The ultrasound fetal nuchal translucency image segmentation method provided by the present invention is used to perform image segmentation of the fetal nuchal translucency (NT) in an ultrasound image. It should be noted that in ultrasound image detection, factors such as signal attenuation, uneven grayscale distribution, artifacts, and speckle noise will lead to a low signal-to-noise ratio of the ultrasound image, making it difficult to detect the NT area in the ultrasound image. The image segmentation method provided by the present invention realizes automatic segmentation of the NT area in the ultrasound image, thereby making the subsequent measurement of the NT thickness more accurate.

[0060] Figure 1 FIG. 4 shows an ultrasonic NT image segmentation method according to some embodiments of the present invention. Figure 1 As shown, the method may include:

[0061] S1. Acquire an ultrasound image containing NT.

[0062] S2. Based on the NT feature, double straight line detection is performed on the ultrasound image to obtain the NT reference area.

[0063] S3. Obtain the target contour of NT based on the reference area.

[0064] Specifically, in step S1, an ultrasonic image containing NT can be obtained by an ultrasonic diagnostic system, for example, an ultrasonic image containing NT area after acoustic amplification generated by an ultrasonic diagnostic instrument. Figure 2In some embodiments, to reduce interference, in the ultrasound image generated by the ultrasound diagnostic instrument, the position and size of the rectangular frame can be adjusted by, for example, a trackball, a mouse, or a keyboard, and a target area containing a complete NT can be further selected and the target area can be cropped as the ultrasound NT image to be tested. For example, please refer to Figure 2 , the ultrasound image of the target area after the frame selection can be Figure 2 As shown in (b), the target area image is used as the ultrasonic image containing NT to be tested, and the NT area is preliminarily located, which can reduce interference information and reduce calculation time. In step S2, the above ultrasonic NT image is subjected to double straight line detection, which can be Figure 2 (b) The NT reference region is further extracted from the ultrasonic NT image. The reference region is constructed by double straight line detection to eliminate the interference of the outside region on the subsequent calculation. At the same time, the reference region can directly provide a basis for the initial contour of the subsequent construction of the active contour model, without the need to perform feature screening and correction on the image, which greatly simplifies the calculation steps and improves the calculation accuracy. In step S3, the target contour of the NT is obtained based on the above-mentioned reference region. Since the reference region is already very close to the target (NT) boundary, the calculation of the target contour is faster and more accurate.

[0065] like Figure 3 As shown, in the ultrasonic NT image segmentation method according to some embodiments of the present invention, step S1 includes:

[0066] S10. Acquire an ultrasound image that meets measurement requirements.

[0067] In the conventional NT measurement requirements, in order to provide more accurate test results, ultrasound images generally require fetal long axis sagittal sections, so when acquiring ultrasound images, ultrasound images of standard sections can be provided. It should be noted that the above measurement requirements are only used as the optimal implementation method. This method is also applicable to NT images with obvious upper and lower high echo features, and there is no limitation on this.

[0068] It should be noted that in the ultrasound NT image, the skin hyperechoic bands on both sides of the NT area and the deep soft tissue hyperechoic bands are brighter than the low echoic bands in the middle. Therefore, in the segmentation method of some embodiments of the present invention, the detection of the NT image adopts feature detection of the hyperechoic area. Figure 3 In some of the embodiments shown, step S2 may include:

[0069] S21. Enhance the features of the NT high-echo feature area to obtain a feature image. In some embodiments, the high-echo region feature of the input ultrasound image can be enhanced by anisotropic filtering. The feature detection algorithm can adopt a method based on the Hessian matrix. The Hessian matrix is ​​a square matrix of the second-order partial derivative of a scalar function, which describes the local curvature of a multivariate function and is defined as:

[0070]

[0071] Among them I xx represents the second-order partial derivative of image I in the x direction, I yy represents the second-order partial derivative of I in the y direction, I xy =I yx Represents the mixed partial derivative of the image in the xy direction, that is:

[0072] Ix x =I(i+1,j)-2*I(i,j)+I(i-1,j)

[0073] I yy =I(i,j+1)-2*I(i,j)+I(i,j-1)

[0074] Ix y =I(i+1,j+1)-I(i,j+1)-I(i+1,j)+I(i,j)

[0075] Since the second-order partial derivative is generally sensitive to noise, in some implementations, before step S21, the following steps may be further included:

[0076] S20, denoising the ultrasound image. In an exemplary embodiment, Gaussian smoothing is used to denoise the ultrasound image.

[0077] The high echo area is feature detected by the above method to obtain a corresponding feature image. It should be noted that in steps S20 and S21, Hessian and Gaussian smoothing are only used as an implementation method of the present invention, and those skilled in the art may also select other suitable implementation methods, which are not limited by the present invention.

[0078] After the enhancement of the characteristics of the high echo area of ​​the ultrasound NT image, the following are also included:

[0079] S22. Based on the above feature image, the image feature area is detected using Radon transform to obtain two parallel baselines.

[0080] In ultrasonic NT images, the skin hyperechoic bands and deep soft tissue hyperechoic bands on both sides of the NT hypoechoic band are bright and approximately parallel. Based on this characteristic, this method uses Radon transform to perform double straight line detection on ultrasonic NT images. Compared with edge detection, the calculation is simpler and the result is more accurate.

[0081] Specifically, the Radon transformation of the method first maps the feature image after the above-mentioned feature enhancement to the Radon space through Radon transformation, and the two high echo feature areas in the image space correspond to the two highlight areas in the Radon space, thereby obtaining the spatial image of NT.

[0082] The Radon transform can actually calculate the projection of the image matrix in all directions. The projection here is actually its line integral. The Radon space describes the coordinates of each straight line in polar coordinates, that is, a straight line in the image space corresponds to a point in the Radon space. The conversion formula is:

[0083] xcosθ+ysinθ=ρ

[0084] Among them, ρ corresponds to the ordinate of the Radon space, which is the distance from the origin to the straight line in the image space, and θ corresponds to the abscissa of the Radon space, which represents the inclination angle of the straight line in the image space.

[0085] Based on the above spatial image, the corresponding curves under different θ values ​​are calculated, and then the angle is determined by constructing a cost function, so as to find the angle with the maximum cost function, and then the inclination angle of NT in the rectangular coordinate system and the positions of the two corresponding maximum peaks can be calculated. Then, after the inverse Radon transform, the points corresponding to the peaks are mapped back to the image space, and the corresponding parallel baseline parameters can be obtained.

[0086] In this embodiment, considering that the inclination angle of NT is generally not parallel to the longitudinal axis, the value range of θ calculated in the present invention is [0,70]∪[110,179]. The value range of ρ is [0,γ), where: W and H correspond to the width and height of the ultrasound image, respectively.

[0087] like Figure 4 As shown in (a), in this embodiment, two baselines passing through two high echo areas can be obtained by the above method, because the angle θ calculated during the inverse Radon transform is one value, but the corresponding ρ is two completely different values, so the two baselines are parallel and do not overlap. Figure 4As shown in (a), the area between the two parallel baselines constitutes the NT reference area, which effectively removes the interference information outside the reference area. At the same time, the reference area can provide a basis for the subsequent construction of the initial contour of the active contour model, without the need to perform feature screening and correction on the image. For the convenience of explanation, Figure 4 (b) shows the result of edge detection on the ultrasonic NT image. It can be seen that there is a lot of fine noise information in the ultrasonic image after edge detection. When constructing the initial contour of the NT area, these fine noises need to be screened out and corrected to further confirm the edge range of the NT area. The result obtained by Radon transform in the present invention Figure 4 As shown in (a), the edge range and tilt angle of the NT can be accurately confirmed, and a relatively accurate NT reference area can be obtained without interference information. At the same time, the confirmed reference area includes the boundary range and tilt angle of the NT, which greatly simplifies the subsequent calculation process.

[0088] Based on the NT reference area obtained above, the segmentation method of the present invention adopts a region-based level set segmentation method to segment the NT contour. The region-based level set method of the present invention is different from the traditional level set method based on global information. The region-based level set method can find relative boundary features based on the region. The reason for adopting such a scheme is that the position of the parallel baseline is calculated in the previous implementation step, and the initial contour constructed is actually closer to the target boundary, which avoids the problem of the global method when the edge feature contrast is uneven, and can enhance the algorithm's detection performance for uneven features. Therefore, step S3 includes:

[0089] S30, constructing the initial contour of the level set based on the parallel baseline. Figure 4 As shown in (a), the area between the two parallel baselines forms the NT reference area, and the initial contour is constructed based on the NT reference area. When constructing the initial contour, there are many methods for constructing the initial contour. Due to the differences in images of different manufacturers, it can be adjusted according to specific circumstances. For example, in one embodiment, the closed area formed by the two parallel baselines and the boundary of the ultrasound image constitutes the initial contour, that is, Figure 4 In another embodiment, a perpendicular line can be drawn from the intersection of one of the baselines and the boundary of the ultrasound image to the other baseline, and the rectangular area formed by the two perpendicular lines and the two baselines and contained in the image constitutes the initial contour. Figure 4(a) is taken as an example, a perpendicular line is drawn from the intersection of the upper baseline and the right side boundary of the image to the lower baseline, and a perpendicular line is drawn from the intersection of the lower baseline and the left side boundary of the image to the upper baseline, and the rectangular area formed by the two perpendicular lines and the two baselines is the rectangular area inside the image. The present invention uses this as the initial contour. In one embodiment, an inscribed ellipse of the rectangle can also be made based on the above-mentioned rectangular area boundary, and the ellipse boundary constitutes the initial contour. It should be noted that the above-mentioned initial contour construction method is only used to illustrate the present invention, and does not limit the present invention. Those skilled in the art may also have other methods suitable for constructing the initial contour.

[0090] S31. Based on the initial contour, a region-based level set segmentation method is used to obtain the target contour of the NT.

[0091] The level set method is a geometric contour model. The main idea is to describe the evolution process of a low-dimensional target curve with a zero-level set section of a level set function with a higher dimension. When the functional of the level set is minimized, the zero-level set of the level set function will shrink to the boundary of the target under the action of internal and external forces. However, the level set method based on global statistical information often does not produce ideal results when segmenting targets with uneven background brightness distribution. The present invention applies the local level-set method to the image segmentation of fetal NT.

[0092] Given an image I, whose domain is Ω, let a curve C be on the two-dimensional image domain Ω, and x is any point on the image (x∈Ω), then its energy function is:

[0093] Lc(f)={(x1,x Z ,…,xn)|f(x1,x Z , …, xn)

[0094] where x1, x2, …, x n represents a curve, c is a high-dimensional constant. c(f) is used to represent a smooth, closed curve. Let k represent the curvature and N represent the normal. The change of the curve is only related to the change of the normal direction. That is, the following relationship exists:

[0095]

[0096] In summary, the target contour of image NT can be obtained.

[0097] In this embodiment, the regional energy of the calculated image is determined by the regional information, so that the defect that the method cannot accurately segment the image with uneven brightness because it only uses global information can be overcome. The correct position of the curve C, that is, the accurate position of the target edge, is obtained by minimizing the energy function L, and the level set method is retained to automatically process any topological changes of the contour curve C, such as the splitting or merging of the curve.

[0098] Secondly, considering that the clinical index of NT is mainly based on the thickest part, the present invention also provides an ultrasonic fetal nuchal transparent layer image detection method, which is suitable for measuring the target contour of the ultrasonic NT image segmentation method in the above embodiment.

[0099] Figure 5 FIG. 4 shows an ultrasonic NT image detection method according to some embodiments of the present invention, such as Figure 5 As shown, the method includes:

[0100] S41. Based on the target contour, determine the upper and lower boundaries of the NT on the ultrasound image, and calculate the midline of the upper and lower boundaries. When determining the upper and lower boundaries, detect the first step point of each column from top to bottom on the ultrasound image as the upper boundary of each column, and then detect the first step point of each column from bottom to top as the lower boundary, and calculate the midline based on the upper and lower boundaries.

[0101] S42, perform straight line fitting on the upper and lower boundaries and the center line, and calculate the distance between the normal line of the center line and the two intersection points of the upper and lower boundary fitting lines. Traverse the target area in the form of a traversal window, calculate the distance between the normal line of the center line and the two intersection points of the upper and lower boundary fitting lines, and use this distance as the upper and lower distances of the traversed point, and so on, to obtain the upper and lower boundary distances of all points.

[0102] In this step, the fitting method used is the least squares method. The least squares method is to calculate by minimizing the residual of the fitted line or curve. Assuming that n points to be fitted are known, the residual formula of the straight line fitting is:

[0103]

[0104] Therefore, the residual i The sum of squares is:

[0105]

[0106] By minimizing Q, that is, taking the extreme value, it is equivalent to finding its partial derivative with respect to the parameters b0 and b1, and setting its partial derivative equal to 0. Then, the obtained parameters b0 and b1 are substituted into the straight line equation (1) to determine the straight line equation. If the curve fitting method is required, the calculation idea is similar, except that the parameters of the curve may be more. By listing the curve expression equation, constructing the Q value according to formula (3), and respectively finding the parameters when taking the extreme value, the curve equation can be determined.

[0107] It should be noted that the straight line or curve fitting method of the present invention can also be adjusted according to the specific image. The least square fitting method given in this embodiment is only an example and is not used to limit the present invention.

[0108] S43, finding the position coordinates of the two points at the maximum distance. Based on the upper and lower boundary distances obtained above, finding the position coordinates of the two points with the maximum distance is what is required.

[0109] S44, converting the position coordinates into physical unit size. By converting the position coordinates into physical unit position coordinates, the position coordinates of the maximum thickness can be determined, and the NT thickness parameters in the target contour can be further obtained.

[0110] In a third aspect, the present invention provides an ultrasonic NT image segmentation device, such as Figure 6 As shown, in some embodiments, the device may include:

[0111] An image acquisition module 10 is used to acquire an ultrasound image including a fetal nuchal transparent layer;

[0112] A straight line detection module 20, configured to perform double straight line detection on the ultrasound image to obtain a fetal nuchal transparent layer reference area; and

[0113] The image segmentation module 30 is used to obtain the target contour of the fetal nuchal translucency layer based on the reference area.

[0114] In a fourth aspect, the present invention provides an ultrasonic NT image detection device for measuring the target contour in the above-mentioned embodiment. Figure 7 As shown, the device may include:

[0115] An acquisition module 40 is used to acquire a target contour of the NT, wherein the target contour is obtained by using the ultrasonic fetal nuchal transparent layer image segmentation method of the above embodiment; and

[0116] The measuring module 50 is used to convert the position coordinates into physical unit sizes.

[0117] In a fifth aspect, the present invention provides an ultrasonic device, comprising:

[0118] Processor; and

[0119] The memory is communicatively connected to the processor and stores computer-readable instructions that can be executed by the processor. When the computer-readable instructions are executed, the processor executes the ultrasound fetal nuchal transparent layer image segmentation method in the above embodiment.

[0120] In a sixth aspect, the present invention provides a storage medium storing computer instructions for enabling a computer to execute the above-mentioned ultrasound NT image segmentation and / or detection method.

[0121] Specifically, Figure 8 A schematic diagram of the structure of an ultrasound device 600 suitable for implementing a method or a processor according to an embodiment of the present invention is shown. Figure 8 The system shown realizes the corresponding functions of the ultrasonic equipment and storage medium.

[0122] like Figure 8 As shown, the ultrasound device 600 includes a central processing unit (CPU) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage part 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for operation are also stored. The CPU 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0123] The following components are connected to the I / O interface 605: an input section 606 including a keyboard, a mouse, etc.; an output section 607 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, a modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the I / O interface 605 as needed. A removable medium 611, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 610 as needed, so that a computer program read therefrom is installed into the storage section 608 as needed.

[0124] In particular, according to the embodiments of the present disclosure, the above reference Figure 1 The described processes may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program tangibly embodied on a machine-readable medium, the computer program comprising a computer program for executing Figure 1In such an embodiment, the computer program may be downloaded and installed from a network via the communication section 609 and / or installed from the removable medium 611.

[0125] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and a module, a program segment or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a different order from the order marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0126] Obviously, the above embodiments are merely examples for the purpose of clear explanation, and are not intended to limit the embodiments. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. The obvious changes or modifications derived therefrom are still within the scope of protection of the present invention.

Claims

1. A method for segmenting an ultrasonic fetal nuchal transparent layer image, characterized in that: include: Obtain ultrasound images that include the fetal nuchal translucency; Based on the fetal nuchal translucency characteristics, a double straight line detection is performed on the ultrasound image to obtain a fetal nuchal translucency reference area; the double straight line detection method is Radon transform detection; A target contour of the fetal nuchal translucency layer is obtained based on the reference area.

2. The method for segmenting an ultrasonic fetal nuchal transparent layer image according to claim 1, characterized in that: The performing double straight line detection on the ultrasonic image comprises: Detecting two high echo band characteristic areas of the fetal nuchal transparent layer from the ultrasound image to obtain two parallel baselines; The reference area is obtained based on the two parallel baselines.

3. The method for segmenting an ultrasonic fetal nuchal transparent layer image according to claim 2, characterized in that: The detecting of two high echogenic band characteristic areas of the fetal nuchal transparent layer from the ultrasound image comprises: Performing feature enhancement on two high echo band characteristic areas of the fetal nuchal transparent layer on the ultrasound image to obtain a high echo band characteristic image; Based on the characteristic image, the two high echo band characteristic regions are detected.

4. The method for segmenting an ultrasonic fetal nuchal transparent layer image according to claim 3, characterized in that: Using Radon transform to detect double lines includes: Mapping the ultrasonic image to Radon space through Radon transformation to obtain a spatial image with two highlight areas; Calculate the angle and position of two maximum cumulative characteristic peaks based on the aerial image; Through inverse Radon transformation, the points corresponding to the peak values ​​are mapped onto the image to obtain the parameters of the parallel baseline.

5. The method for segmenting an ultrasonic fetal nuchal transparent layer image according to claim 3, characterized in that: Before the two high echo band characteristic areas of the fetal nuchal transparent layer on the ultrasound image are enhanced, the method further includes: The ultrasonic image is subjected to denoising processing.

6. The method for segmenting an ultrasonic fetal nuchal transparent layer image according to claim 2, characterized in that: The target contour of the fetal nuchal translucency layer is obtained based on the reference area, comprising: constructing an initial profile based on the two parallel baselines; Based on the initial contour, a target contour of the fetal nuchal translucency layer is obtained by a region-based level set segmentation method.

7. The method for segmenting an ultrasonic fetal nuchal transparent layer image according to claim 6, characterized in that: The constructing of the initial profile based on the two parallel baselines comprises: The closed area formed by the two parallel baselines and the boundary of the ultrasound image constitutes the initial contour; or The initial contour is formed by drawing a perpendicular line from the intersection of one of the parallel baselines and the boundary of the ultrasound image to the other parallel baseline, and the boundary of the rectangular area contained in the image is formed; or A perpendicular line is drawn from the intersection point of one of the parallel baselines and the boundary of the ultrasonic image to the other parallel baseline to form the boundary of the region of the rectangular inscribed ellipse to constitute the initial contour.

8. A method for detecting an ultrasonic fetal nuchal transparent layer image, characterized in that: include: Acquire a target contour of a fetal nuchal translucency layer, wherein the target contour is obtained by using the ultrasonic fetal nuchal translucency layer image segmentation method according to any one of claims 1 to 7; The target profile parameters are measured.

9. The method for detecting the ultrasonic fetal nuchal translucency image according to claim 8, characterized in that: The measuring of the target profile parameters comprises: Based on the target contour, determining the upper boundary and the lower boundary of the fetal nuchal translucency layer on the ultrasound image, and calculating the midline of the upper boundary and the lower boundary; Performing straight line fitting on the upper boundary, the lower boundary and the midline, Calculate the distance between the normal line of the midline and the two intersection points of the upper and lower boundaries using the fitted straight line; Find the position coordinates of the two intersection points at the maximum distance; Convert the position coordinates to physical unit size.

10. An ultrasonic fetal nuchal transparent layer image segmentation device, characterized in that: include: An image acquisition module, used for acquiring an ultrasound image including a fetal nuchal transparent layer; A straight line detection module, used for performing double straight line detection on the ultrasound image to obtain a fetal nuchal transparent layer reference area; the double straight line detection method is Radon transform detection; as well as An image segmentation module is used to obtain a target contour of the fetal nuchal translucency layer based on the reference area.

11. An ultrasonic fetal nuchal transparent layer image detection device, characterized in that: include: An acquisition module, used for acquiring a target contour of a fetal nuchal translucency layer, wherein the target contour is obtained by using the ultrasonic fetal nuchal translucency layer image segmentation method according to any one of claims 1 to 7; The measuring module is used to measure the target profile parameters.

12. An ultrasonic device, characterized in that: include: processor; and A memory is communicatively connected to the processor and stores computer-readable instructions that can be executed by the processor. When the computer-readable instructions are executed, the processor executes the ultrasound fetal nuchal transparent layer image segmentation method according to any one of claims 1 to 7.

Images