WO2022141085A1 - Ultrasonic detection method and ultrasonic imaging system - Google Patents
Ultrasonic detection method and ultrasonic imaging system Download PDFInfo
- Publication number
- WO2022141085A1 WO2022141085A1 PCT/CN2020/140985 CN2020140985W WO2022141085A1 WO 2022141085 A1 WO2022141085 A1 WO 2022141085A1 CN 2020140985 W CN2020140985 W CN 2020140985W WO 2022141085 A1 WO2022141085 A1 WO 2022141085A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- symmetry
- anatomical structure
- region
- image
- ultrasound image
- Prior art date
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 97
- 238000003384 imaging method Methods 0.000 title abstract description 8
- 238000002604 ultrasonography Methods 0.000 claims abstract description 280
- 230000001605 fetal effect Effects 0.000 claims abstract description 139
- 210000004556 brain Anatomy 0.000 claims abstract description 101
- 210000003484 anatomy Anatomy 0.000 claims description 292
- 210000003625 skull Anatomy 0.000 claims description 143
- 238000000034 method Methods 0.000 claims description 136
- 239000003550 marker Substances 0.000 claims description 64
- 239000000523 sample Substances 0.000 claims description 54
- 238000004422 calculation algorithm Methods 0.000 claims description 51
- 238000010801 machine learning Methods 0.000 claims description 37
- 230000008569 process Effects 0.000 claims description 20
- 210000001638 cerebellum Anatomy 0.000 claims description 17
- 210000001103 thalamus Anatomy 0.000 claims description 15
- 238000012285 ultrasound imaging Methods 0.000 claims description 10
- 210000003754 fetus Anatomy 0.000 claims description 9
- 210000003140 lateral ventricle Anatomy 0.000 claims description 9
- 210000002987 choroid plexus Anatomy 0.000 claims description 7
- 238000003709 image segmentation Methods 0.000 claims description 7
- 238000002592 echocardiography Methods 0.000 claims description 6
- 238000011156 evaluation Methods 0.000 abstract description 31
- 238000010586 diagram Methods 0.000 description 26
- 238000012549 training Methods 0.000 description 12
- 238000011158 quantitative evaluation Methods 0.000 description 10
- 210000000145 septum pellucidum Anatomy 0.000 description 10
- 230000011218 segmentation Effects 0.000 description 9
- 230000002490 cerebral effect Effects 0.000 description 8
- 238000003745 diagnosis Methods 0.000 description 8
- 238000012545 processing Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 238000007689 inspection Methods 0.000 description 6
- 238000013135 deep learning Methods 0.000 description 5
- 238000013461 design Methods 0.000 description 5
- 210000001259 mesencephalon Anatomy 0.000 description 5
- 210000001519 tissue Anatomy 0.000 description 5
- 210000000988 bone and bone Anatomy 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000035699 permeability Effects 0.000 description 4
- 229910052704 radon Inorganic materials 0.000 description 4
- SYUHGPGVQRZVTB-UHFFFAOYSA-N radon atom Chemical compound [Rn] SYUHGPGVQRZVTB-UHFFFAOYSA-N 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 241000270295 Serpentes Species 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000004641 brain development Effects 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 210000000877 corpus callosum Anatomy 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000013519 translation Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 210000004720 cerebrum Anatomy 0.000 description 1
- 238000003759 clinical diagnosis Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000002059 diagnostic imaging Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000012854 evaluation process Methods 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 125000001475 halogen functional group Chemical group 0.000 description 1
- 210000005240 left ventricle Anatomy 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 238000004091 panning Methods 0.000 description 1
- 230000035479 physiological effects, processes and functions Effects 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 210000005241 right ventricle Anatomy 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 239000013598 vector Substances 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 230000016776 visual perception Effects 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
Definitions
- the present application relates to the field of medical devices, and more particularly, to an ultrasonic detection method and an ultrasonic imaging system.
- Ultrasound as the most widely used examination method in modern medical imaging technology, is widely used in the diagnosis of human diseases due to its advantages of safe use, low cost, rapid imaging, and high reliability.
- the development of ultrasound-assisted diagnosis technology has further promoted the practical application of ultrasound in clinical diagnosis. It not only expands the application scope of ultrasound technology, but also helps doctors improve the accuracy and efficiency of examinations, and provides patients with more comprehensive and personalized treatment plans. .
- the present application has been made to solve at least one of the above-mentioned problems.
- the present application provides an ultrasonic detection method and an ultrasonic imaging system, which can help the doctor obtain the symmetry assessment guide of the fetal skull after the doctor has collected the ultrasonic image of the fetal skull, and provide the doctor with the ability to evaluate the symmetry of the fetal skull. Reference, or further diagnosis based on symmetry assessment guidelines.
- the embodiments of the present application provide an ultrasonic detection method, comprising:
- the image acquisition step is to acquire an ultrasound image of the fetal skull
- the symmetry assessment guide step is to acquire an ultrasound image located on at least one side of the symmetry mark, and obtain a symmetry assessment guide of the fetal skull with respect to the symmetry mark based on the ultrasound image of the at least one side.
- the embodiments of the present application also provide an ultrasonic detection method, comprising:
- the image acquisition step is to acquire an ultrasound image of the fetal skull
- the step of obtaining a symmetry mark is to obtain the symmetry mark of the fetal skull in the ultrasound image
- the projection step is to obtain the outline of the anatomical structure or a projection mark that is symmetrical on the other side of the symmetry mark with respect to the symmetry mark in the region where the anatomical structure is located;
- a display step displays the projected mark.
- embodiments of the present application also provide an ultrasound imaging system comprising:
- Ultrasound probe used to transmit ultrasound waves and receive ultrasound echoes from the cranial brain of the tested fetus to obtain ultrasound echo signals
- a processor configured to process the ultrasonic echo signal to obtain an ultrasonic image of the biological tissue
- the processor is further configured to execute the ultrasonic detection method according to the first aspect or the second aspect based on the executable program instructions.
- the embodiments of the present application provide an ultrasonic detection method and an ultrasonic imaging system for an anatomical structure. Based on the acquired ultrasonic images of the fetal skull, a guideline for evaluating the symmetry of the fetal skull is obtained, so as to evaluate the symmetry of the fetal skull. Provide a reference, or proceed with further diagnosis based on the symmetry assessment guidelines.
- FIG. 1 shows a schematic flowchart of an ultrasonic detection method according to an embodiment of the present application
- FIG. 2 shows a schematic block diagram of an ultrasound imaging system according to an embodiment of the present application
- FIG. 3 shows a schematic flowchart of a method for calculating symmetry in an ultrasonic testing method according to an embodiment of the present application
- FIG. 4 shows a schematic diagram of an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application
- FIG. 5 shows a schematic diagram of displaying a symmetry mark in an ultrasound image of a fetal skull on a display in an ultrasound detection method according to an embodiment of the present application
- FIG. 6 shows a schematic diagram of acquiring a region of interest in an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application
- FIG. 7 shows a schematic diagram of acquiring an anatomical structure in an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application
- FIG. 8 shows a schematic flowchart of a method for obtaining a symmetry assessment guide in an ultrasonic inspection method according to an embodiment of the present application
- FIG. 9 shows a schematic diagram of displaying a projection marker in an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application
- FIG. 10 shows a schematic diagram of displaying a projection marker in an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application
- FIG. 11 shows a schematic diagram of distinguishing and displaying projection markers in the ultrasound image of the fetal skull displayed on the display in the ultrasound detection method according to an embodiment of the present application;
- FIG. 12 shows a schematic flowchart of an ultrasonic detection method according to an embodiment of the present application
- FIG. 13 shows a schematic diagram of an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application
- FIG. 14 shows a schematic diagram of displaying a symmetry mark in an ultrasound image of a fetal skull on a display in an ultrasound detection method according to an embodiment of the present application
- FIG. 15 shows a schematic diagram of displaying projection markers in the ultrasound image of the fetal skull displayed on the display in the ultrasound detection method according to an embodiment of the present application
- FIG. 16 shows a schematic diagram of displaying a projection marker in an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application
- FIG. 17 shows a schematic diagram of distinguishing and displaying projection markers in the ultrasound image of the fetal skull displayed on the display in the ultrasound detection method according to an embodiment of the present application.
- the first embodiment of the present application provides an ultrasonic detection method, which is used to help the doctor obtain a symmetry assessment guide about the fetal skull after the doctor collects the ultrasonic image of the fetal skull, so as to help the doctor to evaluate the fetal skull. Symmetry of the brain provides a reference.
- FIG. 1 shows an ultrasonic inspection method 100 according to one embodiment of the present application.
- the ultrasonic detection method 100 may include the following steps:
- the image acquisition step S110 is to acquire an ultrasound image of the fetal skull.
- an ultrasonic imaging system is used to acquire an ultrasonic image.
- the ultrasound imaging system 200 includes an ultrasound probe 201 and an image processing unit 202 , and the image processing unit includes a processor 2022 , a memory 2021 and a display 2023 .
- the ultrasonic probe 201 is used to transmit ultrasonic waves to the skull of the tested fetus and receive ultrasonic echoes to obtain ultrasonic echo signals.
- the processor 2022 is configured to process the ultrasonic echo signals to obtain an ultrasonic image of the fetal skull, and perform ultrasonic detection on the fetal skull based on the ultrasonic image.
- the memory 2021 is used to store ultrasound images.
- the display 2023 is used to display ultrasound images.
- the symmetry is identified as the midline of the brain corresponding to the cross-sectional image, and when the ultrasound image is a three-dimensional image of the fetal skull, the symmetry is identified as corresponding to the brain midline. in the midsagittal plane of the 3D image.
- the doctor manually obtains the symmetry identification according to experience.
- the ultrasound image is a two-dimensional image of the fetal skull
- the method of obtaining the midline of the brain of the two-dimensional image adopts the method of connecting two points on the midline of the brain by selecting and marking the two points on the midline of the brain when acquiring the ultrasound image.
- the points form a straight line, which serves as the midline of the brain.
- a processor in the step S120 of obtaining the symmetry identification, is used to process the ultrasound image to obtain the symmetry identification.
- the characteristic anatomical structure is detected in the three-dimensional image by the structure detection algorithm, the detected characteristic anatomical structure is fitted as a plane, and the fitted plane is determined as the midsagittal plane;
- the midsagittal plane is determined from the three-dimensional image by a pretrained machine learning model.
- a straight line detection algorithm is used to detect the midline of the brain, or a plane detection algorithm is used to detect the midsagittal plane. Since the fetal skull is symmetrical about the midline or the three-dimensional image of the fetal skull is symmetrical about the midsagittal plane, the midline or midsagittal plane in the ultrasound image tends to have distinct features. Based on this feature, the line detection method is used to detect the midline of the brain in the two-dimensional image, or the plane detection method is used to detect the midsagittal plane of the three-dimensional image.
- the special anatomical structures located in the midline of the brain include The special anatomical structures on the midline include the thalamus, septum pellucidum, etc.
- the midline of the brain is obtained by straight-line fitting of the points on the obtained cerebellum, septum pellucidum and cerebellar vermis or thalamus and septum pellucidum.
- the characteristic anatomical structures located on the midsagittal plane include the corpus callosum, cerebellar vermis, septum pellucidum, etc.
- a plane fitting was performed to obtain the midsagittal plane.
- Commonly used structure detection methods include, but are not limited to, Otsu threshold algorithm (OSTU), level set (LevelSet) and other algorithms.
- the symmetry identification is displayed on the display that displays the ultrasound image.
- a symbol of symmetry is highlighted in the ultrasound image, eg, the midline or midsagittal plane of the brain is displayed in a distinct color from the ultrasound image.
- the midline of the brain may be displayed as a red straight line in the ultrasound image while the ultrasound image is displayed on the monitor, or the midsagittal plane may be displayed as a plane marked with red shading.
- FIG. 5 a schematic diagram of displaying a symmetry marker in an ultrasound image of a fetal skull in a display according to an embodiment of the present application is shown.
- the mid-brain line 404 is shown in the ultrasound image 400 of the fetal skull as a straight line identified by a dashed line style.
- Symmetry assessment guide step S130 Acquire an ultrasound image located on at least one side of the symmetry mark, and obtain a symmetry assessment guide of the fetal skull with respect to the symmetry mark based on the ultrasound image of the at least one side.
- Symmetry evaluation guide step S130 using the processor based on the symmetry identification of the ultrasound image to obtain the symmetry evaluation guide for the fetal skull, according to the symmetry evaluation guide, the doctor can qualitatively determine whether the fetal skull is symmetrical with respect to the symmetry identification or not.
- Quantitative evaluation can also guide doctors to identify and diagnose key anatomical structures based on the symmetrical features of the fetal brain.
- the symmetry assessment guide includes a degree of symmetry, and the degree of symmetry can be a qualitative evaluation of whether the fetal brain is symmetrical about the symmetry mark, or it can be a quantitative evaluation of whether the fetal brain is symmetrical about the symmetry mark.
- qualitative evaluation according to the degree of symmetry, the doctor can directly know whether the fetal brain is symmetrical or whether the symmetry is good or not, and judge whether the fetal brain development is normal.
- the doctor obtains an evaluation value for evaluating whether the fetal brain is symmetrical with respect to the symmetry mark. Based on this value, the doctor can judge whether the fetal brain development is normal.
- the method 300 for calculating the degree of symmetry includes:
- Step S310 acquiring an ultrasound image of the first region on the side of the symmetry mark
- Step S320 obtaining an ultrasound image of a second region located on the other side of the symmetry mark according to the ultrasound image of the first region, and the shape of the first region and the shape of the second region are symmetrical with respect to the symmetry mark ;
- Step S330 determining the degree of symmetry of the ultrasound image of the first region and the ultrasound image of the second region with respect to the symmetry identifier.
- the method for acquiring the ultrasound image of the first region on the side of the symmetry marker includes: acquiring the skull of the fetal skull from the ultrasound image , wherein the area surrounded by the skull on one side of the symmetry marker is used as the first area ultrasound image; then, steps S320 and S330 are performed.
- the method for acquiring the skull of the fetal skull in the ultrasound image includes: detecting the skull of the fetal skull based on a grayscale detection or segmentation method, detecting the skull of the fetal skull by a method based on machine learning, Or a method based on deep learning to detect the skull of the fetal brain.
- a skull halo is displayed in the ultrasound image after obtaining the skull of the fetal skull. For example, as shown in FIG.
- the skull 401 of the fetal skull is acquired, and the region surrounded by the skull 401 below the midline 404 of the brain is taken as the first region ultrasound image; the first region above the midline 404 is obtained according to the first region ultrasound image Two-region ultrasound image; finally, determine the degree of symmetry of the first-region ultrasound image and the second-region ultrasound image with respect to the symmetry identification.
- the ultrasound image acquired in the image acquisition step S110 is not clear enough (for example, the image of the fetal skull on the side close to the probe is not clear) so that it is not meaningful to obtain the entire area of the fetal skull to calculate the symmetry degree
- Symmetry evaluation can be performed by identifying at least part of the region in the ultrasound image as the region of interest, or the doctor needs to select at least part of the region in the ultrasound image as the region of interest for symmetry evaluation, and then calculate the symmetry of the region of interest. .
- step S310 a region of interest is acquired in the ultrasound image, the region of interest is at least partially located on one side of the symmetry mark, and the region of interest is located at The area on one side of the symmetry mark is the ultrasound image of the first area; then, step S320 and step S330 are performed.
- the method of obtaining the region of interest can be determined automatically by the processor.
- at least a part of a relatively clear region in the ultrasound image is identified as the region of interest by the processor.
- the processor identifies a region with a sharpness value higher than a certain value in the ultrasound image of the side of the skull near the acquisition end of the probe as the region of interest by identifying the symmetry.
- the region where the characteristic anatomical structure may be located in the ultrasound image is intercepted by the processor as the region of interest.
- the area of the first anatomical structure on one side of the symmetry marker includes an area surrounded by an outline of the first anatomical structure, or The area enclosed by the border surrounding the first anatomical structure.
- step S310 in the method for acquiring the ultrasound image of the first region on the side of the symmetry marker by identifying at least one first anatomical structure in the ultrasound image, when the symmetry marker is far from the ultrasound
- the first anatomical structure is identified in the ultrasonic image on one side of the probe acquisition end, and the region of the first anatomical structure located on the side of the symmetry mark is the ultrasonic image of the first region, so that the contour or the contour of the first anatomical structure or the region of the first anatomical structure can be identified.
- the region where the first anatomical structure is located is accurately and clearly identified, so as to obtain the second region based on the first region, and determine the degree of symmetry based on the first region and the second region.
- the method for determining the degree of symmetry of the ultrasound image of the first region and the ultrasound image of the second region with respect to the symmetry identifier includes: calculating the The absolute difference value and/or the squared difference value of the first region ultrasound image and the second region ultrasound image with respect to the symmetry point identified by the symmetry, based on the absolute grayscale difference value and/or the squared difference value The degree of symmetry is determined.
- the obtained absolute grayscale difference and/or squared difference is directly displayed in the ultrasound image as a numerical value of the degree of symmetry for quantitative evaluation.
- the obtained structural similarity value is scaled with the quantitative evaluation value corresponding to the symmetry degree of 0-10 (or 0-100, 0-100%).
- the obtained structural similarity value is compared with a predetermined structural similarity threshold, and when the structural similarity value is smaller than the structural similarity threshold, a qualitative evaluation of good symmetry is output; otherwise, the output is poor. qualitative evaluation.
- the calculation of the symmetry degree mainly uses the similarity measurement method.
- the L1 and L2 norms use the absolute difference and squared error between pixels to measure the similarity;
- Correlation Ratio measures the similarity by measuring the functional dependency between two vectors;
- the structure is similar Degree (SSIM) measures the similarity of ultrasound images from three aspects: brightness, contrast and structure according to the principle of human visual perception.
- the degree of symmetry may also be calculated by using other methods for measuring similarity, such as peak signal-to-noise ratio (PSNR), mutual information (MI), and cross correlation (Cross Correlation).
- PSNR peak signal-to-noise ratio
- MI mutual information
- Cross Correlation cross correlation
- the doctor often needs to individually evaluate the symmetry of the position of the specific anatomical structure located on both sides of the symmetry mark in the fetal skull.
- identifying the anatomical structure on one side of the symmetry mark in the ultrasound image, and projecting it on the other side of the symmetry mark with respect to the symmetry mark can guide the doctor
- Corresponding anatomical structures are identified on the other side of the symmetry marker, providing the physician with a symmetry assessment guide as to whether the positions of the anatomical structures on both sides of the symmetry marker are symmetrical.
- the symmetry assessment guide includes a projection marker that is symmetrical on the other side of the symmetry marker with respect to the second anatomical structure on one side of the symmetry marker. Due to the intuitive display of ultrasound images, after acquiring the anatomical structure on one side of the symmetry mark, the anatomical structure is projected on the other side of the symmetry mark with respect to the symmetry mark. Whether the third anatomical structure on the other side of the symmetry mark (the third anatomical structure can be recognized by the doctor through the human eye or by the processor machine) is coincident, and directly obtain whether the positions of the anatomical structures on both sides of the symmetry mark are symmetrical. Symmetry evaluation.
- the second anatomical structure and the third anatomical structure are two anatomical structures that are symmetrical with respect to the symmetry sign in craniocerebral physiology, for example, the second anatomical structure is the left cerebellum, and the third anatomical structure is the right cerebellum, etc.
- the doctor observes that the projection mark of the second anatomical structure coincides with the third anatomical structure it can be directly judged that the second anatomical structure and the third anatomical structure are symmetrical with respect to the symmetry mark, so that it can be judged that the symmetry of the fetal skull is good
- the doctor observes that the projection mark of the second anatomical structure does not coincide with the third anatomical structure it can be judged that the second anatomical structure and the third anatomical structure are asymmetrical with respect to the symmetry mark, so that it can be judged that the symmetry of the fetal skull is not good , or, a judgment may be made that further testing is required.
- the above method makes the symmetry evaluation process intuitive and visual.
- Projecting step S8302 obtaining the outline of the second anatomical structure or the projection mark that is symmetrical with respect to the symmetry marker in the region where the second anatomical structure is located;
- Step S8303 is displayed, and the projection mark is displayed.
- machine learning models may include traditional machine learning models and deep learning models.
- the structure detection method based on machine learning first builds a sample learning library; then uses methods such as PCA, LDA, Haar to extract image features, and matches the extracted features with the learning library; finally, using SVM, KNN, PCANet and other methods are used for classification, so as to obtain ROI and determine the anatomical structure category, and realize the detection of key anatomical structures in the far field of the fetal skull.
- the first case can be based on deep learning
- Bounding-Box detection method this method inputs an image and directly returns the Bounding-Box of the ROI through the network.
- the structure category of the ROI can be obtained, so as to realize the detection of key anatomical structures.
- Common networks include RCNN, Fast RCNN, etc.
- the second case is an end-to-end semantic segmentation network method based on deep learning. This method removes the fully connected layer, and makes the size of the input and output images the same through upsampling or deconvolution, and the ROI of the input image can be directly obtained.
- the corresponding categories, common networks are FCN, UNet, etc.
- the third case is to first locate the target and extract the features of the ROI.
- contour shadow marks 4023 and 4033 wherein , the shading patterns of the outline shading mark 4023 and the outline shading mark 4033 are different.
- the symmetry assessment guide includes a second anatomical structure on one side of the symmetry marker and a third anatomical structure on the other side of the symmetry marker with respect to the symmetry marker Symmetry evaluation of whether it is symmetrical, wherein the second anatomical structure and the third anatomical structure are anatomical structure pairs that are physiologically symmetrical with respect to the symmetry identification of the fetal brain.
- the symmetry evaluation is to compare the outline of the second anatomical structure or the region where the second anatomical structure is located with respect to the symmetry marker, and the projection mark that is symmetric with the outline of the third anatomical structure or the third anatomy
- the regions where the structures are located are compared and made, which is a symmetry evaluation for whether the second anatomical structure and the third anatomical structure are symmetrical in position.
- the third anatomical structure located on the other side of the symmetry marker is also identified, and the outline of the third anatomical structure or the region where the third anatomical structure is located is obtained, and the third anatomical structure is The structures are in one-to-one correspondence with each of the at least one second anatomical structure; in the displaying step S8303, the outline of the third anatomical structure or the region where the third anatomical structure is located is also displayed.
- the cerebellum, lateral ventricle, choroid plexus, sylvian fissure and thalamus are all a general anatomical structure name, one of which includes the left and right two parts symmetrical structure, here the first anatomical structure, the second anatomical structure and the third anatomy
- the structure can refer to any part of the left and right symmetrical structures, for example, the first anatomical structure can be the left cerebellum, the second anatomical structure can be the left ventricle, the third anatomical structure can be the right ventricle, etc.;
- An anatomical structure, a second anatomical structure and a third anatomical structure can also refer to the whole of the above-mentioned anatomical structures, for example, the first anatomical structure can be the cerebellum, the second anatomical structure can be the Sylvian fissure, and the third anatomical structure can also be the lateral cerebrum crack.
- FIG. 13 a schematic diagram of an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application is shown. As shown in FIG.
- an ultrasound image 1300 of the fetal skull a skull 1301 of the fetal skull and an anatomical structure located inside the skull 1301 are displayed, and the top of the ultrasound image is a direction close to the ultrasound probe. Due to the poor sound permeability of the bony structure of the skull, the tissue structure near the back of the bone is often not clearly displayed.
- a second embodiment of the present application provides an ultrasonic detection method, which, based on the symmetry of the skull, determines the corresponding anatomical structure in the ultrasonic image on the side close to the ultrasonic probe according to the position of the anatomical structure in the ultrasonic image collected on the side away from the ultrasonic probe where anatomical structures may be present, thus providing guidance to physicians for further testing.
- FIG. 12 shows an ultrasonic inspection method 1200 according to one embodiment of the present application. As shown in FIG. 12, the ultrasonic detection method 1200 may include the following steps:
- the image acquisition step S1210 is to acquire an ultrasound image of the fetal skull.
- an ultrasonic imaging system is used to acquire an ultrasonic image.
- the ultrasound imaging system 200 includes an ultrasound probe 201 and an image processing unit 202 , and the image processing unit includes a processor 2022 , a memory 2021 and a display 2023 .
- the ultrasonic probe 201 is used to transmit ultrasonic waves to the skull of the tested fetus and receive ultrasonic echoes to obtain ultrasonic echo signals.
- the processor 2022 is configured to process the ultrasonic echo signals to obtain an ultrasonic image of the fetal skull, and to examine the fetal skull based on the ultrasonic image.
- the memory 2021 is used to store ultrasound images.
- the display 2023 is used to display ultrasound images.
- the image acquisition step acquires an ultrasound image of the fetal skull, wherein the ultrasound image includes a two-dimensional image of the fetal skull or a three-dimensional image of the fetal skull.
- the two-dimensional image of the fetal skull is obtained by adjusting the fetal position, the position of the ultrasound probe, etc. to obtain the sliced images of the fetal skull, such as the cerebellar slice, the lateral ventricle brain slice, and the thalamus slice.
- Symmetry identification acquisition step S1220 acquiring the symmetry identification of the fetal skull in the ultrasound image.
- the symmetry is identified as the midline of the brain corresponding to the sectional image, and when the ultrasound image is a three-dimensional image of the fetal skull, the symmetry is identified as corresponding to the brain midline. in the midsagittal plane of the 3D image.
- the doctor manually obtains the symmetry identification according to experience.
- the method of obtaining the midline of the brain in the two-dimensional image is to select two points on the midline of the brain during the acquisition of the ultrasound image and mark them to connect the two points.
- the two points form a straight line, which serves as the midline of the brain.
- the method for obtaining the midsagittal plane is to transform the doctor through manual rotation, translation, etc., to obtain three mutually orthogonal cross-sectional images in the three-dimensional image. Obtain the midsagittal plane.
- a processor is used to process the ultrasound image to obtain the symmetry identification.
- the image acquisition step includes acquiring a two-dimensional image of the fetal skull; in the symmetry marker acquisition step, the method for acquiring the midline of the brain includes at least one of the following:
- the characteristic anatomical structure is detected in the two-dimensional image by a structure detection algorithm, the detected characteristic anatomical structure is fitted as a straight line, and the fitted straight line is determined as the midline of the brain;
- the midline of the brain is determined from the two-dimensional images by a pretrained machine learning model.
- the image acquisition step includes: acquiring a three-dimensional image of the fetal skull;
- the method for obtaining the midsagittal plane includes at least one of the following:
- a straight line detection algorithm is used to detect the midline of the brain, or a plane detection algorithm is used to detect the midsagittal plane. Since the fetal skull is symmetrical about the midline or the three-dimensional image of the fetal skull is symmetrical about the midsagittal plane, the midline or midsagittal plane in the ultrasound image tends to have distinct features. Based on this feature, the line detection method is used to detect the midline of the brain in the two-dimensional image, or the plane detection method is used to detect the midsagittal plane of the three-dimensional image.
- the characteristic anatomical structures located on the midline of the brain are detected based on a structure detection algorithm, and the midline of the brain is obtained by performing straight line fitting on the characteristic anatomical structures, or the characteristic anatomical structures located on the midsagittal plane are detected based on the structure detection algorithm, and the characteristic anatomical structures are detected.
- the anatomical structure was plane fitted to obtain the midsagittal plane.
- the special anatomical structures located on the midline of the brain include the cerebellum, septum pellucidum, cerebellar vermis, etc.
- the special anatomical structures located in the midline of the brain include The special anatomical structures on the midline include the thalamus, septum pellucidum, etc.
- the midline of the brain is obtained by straight-line fitting of the points on the obtained cerebellum, septum pellucidum and cerebellar vermis or thalamus and septum pellucidum.
- the characteristic anatomical structures located on the midsagittal plane include the corpus callosum, cerebellar vermis, septum pellucidum, etc.
- a plane fitting was performed to obtain the midsagittal plane.
- Commonly used structure detection methods include, but are not limited to, Otsu threshold algorithm (OSTU), level set (LevelSet) and other algorithms.
- the process of determining the midsagittal plane through the pre-trained machine learning model is similar to the process of determining the midline of the brain through the pre-trained machine learning model.
- the plane equation of the midsagittal plane is obtained as output data. For the specific manner, please refer to the above related description, which will not be repeated here.
- identifying at least one anatomical structure at least partially located on one side of the symmetry mark includes: ultrasonic waves on a side of the symmetry mark far from the acquisition end of the ultrasound probe Identify the anatomical structure in the image.
- at least one anatomical structure in the ultrasound image 1400 on the side away from arrow B is identified symmetrically, ie, anatomical structures located in the area below the midline 1404 of the brain, such as anatomical structures 1402 and 1403 .
- the ultrasonic images are mostly fan-shaped images, in which the ultrasonic image has a smaller arc at the probe acquisition end, and a larger arc away from the probe acquisition end.
- the two sides of the ultrasound image that is, the left and right hypotenuses of the fan-shaped ultrasound image in Figure 13
- the end of the ultrasound image close to this point is the ultrasound probe
- the side of the symmetry mark close to the point is the side of the symmetry mark close to the acquisition end of the ultrasound probe
- the side of the symmetry mark away from this point is the side of the symmetry mark away from the acquisition end of the ultrasound probe.
- the probe acquisition end of the ultrasonic image can also be determined by the related file generated together with the ultrasonic image.
- various data for generating the ultrasonic image are stored, including the ultrasonic probe.
- the position information in the file can be directly retrieved, the probe acquisition end of the ultrasonic image can be determined, and the ultrasonic image with the symmetry mark on the side away from the ultrasonic probe acquisition end can be further determined, so as to measure the ultrasonic image. Images for automatic identification of key anatomical structures.
- the method for identifying at least one anatomical structure includes, but is not limited to, obtaining at least one contour of the second anatomical structure by segmenting the ultrasound image through an image segmentation algorithm; The contour of the second anatomical structure or the region where the second anatomical structure is located.
- the image segmentation algorithm includes detecting anatomical structures through grayscale detection or segmentation methods, and machine learning includes detecting anatomical structures based on traditional machine learning methods and methods based on deep learning.
- Exemplary, grayscale detection or segmentation methods include, but are not limited to, Otsu threshold algorithm (OSTU), level set method (Level Set), active contour model (Snake), Graph Cut (Graph Cut), Region Growing (Region Growing) and other methods.
- Machine learning models include: SVM, KNN, PCANet, etc.
- Deep learning networks include: RCNN, Fast RCNN, FCN, UNet, etc.
- the method for obtaining the outline of the anatomical structure or the region where the anatomical structure is located includes, but is not limited to, extracting the boundary of the anatomical structure or returning the anatomical structure containing the anatomical structure. ROI area, etc.
- Projecting step S1240 acquiring the outline of the anatomical structure or a projection mark that is symmetrical on the other side of the symmetry marker with respect to the symmetry marker in the region where the anatomical structure is located.
- the processor is used to project the anatomical structure in the ultrasound image on one side of the symmetry mark identified in the identification step 1230 to the other side of the symmetry mark.
- the display step S1250 is to display the projection mark.
- the projection marker obtained in the projection step S1240 is displayed on the ultrasound image using a display.
- the contour marks include contour boundary marks, contour shadow marks and/or contour arrow marks; the area marks include area border marks, area shadow marks and/or area arrow marks.
- Figures 15 and 16 illustrate a method of ultrasound inspection according to one embodiment of the present application, the method of ultrasound inspection comprising a schematic diagram showing projection markers in an ultrasound image of a fetal skull. As shown in FIG. 15 , corresponding to the outline of anatomy 1402 , the projected marker of anatomy 1402 on the other side of midline 1404 is shown as outline boundary marker 14021 , corresponding to the outline of anatomy 1403 , which is on midline 1404 The projected markers on the other side of the are shown as outline arrow markers 14031.
- contour shadow mark 14023 and contour shadow mark 14033 the projection marks of the contours of anatomical structure 1402 and anatomical structure 1403 on the other side of the midline of the brain 1404 are shown as contour shadow mark 14023 and contour shadow mark 14033 , wherein, The shading patterns of the outline shading mark 14023 and the outline shading mark 14033 are different.
- the doctor may continue to perform ultrasonic detection on the area covered by the projection marker according to the projection marker.
- the ultrasonic probe is used to transmit ultrasonic waves to the brain of the tested fetus and receive ultrasonic echoes to obtain ultrasonic echo signals;
- the ultrasound imaging system 200 includes an ultrasound probe 201 and an image processing unit 202 , and the image processing unit includes a processor 2022 , a memory 2021 and a display 2023 .
- the ultrasonic probe 201 is used to transmit ultrasonic waves to the skull of the tested fetus and receive ultrasonic echoes to obtain ultrasonic echo signals.
- the processor 2022 is configured to process the ultrasonic echo signals to obtain an ultrasonic image of the fetal skull, and perform ultrasonic detection on the fetal skull based on the ultrasonic image.
- the memory 2021 is used to store ultrasound images.
- the display 2023 is used to display ultrasound images.
- the memory 2021 is further configured to store executable program instructions, and the processor 2022 executes the methods of the embodiments of the present application based on the executable program instructions.
- the disclosed apparatus and method may be implemented in other manners.
- the device embodiments described above are only illustrative.
- the division of the units is only a logical function division. In actual implementation, there may be other division methods.
- multiple units or components may be combined or May be integrated into another device, or some features may be omitted, or not implemented.
Landscapes
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Pathology (AREA)
- Radiology & Medical Imaging (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Ultra Sonic Daignosis Equipment (AREA)
Abstract
An ultrasonic detection method and an ultrasonic imaging system. The ultrasonic detection method comprises: an image acquisition step (S110) of acquiring an ultrasound image (400, 1400) of a fetal brain; a symmetry identifier acquisition step (S120) of acquiring a symmetry identifier of the fetal brain on the basis of the ultrasound image (400, 1400); and a symmetry evaluation guiding step (S130) of acquiring an ultrasound image from at least one side of the symmetry identifier, and acquiring a symmetry evaluation guidance for the symmetry identifier of the fetal brain on the basis of the ultrasound image from the at least one side. According to the acquired ultrasound image (400, 1400) of the fetal brain, the symmetry evaluation guidance for the fetal brain is provided to a doctor, so as to help the doctor to evaluate the symmetry of the fetal brain.
Description
说明书manual
本申请涉及医疗器械领域,更具体地涉及一种超声检测方法和超声成像***。The present application relates to the field of medical devices, and more particularly, to an ultrasonic detection method and an ultrasonic imaging system.
超声作为现代医学影像技术中使用最广的检查手段,由于其使用安全、成本低廉、成像快速、可靠性高等优点被广泛应用于人类疾病的诊断中。超声辅助诊断技术的发展,进一步推动了超声在临床诊断中的实际应用,不仅扩大了超声技术的用途范围,更可以帮助医生提高检查精度以及效率,为患者提供了更加全面、个性化的治疗方案。Ultrasound, as the most widely used examination method in modern medical imaging technology, is widely used in the diagnosis of human diseases due to its advantages of safe use, low cost, rapid imaging, and high reliability. The development of ultrasound-assisted diagnosis technology has further promoted the practical application of ultrasound in clinical diagnosis. It not only expands the application scope of ultrasound technology, but also helps doctors improve the accuracy and efficiency of examinations, and provides patients with more comprehensive and personalized treatment plans. .
近年来,随着二胎政策的开放,新生儿的数量渐渐增长,为了保证胎儿的健康成长,早期检测就显得尤为重要。借助于超声技术,医生可以获取到胎儿颅脑的数据并作显示。由于颅脑是一种高度对称的结构,对称性是判断胎儿颅脑是否正常的一项重要指标,对异常胎儿的诊断具有重要的作用。另外,由于骨性结构的透声性较差,靠近骨头后方组织结构往往显示不清,使得颅脑图像近场(靠近超声探头一侧)信息丢失严重,在胎儿颅脑近场关键解剖结构检查时,医生只能通过临床经验并借助颅脑图像的远场(远离超声探头一侧)信息进行诊断,有很大的局限性和弊端,易造成误诊。In recent years, with the opening of the two-child policy, the number of newborns has gradually increased. In order to ensure the healthy growth of the fetus, early detection is particularly important. With the help of ultrasound technology, doctors can obtain and display fetal brain data. Since the brain is a highly symmetrical structure, symmetry is an important indicator to judge whether the fetal brain is normal, and it plays an important role in the diagnosis of abnormal fetuses. In addition, due to the poor sound permeability of the bony structure, the tissue structure near the back of the bone is often not clearly displayed, resulting in serious loss of information in the near-field (closer to the ultrasound probe) of the cranial image. At that time, doctors can only make a diagnosis through clinical experience and with the help of the far-field (far away from the ultrasound probe) information of the cranial image, which has great limitations and drawbacks, and is easy to cause misdiagnosis.
发明内容SUMMARY OF THE INVENTION
为了解决上述问题中的至少一个而提出了本申请。本申请提供一种超声检测方法和超声成像***,其可以在医生采集完胎儿颅脑的超声图像之后,帮助医生获得关于胎儿颅脑的对称性评估指引,为医生评估胎儿颅脑的对称性提供参考,或者基于对称性评估指引进行进一步的诊断。The present application has been made to solve at least one of the above-mentioned problems. The present application provides an ultrasonic detection method and an ultrasonic imaging system, which can help the doctor obtain the symmetry assessment guide of the fetal skull after the doctor has collected the ultrasonic image of the fetal skull, and provide the doctor with the ability to evaluate the symmetry of the fetal skull. Reference, or further diagnosis based on symmetry assessment guidelines.
第一方面,本申请的实施例提供一种超声检测方法,包括:In a first aspect, the embodiments of the present application provide an ultrasonic detection method, comprising:
图像获取步骤,获取胎儿颅脑的超声图像;The image acquisition step is to acquire an ultrasound image of the fetal skull;
对称性标识获取步骤,基于所述超声图像获取所述胎儿颅脑的对称性标识;The step of obtaining a symmetry mark, obtaining a symmetry mark of the fetal skull based on the ultrasound image;
对称性评估指引步骤,获取位于所述对称性标识至少一侧的超声图像,基于所述至少一侧的超声图像获取所述胎儿颅脑关于所述对称性标识的对称性评估指引。The symmetry assessment guide step is to acquire an ultrasound image located on at least one side of the symmetry mark, and obtain a symmetry assessment guide of the fetal skull with respect to the symmetry mark based on the ultrasound image of the at least one side.
第二方面,本申请的实施例还提供一种超声检测方法,包括:In a second aspect, the embodiments of the present application also provide an ultrasonic detection method, comprising:
图像获取步骤,获取关于胎儿颅脑的超声图像;The image acquisition step is to acquire an ultrasound image of the fetal skull;
对称性标识获取步骤,获取所述超声图像中所述胎儿颅脑的对称性标识;The step of obtaining a symmetry mark is to obtain the symmetry mark of the fetal skull in the ultrasound image;
识别步骤,在所述超声图像上识别至少一个至少部分位于所述对称性标识的一侧的解剖结构,获取所述解剖结构的轮廓或所述解剖结构所在的区域;Identifying, on the ultrasound image, identifying at least one anatomical structure at least partially located on one side of the symmetry mark, and acquiring an outline of the anatomical structure or a region where the anatomical structure is located;
投影步骤,获取所述解剖结构的轮廓或所述解剖结构所在的区域关于所述对称性标识在所述对称性标识另一侧对称的投影标记;The projection step is to obtain the outline of the anatomical structure or a projection mark that is symmetrical on the other side of the symmetry mark with respect to the symmetry mark in the region where the anatomical structure is located;
显示步骤,显示所述投影标记。A display step displays the projected mark.
第三方面,本申请的实施例还提供一种超声成像***包括:In a third aspect, embodiments of the present application also provide an ultrasound imaging system comprising:
超声探头,用于受测胎儿颅脑发射超声波并接收超声回波,得到超声回波信号;Ultrasound probe, used to transmit ultrasound waves and receive ultrasound echoes from the cranial brain of the tested fetus to obtain ultrasound echo signals;
处理器,用于对超声回波信号进行处理,得到所述生物体组织的超声图像;a processor, configured to process the ultrasonic echo signal to obtain an ultrasonic image of the biological tissue;
显示器,用于显示所述超声图像;a display for displaying the ultrasound image;
存储器,用于存储可执行的程序指令;memory for storing executable program instructions;
所述处理器,还用于基于所述可执行的程序指令执行上述第一方面或第二方面所述的超声检测方法。The processor is further configured to execute the ultrasonic detection method according to the first aspect or the second aspect based on the executable program instructions.
本申请的实施例提供了一种解剖结构的超声检测方法、超声成像***,基于采集的关于胎儿颅脑的超声图像,获得关于胎儿颅脑的对称性评估指引,为评估胎儿颅脑的对称性提供参考,或者基于对称性评估指引进行进一步的诊断。The embodiments of the present application provide an ultrasonic detection method and an ultrasonic imaging system for an anatomical structure. Based on the acquired ultrasonic images of the fetal skull, a guideline for evaluating the symmetry of the fetal skull is obtained, so as to evaluate the symmetry of the fetal skull. Provide a reference, or proceed with further diagnosis based on the symmetry assessment guidelines.
图1示出根据本申请的一个实施例的超声检测方法的示意性流程图FIG. 1 shows a schematic flowchart of an ultrasonic detection method according to an embodiment of the present application
图2示出根据本申请的一个实施例的超声成像***的示意性框图;FIG. 2 shows a schematic block diagram of an ultrasound imaging system according to an embodiment of the present application;
图3示出根据本申请的一个实施例的超声检测方法中计算对称度的方法的示意性流程图;3 shows a schematic flowchart of a method for calculating symmetry in an ultrasonic testing method according to an embodiment of the present application;
图4示出根据本申请的一个实施例的超声检测方法中在显示器中显示的胎儿颅脑的超声图像的示意图;4 shows a schematic diagram of an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application;
图5示出根据本申请的一个实施例的超声检测方法中在显示器中将对称性标识显示在胎儿颅脑的超声图像中的示意图;5 shows a schematic diagram of displaying a symmetry mark in an ultrasound image of a fetal skull on a display in an ultrasound detection method according to an embodiment of the present application;
图6示出根据本申请的一个实施例的超声检测方法中在显示器中显示的胎儿颅脑的超声图像中获取感兴趣区域的示意图;6 shows a schematic diagram of acquiring a region of interest in an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application;
图7示出根据本申请的一个实施例的超声检测方法中在显示器中显示的胎儿颅脑的超声图像中获取解剖结构的示意图;7 shows a schematic diagram of acquiring an anatomical structure in an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application;
图8示出根据本申请的一个实施例的超声检测方法中获取对称性评估指引的方法的示意性流程图;8 shows a schematic flowchart of a method for obtaining a symmetry assessment guide in an ultrasonic inspection method according to an embodiment of the present application;
图9示出根据本申请的一个实施例的超声检测方法中在显示器中显示的胎儿颅脑的超声图像中显示投影标记的示意图;9 shows a schematic diagram of displaying a projection marker in an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application;
图10示出根据本申请的一个实施例的超声检测方法中在显示器中显示的胎儿颅脑的超声图像中显示投影标记的示意图;10 shows a schematic diagram of displaying a projection marker in an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application;
图11示出根据本申请的一个实施例的超声检测方法中在显示器中显示的胎儿颅脑的超声图像中区分显示投影标记的示意图;11 shows a schematic diagram of distinguishing and displaying projection markers in the ultrasound image of the fetal skull displayed on the display in the ultrasound detection method according to an embodiment of the present application;
图12示出根据本申请的一个实施例的超声检测方法的示意性流程图;FIG. 12 shows a schematic flowchart of an ultrasonic detection method according to an embodiment of the present application;
图13示出根据本申请的一个实施例的超声检测方法中在显示器中显示的胎儿颅脑的超声图像的示意图;13 shows a schematic diagram of an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application;
图14示出根据本申请的一个实施例的超声检测方法中在显示器中将对称性标识显示在胎儿颅脑的超声图像中的示意图;14 shows a schematic diagram of displaying a symmetry mark in an ultrasound image of a fetal skull on a display in an ultrasound detection method according to an embodiment of the present application;
图15示出根据本申请的一个实施例的超声检测方法中在显示器中显示的胎儿颅脑的超声图像中显示投影标记的示意图;15 shows a schematic diagram of displaying projection markers in the ultrasound image of the fetal skull displayed on the display in the ultrasound detection method according to an embodiment of the present application;
图16示出根据本申请的一个实施例的超声检测方法中在显示器中显 示的胎儿颅脑的超声图像中显示投影标记的示意图;16 shows a schematic diagram of displaying a projection marker in an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application;
图17示出根据本申请的一个实施例的超声检测方法中在显示器中显示的胎儿颅脑的超声图像中区分显示投影标记的示意图。FIG. 17 shows a schematic diagram of distinguishing and displaying projection markers in the ultrasound image of the fetal skull displayed on the display in the ultrasound detection method according to an embodiment of the present application.
为了使得本申请的目的、技术方案和优点更为明显,下面将参照附图详细描述根据本申请的示例实施例。显然,所描述的实施例仅仅是本申请的一部分实施例,而不是本申请的全部实施例,应理解,本申请不受这里描述的示例实施例的限制。基于本申请中描述的本申请实施例,本领域技术人员在没有付出创造性劳动的情况下所得到的所有其它实施例都应落入本申请的保护范围之内。In order to make the objectives, technical solutions and advantages of the present application more apparent, the exemplary embodiments according to the present application will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein. Based on the embodiments of the present application described in the present application, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the protection scope of the present application.
在下文的描述中,给出了大量具体的细节以便提供对本申请更为彻底的理解。然而,对于本领域技术人员而言显而易见的是,本申请可以无需一个或多个这些细节而得以实施。在其他的例子中,为了避免与本申请发生混淆,对于本领域公知的一些技术特征未进行描述。In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the present application. It will be apparent, however, to one skilled in the art that the present application may be practiced without one or more of these details. In other instances, some technical features known in the art have not been described in order to avoid confusion with the present application.
应当理解的是,本申请能够以不同形式实施,而不应当解释为局限于这里提出的实施例。相反地,提供这些实施例将使公开彻底和完全,并且将本申请的范围完全地传递给本领域技术人员。It should be understood that the application may be embodied in different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of this application to those skilled in the art.
在此使用的术语的目的仅在于描述具体实施例并且不作为本申请的限制。在此使用时,单数形式的“一”、“一个”和“所述/该”也意图包括复数形式,除非上下文清楚指出另外的方式。还应明白术语“组成”和/或“包括”,当在该说明书中使用时,确定所述特征、整数、步骤、操作、元件和/或部件的存在,但不排除一个或更多其它的特征、整数、步骤、操作、元件、部件和/或组的存在或添加。在此使用时,术语“和/或”包括相关所列项目的任何及所有组合。The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used herein, the singular forms "a," "an," and "the/the" are intended to include the plural forms as well, unless the context clearly dictates otherwise. It should also be understood that the terms "compose" and/or "include", when used in this specification, identify the presence of stated features, integers, steps, operations, elements and/or components, but do not exclude one or more other The presence or addition of features, integers, steps, operations, elements, parts and/or groups. As used herein, the term "and/or" includes any and all combinations of the associated listed items.
为了彻底理解本申请,将在下列的描述中提出详细的步骤以及详细的结构,以便阐释本申请提出的技术方案,然而除了这些详细描述外,本申请还可以具有其他实施方式。For a thorough understanding of the present application, detailed steps and detailed structures will be presented in the following description to explain the technical solutions proposed by the present application, but in addition to these detailed descriptions, the present application may also have other embodiments.
由于颅脑是一种高度对称的结构,对称性是判断胎儿颅脑是否正常的 一项重要指标,对异常胎儿的诊断具有重要的作用。据此,本申请的第一个实施例提供了一种超声检测方法,用于在医生采集完胎儿颅脑的超声图像之后,帮助医生获得关于胎儿颅脑的对称性评估指引,为医生评估胎儿颅脑的对称性提供参考。Since the brain is a highly symmetrical structure, symmetry is an important indicator to judge whether the fetal brain is normal, and it plays an important role in the diagnosis of abnormal fetuses. Accordingly, the first embodiment of the present application provides an ultrasonic detection method, which is used to help the doctor obtain a symmetry assessment guide about the fetal skull after the doctor collects the ultrasonic image of the fetal skull, so as to help the doctor to evaluate the fetal skull. Symmetry of the brain provides a reference.
图1示出了根据本申请的一个实施例的超声检测方法100。如图1所示,超声检测方法100可以包括如下步骤:FIG. 1 shows an ultrasonic inspection method 100 according to one embodiment of the present application. As shown in FIG. 1, the ultrasonic detection method 100 may include the following steps:
图像获取步骤S110,获取胎儿颅脑的超声图像。The image acquisition step S110 is to acquire an ultrasound image of the fetal skull.
图像获取步骤S110中,采用超声成像***采集超声图像。参看图2示出了根据一个实施例的超声成像***的示意性框图。其中,超声成像***200包括超声探头201和图像处理单元202,图像处理单元包括处理器2022、存储器2021以及显示器2023。超声探头201用于向受测胎儿颅脑发射超声波并接收超声回波,得到超声回波信号。处理器2022用于对超声回波信号进行处理,得到关于胎儿颅脑的超声图像,并基于该超声图像对胎儿颅脑进行超声检测。存储器2021用于存储超声图像。显示器2023用于显示超声图像。In the image acquisition step S110, an ultrasonic imaging system is used to acquire an ultrasonic image. Referring to FIG. 2, a schematic block diagram of an ultrasound imaging system according to one embodiment is shown. The ultrasound imaging system 200 includes an ultrasound probe 201 and an image processing unit 202 , and the image processing unit includes a processor 2022 , a memory 2021 and a display 2023 . The ultrasonic probe 201 is used to transmit ultrasonic waves to the skull of the tested fetus and receive ultrasonic echoes to obtain ultrasonic echo signals. The processor 2022 is configured to process the ultrasonic echo signals to obtain an ultrasonic image of the fetal skull, and perform ultrasonic detection on the fetal skull based on the ultrasonic image. The memory 2021 is used to store ultrasound images. The display 2023 is used to display ultrasound images.
根据本申请的实施例中,图像获取步骤获取胎儿颅脑的超声图像,其中超声图像包括胎儿颅脑的二维图像,或者胎儿颅脑的三维图像。胎儿颅脑的二维图像通过调整胎儿***、超声探头的位置等获得通过的胎儿颅脑的切面图像,例如小脑切面、侧室脑切面、丘脑切面等。参看图4,示出了根据本申请的一个实施例的超声检测方法中在显示器中显示的胎儿颅脑的超声图像的示意图。如图4所示,在胎儿颅脑的超声图像400中显示有胎儿颅脑的颅骨401和位于颅骨401内侧的解剖结构,其中包括:解剖结构402与解剖结构403,其中,箭头A显示超声探头发射超声波的方向。According to the embodiment of the present application, the image acquisition step acquires an ultrasound image of the fetal skull, wherein the ultrasound image includes a two-dimensional image of the fetal skull or a three-dimensional image of the fetal skull. The two-dimensional image of the fetal skull is obtained by adjusting the fetal position, the position of the ultrasound probe, etc. to obtain the sliced images of the fetal skull, such as the cerebellar slice, the lateral ventricle brain slice, and the thalamus slice. Referring to FIG. 4 , a schematic diagram of an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application is shown. As shown in FIG. 4 , in the ultrasound image 400 of the fetal skull, the skull 401 of the fetal skull and the anatomical structures located inside the skull 401 are displayed, including: the anatomical structure 402 and the anatomical structure 403 , wherein the arrow A shows the ultrasound probe The direction in which ultrasonic waves are emitted.
对称性标识获取步骤S120,基于所述超声图像获取所述胎儿颅脑的对称性标识。In the step S120 of obtaining the symmetry identification, the symmetry identification of the fetal skull is obtained based on the ultrasound image.
对称性标识获取步骤S120中,当超声图像为胎儿颅脑的二维图像时,对称性标识为对应于切面图像的脑中线,当超声图像为胎儿颅脑的三维图像时,对称性标识为对应于三维图像的正中矢状面。In the step S120 of obtaining the symmetry identification, when the ultrasound image is a two-dimensional image of the fetal skull, the symmetry is identified as the midline of the brain corresponding to the cross-sectional image, and when the ultrasound image is a three-dimensional image of the fetal skull, the symmetry is identified as corresponding to the brain midline. in the midsagittal plane of the 3D image.
在本申请的一个实施例中,在对称性标识获取步骤S120中,医生根据经验手动获取所述对称性标识。在一个示例中,当超声图像为胎儿颅脑的 二维图像时,获得二维图像的脑中线的方法采用通过在获取超声图像选取位于脑中线上的两个点并进行标记,连接这两个点形成直线,该直线作为脑中线。在另一个示例中,当超声图像为胎儿颅脑的三维图像时,获取正中矢状面的方法采用将医生通过手动旋转、平移等集合变换,从三维图像中三个互相正交的剖面图像中得到正中矢状面。In an embodiment of the present application, in the step S120 of obtaining the symmetry identification, the doctor manually obtains the symmetry identification according to experience. In one example, when the ultrasound image is a two-dimensional image of the fetal skull, the method of obtaining the midline of the brain of the two-dimensional image adopts the method of connecting two points on the midline of the brain by selecting and marking the two points on the midline of the brain when acquiring the ultrasound image. The points form a straight line, which serves as the midline of the brain. In another example, when the ultrasound image is a three-dimensional image of the fetal skull, the method for obtaining the midsagittal plane is to transform the doctor through manual rotation, translation, etc., to obtain three mutually orthogonal cross-sectional images in the three-dimensional image. Obtain the midsagittal plane.
在本申请的一个实施例中,在对称性标识获取步骤S120中,采用处理器处理所述超声图像获得对称性标识。In an embodiment of the present application, in the step S120 of obtaining the symmetry identification, a processor is used to process the ultrasound image to obtain the symmetry identification.
在一个示例中,所述图像获取步骤包括获取胎儿颅脑的二维图像;在所述对称性标识获取步骤中,获取所述脑中线的方法包括以下至少一项:In one example, the image acquisition step includes acquiring a two-dimensional image of the fetal skull; in the symmetry marker acquisition step, the method for acquiring the midline of the brain includes at least one of the following:
通过直线检测算法在所述二维图像中确定脑中线;determining the midline of the brain in the two-dimensional image by a line detection algorithm;
通过结构检测算法在所述二维图像中检测特征解剖结构,将检测到的特征解剖结构拟合为直线,将拟合的所述直线确定为脑中线;The characteristic anatomical structure is detected in the two-dimensional image by a structure detection algorithm, the detected characteristic anatomical structure is fitted as a straight line, and the fitted straight line is determined as the midline of the brain;
通过预训练的机器学习模型从所述二维图像中确定脑中线。The midline of the brain is determined from the two-dimensional images by a pretrained machine learning model.
在一个示例中,所述图像获取步骤包括:获取胎儿颅脑的三维图像;In one example, the image acquisition step includes: acquiring a three-dimensional image of the fetal skull;
在所述对称性标识获取步骤中,获取所述正中矢状面的方法包括以下至少一项:In the step of obtaining the symmetry mark, the method for obtaining the midsagittal plane includes at least one of the following:
通过面检测算法在所述三维图像中确定正中矢状面;determining a midsagittal plane in the three-dimensional image by a plane detection algorithm;
通过结构检测算法所述三维图像中检测特征解剖结构,将检测到的特征解剖结构拟合为平面,将所述拟合的所述平面确定为正中矢状面;The characteristic anatomical structure is detected in the three-dimensional image by the structure detection algorithm, the detected characteristic anatomical structure is fitted as a plane, and the fitted plane is determined as the midsagittal plane;
通过预训练的机器学习模型从所述三维图像中确定正中矢状面。The midsagittal plane is determined from the three-dimensional image by a pretrained machine learning model.
示例性的,基于超声图像的图像特征,采用直线检测算法检测所述脑中线,或者采用平面检测算法检测所述正中矢状面。由于胎儿颅脑关于脑中线对称或胎儿颅脑的三维图像关于正中矢状面对称,在超声图像中的脑中线或正中矢状面往往具有明显的特征。基于该特征,采用直线检测方法检测二维图像的脑中线或者采用平面检测的方法检测三维图像的正中矢状面。例如,在获取脑中线的过程中,提取代表满足线上的灰度值大于两侧的灰度值的线的脑中线特征区域,根据提取的特征区域上的点进行直线检测;同样在获取正中矢状面的过程中,可以提取代表满足平面上的灰度值大于两侧的灰度值的面的正中矢状面特征区域,再根据平面检测算法获得正中矢状面。直线检测算法包括但不限于,Hough变换、LSD快速直线 检测算法、EDlines直线检测算法、Radon变换、LSWMS直线检测算法、CannyLines直线检测算法等。平面检测算法包括但不限于,加权Hough变换、随机Hough变换、Radon变换等。Exemplarily, based on the image features of the ultrasound image, a straight line detection algorithm is used to detect the midline of the brain, or a plane detection algorithm is used to detect the midsagittal plane. Since the fetal skull is symmetrical about the midline or the three-dimensional image of the fetal skull is symmetrical about the midsagittal plane, the midline or midsagittal plane in the ultrasound image tends to have distinct features. Based on this feature, the line detection method is used to detect the midline of the brain in the two-dimensional image, or the plane detection method is used to detect the midsagittal plane of the three-dimensional image. For example, in the process of acquiring the midline of the brain, extract the midline feature area representing the line that satisfies the gray value of the line is greater than the gray value of the two sides, and perform straight line detection according to the points on the extracted feature area; also in the acquisition center In the process of sagittal plane, the midsagittal plane feature area representing the plane whose gray value is greater than the gray value on both sides can be extracted, and then the midsagittal plane can be obtained according to the plane detection algorithm. Line detection algorithms include, but are not limited to, Hough transform, LSD fast line detection algorithm, EDlines line detection algorithm, Radon transform, LSWMS line detection algorithm, CannyLines line detection algorithm, etc. Plane detection algorithms include, but are not limited to, weighted Hough transform, random Hough transform, Radon transform, and the like.
示例性的,基于结构检测算法检测位于脑中线的特征解剖结构,对特征解剖结构进行直线拟合得到脑中线,或基于结构检测算法检测位于正中矢状面上的特征解剖结构,并对进行特征解剖结构进行平面拟合得到正中矢状面。例如,当获取的超声图像是胎儿颅脑的小脑切面时,位于脑中线上的特殊解剖结构包括小脑、透明隔腔、小脑蚓部等,当获取的超声图像是丘脑的切面图像时,位于脑中线上的特殊解剖结构包括丘脑、透明隔腔等,根据获得的小脑、透明隔腔和小脑蚓部或丘脑和透明隔腔对其上的点进行直线拟合得到脑中线。当获取的超声图像是胎儿颅脑的三维图像时,位于正中矢状面上的特征解剖结构包括胼胝体、小脑蚓部、透明隔腔等,通过对胼胝体、小脑蚓部、透明隔腔上的点进行平面拟合得到正中矢状面。常用的结构检测方法有包括但不限于,大津阈值算法(OSTU)、水平集(LevelSet)等算法。Exemplarily, the characteristic anatomical structures located on the midline of the brain are detected based on a structure detection algorithm, and the midline of the brain is obtained by performing straight line fitting on the characteristic anatomical structures, or the characteristic anatomical structures located on the midsagittal plane are detected based on the structure detection algorithm, and the characteristic anatomical structures are detected. The anatomical structure was plane fitted to obtain the midsagittal plane. For example, when the acquired ultrasound image is the cerebellum section of the fetal skull, the special anatomical structures located on the midline of the brain include the cerebellum, septum pellucidum, cerebellar vermis, etc. When the acquired ultrasound image is the section image of the thalamus, the special anatomical structures located in the midline of the brain include The special anatomical structures on the midline include the thalamus, septum pellucidum, etc. The midline of the brain is obtained by straight-line fitting of the points on the obtained cerebellum, septum pellucidum and cerebellar vermis or thalamus and septum pellucidum. When the acquired ultrasound image is a three-dimensional image of the fetal skull, the characteristic anatomical structures located on the midsagittal plane include the corpus callosum, cerebellar vermis, septum pellucidum, etc. A plane fitting was performed to obtain the midsagittal plane. Commonly used structure detection methods include, but are not limited to, Otsu threshold algorithm (OSTU), level set (LevelSet) and other algorithms.
示例性的,通过预训练的机器学习模型确定脑中线或正中矢状面。例如,通过预训练的机器学习确定脑中线的方法包括将所述超声图像作为输入数据输入至预训练的机器学习模型中进行计算以获得脑中线的直线方程作为输出数据,其中,预训练的机器学习模型采用包括至少一个已经标定了脑中线的直线方程的标定超声图像的预设数据库进行训练,在所述训练过程中,标定超声图像作为输入数据,所述标定超声图像的脑中线的直线方程作为输出数据,并且在训练机器学习模型的过程中对机器学习模型的模型参数进行优化。Exemplarily, the midline or midsagittal plane of the brain is determined by a pretrained machine learning model. For example, the method for determining the midline of the brain through pre-trained machine learning includes inputting the ultrasound image as input data into a pre-trained machine learning model for calculation to obtain a straight line equation of the midline of the brain as output data, wherein the pre-trained machine The learning model is trained using a preset database of calibrated ultrasound images including at least one linear equation of the midline of the brain that has been calibrated. As the output data, and the model parameters of the machine learning model are optimized in the process of training the machine learning model.
以在二维图像中确定脑中线为例,训练机器学习模型以得到预训练的机器学习模型的步骤可以包括:1、构建训练样本库步骤:在极坐标系中,假设一条直线过点P(r,α),与极轴夹角为β,则该直线方程可表示为ρsin(β-θ)=rsin(β-α)。由此,可以构建颅脑切面与直线方程参数(r,α,β)一一对应的训练样本。其中,颅脑切面为输入数据,直线方程参数为输出的真值。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,回归拟合直线方程参数。网络的设计主要包括卷积层、池化层、激励层、 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一幅图像,输入所设计的网络模型,输出直线方程参数的估计值,利用估计值、真值以及所设计的损失函数对模型参数进行优化,从而训练模型,得到预训练的机器学习模型。常用的回归网络模型有AlexNet、VGGNet、GoogLeNet等,该方法包含但不仅限于上述网络模型结构。Taking the determination of the midline of the brain in a two-dimensional image as an example, the steps of training a machine learning model to obtain a pre-trained machine learning model may include: 1. Steps of constructing a training sample library: in the polar coordinate system, suppose a straight line passes through the point P ( r, α), and the included angle with the polar axis is β, then the equation of the straight line can be expressed as ρsin(β-θ)=rsin(β-α). Thus, the training samples corresponding to the craniocerebral section and the linear equation parameters (r, α, β) one-to-one can be constructed. Among them, the craniocerebral section is the input data, and the parameters of the straight line equation are the true values of the output. 2. Network design and training steps: After building the training sample library, design a network model to regress and fit the parameters of the linear equation. The design of the network mainly includes convolutional layers, pooling layers, excitation layers, fully connected layers, and the set loss function. The features in the training samples are learned by combining and stacking these layers. For any image in the training sample library, input the designed network model, output the estimated value of the parameters of the linear equation, and use the estimated value, the true value and the designed loss function to optimize the model parameters, so as to train the model and obtain the predicted value. Trained machine learning model. Commonly used regression network models include AlexNet, VGGNet, GoogLeNet, etc. This method includes but is not limited to the above network model structure.
在推理阶段,将待直线检测的颅脑切片图像输入至上述预训练的机器学习模型中,可实时得到直线方程参数,从而实现了二维图像中脑中线的检测。In the inference stage, the brain slice image to be detected in a straight line is input into the above-mentioned pre-trained machine learning model, and the parameters of the straight line equation can be obtained in real time, thereby realizing the detection of the midbrain in the two-dimensional image.
在三维图像中检测正中矢状面的预训练的机器学***面方程参数,通过优化模型参数从而训练网络。在推理阶段,输入三维图像可实时得到对应平面方程参数,从而实现正中矢状面的检测。The pre-trained machine learning model for detecting the midsagittal plane in 3D images is similar to the pre-training machine learning model for detecting the midline of the brain in 2D images. The training process can be as follows: first build a training sample library, then design a regression The network model, the input of the model is the three-dimensional data of the brain, the output is the plane equation parameters, and the network is trained by optimizing the model parameters. In the inference stage, the input 3D image can obtain the corresponding plane equation parameters in real time, so as to realize the detection of the midsagittal plane.
在本申请的一个实施例中,获取对称性标识之后,在显示超声图像的显示器中显示对称性标识。例如,在一个示例中,在超声图像中突出显示对称性标识,例如采用明显区别于超声图像的颜色显示脑中线或正中矢状面。例如,在显示器显示超声图像的同时,在超声图像中将脑中线显示为红色直线,或者,将正中矢状面显示为红色阴影标记的平面。参看图5,示出了根据本申请的一个实施例中在显示器中将对称性标识显示在胎儿颅脑的超声图像中的示意图。如图5所示,在胎儿颅脑的超声图像400中将脑中线404显示为以虚线线型标识的直线。In an embodiment of the present application, after the symmetry identification is acquired, the symmetry identification is displayed on the display that displays the ultrasound image. For example, in one example, a symbol of symmetry is highlighted in the ultrasound image, eg, the midline or midsagittal plane of the brain is displayed in a distinct color from the ultrasound image. For example, the midline of the brain may be displayed as a red straight line in the ultrasound image while the ultrasound image is displayed on the monitor, or the midsagittal plane may be displayed as a plane marked with red shading. Referring to FIG. 5 , a schematic diagram of displaying a symmetry marker in an ultrasound image of a fetal skull in a display according to an embodiment of the present application is shown. As shown in FIG. 5, the mid-brain line 404 is shown in the ultrasound image 400 of the fetal skull as a straight line identified by a dashed line style.
对称性评估指引步骤S130,获取位于所述对称性标识至少一侧的超声图像,基于所述至少一侧的超声图像获取所述胎儿颅脑关于所述对称性标识的对称性评估指引。Symmetry assessment guide step S130: Acquire an ultrasound image located on at least one side of the symmetry mark, and obtain a symmetry assessment guide of the fetal skull with respect to the symmetry mark based on the ultrasound image of the at least one side.
对称性评估指引步骤S130,采用处理器基于超声图像的对称性标识,获得对胎儿颅脑的对称性评估指引,根据对称性评估指引,医生可以对胎儿颅脑关于对称性标识是否对称进行定性或者定量的评价,也可根据胎儿颅脑的对称性特征,指导医生进行关键解剖结构的识别和诊断。Symmetry evaluation guide step S130, using the processor based on the symmetry identification of the ultrasound image to obtain the symmetry evaluation guide for the fetal skull, according to the symmetry evaluation guide, the doctor can qualitatively determine whether the fetal skull is symmetrical with respect to the symmetry identification or not. Quantitative evaluation can also guide doctors to identify and diagnose key anatomical structures based on the symmetrical features of the fetal brain.
在本申请的一个实施例中,对称性评估指引包括对称度,对称度可以 是胎儿颅脑关于对称性标识是否对称的定性评价,也可以是胎儿颅脑关于对称性标识是否对称的定量评价。在定性评价中,根据该对称度,医生可以直接知道胎儿颅脑是否对称或者对称性好或者不好,并判断胎儿颅脑发育是否正常。在定量评价中,医生获得评价胎儿颅脑关于对称性标识是否对称的评估数值,基于这一数值,医生可以判断胎儿颅脑发育是否正常。In an embodiment of the present application, the symmetry assessment guide includes a degree of symmetry, and the degree of symmetry can be a qualitative evaluation of whether the fetal brain is symmetrical about the symmetry mark, or it can be a quantitative evaluation of whether the fetal brain is symmetrical about the symmetry mark. In qualitative evaluation, according to the degree of symmetry, the doctor can directly know whether the fetal brain is symmetrical or whether the symmetry is good or not, and judge whether the fetal brain development is normal. In the quantitative evaluation, the doctor obtains an evaluation value for evaluating whether the fetal brain is symmetrical with respect to the symmetry mark. Based on this value, the doctor can judge whether the fetal brain development is normal.
在本申请的一个实施例中,如图3所示,计算对称度的方法300包括:In an embodiment of the present application, as shown in FIG. 3 , the method 300 for calculating the degree of symmetry includes:
步骤S310,在所述对称性标识一侧获取第一区域超声图像;Step S310, acquiring an ultrasound image of the first region on the side of the symmetry mark;
步骤S320,根据所述第一区域超声图像获取位于所述对称性标识另一侧的第二区域超声图像,所述第一区域的形状和所述第二区域的形状关于所述对称性标识对称;Step S320, obtaining an ultrasound image of a second region located on the other side of the symmetry mark according to the ultrasound image of the first region, and the shape of the first region and the shape of the second region are symmetrical with respect to the symmetry mark ;
步骤S330,确定所述第一区域超声图像和所述第二区域超声图像关于所述对称性标识的对称度。Step S330, determining the degree of symmetry of the ultrasound image of the first region and the ultrasound image of the second region with respect to the symmetry identifier.
在一个示例中,在图像获取步骤S110中获取的超声图像足够清晰,而可以识别胎儿颅脑的全部区域的情况下,可以评价胎儿颅脑的全部区域的对称度。具体的,在计算对称度的过程中,在步骤S310中,在所述对称性标识一侧获取所述第一区域超声图像的方法包括:在所述超声图像中获取所述胎儿颅脑的颅骨,其中,将所述对称性标识一侧的所述颅骨包围的区域作为所述第一区域超声图像;接着,执行步骤S320和步骤S330。In one example, when the ultrasound image acquired in the image acquisition step S110 is clear enough to identify the entire region of the fetal brain, the symmetry of the entire region of the fetal brain can be evaluated. Specifically, in the process of calculating the degree of symmetry, in step S310, the method for acquiring the ultrasound image of the first region on the side of the symmetry marker includes: acquiring the skull of the fetal skull from the ultrasound image , wherein the area surrounded by the skull on one side of the symmetry marker is used as the first area ultrasound image; then, steps S320 and S330 are performed.
在一个示例中,在所述超声图像中获取所述胎儿颅脑的颅骨的方法包括:基于灰度检测或分割的方法检测胎儿颅脑的颅骨,基于机器学习的方法检测胎儿颅脑的颅骨,或者基于深度学习的方法检测胎儿颅脑的颅骨。在一个示例中,在获取胎儿颅脑的颅骨之后在超声图像中显示颅骨光环。例如,如图5所示,获取所述胎儿颅脑的颅骨401,将颅骨401在脑中线404下方包围的区域作为第一区域超声图像;根据第一区域超声图像获取位于脑中线404上方的第二区域超声图像;最后,确定第一区域超声图像和第二区域超声图像关于所述对称性标识的对称度。In one example, the method for acquiring the skull of the fetal skull in the ultrasound image includes: detecting the skull of the fetal skull based on a grayscale detection or segmentation method, detecting the skull of the fetal skull by a method based on machine learning, Or a method based on deep learning to detect the skull of the fetal brain. In one example, a skull halo is displayed in the ultrasound image after obtaining the skull of the fetal skull. For example, as shown in FIG. 5 , the skull 401 of the fetal skull is acquired, and the region surrounded by the skull 401 below the midline 404 of the brain is taken as the first region ultrasound image; the first region above the midline 404 is obtained according to the first region ultrasound image Two-region ultrasound image; finally, determine the degree of symmetry of the first-region ultrasound image and the second-region ultrasound image with respect to the symmetry identification.
在一个示例中,在图像获取步骤S110中获取的超声图像不够清晰(例如,靠近探头一侧的胎儿颅脑的图像不清晰)以至于获取胎儿颅脑的全部区域计算对称度不具有评价意义,可以通过识别超声图像中至少部分区域为感兴趣的区域进行对称性评估,或者医生需要根据需要选择超声图像中 至少部分区域为感兴趣的区域进行对称性评估,进而计算感兴趣的区域的对称度。具体的,在计算对称度的过程中,在步骤S310中,在所述超声图像中获取感兴趣区域,所述感兴趣区域至少部分位于所述对称性标识的一侧,所述感兴趣区域位于所述对称性标识一侧的区域为所述第一区域超声图像;接着,执行步骤S320和步骤S330。In one example, the ultrasound image acquired in the image acquisition step S110 is not clear enough (for example, the image of the fetal skull on the side close to the probe is not clear) so that it is not meaningful to obtain the entire area of the fetal skull to calculate the symmetry degree, Symmetry evaluation can be performed by identifying at least part of the region in the ultrasound image as the region of interest, or the doctor needs to select at least part of the region in the ultrasound image as the region of interest for symmetry evaluation, and then calculate the symmetry of the region of interest. . Specifically, in the process of calculating the symmetry degree, in step S310, a region of interest is acquired in the ultrasound image, the region of interest is at least partially located on one side of the symmetry mark, and the region of interest is located at The area on one side of the symmetry mark is the ultrasound image of the first area; then, step S320 and step S330 are performed.
在一个示例中,获取感兴趣区域的方法可以通过处理器自动确定。示例性的,通过处理器识别超声图像中至少部分较为清晰的区域为感兴趣的区域。例如,处理器识别对称性标识靠近探头采集端的一侧颅脑的超声图像中清晰度的值高于一特定值的区域为感兴趣区域。示例性的,通过处理器截取超声图像中特征解剖结构可能位于的区域为感兴趣的区域。例如,基于特征解剖结构在胎儿颅脑中的区域的统计值,截取超声图像中对应于该统计值的区域作为感兴趣的区域,其中,特征解剖结构在颅脑中的区域的统计值包括但不限于特征解剖结构在胎儿颅脑中所在区域的最大概率统计值,以及其中,特征解剖结构在胎儿颅脑中所在区域至少一部分位于对称性标识的一侧。In one example, the method of obtaining the region of interest can be determined automatically by the processor. Exemplarily, at least a part of a relatively clear region in the ultrasound image is identified as the region of interest by the processor. For example, the processor identifies a region with a sharpness value higher than a certain value in the ultrasound image of the side of the skull near the acquisition end of the probe as the region of interest by identifying the symmetry. Exemplarily, the region where the characteristic anatomical structure may be located in the ultrasound image is intercepted by the processor as the region of interest. For example, based on the statistical value of the region of the characteristic anatomical structure in the fetal skull, the region corresponding to the statistical value in the ultrasound image is intercepted as the region of interest, wherein the statistical value of the region of the characteristic anatomical structure in the brain includes but It is not limited to the maximum probability statistic value of the region where the characteristic anatomical structure is located in the fetal skull, and wherein at least a part of the region where the characteristic anatomical structure is located in the fetal skull is located on one side of the symmetry mark.
在一个示例中,获取感兴趣区域的方法可以是医生手动截取感兴趣区域。图6示出了根据本申请的一个实施例中在显示器中显示的胎儿颅脑的超声图像中获取感兴趣区域的示意图。例如,如图6示,采用矩形框在胎儿颅脑的超声图像400中框定感兴趣区域405,感兴趣区域405部分位于脑中线404上方、部分位于脑中线404下方,将感兴趣区域405位于脑中线下方的部分设定为第一区域超声图像;接着,根据第一区域超声图像获取位于脑中线404上方的第二区域超声图像;最后,确定第一区域超声图像和第二区域超声图像关于所述对称性标识的对称度。In one example, the method of obtaining the region of interest may be that the doctor manually intercepts the region of interest. FIG. 6 shows a schematic diagram of acquiring a region of interest in an ultrasound image of a fetal skull displayed on a display according to an embodiment of the present application. For example, as shown in FIG. 6 , a rectangular frame is used to frame a region of interest 405 in an ultrasound image 400 of the fetal skull, and the region of interest 405 is partially located above the midline 404 of the brain and partially located below the midline 404 of the brain, and the region of interest 405 is located in the brain The part below the midline is set as the ultrasound image of the first region; then, the ultrasound image of the second region located above the midline 404 of the brain is acquired according to the ultrasound image of the first region; finally, it is determined that the ultrasound image of the first region and the ultrasound image of the second region are related to each other. Describe the symmetry of the symmetry logo.
在一个示例中,在医生需要针对胎儿颅脑的解剖结构评价其关于对称性标识是否对称的情况下,选择超声图像中的特征解剖结构的区域计算对称度。具体的,在计算对称度的过程中,在步骤S310中,在所述对称性标识一侧获取所述第一区域超声图像的方法包括:在所述超声图像中识别至少一个第一解剖结构,所述至少一个第一解剖结构的至少部分位于所述对称性标识的一侧,所述至少一个第一解剖结构位于所述对称性标识一侧的区域为所述第一区域超声图像;接着,执行步骤S320和步骤S330。在一 个示例中,第一解剖结构位于对称性标识一侧。在另一个示例中,第一解剖结构部分位于对称性标识一侧,部分位于对称性标识另一侧。在一个示例中,第一解剖结构包括:小脑、侧脑室、脉络丛、大脑外侧裂、丘脑。在一个示例中,在所述超声图像中获取所述第一解剖结构的方法包括:基于灰度检测或分割的方法检测第一解剖结构,基于机器学习的方法检测第一解剖结构,或者基于深度学习的方法识别第一解剖结构。在一个示例中,基于在超声图像中获取所述第一解剖结构的方法,所述第一解剖结构位于所述对称性标识一侧的区域包括所述第一解剖结构的轮廓包围的区域,或包围所述第一解剖结构的边框包围的区域。In one example, where a physician needs to evaluate whether the anatomy of the fetal skull is symmetric with respect to a marker of symmetry, a region of the characteristic anatomy in the ultrasound image is selected to calculate the degree of symmetry. Specifically, in the process of calculating the degree of symmetry, in step S310, the method for acquiring the ultrasound image of the first region on the side of the symmetry identification includes: identifying at least one first anatomical structure in the ultrasound image, At least part of the at least one first anatomical structure is located on one side of the symmetry mark, and the region where the at least one first anatomical structure is located on one side of the symmetry mark is an ultrasound image of the first region; then, Steps S320 and S330 are performed. In one example, the first anatomical structure is on one side of the symmetry marker. In another example, the first anatomical structure is partially on one side of the symmetry marker and partially on the other side of the symmetry marker. In one example, the first anatomical structure includes: cerebellum, lateral ventricle, choroid plexus, sylvian fissure, thalamus. In one example, the method for acquiring the first anatomical structure in the ultrasound image includes: detecting the first anatomical structure based on a grayscale detection or segmentation method, detecting the first anatomical structure based on a machine learning method, or detecting the first anatomical structure based on a depth-based method Learn methods to identify first anatomical structures. In one example, based on the method of acquiring the first anatomical structure in an ultrasound image, the area of the first anatomical structure on one side of the symmetry marker includes an area surrounded by an outline of the first anatomical structure, or The area enclosed by the border surrounding the first anatomical structure.
图7示出了根据本申请的一个实施例的超声检测方法中在显示器中显示的胎儿颅脑的超声图像中获取解剖结构的示意图。例如,如图7示,在胎儿颅脑的超声图像400中识别解剖结构402和403,解剖结构402和403均位于脑中线的下方,解剖结构402和403所在区域为第一区域超声图像;接着,根据第一区域超声图像获取位于脑中线404上方的第二区域超声图像;最后,确定第一区域超声图像和第二区域超声图像关于所述对称性标识的对称度。7 shows a schematic diagram of acquiring an anatomical structure in an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application. For example, as shown in FIG. 7 , the anatomical structures 402 and 403 are identified in the ultrasound image 400 of the fetal skull, the anatomical structures 402 and 403 are both located below the midline of the brain, and the region where the anatomical structures 402 and 403 are located is the ultrasound image of the first region; then , acquiring the ultrasound image of the second region above the midline 404 of the brain according to the ultrasound image of the first region; finally, determining the degree of symmetry of the ultrasound image of the first region and the ultrasound image of the second region with respect to the symmetry mark.
在本申请的一个实施例中,若所采集的超声图像足够清晰,则可以在超声图像中位于对称性标识的两侧中任一侧的区域均可以作为第一区域超声图像。In an embodiment of the present application, if the acquired ultrasound image is clear enough, the region that can be located on either side of the symmetry mark in the ultrasound image can be used as the first region ultrasound image.
在本申请的一个实施例中,若所采集的超声图像不够清晰。具体的,由于颅骨的骨性结构的透声性较差,在采集超声图像时,对应于靠近超声探头一侧的靠近骨头后方组织结构往往显示不清,表现为超声图像中对应于靠近超声探头一侧的信息丢失严重,而对应于远离超声探头一侧的信息较为清晰,为此,在所述对称性标识一侧获取第一区域超声图像包括在所述对称性标识远离超声探头采集端的一侧的超声图像中获取第一区域超声图像。例如,在步骤S310中,通过在所述超声图像中识别至少一个第一解剖结构以在所述对称性标识一侧获取所述第一区域超声图像的方法中,在所述对称性标识远离超声探头采集端的一侧的超声图像中识别第一解剖结构,所述第一解剖结构位于所述对称性标识一侧的区域为所述第一区域超声图像,这样可以对第一解剖结构的轮廓或第一解剖结构所在的区域进行 准确和清晰的识别,便于基于第一区域获取第二区域,并基于第一区域和第二区域确定对称度。In an embodiment of the present application, if the acquired ultrasound image is not clear enough. Specifically, due to the poor sound permeability of the bony structure of the skull, when acquiring an ultrasound image, the tissue structure behind the bone corresponding to the side close to the ultrasound probe is often unclear, which is manifested in the ultrasound image corresponding to the side close to the ultrasound probe. The information on one side is seriously lost, while the information corresponding to the side away from the ultrasound probe is relatively clear. For this reason, acquiring the ultrasound image of the first region on the side of the symmetry mark includes a position where the symmetry mark is far from the acquisition end of the ultrasound probe. The ultrasound image of the first region is acquired in the ultrasound image of the side. For example, in step S310, in the method for acquiring the ultrasound image of the first region on the side of the symmetry marker by identifying at least one first anatomical structure in the ultrasound image, when the symmetry marker is far from the ultrasound The first anatomical structure is identified in the ultrasonic image on one side of the probe acquisition end, and the region of the first anatomical structure located on the side of the symmetry mark is the ultrasonic image of the first region, so that the contour or the contour of the first anatomical structure or the region of the first anatomical structure can be identified. The region where the first anatomical structure is located is accurately and clearly identified, so as to obtain the second region based on the first region, and determine the degree of symmetry based on the first region and the second region.
在本申请的一个实施例中,在对称性评估指引步骤S130中,确定所述第一区域超声图像和所述第二区域超声图像关于所述对称性标识的对称度的方法包括:计算所述第一区域超声图像和所述第二区域超声图像关于所述对称性标识的对称点处的灰度绝对差值和/或平方差值,基于所述灰度绝对差值和/或平方差值确定所述对称度。在一个示例中,直接将获得的灰度绝对差值和/或平方差值作为定量评价的对称度的数值显示在超声图像中。在一个示例中,将获得的灰度绝对差值和/或平方差值进行换算后得到作为定量评价的对称度的数值显示在超声图像中。例如,将获得的灰度绝对差值和/或平方差值与对应于对称度0-10(或者0-100、0-100%)的定量评价数值进行成比例换算。在一个示例中,将获得的灰度绝对差值和/或平方差值与预定的灰度绝对差阈值和/或平方差阈值进行对比,当灰度绝对差值和/或平方差值小于灰度绝对差阈值和/或平方差阈值时,输出对称度好的定性评价;反之,则输出对称度不好的定性评价。In an embodiment of the present application, in the symmetry evaluation guide step S130, the method for determining the degree of symmetry of the ultrasound image of the first region and the ultrasound image of the second region with respect to the symmetry identifier includes: calculating the The absolute difference value and/or the squared difference value of the first region ultrasound image and the second region ultrasound image with respect to the symmetry point identified by the symmetry, based on the absolute grayscale difference value and/or the squared difference value The degree of symmetry is determined. In one example, the obtained absolute grayscale difference and/or squared difference is directly displayed in the ultrasound image as a numerical value of the degree of symmetry for quantitative evaluation. In one example, the obtained absolute difference value and/or squared difference value of gray scale is converted to obtain a numerical value of the degree of symmetry as a quantitative evaluation, which is displayed in the ultrasound image. For example, the obtained grayscale absolute difference and/or squared difference is scaled with a quantitative evaluation value corresponding to a symmetry degree of 0-10 (or 0-100, 0-100%). In one example, the obtained grayscale absolute difference and/or squared difference is compared with a predetermined grayscale absolute difference threshold and/or squared difference threshold, and when the grayscale absolute difference and/or squared difference is smaller than the grayscale absolute difference and/or squared difference When the absolute difference threshold and/or the squared difference threshold is set, a qualitative evaluation of good symmetry is output; otherwise, a qualitative evaluation of poor symmetry is output.
在本申请的一个实施例中,在对称性评估指引步骤S130中,确定所述第一区域超声图像和所述第二区域超声图像关于所述对称性标识的对称度的方法包括:计算所述第一区域超声图像和所述第二区域超声图像的结构相似度,基于所述结构相似度确定所述对称度。在一个示例中,直接将获得的结构相似度值作为定量评价的对称度的数值显示在超声图像中。在一个示例中,将获得的结构相似度值进行换算后得到作为定量评价的对称度的数值显示在超声图像中。例如,将获得的结构相似度值与对应于对称度0-10(或者0-100、0-100%)的定量评价数值进行成比例换算。在一个示例中,将获得的结构相似度值与预定的结构相似度阈值进行对比,当结构相似度值小于结构相似度阈值时,输出对称度好的定性评价;反之,则输出对称度不好的定性评价。In an embodiment of the present application, in the symmetry evaluation guide step S130, the method for determining the degree of symmetry of the ultrasound image of the first region and the ultrasound image of the second region with respect to the symmetry identifier includes: calculating the The degree of structural similarity between the ultrasonic image of the first region and the ultrasonic image of the second region, and the degree of symmetry is determined based on the structural similarity. In one example, the obtained structural similarity value is directly displayed in the ultrasound image as the numerical value of the quantitatively evaluated symmetry. In one example, after converting the obtained structural similarity value, a numerical value of the symmetry degree as a quantitative evaluation is obtained and displayed in the ultrasound image. For example, the obtained structural similarity value is scaled with the quantitative evaluation value corresponding to the symmetry degree of 0-10 (or 0-100, 0-100%). In one example, the obtained structural similarity value is compared with a predetermined structural similarity threshold, and when the structural similarity value is smaller than the structural similarity threshold, a qualitative evaluation of good symmetry is output; otherwise, the output is poor. qualitative evaluation.
在本申请的一个实施例中,在对称性评估指引步骤S130中,确定所述第一区域超声图像和所述第二区域超声图像关于所述对称性标识的对称度的方法包括:计算所述第一区域超声图像和所述第二区域超声图像的相关比,基于所述相关比确定所述对称度。在一个示例中,直接将获得的相关 比值作为定量评价的对称度的数值显示在超声图像中。在一个示例中,将获得的相关比值进行换算后得到作为定量评价的对称度的数值显示在超声图像中。例如,将获得的相关比值与对应于对称度0-10(或者0-100、0-100%)的定量评价数值进行成比例换算。在一个示例中,将获得的相关比值与预定的相关比阈值进行对比,当相关比值小于相关比阈值时,输出对称度好的定性评价;反之,则输出对称度不好的定性评价。In an embodiment of the present application, in the symmetry evaluation guide step S130, the method for determining the degree of symmetry of the ultrasound image of the first region and the ultrasound image of the second region with respect to the symmetry identifier includes: calculating the A correlation ratio of the ultrasound image of the first region and the ultrasound image of the second region, and the degree of symmetry is determined based on the correlation ratio. In one example, the obtained correlation ratio is directly displayed in the ultrasound image as a numerical value for a quantitative assessment of the degree of symmetry. In one example, the obtained correlation ratio is converted to obtain a numerical value of the degree of symmetry as a quantitative evaluation and displayed in the ultrasound image. For example, the obtained correlation ratio is scaled with the quantitative evaluation value corresponding to the degree of symmetry 0-10 (or 0-100, 0-100%). In one example, the obtained correlation ratio is compared with a predetermined correlation ratio threshold. When the correlation ratio is smaller than the correlation ratio threshold, a qualitative evaluation of good symmetry is output; otherwise, a qualitative evaluation of poor symmetry is output.
在本申请的实施例中,对称度的计算主要利用了相似性度量方法。例如,L1和L2范数利用像素间的灰度绝对差值和平方误差进行相似性的度量;相关比(Correlation Ratio)通过衡量两个向量间的函数依赖关系,进行相似性的度量;结构相似度(SSIM)根据人类视觉感知原理,从亮度、对比度以及结构三个方面对超声图像进行相似性的度量。在本申请的其他实施例中,还可以利用其它进行相似性的度量的方法计算对称度,例如通过峰值信噪比(PSNR)、互信息(MI)、互相关(Cross Correlation)等的方法。In the embodiments of the present application, the calculation of the symmetry degree mainly uses the similarity measurement method. For example, the L1 and L2 norms use the absolute difference and squared error between pixels to measure the similarity; Correlation Ratio measures the similarity by measuring the functional dependency between two vectors; the structure is similar Degree (SSIM) measures the similarity of ultrasound images from three aspects: brightness, contrast and structure according to the principle of human visual perception. In other embodiments of the present application, the degree of symmetry may also be calculated by using other methods for measuring similarity, such as peak signal-to-noise ratio (PSNR), mutual information (MI), and cross correlation (Cross Correlation).
由于在对称性评估指引步骤S130中,医生往往需要针对位于对称性标识两侧的特定的解剖结构的位置,单独评价其在胎儿颅脑中位置上的对称性。在本申请的一个实施例中,在获取超声图像后,识别超声图像中位于对称性标识一侧的解剖结构,并将其关于对称性标识在对称性标识的另一侧进行投影,可以引导医生在对称性标识的另一侧识别相对应的解剖结构,为医生提供关于位于对称性标识两侧的解剖结构的位置是否对称的对称性评估指引。Because in the symmetry evaluation guide step S130, the doctor often needs to individually evaluate the symmetry of the position of the specific anatomical structure located on both sides of the symmetry mark in the fetal skull. In one embodiment of the present application, after acquiring the ultrasound image, identifying the anatomical structure on one side of the symmetry mark in the ultrasound image, and projecting it on the other side of the symmetry mark with respect to the symmetry mark, can guide the doctor Corresponding anatomical structures are identified on the other side of the symmetry marker, providing the physician with a symmetry assessment guide as to whether the positions of the anatomical structures on both sides of the symmetry marker are symmetrical.
在本申请的一个实施例中,所述对称性评估指引包括位于所述对称性标识一侧的第二解剖结构关于所述对称性标识在所述对称性标识另一侧对称的投影标记。由于超声图像具有显示的直观性,通过获取位于对称性标识一侧的解剖结构之后,将该解剖结构关于对称性标识在所述对称性标识另一侧进行投影,医生通过观察投影标记与位于所述对称性标识另一侧的第三解剖结构(第三解剖结构可以是医生通过人眼识别也可以通过处理器机器识别)是否重合,直接获得位于对称性标识两侧的解剖结构的位置是否对称的对称性评价。需要说明的是,第二解剖结构和第三解剖结构为颅脑生理学上关于对称性标识对称的两个解剖结构,例如第二解剖结构为左 侧小脑,第三解剖结构为右侧小脑等。其中,当医生观察到第二解剖结构的该投影标记与第三解剖结构重合时,可以直接判断第二解剖结构和第三解剖结构关于对称性标识对称,从而可以判断胎儿颅脑的对称性良好;当医生观察到第二解剖结构的该投影标记与第三解剖结构不重合时,可以判断第二解剖结构和第三解剖结构关于对称性标识不对称从而可以判断胎儿颅脑的对称性不好,或者,可以做出需要进一步检测的判断。上述方法使得对称性评估过程直观、可视。在超声图像的靠近探头采集端的一侧部分显示不清楚的情况中,当医生观察到第二解剖结构的该投影标记,还可以根据第二解剖结构的该投影标记对第三解剖结构的区域进行辅助定位,参看图8,示出了根据本申请的一个实施例的超声检测方法中获取对称性评估指引的方法的示意性流程图。如图8所示,所述对称性评估指引步骤S130包括:In one embodiment of the present application, the symmetry assessment guide includes a projection marker that is symmetrical on the other side of the symmetry marker with respect to the second anatomical structure on one side of the symmetry marker. Due to the intuitive display of ultrasound images, after acquiring the anatomical structure on one side of the symmetry mark, the anatomical structure is projected on the other side of the symmetry mark with respect to the symmetry mark. Whether the third anatomical structure on the other side of the symmetry mark (the third anatomical structure can be recognized by the doctor through the human eye or by the processor machine) is coincident, and directly obtain whether the positions of the anatomical structures on both sides of the symmetry mark are symmetrical. Symmetry evaluation. It should be noted that the second anatomical structure and the third anatomical structure are two anatomical structures that are symmetrical with respect to the symmetry sign in craniocerebral physiology, for example, the second anatomical structure is the left cerebellum, and the third anatomical structure is the right cerebellum, etc. Wherein, when the doctor observes that the projection mark of the second anatomical structure coincides with the third anatomical structure, it can be directly judged that the second anatomical structure and the third anatomical structure are symmetrical with respect to the symmetry mark, so that it can be judged that the symmetry of the fetal skull is good When the doctor observes that the projection mark of the second anatomical structure does not coincide with the third anatomical structure, it can be judged that the second anatomical structure and the third anatomical structure are asymmetrical with respect to the symmetry mark, so that it can be judged that the symmetry of the fetal skull is not good , or, a judgment may be made that further testing is required. The above method makes the symmetry evaluation process intuitive and visual. In the case that the side part of the ultrasound image close to the acquisition end of the probe is not clearly displayed, when the doctor observes the projected mark of the second anatomical structure, the doctor can also perform the operation on the region of the third anatomical structure according to the projected mark of the second anatomical structure. Assisted positioning, referring to FIG. 8 , it shows a schematic flowchart of a method for obtaining a symmetry evaluation guide in an ultrasonic inspection method according to an embodiment of the present application. As shown in FIG. 8 , the symmetry evaluation guide step S130 includes:
识别步骤S8301,在所述超声图像上识别至少一个所述第二解剖结构,获取所述第二解剖结构的轮廓或所述第二解剖结构的所在区域;Identifying step S8301, identifying at least one second anatomical structure on the ultrasound image, and acquiring the contour of the second anatomical structure or the region where the second anatomical structure is located;
投影步骤S8302,获取所述第二解剖结构的轮廓或所述第二解剖结构的所在区域关于所述对称性标识对称的所述投影标记;Projecting step S8302, obtaining the outline of the second anatomical structure or the projection mark that is symmetrical with respect to the symmetry marker in the region where the second anatomical structure is located;
显示步骤S8303,显示所述投影标记。Step S8303 is displayed, and the projection mark is displayed.
在所述识别步骤S8301中,识别至少一个所述第二解剖结构的方法包括以下至少一项:In the identifying step S8301, the method for identifying at least one of the second anatomical structures includes at least one of the following:
通过图像分割算法,在所述超声图像上分割得到至少一个所述第二解剖结构的轮廓;Segmenting the ultrasound image to obtain at least one contour of the second anatomical structure through an image segmentation algorithm;
通过预训练的机器学习模型,获取所述超声图像上的所述第二解剖结构的轮廓或所述第二解剖结构的所在区域。The contour of the second anatomical structure or the region where the second anatomical structure is located on the ultrasound image is obtained by using the pre-trained machine learning model.
示例性的,通过图像分割算法,在超声图像上分割得到至少一个所述第二解剖结构的轮廓,其中,图像分割算法可以包括传统的灰度检测或分割方法,利用传统的灰度检测或分割方法可以有效检测颅脑的关键解剖结构。例如,大津阈值算法(OSTU)通过最大化前景与背景区域的类间方差求得阈值,实现图像的二值化分割;水平集方法(Level Set)和主动轮廓模型(Snake)通过最小化能量泛函从而实现图像分割。另外,还有如图割(Graph Cut)、区域生长(Region Growing)等方法也可以达到关键解剖 结构分割的目的。Exemplarily, the contour of at least one of the second anatomical structures is obtained by segmenting the ultrasound image through an image segmentation algorithm, wherein the image segmentation algorithm may include a traditional grayscale detection or segmentation method, using traditional grayscale detection or segmentation. The method can effectively detect the key anatomical structures of the brain. For example, the Otsu thresholding algorithm (OSTU) obtains the threshold by maximizing the inter-class variance of the foreground and background regions, and realizes the binary segmentation of the image; the level set method (Level Set) and the active contour model (Snake) by minimizing the energy panning function to achieve image segmentation. In addition, there are also methods such as Graph Cut and Region Growing, which can also achieve the purpose of key anatomical structure segmentation.
示例性的,机器学习模型可以包括传统的机器学习模型和深度学习模型。对于传统的机器学习模型,基于机器学习的结构检测方法首先构建样本学习库;之后,利用如PCA、LDA、Haar等方法提取图像特征,并将提取到的特征与学习库进行匹配;最后,利用SVM、KNN、PCANet等方法进行分类,从而获取ROI并判断解剖结构类别,实现了胎儿颅脑远场关键解剖结构的检测。Exemplarily, machine learning models may include traditional machine learning models and deep learning models. For traditional machine learning models, the structure detection method based on machine learning first builds a sample learning library; then uses methods such as PCA, LDA, Haar to extract image features, and matches the extracted features with the learning library; finally, using SVM, KNN, PCANet and other methods are used for classification, so as to obtain ROI and determine the anatomical structure category, and realize the detection of key anatomical structures in the far field of the fetal skull.
对于深度学习模型,第一种情况可以采用基于深度学习的For deep learning models, the first case can be based on deep learning
Bounding-Box检测方法,该方法输入一副图像,通过网络直接回归出ROI的Bounding-Box。与此同时,可以获取到ROI的结构类别,从而实现关键解剖结构的检测,常见的网络有RCNN、Fast RCNN等。第二种情况为基于深度学习的端到端的语义分割网络方法,该方法去除了全连接层,通过上采样或反卷积的形式使得输入和输出图像的尺寸相同,可以直接得到输入图像的ROI以及对应的类别,常见的网络有FCN、UNet等。第三种情况是首先对目标进行定位,提取ROI的特征,特征提取方法可以是传统的PCA、LDA等,也可利用卷积神经网络进行特征提取。之后,基于提取到的特征,利用神经网络判别器进行分类,从而实现了远场关键解剖结构的检测。Bounding-Box detection method, this method inputs an image and directly returns the Bounding-Box of the ROI through the network. At the same time, the structure category of the ROI can be obtained, so as to realize the detection of key anatomical structures. Common networks include RCNN, Fast RCNN, etc. The second case is an end-to-end semantic segmentation network method based on deep learning. This method removes the fully connected layer, and makes the size of the input and output images the same through upsampling or deconvolution, and the ROI of the input image can be directly obtained. And the corresponding categories, common networks are FCN, UNet, etc. The third case is to first locate the target and extract the features of the ROI. The feature extraction method can be traditional PCA, LDA, etc., or a convolutional neural network can be used for feature extraction. Afterwards, based on the extracted features, a neural network discriminator is used for classification, thereby realizing the detection of key anatomical structures in the far field.
在投影步骤S8302中,获取所述第二解剖结构的轮廓或所述第二解剖结构的所在区域关于所述对称性标识对称的所述投影标记的方法包括:将第二解剖结构的轮廓或所述第二解剖结构的所在区域关于对称性标识进行对称变换。在显示步骤中,将投影标记显示在超声图像中。In the projecting step S8302, the method for acquiring the contour of the second anatomical structure or the projection mark that is symmetrical with respect to the symmetry marker in the region where the second anatomical structure is located includes: converting the contour or all of the second anatomical structure into The region where the second anatomical structure is located is symmetrically transformed with respect to the symmetry marker. In the displaying step, the projection marker is displayed in the ultrasound image.
在本申请的一个实施例中,所述投影标记包括对应于所述第二解剖结构的轮廓的轮廓标记,和/或对应于所述第二解剖结构的所在区域的区域标记。在一个示例中,在投影步骤S8302中,获取所述第二解剖结构的轮廓或所述第二解剖结构的所在区域关于所述对称性标识对称的所述投影标记的方法包括:将第二解剖结构的轮廓或第二解剖结构的所在区域关于对称性标识进行对称变换,获得与第二解剖结构的轮廓或第二解剖结构的所在区域相对应的轮廓或者区域。其中,第二解剖结构的轮廓指第二解剖结构的实际轮廓边界,第二解剖结构的所在区域为包围有第二解剖结构的框型 区域,例如包围了第二解剖结构的方形框区域,或包围第二解剖结构并与第二解剖结构的实际轮廓相切的圆形框区域。In one embodiment of the present application, the projection markers include contour markers corresponding to the contour of the second anatomical structure, and/or area markers corresponding to the region where the second anatomical structure is located. In one example, in the projecting step S8302, the method for acquiring the outline of the second anatomical structure or the projection mark that is symmetrical with respect to the symmetry marker in the region where the second anatomical structure is located includes: mapping the second anatomical structure to the The contour of the structure or the region where the second anatomical structure is located is symmetrically transformed with respect to the symmetry flag to obtain a contour or region corresponding to the contour of the second anatomical structure or the region where the second anatomical structure is located. Wherein, the contour of the second anatomical structure refers to the actual contour boundary of the second anatomical structure, and the region where the second anatomical structure is located is a box-shaped region surrounding the second anatomical structure, such as a square box region surrounding the second anatomical structure, or A circular box area that encloses the second anatomical structure and is tangent to the actual contour of the second anatomical structure.
在一个示例中,所述轮廓标记包括轮廓边界标记、轮廓阴影标记和/或轮廓箭头标记;所述区域标记包括区域边界标记、区域阴影标记和/或区域箭头标记。图9和图10示出了根据本申请的一个实施例的超声检测方法中在显示器中显示的胎儿颅脑的超声图像中显示投影标记的示意图。如图9所示,对应于解剖结构402的轮廓,解剖结构402在脑中线404的另一侧的投影标记显示为轮廓边界标记4021,对应于解剖结构403的轮廓,解剖结构403在脑中线404的另一侧的投影标记显示为轮廓箭头标记4031。如图10所示,对应于解剖结构402的轮廓,解剖结构402在脑中线404的另一侧的投影标记显示为轮廓边界标记4022,对应于解剖结构403所在的区域,解剖结构403在脑中线404的另一侧的投影标记显示为轮廓阴影标记4032。In one example, the contour markers include contour boundary markers, contour shadow markers, and/or contour arrow markers; and the area markers include area border markers, area shadow markers, and/or area arrow markers. 9 and 10 are schematic diagrams showing projection markers displayed in the ultrasound image of the fetal skull displayed on the display in the ultrasound detection method according to an embodiment of the present application. As shown in FIG. 9 , corresponding to the outline of anatomical structure 402 , the projected marker of anatomical structure 402 on the other side of midline 404 is shown as outline boundary marker 4021 , corresponding to the outline of anatomical structure 403 , which is on midline 404 of the brain The projected markers on the other side of the are shown as outline arrow markers 4031. As shown in Figure 10, corresponding to the outline of anatomical structure 402, the projected marking of anatomical structure 402 on the other side of mid-brain line 404 is shown as outline boundary marking 4022, corresponding to the area where anatomical structure 403 is located on the mid-brain line The projected indicia on the other side of 404 are shown as outline shadow indicia 4032.
在一个示例中,当所述投影标记为多个时,区分显示所述多个投影标记。示例性的,所述区分显示的方法包括:对应于每一所述第二解剖结构将所述投影标记进行颜色区分;或者对应于每一所述第二解剖结构将所述投影标记进行线型区分;或者对应于每一所述第二解剖结构将所述投影标记进行阴影区分。图11示出了根据本申请的一个实施例的超声检测方法中在显示器中显示的胎儿颅脑的超声图像中区分显示投影标记的示意图。如图11所示,对应于解剖结构402和解剖结构403,解剖结构402和解剖结构403所在的轮廓在脑中线404的另一侧的投影标记显示为轮廓阴影标记4023和轮廓阴影标记4033,其中,轮廓阴影标记4023和轮廓阴影标记4033的阴影图案不同。In one example, when there are multiple projection marks, the multiple projection marks are displayed in a differentiated manner. Exemplarily, the method for differentiating and displaying includes: color-distinguishing the projection marks corresponding to each of the second anatomical structures; or performing line-typing of the projection marks corresponding to each of the second anatomical structures. distinguishing; or shadow distinguishing the projection markers corresponding to each of the second anatomical structures. FIG. 11 shows a schematic diagram of distinguishing and displaying projection markers in the ultrasound image of the fetal skull displayed on the display in the ultrasound detection method according to an embodiment of the present application. As shown in FIG. 11 , corresponding to the anatomical structures 402 and 403 , the projection marks on the other side of the midline 404 of the brain where the contours of the anatomical structures 402 and 403 are located are shown as contour shadow marks 4023 and 4033 , wherein , the shading patterns of the outline shading mark 4023 and the outline shading mark 4033 are different.
在本申请的一个实施例中,所述对称性评估指引包括位于所述对称性标识一侧的第二解剖结构与位于所述对称性标识另一侧的第三解剖结构关于所述对称性标识是否对称的对称性评价,其中,所述第二解剖结构和所述第三解剖结构是生理学上胎儿颅脑关于所述对称性标识对称的解剖结构对。在本实施例中,对称性评价是将第二解剖结构的轮廓或所述第二解剖结构的所在区域关于所述对称性标识对称的所述投影标记与第三解剖结构的轮廓或第三解剖结构所在的区域进行对比做出的,其是针对第二解剖结 构和第三解剖结构在位置上是否对称的对称性评价。因此,在识别步骤S8301中,还识别位于所述对称性标识另一侧的第三解剖结构,获取所述第三解剖结构的轮廓或所述第三解剖结构所在的区域,所述第三解剖结构与所述至少一个所述第二解剖结构中的每一个一一对应;在显示步骤S8303中,还显示所述第三解剖结构的轮廓或所述第三解剖结构所在的区域。通过将在投影步骤S8302中获取的第二解剖结构的轮廓或所述第二解剖结构所在的区域关于所述对称性标识在在所述对称性标识另一侧的投影标记与第三解剖结构的轮廓或所述第三解剖结构所在的区域进行对比,得到对称性评价。In one embodiment of the present application, the symmetry assessment guide includes a second anatomical structure on one side of the symmetry marker and a third anatomical structure on the other side of the symmetry marker with respect to the symmetry marker Symmetry evaluation of whether it is symmetrical, wherein the second anatomical structure and the third anatomical structure are anatomical structure pairs that are physiologically symmetrical with respect to the symmetry identification of the fetal brain. In this embodiment, the symmetry evaluation is to compare the outline of the second anatomical structure or the region where the second anatomical structure is located with respect to the symmetry marker, and the projection mark that is symmetric with the outline of the third anatomical structure or the third anatomy The regions where the structures are located are compared and made, which is a symmetry evaluation for whether the second anatomical structure and the third anatomical structure are symmetrical in position. Therefore, in the identifying step S8301, the third anatomical structure located on the other side of the symmetry marker is also identified, and the outline of the third anatomical structure or the region where the third anatomical structure is located is obtained, and the third anatomical structure is The structures are in one-to-one correspondence with each of the at least one second anatomical structure; in the displaying step S8303, the outline of the third anatomical structure or the region where the third anatomical structure is located is also displayed. By comparing the contour of the second anatomical structure obtained in the projection step S8302 or the region where the second anatomical structure is located with respect to the symmetry mark on the projection mark on the other side of the symmetry mark and the third anatomical structure Contours or regions where the third anatomical structure is located are compared to obtain a symmetry assessment.
当所述投影标记与所述第三解剖结构的轮廓或所述第三解剖结构所在的区域相同时,获得第二解剖结构和第三解剖结构对称的对称性评价;当所述投影标记与所述第三解剖结构的轮廓或所述第三解剖结构所在的区域不同时,获得第二解剖结构和第三解剖结构不对称的对称性评价。在一个示例中,在显示器上显示所述对称性评价。在一个示例中,通过语音播报的形式输出所述对称性评价。When the projected marker is the same as the outline of the third anatomical structure or the region where the third anatomical structure is located, a symmetry evaluation of the symmetry of the second anatomical structure and the third anatomical structure is obtained; when the projected marker is the same as the projected marker When the contour of the third anatomical structure or the region where the third anatomical structure is located is different, a symmetry evaluation of the asymmetry of the second anatomical structure and the third anatomical structure is obtained. In one example, the symmetry evaluation is displayed on a display. In one example, the symmetry evaluation is output in the form of a voice announcement.
需要说明的是,第一解剖结构可以包括小脑、侧脑室、脉络丛、大脑外侧裂和丘脑中的至少一项,第二解剖结构也可以包括小脑、侧脑室、脉络丛、大脑外侧裂和丘脑中的至少一项,第三解剖结构也可以包括小脑、侧脑室、脉络丛、大脑外侧裂和丘脑的一项。其中,小脑、侧脑室、脉络丛、大脑外侧裂和丘脑都是一个概括的解剖结构名称,其中一者包括左右两部分对称结构时,这里的第一解剖结构、第二解剖结构和第三解剖结构可以指左右两部分对称结构中的任一一部分,例如第一解剖结构可以为左小脑,第二解剖结构可以为左侧脑室,第三解剖结构可以为右侧脑室等等;当然这里的第一解剖结构、第二解剖结构和第三解剖结构也可以指上述解剖结构的整体,例如第一解剖结构可以为小脑,第二解剖结构可以为大脑外侧裂,第三解剖结构也可以为大脑外侧裂。It should be noted that the first anatomical structure may include at least one of the cerebellum, lateral ventricle, choroid plexus, sylvian fissure and thalamus, and the second anatomical structure may also include cerebellum, lateral ventricle, choroid plexus, sylvian fissure and thalamus At least one of the third anatomical structures may also include one of the cerebellum, the lateral ventricle, the choroid plexus, the sylvian fissure, and the thalamus. Among them, the cerebellum, lateral ventricle, choroid plexus, sylvian fissure and thalamus are all a general anatomical structure name, one of which includes the left and right two parts symmetrical structure, here the first anatomical structure, the second anatomical structure and the third anatomy The structure can refer to any part of the left and right symmetrical structures, for example, the first anatomical structure can be the left cerebellum, the second anatomical structure can be the left ventricle, the third anatomical structure can be the right ventricle, etc.; An anatomical structure, a second anatomical structure and a third anatomical structure can also refer to the whole of the above-mentioned anatomical structures, for example, the first anatomical structure can be the cerebellum, the second anatomical structure can be the Sylvian fissure, and the third anatomical structure can also be the lateral cerebrum crack.
由于颅骨的骨性结构的透声性较差,靠近骨头后方组织结构往往显示不清,使得颅脑图像近场(靠近超声探头一侧)信息丢失严重,在胎儿颅脑近场关键解剖结构检查时,医生只能通过临床经验并借助颅脑图像的远场(远离超声探头一侧)信息进行诊断,有很大的局限性和弊端,易造成 误诊。参看图13,示出了根据本申请的一个实施例的超声检测方法中在显示器中显示的胎儿颅脑的超声图像的示意图。如图13所示,在胎儿颅脑的超声图像1300中显示有胎儿颅脑的颅骨1301和位于颅骨1301内侧的解剖结构,超声图像上方为靠近超声探头的方向。由于颅骨的骨性结构的透声性较差,靠近骨头后方组织结构往往显示不清,其中,超声图像靠近超声探头一侧信息丢失严重,而远离超声探头一侧的信息较为清晰,如图14中显示的对应于靠近超声探头一侧的超声图像1300中颅骨1301包围的上部区域的超声图像显示不全,对应于远离超声探头一侧的超声图像1300中颅骨1301包围的下部区域的超声图像较为完整。本申请的第二个实施例提供了一种超声检测方法,基于颅骨的对称性,根据远离超声探头一侧采集的超声图像中的解剖结构的位置确定靠近超声探头一侧的超声图像中相应的解剖结构可能存在的位置,从而为医生提供进行进一步检测的指引。Due to the poor sound permeability of the bony structure of the skull, the tissue structures near the back of the bone are often not clearly displayed, resulting in serious loss of information in the near field of the cranial image (closer to the ultrasound probe side). At that time, doctors can only make a diagnosis through clinical experience and with the help of the far-field (far away from the ultrasound probe) information of the cranial image, which has great limitations and drawbacks, and is easy to cause misdiagnosis. Referring to FIG. 13 , a schematic diagram of an ultrasound image of a fetal skull displayed on a display in an ultrasound detection method according to an embodiment of the present application is shown. As shown in FIG. 13 , in an ultrasound image 1300 of the fetal skull, a skull 1301 of the fetal skull and an anatomical structure located inside the skull 1301 are displayed, and the top of the ultrasound image is a direction close to the ultrasound probe. Due to the poor sound permeability of the bony structure of the skull, the tissue structure near the back of the bone is often not clearly displayed. Among them, the information on the side of the ultrasound image close to the ultrasound probe is seriously lost, while the information on the side far from the ultrasound probe is clearer, as shown in Figure 14 The ultrasound image corresponding to the upper region surrounded by the skull 1301 in the ultrasound image 1300 on the side close to the ultrasound probe shown in is incomplete, and the ultrasound image corresponding to the lower region surrounded by the skull 1301 in the ultrasound image 1300 on the side away from the ultrasound probe is relatively complete. . A second embodiment of the present application provides an ultrasonic detection method, which, based on the symmetry of the skull, determines the corresponding anatomical structure in the ultrasonic image on the side close to the ultrasonic probe according to the position of the anatomical structure in the ultrasonic image collected on the side away from the ultrasonic probe where anatomical structures may be present, thus providing guidance to physicians for further testing.
图12示出了根据本申请的一个实施例的超声检测方法1200。如图12所示,超声检测方法1200可以包括如下步骤:FIG. 12 shows an ultrasonic inspection method 1200 according to one embodiment of the present application. As shown in FIG. 12, the ultrasonic detection method 1200 may include the following steps:
图像获取步骤S1210,获取胎儿颅脑的超声图像。The image acquisition step S1210 is to acquire an ultrasound image of the fetal skull.
图像获取步骤S1210中,采用超声成像***采集超声图像。参看图2示出了根据一个实施例的超声成像***的示意性框图。其中,超声成像***200包括超声探头201和图像处理单元202,图像处理单元包括处理器2022、存储器2021以及显示器2023。超声探头201用于向受测胎儿颅脑发射超声波并接收超声回波,得到超声回波信号。处理器2022用于对超声回波信号进行处理,得到关于胎儿颅脑的超声图像,并基于该超声图像对胎儿颅脑进行检查。存储器2021用于存储超声图像。显示器2023用于显示超声图像。In the image acquisition step S1210, an ultrasonic imaging system is used to acquire an ultrasonic image. Referring to FIG. 2, a schematic block diagram of an ultrasound imaging system according to one embodiment is shown. The ultrasound imaging system 200 includes an ultrasound probe 201 and an image processing unit 202 , and the image processing unit includes a processor 2022 , a memory 2021 and a display 2023 . The ultrasonic probe 201 is used to transmit ultrasonic waves to the skull of the tested fetus and receive ultrasonic echoes to obtain ultrasonic echo signals. The processor 2022 is configured to process the ultrasonic echo signals to obtain an ultrasonic image of the fetal skull, and to examine the fetal skull based on the ultrasonic image. The memory 2021 is used to store ultrasound images. The display 2023 is used to display ultrasound images.
根据本申请的一个实施例中,图像获取步骤获取胎儿颅脑的超声图像,其中超声图像包括胎儿颅脑的二维图像,或者胎儿颅脑的三维图像。胎儿颅脑的二维图像通过调整胎儿***、超声探头的位置等获得通过的胎儿颅脑的切面图像,例如小脑切面、侧室脑切面、丘脑切面等。According to an embodiment of the present application, the image acquisition step acquires an ultrasound image of the fetal skull, wherein the ultrasound image includes a two-dimensional image of the fetal skull or a three-dimensional image of the fetal skull. The two-dimensional image of the fetal skull is obtained by adjusting the fetal position, the position of the ultrasound probe, etc. to obtain the sliced images of the fetal skull, such as the cerebellar slice, the lateral ventricle brain slice, and the thalamus slice.
对称性标识获取步骤S1220,获取所述超声图像中所述胎儿颅脑的对称性标识。Symmetry identification acquisition step S1220, acquiring the symmetry identification of the fetal skull in the ultrasound image.
对称性标识获取步骤S1220中,当超声图像为胎儿颅脑的二维图像时, 对称性标识为对应于切面图像的脑中线,当超声图像为胎儿颅脑的三维图像时,对称性标识为对应于三维图像的正中矢状面。In the step S1220 of obtaining the symmetry identification, when the ultrasound image is a two-dimensional image of the fetal skull, the symmetry is identified as the midline of the brain corresponding to the sectional image, and when the ultrasound image is a three-dimensional image of the fetal skull, the symmetry is identified as corresponding to the brain midline. in the midsagittal plane of the 3D image.
在本申请的一个实施例中,在对称性标识获取步骤S1220中,医生根据经验手动获取所述对称性标识。例如,在一个示例中,当超声图像为胎儿颅脑的二维图像时,获得二维图像的脑中线的方法采用通过在获取超声图像选取位于脑中线上的两个点并进行标记,连接这两个点形成直线,该直线作为脑中线。在另一个示例中,当超声图像为胎儿颅脑的三维图像时,获取正中矢状面的方法采用将医生通过手动旋转、平移等集合变换,从三维图像中三个互相正交的剖面图像中得到正中矢状面。In an embodiment of the present application, in the step S1220 of obtaining the symmetry identification, the doctor manually obtains the symmetry identification according to experience. For example, in one example, when the ultrasound image is a two-dimensional image of the fetal skull, the method of obtaining the midline of the brain in the two-dimensional image is to select two points on the midline of the brain during the acquisition of the ultrasound image and mark them to connect the two points. The two points form a straight line, which serves as the midline of the brain. In another example, when the ultrasound image is a three-dimensional image of the fetal skull, the method for obtaining the midsagittal plane is to transform the doctor through manual rotation, translation, etc., to obtain three mutually orthogonal cross-sectional images in the three-dimensional image. Obtain the midsagittal plane.
在本申请的一个实施例中,在对称性标识获取步骤S1220中,采用处理器处理所述超声图像获得对称性标识。In an embodiment of the present application, in the step S1220 of obtaining the symmetry identification, a processor is used to process the ultrasound image to obtain the symmetry identification.
在一个示例中,所述图像获取步骤包括获取胎儿颅脑的二维图像;在所述对称性标识获取步骤中,获取所述脑中线的方法包括以下至少一项:In one example, the image acquisition step includes acquiring a two-dimensional image of the fetal skull; in the symmetry marker acquisition step, the method for acquiring the midline of the brain includes at least one of the following:
通过直线检测算法在所述二维图像中确定脑中线;determining the midline of the brain in the two-dimensional image by a line detection algorithm;
通过结构检测算法在所述二维图像中检测特征解剖结构,将检测到的特征解剖结构拟合为直线,将拟合的所述直线确定为脑中线;The characteristic anatomical structure is detected in the two-dimensional image by a structure detection algorithm, the detected characteristic anatomical structure is fitted as a straight line, and the fitted straight line is determined as the midline of the brain;
通过预训练的机器学习模型从所述二维图像中确定脑中线。The midline of the brain is determined from the two-dimensional images by a pretrained machine learning model.
在一个示例中,所述图像获取步骤包括:获取胎儿颅脑的三维图像;In one example, the image acquisition step includes: acquiring a three-dimensional image of the fetal skull;
在所述对称性标识获取步骤中,获取所述正中矢状面的方法包括以下至少一项:In the step of obtaining the symmetry mark, the method for obtaining the midsagittal plane includes at least one of the following:
通过面检测算法在所述三维图像中确定正中矢状面;determining a midsagittal plane in the three-dimensional image by a plane detection algorithm;
通过结构检测算法所述三维图像中检测特征解剖结构,将检测到的特征解剖结构拟合为平面,将所述拟合的所述平面确定为正中矢状面;The characteristic anatomical structure is detected in the three-dimensional image by the structure detection algorithm, the detected characteristic anatomical structure is fitted as a plane, and the fitted plane is determined as the midsagittal plane;
通过预训练的机器学习模型从所述三维图像中确定正中矢状面。The midsagittal plane is determined from the three-dimensional image by a pretrained machine learning model.
示例性的,基于超声图像的图像特征,采用直线检测算法检测所述脑中线,或者采用平面检测算法检测所述正中矢状面。由于胎儿颅脑关于脑中线对称或胎儿颅脑的三维图像关于正中矢状面对称,在超声图像中中的脑中线或正中矢状面往往具有明显的特征。基于该特征,采用直线检测方法检测二维图像的脑中线或者采用平面检测的方法检测三维图像的正中矢 状面。例如,在获取脑中线的过程中,提取代表满足线上的灰度值大于两侧的灰度值的线的脑中线特征区域,根据提取的特征区域上的点进行直线检测;同样在获取正中矢状面的过程中,可以提取代表满足平面上的灰度值大于两侧的灰度值的面的正中矢状面特征区域,再根据平面检测算法获得正中矢状面。直线检测算法包括但不限于,Hough变换、LSD快速之间检测算法、EDlines直线检测算法、Radon变换、LSWMS直线检测算法、CannyLines直线检测算法等。平面检测算法包括但不限于,加权Hough变换、随机Hough变换、Radon变换等。Exemplarily, based on the image features of the ultrasound image, a straight line detection algorithm is used to detect the midline of the brain, or a plane detection algorithm is used to detect the midsagittal plane. Since the fetal skull is symmetrical about the midline or the three-dimensional image of the fetal skull is symmetrical about the midsagittal plane, the midline or midsagittal plane in the ultrasound image tends to have distinct features. Based on this feature, the line detection method is used to detect the midline of the brain in the two-dimensional image, or the plane detection method is used to detect the midsagittal plane of the three-dimensional image. For example, in the process of acquiring the midline of the brain, extract the midline feature area representing the line that satisfies the gray value of the line is greater than the gray value of the two sides, and perform straight line detection according to the points on the extracted feature area; also in the acquisition center In the process of sagittal plane, the midsagittal plane feature area representing the plane whose gray value is greater than the gray value on both sides can be extracted, and then the midsagittal plane can be obtained according to the plane detection algorithm. Straight line detection algorithms include, but are not limited to, Hough transform, LSD fast inter-detection algorithm, EDlines straight line detection algorithm, Radon transform, LSWMS straight line detection algorithm, CannyLines straight line detection algorithm, and the like. Plane detection algorithms include, but are not limited to, weighted Hough transform, random Hough transform, Radon transform, and the like.
示例性的,基于结构检测算法检测位于脑中线的特征解剖结构,对特征解剖结构进行直线拟合得到脑中线,或基于结构检测算法检测位于正中矢状面上的特征解剖结构,并对进行特征解剖结构进行平面拟合得到正中矢状面。例如,当获取的超声图像是胎儿颅脑的小脑切面时,位于脑中线上的特殊解剖结构包括小脑、透明隔腔、小脑蚓部等,当获取的超声图像是丘脑的切面图像时,位于脑中线上的特殊解剖结构包括丘脑、透明隔腔等,根据获得的小脑、透明隔腔和小脑蚓部或丘脑和透明隔腔对其上的点进行直线拟合得到脑中线。当获取的超声图像是胎儿颅脑的三维图像时,位于正中矢状面上的特征解剖结构包括胼胝体、小脑蚓部、透明隔腔等,通过对胼胝体、小脑蚓部、透明隔腔上的点进行平面拟合得到正中矢状面。常用的结构检测方法有包括但不限于,大津阈值算法(OSTU)、水平集(LevelSet)等算法。Exemplarily, the characteristic anatomical structures located on the midline of the brain are detected based on a structure detection algorithm, and the midline of the brain is obtained by performing straight line fitting on the characteristic anatomical structures, or the characteristic anatomical structures located on the midsagittal plane are detected based on the structure detection algorithm, and the characteristic anatomical structures are detected. The anatomical structure was plane fitted to obtain the midsagittal plane. For example, when the acquired ultrasound image is the cerebellum section of the fetal skull, the special anatomical structures located on the midline of the brain include the cerebellum, septum pellucidum, cerebellar vermis, etc. When the acquired ultrasound image is the section image of the thalamus, the special anatomical structures located in the midline of the brain include The special anatomical structures on the midline include the thalamus, septum pellucidum, etc. The midline of the brain is obtained by straight-line fitting of the points on the obtained cerebellum, septum pellucidum and cerebellar vermis or thalamus and septum pellucidum. When the acquired ultrasound image is a three-dimensional image of the fetal skull, the characteristic anatomical structures located on the midsagittal plane include the corpus callosum, cerebellar vermis, septum pellucidum, etc. A plane fitting was performed to obtain the midsagittal plane. Commonly used structure detection methods include, but are not limited to, Otsu threshold algorithm (OSTU), level set (LevelSet) and other algorithms.
示例性的,通过预训练的机器学*** 面方程作为输出数据。具体方式请参见上文的相关叙述,在此不再赘述。Exemplarily, the midline or midsagittal plane of the brain is determined by a pretrained machine learning model. For example, the method for determining the midline of the brain by pre-trained machine learning includes inputting the ultrasound image as input data into a pre-trained machine learning model for calculation to obtain a straight line equation of the midline of the brain as output data, wherein the pre-trained machine The learning model is trained using a preset database of calibrated ultrasound images including at least one linear equation of the midline of the brain that has been calibrated. As the output data, and the model parameters of the machine learning model are optimized in the process of training the machine learning model. The process of determining the midsagittal plane through the pre-trained machine learning model is similar to the process of determining the midline of the brain through the pre-trained machine learning model. The plane equation of the midsagittal plane is obtained as output data. For the specific manner, please refer to the above related description, which will not be repeated here.
在本申请的一个实施例中,获取对称性标识之后,在显示超声图像的显示器中显示对称性标识。例如,在一个实施例中,采用明显区别于超声图像的颜色显示脑中线或正中矢状面。例如,在显示器显示超声图像的同时,在超声图像中将脑中线显示为红色直线,或者,将正中矢状面显示为红色阴影标记的平面。参看图14,示出了根据本申请的一个实施例中在显示器中将对称性标识显示在胎儿颅脑的超声图像中的示意图。如图14所示,在胎儿颅脑的超声图像1400中将脑中线1404显示为以实线线型标识的直线。In an embodiment of the present application, after the symmetry identification is acquired, the symmetry identification is displayed on the display that displays the ultrasound image. For example, in one embodiment, the midline or midsagittal plane of the brain is displayed in a distinct color from the ultrasound image. For example, the midline of the brain may be displayed as a red straight line in the ultrasound image while the ultrasound image is displayed on the monitor, or the midsagittal plane may be displayed as a plane marked with red shading. Referring to FIG. 14, a schematic diagram of displaying a symmetry marker in an ultrasound image of a fetal skull in a display according to an embodiment of the present application is shown. As shown in FIG. 14, the mid-brain line 1404 is shown as a straight line identified in a solid line style in an ultrasound image 1400 of the fetal skull.
识别步骤S1230,在所述超声图像上识别至少一个至少部分位于所述对称性标识的一侧的解剖结构,获取所述解剖结构的轮廓或所述解剖结构所在的区域。Identifying step S1230, identifying at least one anatomical structure at least partially located on one side of the symmetry mark on the ultrasound image, and acquiring the outline of the anatomical structure or the region where the anatomical structure is located.
在识别步骤S1230中,采用处理器识别超声图像中的解剖结构。In the identifying step S1230, a processor is used to identify anatomical structures in the ultrasound image.
在本申请的一个实施例中,所述识别步骤S1230中,识别至少一个至少部分位于所述对称性标识的一侧的解剖结构包括:在所述对称性标识远离超声探头采集端的一侧的超声图像中识别所述解剖结构。如图14所示,识别超声图像1400中在对称性标识远离箭头B的一侧的至少一个解剖结构,即识别位于脑中线1404下方区域的解剖结构,如解剖结构1402和解剖结构1403。In an embodiment of the present application, in the identifying step S1230, identifying at least one anatomical structure at least partially located on one side of the symmetry mark includes: ultrasonic waves on a side of the symmetry mark far from the acquisition end of the ultrasound probe Identify the anatomical structure in the image. As shown in FIG. 14 , at least one anatomical structure in the ultrasound image 1400 on the side away from arrow B is identified symmetrically, ie, anatomical structures located in the area below the midline 1404 of the brain, such as anatomical structures 1402 and 1403 .
在对称性标识远离超声探头采集端的一侧的超声图像中识别所述解剖结构,可以通过用户手动识别,也可以通过超声设备自动识别。对于超声设备自动识别对称性标识远离超声探头采集端的一侧的超声图像中的解剖结构,首先,可以自动识别超声图像中的超声探头采集端,其后,自动在对称性标识远离超声探头采集端的一侧的超声图像中识别解剖结构。一种方式中,可以通过超声图像的形状确定超声探头采集端,例如图13所示,超声图像多为扇形图像,其中超声图像在探头采集端的圆弧较小,远离探头采集端的圆弧较大,可以通过延长超声图像的两侧边(即图13中,扇形超声图像的左右两条斜边),使两侧边的延长线交于一点,则该超声图像靠近该点的一端为超声探头采集端,对称性标识靠近该点的一侧即为对称性标识靠近超声探头采集端的一侧,相应的,对称性标识远离该点的一侧即 为对称性标识远离超声探头采集端的一侧。其他方式中,也可以通过与超声图像一同生成的相关文件确定超声图像的探头采集端,在与超声图像一同生成的相关文件中,存储有生成该超声图像的各种数据,其中包括了超声探头相对于该超声图像的位置信息,可以直接调取该文件中的位置信息,确定超声图像的探头采集端,并进一步确定对称性标识远离超声探头采集端的一侧的超声图像,以对该测超声图像进行关键解剖结构的自动识别。The anatomical structure is identified in the ultrasound image with the symmetry marking the side away from the acquisition end of the ultrasound probe, which can be manually identified by a user or automatically identified by an ultrasound device. For the ultrasonic equipment to automatically identify the anatomical structure in the ultrasonic image on the side with the symmetry mark away from the acquisition end of the ultrasonic probe, firstly, it can automatically identify the acquisition end of the ultrasonic probe in the ultrasonic image, and then automatically identify the anatomical structure in the ultrasonic image on the side where the symmetry mark is far away from the acquisition end of the ultrasonic probe. Identify anatomical structures in an ultrasound image of one side. In one way, the ultrasonic probe acquisition end can be determined by the shape of the ultrasonic image. For example, as shown in Figure 13, the ultrasonic images are mostly fan-shaped images, in which the ultrasonic image has a smaller arc at the probe acquisition end, and a larger arc away from the probe acquisition end. , by extending the two sides of the ultrasound image (that is, the left and right hypotenuses of the fan-shaped ultrasound image in Figure 13), so that the extension lines of the two sides intersect at a point, then the end of the ultrasound image close to this point is the ultrasound probe At the acquisition end, the side of the symmetry mark close to the point is the side of the symmetry mark close to the acquisition end of the ultrasound probe, and correspondingly, the side of the symmetry mark away from this point is the side of the symmetry mark away from the acquisition end of the ultrasound probe. In other ways, the probe acquisition end of the ultrasonic image can also be determined by the related file generated together with the ultrasonic image. In the related file generated together with the ultrasonic image, various data for generating the ultrasonic image are stored, including the ultrasonic probe. Relative to the position information of the ultrasonic image, the position information in the file can be directly retrieved, the probe acquisition end of the ultrasonic image can be determined, and the ultrasonic image with the symmetry mark on the side away from the ultrasonic probe acquisition end can be further determined, so as to measure the ultrasonic image. Images for automatic identification of key anatomical structures.
识别至少一个解剖结构的方法包括但不限于,通过图像分割算法,在所述超声图像上分割得到至少一个所述第二解剖结构的轮廓;通过预训练的机器学***集方法(Level Set)、主动轮廓模型(Snake)、图割(Graph Cut)、区域生长(Region Growing)等方法。机器学习模型包括:SVM、KNN、PCANet等。深度学习网络包括:RCNN、Fast RCNN、FCN、UNet等。在本申请的一个实施例中,在所述识别步骤S1230中,识别解剖结构之后,获取解剖结构的轮廓或解剖结构所在的区域的方法包括但不限于提取解剖结构的边界或者回归包含解剖结构的ROI区域等。The method for identifying at least one anatomical structure includes, but is not limited to, obtaining at least one contour of the second anatomical structure by segmenting the ultrasound image through an image segmentation algorithm; The contour of the second anatomical structure or the region where the second anatomical structure is located. The image segmentation algorithm includes detecting anatomical structures through grayscale detection or segmentation methods, and machine learning includes detecting anatomical structures based on traditional machine learning methods and methods based on deep learning. Exemplary, grayscale detection or segmentation methods include, but are not limited to, Otsu threshold algorithm (OSTU), level set method (Level Set), active contour model (Snake), Graph Cut (Graph Cut), Region Growing (Region Growing) and other methods. Machine learning models include: SVM, KNN, PCANet, etc. Deep learning networks include: RCNN, Fast RCNN, FCN, UNet, etc. In an embodiment of the present application, in the identifying step S1230, after identifying the anatomical structure, the method for obtaining the outline of the anatomical structure or the region where the anatomical structure is located includes, but is not limited to, extracting the boundary of the anatomical structure or returning the anatomical structure containing the anatomical structure. ROI area, etc.
投影步骤S1240,获取所述解剖结构的轮廓或所述解剖结构所在的区域关于所述对称性标识在所述对称性标识另一侧对称的投影标记。Projecting step S1240, acquiring the outline of the anatomical structure or a projection mark that is symmetrical on the other side of the symmetry marker with respect to the symmetry marker in the region where the anatomical structure is located.
投影步骤S1240中,采用处理器将在识别步骤1230中识别的位于对称性标识一侧的超声图像中的解剖结构投影到对称性标识的另一侧。In the projection step S1240, the processor is used to project the anatomical structure in the ultrasound image on one side of the symmetry mark identified in the identification step 1230 to the other side of the symmetry mark.
在本申请的一个实施例中,所述投影标记包括对应于所述第二解剖结构的轮廓的轮廓标记,和/或对应于所述第二解剖结构的所在区域的区域标记。在一个示例中,在投影步骤S1240中,获取所述解剖结构的轮廓或所述解剖结构的所在区域关于所述对称性标识对称的所述投影标记的方法包括:将解剖结构的轮廓或解剖结构的所在区域关于对称性标识进行对称变换,获得与解剖结构的轮廓或解剖结构的所在区域相对应的轮廓或者区域。In one embodiment of the present application, the projection markers include contour markers corresponding to the contour of the second anatomical structure, and/or area markers corresponding to the region where the second anatomical structure is located. In one example, in the projecting step S1240, the method for obtaining the outline of the anatomical structure or the projection mark that is symmetrical with respect to the symmetry marker in the region where the anatomical structure is located includes: converting the outline or the anatomical structure of the anatomical structure Symmetric transformation is performed on the region where the anatomical structure is located, and the contour or region corresponding to the contour of the anatomical structure or the region where the anatomical structure is located is obtained.
显示步骤S1250,显示所述投影标记。The display step S1250 is to display the projection mark.
在显示步骤S1250中,采用显示器将投影步骤S1240中获得的投影标记显示在超声图像上。In the display step S1250, the projection marker obtained in the projection step S1240 is displayed on the ultrasound image using a display.
在本申请一个实施例中,所述轮廓标记包括轮廓边界标记、轮廓阴影标记和/或轮廓箭头标记;所述区域标记包括区域边界标记、区域阴影标记和/或区域箭头标记。图15和图16示出了根据本申请的一个实施例的超声检测方法,该超声检查方法包括在胎儿颅脑的超声图像中显示投影标记的示意图。如图15所示,对应于解剖结构1402的轮廓,解剖结构1402在脑中线1404的另一侧的投影标记显示为轮廓边界标记14021,对应于解剖结构1403的轮廓,解剖结构1403在脑中线1404的另一侧的投影标记显示为轮廓箭头标记14031。如图16所示,对应于解剖结构1402的轮廓,解剖结构1402在脑中线1404的另一侧的投影标记显示为轮廓边界标记14022,对应于解剖结构1403所在的轮廓,解剖结构1403在脑中线1404的另一侧的投影标记显示为轮廓阴影标记14032。In an embodiment of the present application, the contour marks include contour boundary marks, contour shadow marks and/or contour arrow marks; the area marks include area border marks, area shadow marks and/or area arrow marks. Figures 15 and 16 illustrate a method of ultrasound inspection according to one embodiment of the present application, the method of ultrasound inspection comprising a schematic diagram showing projection markers in an ultrasound image of a fetal skull. As shown in FIG. 15 , corresponding to the outline of anatomy 1402 , the projected marker of anatomy 1402 on the other side of midline 1404 is shown as outline boundary marker 14021 , corresponding to the outline of anatomy 1403 , which is on midline 1404 The projected markers on the other side of the are shown as outline arrow markers 14031. As shown in Figure 16, corresponding to the outline of anatomical structure 1402, the projected marker of anatomical structure 1402 on the other side of midline 1404 is shown as outline boundary marker 14022, corresponding to the outline of anatomical structure 1403 on the midline of the brain The projected indicia on the other side of 1404 are shown as outline shadow indicia 14032.
在本申请一个实施例中,当所述投影标记为多个时,区分显示所述多个投影标记。在一个示例中,所述区分显示的方法包括:对应于每一所述第二解剖结构将所述投影标记进行颜色区分;或者对应于每一所述第二解剖结构将所述投影标记进行线型区分;或者对应于每一所述第二解剖结构将所述投影标记进行阴影区分。图17示出了根据本申请的一个实施例中在显示器中显示的胎儿颅脑的超声图像中区分显示投影标记的示意图。如图17所示,对应于解剖结构1402和解剖结构1403,解剖结构1402和解剖结构1403的轮廓在脑中线1404的另一侧的投影标记显示为轮廓阴影标记14023和轮廓阴影标记14033,其中,轮廓阴影标记14023和轮廓阴影标记14033的阴影图案不同。In an embodiment of the present application, when there are multiple projection marks, the multiple projection marks are displayed in a differentiated manner. In one example, the method for differentiating display includes: color-distinguishing the projection marks corresponding to each of the second anatomical structures; or lining the projection marks corresponding to each of the second anatomical structures type distinction; or shadow distinction is performed on the projection markers corresponding to each of the second anatomical structures. FIG. 17 shows a schematic diagram of distinguishing and displaying projection markers in an ultrasound image of a fetal skull displayed on a display according to an embodiment of the present application. As shown in FIG. 17 , corresponding to anatomical structure 1402 and anatomical structure 1403 , the projection marks of the contours of anatomical structure 1402 and anatomical structure 1403 on the other side of the midline of the brain 1404 are shown as contour shadow mark 14023 and contour shadow mark 14033 , wherein, The shading patterns of the outline shading mark 14023 and the outline shading mark 14033 are different.
在本申请一个实施例中,在显示步骤S1250之后,医生可以根据投影标记,继续对投影标记覆盖的区域进行超声检测。In an embodiment of the present application, after step S1250 is displayed, the doctor may continue to perform ultrasonic detection on the area covered by the projection marker according to the projection marker.
本申请的第三个实施例提供了一种超声成像***,包括:A third embodiment of the present application provides an ultrasound imaging system, including:
超声探头,用于对受测胎儿颅脑发射超声波并接收超声回波,得到超声回波信号;The ultrasonic probe is used to transmit ultrasonic waves to the brain of the tested fetus and receive ultrasonic echoes to obtain ultrasonic echo signals;
处理器,用于对超声回波信号进行处理,得到所述胎儿颅脑的超声图 像;a processor for processing the ultrasonic echo signals to obtain an ultrasonic image of the fetal skull;
显示器,用于显示所述超声图像;a display for displaying the ultrasound image;
存储器,用于存储可执行的程序指令;memory for storing executable program instructions;
所述处理器,还用于基于所述可执行的程序指令执行本申请实施The processor is further configured to execute the implementation of the present application based on the executable program instructions
例的方法。example method.
参看图2示出了根据一个实施例的超声成像***的示意性框图。其中,超声成像***200包括超声探头201和图像处理单元202,图像处理单元包括处理器2022、存储器2021以及显示器2023。超声探头201用于向受测胎儿颅脑发射超声波并接收超声回波,得到超声回波信号。处理器2022用于对超声回波信号进行处理,得到关于胎儿颅脑的超声图像,并基于该超声图像对胎儿颅脑进行超声检测。存储器2021用于存储超声图像。显示器2023用于显示超声图像。Referring to FIG. 2, a schematic block diagram of an ultrasound imaging system according to one embodiment is shown. The ultrasound imaging system 200 includes an ultrasound probe 201 and an image processing unit 202 , and the image processing unit includes a processor 2022 , a memory 2021 and a display 2023 . The ultrasonic probe 201 is used to transmit ultrasonic waves to the skull of the tested fetus and receive ultrasonic echoes to obtain ultrasonic echo signals. The processor 2022 is configured to process the ultrasonic echo signals to obtain an ultrasonic image of the fetal skull, and perform ultrasonic detection on the fetal skull based on the ultrasonic image. The memory 2021 is used to store ultrasound images. The display 2023 is used to display ultrasound images.
其中,存储器2021还用于存储可执行的程序指令,处理器2022基于可执行的程序指令执行本申请实施例的的方法。The memory 2021 is further configured to store executable program instructions, and the processor 2022 executes the methods of the embodiments of the present application based on the executable program instructions.
尽管这里已经参考附图描述了示例实施例,应理解上述示例实施例仅仅是示例性的,并且不意图将本申请的范围限制于此。本领域普通技术人员可以在其中进行各种改变和修改,而不偏离本申请的范围和精神。所有这些改变和修改意在被包括在所附权利要求所要求的本申请的范围之内。Although example embodiments have been described herein with reference to the accompanying drawings, it should be understood that the above-described example embodiments are exemplary only, and are not intended to limit the scope of the application thereto. Various changes and modifications may be made therein by those of ordinary skill in the art without departing from the scope and spirit of the present application. All such changes and modifications are intended to be included within the scope of this application as claimed in the appended claims.
本领域普通技术人员可以意识到,结合本文中所公开的实施例描述的各示例的单元及算法步骤,能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟以硬件还是软件方式来执行,取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能,但是这种实现不应认为超出本申请的范围。Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Skilled artisans may implement the described functionality using different methods for each particular application, but such implementations should not be considered beyond the scope of this application.
在本申请所提供的几个实施例中,应该理解到,所揭露的设备和方法,可以通过其它的方式实现。例如,以上所描述的设备实施例仅仅是示意性的,例如,所述单元的划分,仅仅为一种逻辑功能划分,实际实现时可以有另外的划分方式,例如多个单元或组件可以结合或者可以集成到另一个设备,或一些特征可以忽略,或不执行。In the several embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or May be integrated into another device, or some features may be omitted, or not implemented.
在此处所提供的说明书中,说明了大量具体细节。然而,能够理解,本申请的实施例可以在没有这些具体细节的情况下实践。在一些实例中,并未详细示出公知的方法、结构和技术,以便不模糊对本说明书的理解。In the description provided herein, numerous specific details are set forth. It will be understood, however, that the embodiments of the present application may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
类似地,应当理解,为了精简本申请并帮助理解各个发明方面中的一个或多个,在对本申请的示例性实施例的描述中,本申请的各个特征有时被一起分组到单个实施例、图、或者对其的描述中。然而,并不应将该本申请的方法解释成反映如下意图:即所要求保护的本申请要求比在每个权利要求中所明确记载的特征更多的特征。更确切地说,如相应的权利要求书所反映的那样,其发明点在于可以用少于某个公开的单个实施例的所有特征的特征来解决相应的技术问题。因此,遵循具体实施方式的权利要求书由此明确地并入该具体实施方式,其中每个权利要求本身都作为本申请的单独实施例。Similarly, it is to be understood that in the description of the exemplary embodiments of the present application, various features of the present application are sometimes grouped together into a single embodiment, FIG. , or in its description. However, this method of application should not be construed as reflecting an intention that the claimed application requires more features than are expressly recited in each claim. Rather, as the corresponding claims reflect, the invention lies in the fact that the corresponding technical problem may be solved with less than all features of a single disclosed embodiment. Thus, the claims following the Detailed Description are hereby expressly incorporated into this Detailed Description, with each claim standing on its own as a separate embodiment of this application.
本领域的技术人员可以理解,除了特征之间相互排斥之外,可以采用任何组合对本说明书(包括伴随的权利要求、摘要和附图)中公开的所有特征以及如此公开的任何方法或者设备的所有过程或单元进行组合。除非另外明确陈述,本说明书(包括伴随的权利要求、摘要和附图)中公开的每个特征可以由提供相同、等同或相似目的替代特征来代替。It will be understood by those skilled in the art that all features disclosed in this specification (including the accompanying claims, abstract and drawings) and any method or apparatus so disclosed may be used in any combination, except that the features are mutually exclusive. Processes or units are combined. Each feature disclosed in this specification (including accompanying claims, abstract and drawings) may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.
此外,本领域的技术人员能够理解,尽管在此所述的一些实施例包括其它实施例中所包括的某些特征而不是其它特征,但是不同实施例的特征的组合意味着处于本申请的范围之内并且形成不同的实施例。例如,在权利要求书中,所要求保护的实施例的任意之一都可以以任意的组合方式来使用。Furthermore, those skilled in the art will appreciate that although some of the embodiments described herein include certain features, but not others, included in other embodiments, that combinations of features of different embodiments are intended to be within the scope of the present application within and form different embodiments. For example, in the claims, any of the claimed embodiments may be used in any combination.
本申请的各个部件实施例可以以硬件实现,或者以在一个或者多个处理器上运行的软件模块实现,或者以它们的组合实现。本领域的技术人员应当理解,可以在实践中使用微处理器或者数字信号处理器(Digital Signal Processor,DSP)来实现根据本申请实施例的一些模块的一些或者全部功能。本申请还可以实现为用于执行这里所描述的方法的一部分或者全部的装置程序(例如,计算机程序和计算机程序产品)。这样的实现本申请的程序可以存储在计算机可读介质上,或者可以具有一个或者多个信号的形式。这样的信号可以从因特网网站上下载得到,或者在载体信号上提供,或者 以任何其他形式提供。Various component embodiments of the present application may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art should understand that a microprocessor or a digital signal processor (Digital Signal Processor, DSP) may be used in practice to implement some or all of the functions of some modules according to the embodiments of the present application. The present application can also be implemented as a program of apparatus (eg, computer programs and computer program products) for performing part or all of the methods described herein. Such a program implementing the present application may be stored on a computer-readable medium, or may be in the form of one or more signals. Such signals may be downloaded from Internet sites, or provided on carrier signals, or in any other form.
应该注意的是上述实施例对本申请进行说明而不是对本申请进行限制,并且本领域技术人员在不脱离所附权利要求的范围的情况下可设计出替换实施例。在权利要求中,不应将位于括号之间的任何参考符号构造成对权利要求的限制。本申请可以借助于包括有若干不同元件的硬件以及借助于适当编程的计算机来实现。在列举了若干装置的单元权利要求中,这些装置中的若干个可以是通过同一个硬件项来具体体现。单词第一、第二、以及第三等的使用不表示任何顺序。可将这些单词解释为名称。It should be noted that the above-described embodiments illustrate rather than limit the application, and alternative embodiments may be devised by those skilled in the art without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The application can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by one and the same item of hardware. The use of the words first, second, and third, etc. do not denote any order. These words can be interpreted as names.
以上所述,仅为本申请的具体实施方式或对具体实施方式的说明,本申请的保护范围并不局限于此,任何熟悉本技术领域的技术人员在本申请揭露的技术范围内,可轻易想到变化或替换,都应涵盖在本申请的保护范围之内。本申请的保护范围应以权利要求的保护范围为准。The above are only specific embodiments of the present application or descriptions of the specific embodiments, and the protection scope of the present application is not limited thereto. Any changes or substitutions should be included within the protection scope of the present application. The protection scope of the present application shall be subject to the protection scope of the claims.
Claims (34)
- 一种超声检测方法,其特征在于,包括:An ultrasonic detection method, characterized in that, comprising:图像获取步骤,获取胎儿颅脑的超声图像;The image acquisition step is to acquire an ultrasound image of the fetal skull;对称性标识获取步骤,基于所述超声图像获取所述胎儿颅脑的对称性标识;The step of obtaining a symmetry mark, obtaining a symmetry mark of the fetal skull based on the ultrasound image;对称性评估指引步骤,获取位于所述对称性标识至少一侧的超声图像,基于所述至少一侧的超声图像获取所述胎儿颅脑关于所述对称性标识的对称性评估指引。The symmetry assessment guide step is to acquire an ultrasound image located on at least one side of the symmetry mark, and obtain a symmetry assessment guide of the fetal skull with respect to the symmetry mark based on the ultrasound image of the at least one side.
- 根据权利要求1所述的方法,其特征在于,所述对称性评估指引包括对称度。The method of claim 1, wherein the symmetry assessment guideline includes a degree of symmetry.
- 根据权利要求2所述的方法,其特征在于,所述对称性评估指引步骤包括:The method according to claim 2, wherein the symmetry assessment guide step comprises:在所述对称性标识一侧获取第一区域超声图像;acquiring an ultrasound image of the first region on one side of the symmetry marker;根据所述第一区域超声图像获取位于所述对称性标识另一侧的第二区域超声图像,所述第一区域的形状和所述第二区域的形状关于所述对称性标识对称;Acquiring an ultrasound image of a second region on the other side of the symmetry marker according to the first region ultrasound image, where the shape of the first region and the shape of the second region are symmetrical with respect to the symmetry marker;确定所述第一区域超声图像和所述第二区域超声图像关于所述对称性标识的对称度。A degree of symmetry of the ultrasound image of the first region and the ultrasound image of the second region with respect to the symmetry marker is determined.
- 根据权利要求3所述的方法,其特征在于,在所述对称性标识一侧获取所述第一区域超声图像的方法包括:The method according to claim 3, wherein the method for acquiring the ultrasound image of the first region on the side of the symmetry marker comprises:在所述超声图像中获取所述胎儿颅脑的颅骨,其中,将所述对称性标识一侧的所述颅骨包围的区域作为所述第一区域超声图像。The skull of the fetal skull is acquired in the ultrasound image, wherein a region surrounded by the skull on one side of the symmetry mark is used as the first region ultrasound image.
- 根据权利要求3所述的方法,其特征在于,在所述对称性标识一侧获取所述第一区域超声图像的方法包括:The method according to claim 3, wherein the method for acquiring the ultrasound image of the first region on the side of the symmetry marker comprises:在所述超声图像中获取感兴趣区域,所述感兴趣区域至少部分位于所述对称性标识的一侧,所述感兴趣区域位于所述对称性标识一侧的区域为所述第一区域超声图像。A region of interest is acquired in the ultrasound image, the region of interest is at least partially located on one side of the symmetry mark, and the region of the region of interest located on one side of the symmetry mark is the ultrasound of the first region image.
- 根据权利要求3所述的方法,其特征在于,在所述对称性标识一侧 获取所述第一区域超声图像的方法包括:The method according to claim 3, wherein the method for acquiring the ultrasonic image of the first region on the side of the symmetry mark comprises:在所述超声图像中识别至少一个第一解剖结构,所述至少一个第一解剖结构的至少部分位于所述对称性标识的一侧,所述至少一个第一解剖结构位于所述对称性标识一侧的区域为所述第一区域超声图像。At least one first anatomical structure is identified in the ultrasound image, at least a portion of the at least one first anatomical structure is located on one side of the symmetry marker, the at least one first anatomical structure is located on the side of the symmetry marker The area on the side is the ultrasound image of the first area.
- 根据权利要求3-6中任意一项所述的方法,其特征在于,在所述对称性标识一侧获取第一区域超声图像包括在所述对称性标识远离探头采集端的一侧获取第一区域超声图像。The method according to any one of claims 3-6, wherein acquiring the ultrasound image of the first region on the side of the symmetry marker comprises acquiring the first region on the side of the symmetry marker far from the probe acquisition end Ultrasound image.
- 根据权利要求2-6中任意一项所述的方法,其特征在于,所述对称性评估指引步骤包括:The method according to any one of claims 2-6, wherein the symmetry assessment guide step comprises:计算所述第一区域超声图像和所述第二区域超声图像关于所述对称性标识的对称点处的灰度绝对差值和/或平方差值,基于所述灰度绝对差值和/或平方差值确定所述对称度;和/或,calculating the absolute difference value and/or the square difference value of the grayscale at the point of symmetry of the ultrasound image of the first region and the ultrasound image of the second region with respect to the symmetry mark, based on the absolute grayscale difference value and/or The squared difference determines the degree of symmetry; and/or,计算所述第一区域超声图像和所述第二区域超声图像的结构相似度,基于所述结构相似度确定所述对称度;和/或,calculating the structural similarity between the ultrasound image of the first region and the ultrasound image of the second region, and determining the symmetry based on the structural similarity; and/or,计算所述第一区域超声图像和所述第二区域超声图像的相关比,基于所述相关比确定所述对称度。A correlation ratio of the ultrasound image of the first region and the ultrasound image of the second region is calculated, and the degree of symmetry is determined based on the correlation ratio.
- 根据权利要求1所述的方法,其特征在于,所述对称性评估指引包括至少部分位于所述对称性标识一侧的第二解剖结构关于所述对称性标识在所述对称性标识另一侧对称的投影标记。10. The method of claim 1, wherein the symmetry assessment guidelines include a second anatomical structure located at least partially on one side of the symmetry marker on the other side of the symmetry marker with respect to the symmetry marker Symmetrical projection markers.
- 根据权利要求9所述的方法,其特征在于,所述对称性评估指引步骤包括:The method according to claim 9, wherein the symmetry assessment guide step comprises:识别步骤,在所述超声图像上识别至少一个所述第二解剖结构,获取所述第二解剖结构的轮廓或所述第二解剖结构的所在区域;Identifying step, identifying at least one second anatomical structure on the ultrasound image, and acquiring the outline of the second anatomical structure or the region where the second anatomical structure is located;投影步骤,获取所述第二解剖结构的轮廓或所述第二解剖结构的所在区域关于所述对称性标识对称的所述投影标记;Projecting step, obtaining the outline of the second anatomical structure or the projection mark that is symmetrical with respect to the symmetry marker in the region where the second anatomical structure is located;显示步骤,显示所述投影标记。A display step displays the projected mark.
- 根据权利要求10所述的方法,其特征在于,所述投影标记包括对应于所述第二解剖结构的轮廓的轮廓标记,和/或对应于所述第二解剖结构的所在区域的区域标记。11. The method of claim 10, wherein the projection markers include contour markers corresponding to the contour of the second anatomical structure, and/or area markers corresponding to the region where the second anatomical structure is located.
- 根据权利要求11述的方法,其特征在于,所述轮廓标记包括轮廓 边界标记、轮廓阴影标记和/或轮廓箭头标记;The method according to claim 11, wherein the outline markings comprise outline boundary markings, outline shadow markings and/or outline arrow markings;所述区域标记包括区域边界标记、区域阴影标记和/或区域箭头标记。The area markers include area border markers, area shadow markers and/or area arrow markers.
- 根据权利要求9-12中任意一项所述的方法,其特征在于,当所述投影标记为多个时,区分显示所述多个投影标记。The method according to any one of claims 9-12, wherein when there are multiple projection marks, the multiple projection marks are displayed in a differentiated manner.
- 根据权利要求13所述的方法,其特征在于,所述区分显示的方法包括:The method according to claim 13, wherein the method for distinguishing display comprises:对应于每一所述第二解剖结构将所述投影标记进行颜色区分;或者color-coding the projection markers corresponding to each of the second anatomical structures; or对应于每一所述第二解剖结构将所述投影标记进行线型区分;或者linearly distinguishing the projection markers corresponding to each of the second anatomical structures; or对应于每一所述第二解剖结构将所述投影标记进行阴影区分。The projection markers are shaded corresponding to each of the second anatomical structures.
- 根据权利要求10-14中任意一项所述的方法,其特征在于,所述识别步骤包括以下至少一项:The method according to any one of claims 10-14, wherein the identifying step comprises at least one of the following:通过图像分割算法,在所述超声图像上分割得到至少一个所述第二解剖结构的轮廓;Segmenting the ultrasound image to obtain at least one contour of the second anatomical structure through an image segmentation algorithm;通过预训练的机器学习模型,获取所述超声图像上的所述第二解剖结构的轮廓或所述第二解剖结构的所在区域。The contour of the second anatomical structure or the region where the second anatomical structure is located on the ultrasound image is obtained by using the pre-trained machine learning model.
- 根据权利要求10-15中任意一项所述的方法,其特征在于,所述识别步骤中识别至少一个所述第二解剖结构:识别所述对称性标识远离探头采集端的一侧的超声图像中的至少一个所述第二解剖结构。The method according to any one of claims 10-15, wherein in the identifying step, at least one second anatomical structure is identified: identifying the symmetry mark in the ultrasound image on the side away from the probe acquisition end of at least one of said second anatomical structures.
- 根据权利要求10所述的方法,其特征在于,所述识别步骤还包括:识别位于所述对称性标识另一侧的第三解剖结构,获取所述第三解剖结构的轮廓或所述第三解剖结构所在的区域,所述第三解剖结构与所述至少一个所述第二解剖结构中的每一个一一对应;The method according to claim 10, wherein the identifying step further comprises: identifying a third anatomical structure located on the other side of the symmetry mark, obtaining the outline of the third anatomical structure or the third anatomical structure. an area where an anatomical structure is located, the third anatomical structure is in one-to-one correspondence with each of the at least one second anatomical structure;所述显示步骤还包括:显示所述第三解剖结构的轮廓或所述第三解剖结构所在的区域。The displaying step further includes: displaying an outline of the third anatomical structure or a region where the third anatomical structure is located.
- 根据权利要求6、9-17中任意一项所述的方法,其特征在于,所述解剖结构包括至少以下一项:小脑、侧脑室、脉络丛、大脑外侧裂和丘脑。The method of any one of claims 6, 9-17, wherein the anatomical structure comprises at least one of the following: cerebellum, lateral ventricle, choroid plexus, sylvian fissure, and thalamus.
- 根据权利要求1-18中任意一项所述的方法,其特征在于,所述超声图像包括所述胎儿颅脑的二维图像或三维图像,所述对称性标识包括对应于所述二维图像的脑中线或对应于所述三维图像的正中矢状面。The method according to any one of claims 1-18, wherein the ultrasound image includes a two-dimensional image or a three-dimensional image of the fetal skull, and the symmetry marker includes a corresponding image corresponding to the two-dimensional image. The midline of the brain or the midsagittal plane corresponding to the 3D image.
- 根据权利要求1-19中任意一项所述的方法,其特征在于,还包括:突出显示所述对称性标识。The method according to any one of claims 1-19, further comprising: highlighting the symmetry identification.
- 根据权利要求19或20所述的方法,其特征在于,所述图像获取步骤包括获取胎儿颅脑的二维图像;The method according to claim 19 or 20, wherein the image acquisition step comprises acquiring a two-dimensional image of the fetal skull;在所述对称性标识获取步骤中,获取所述脑中线的方法包括以下至少一项:In the step of obtaining the symmetry mark, the method for obtaining the midline of the brain includes at least one of the following:通过直线检测算法在所述二维图像中确定脑中线;determining the midline of the brain in the two-dimensional image by a line detection algorithm;通过结构检测算法在所述二维图像中检测特征解剖结构,将检测到的特征解剖结构拟合为直线,将拟合的所述直线确定为脑中线;The characteristic anatomical structure is detected in the two-dimensional image by a structure detection algorithm, the detected characteristic anatomical structure is fitted as a straight line, and the fitted straight line is determined as the midline of the brain;通过预训练的机器学习模型从所述二维图像中确定脑中线。The midline of the brain is determined from the two-dimensional images by a pretrained machine learning model.
- 根据权利要求19或20所述的方法,其特征在于,The method according to claim 19 or 20, wherein,所述图像获取步骤包括:获取胎儿颅脑的三维图像;The image acquisition step includes: acquiring a three-dimensional image of the fetal skull;在所述对称性标识获取步骤中,获取所述正中矢状面的方法包括以下至少一项:In the step of obtaining the symmetry mark, the method for obtaining the midsagittal plane includes at least one of the following:通过面检测算法在所述三维图像中确定正中矢状面;determining a midsagittal plane in the three-dimensional image by a plane detection algorithm;通过结构检测算法在所述三维图像中检测特征解剖结构,将检测到的特征解剖结构拟合为平面,将所述拟合的所述平面确定为正中矢状面;The characteristic anatomical structure is detected in the three-dimensional image by a structure detection algorithm, the detected characteristic anatomical structure is fitted as a plane, and the fitted plane is determined as a midsagittal plane;通过预训练的机器学习模型从所述三维图像中确定正中矢状面。The midsagittal plane is determined from the three-dimensional image by a pretrained machine learning model.
- 一种超声检测方法,其特征在于,包括:An ultrasonic detection method, characterized in that, comprising:图像获取步骤,获取关于胎儿颅脑的超声图像;The image acquisition step is to acquire an ultrasound image of the fetal skull;对称性标识获取步骤,获取所述超声图像中所述胎儿颅脑的对称性标识;The step of obtaining a symmetry mark is to obtain the symmetry mark of the fetal skull in the ultrasound image;识别步骤,在所述超声图像上识别至少一个至少部分位于所述对称性标识的一侧的解剖结构,获取所述解剖结构的轮廓或所述解剖结构所在的区域;Identifying, on the ultrasound image, identifying at least one anatomical structure at least partially located on one side of the symmetry mark, and acquiring an outline of the anatomical structure or a region where the anatomical structure is located;投影步骤,获取所述解剖结构的轮廓或所述解剖结构所在的区域关于所述对称性标识在所述对称性标识另一侧对称的投影标记;The projection step is to obtain the outline of the anatomical structure or a projection mark that is symmetrical on the other side of the symmetry mark with respect to the symmetry mark in the region where the anatomical structure is located;显示步骤,显示所述投影标记。A display step displays the projected mark.
- 根据权利要求23所述的方法,其特征在于,所述识别步骤中,识别至少一个至少部分位于所述对称性标识的一侧的解剖结构包括:在所述 对称性标识远离超声探头采集端的一侧的超声图像中识别所述解剖结构。24. The method of claim 23, wherein in the identifying step, identifying at least one anatomical structure at least partially located on one side of the symmetry marker comprises: on a side of the symmetry marker remote from the acquisition end of the ultrasound probe Identify the anatomical structure in the ultrasound image of the side.
- 根据权利要求23所述的方法,其特征在于,所述投影标记对应于所述解剖结构的轮廓的轮廓标记,和/或对应于所述解剖结构的所在区域的区域标记。24. The method of claim 23, wherein the projected markers correspond to contour markers of the contour of the anatomical structure, and/or corresponding to area markers of the region where the anatomical structure is located.
- 据权利要求25述的方法,其特征在于,The method of claim 25, wherein:所述轮廓标记包括轮廓边界标记、轮廓阴影标记和/或轮廓箭头标记;The outline markings include outline boundary markings, outline shadow markings and/or outline arrow markings;所述区域标记包括区域边界标记、区域阴影标记和/或区域箭头标记。The area markers include area border markers, area shadow markers and/or area arrow markers.
- 据权利要求23述的方法,其特征在于,The method of claim 23, wherein:当所述投影标记为多个时,区分显示所述多个投影标记。When there are a plurality of projection marks, the plurality of projection marks are displayed in a differentiated manner.
- 据权利要求27所述的方法,其特征在于,所述区分显示的方法包括:The method according to claim 27, wherein the method for distinguishing display comprises:对应于每一所述所述解剖结构将所述投影标记进行颜色区分;或者color-coding the projection markers corresponding to each of the anatomical structures; or对应于每一所述所述解剖结构将所述投影标记进行线型区分;或者linearly distinguishing said projection markers corresponding to each of said anatomical structures; or对应于每一所述所述解剖结构将所述投影标记进行阴影区分。The projection markers are shaded for each of the anatomical structures.
- 根据权利要求23所述的方法,其特征在于,所述超声图像包括所述胎儿颅脑的二维图像或三维图像,所述对称性标识包括对应于所述二维图像的脑中线或对应于所述三维图像的正中矢状面。24. The method of claim 23, wherein the ultrasound image comprises a two-dimensional image or a three-dimensional image of the fetal skull, and the symmetry marker comprises a midline corresponding to the two-dimensional image or a midline corresponding to the two-dimensional image. The midsagittal plane of the three-dimensional image.
- 根据权利要求23所述的方法,其特征在于,在所述显示步骤中还包括:突出显示所述对称性标识。The method according to claim 23, wherein the displaying step further comprises: highlighting the symmetry mark.
- 根据权利要求29所述的方法,其特征在于,The method of claim 29, wherein:所述图像获取步骤包括获取胎儿颅脑的二维图像;The image acquisition step includes acquiring a two-dimensional image of the fetal skull;在所述对称性标识获取步骤中,获取所述脑中线的方法包括以下至少一项:In the step of obtaining the symmetry mark, the method for obtaining the midline of the brain includes at least one of the following:通过直线检测算法在所述二维图像中确定脑中线;determining the midline of the brain in the two-dimensional image by a line detection algorithm;通过结构检测算法在所述二维图像中检测特征解剖结构,将检测到的特征解剖结构拟合为直线,将拟合的所述直线确定为脑中线;The characteristic anatomical structure is detected in the two-dimensional image by a structure detection algorithm, the detected characteristic anatomical structure is fitted as a straight line, and the fitted straight line is determined as the midline of the brain;通过预训练的机器学习模型从所述二维图像中确定脑中线。The midline of the brain is determined from the two-dimensional images by a pretrained machine learning model.
- 根据权利要求29所述的方法,其特征在于,The method of claim 29, wherein:所述图像获取步骤包括:获取胎儿颅脑的三维图像;The image acquisition step includes: acquiring a three-dimensional image of the fetal skull;在所述对称性标识获取步骤中,获取所述正中矢状面的方法包括以下 至少一项:In the step of obtaining the symmetry mark, the method for obtaining the midsagittal plane includes at least one of the following:通过面检测算法在所述三维图像中确定正中矢状面;determining a midsagittal plane in the three-dimensional image by a plane detection algorithm;通过目标检测算法在所述三维图像中检测特征解剖结构,将检测到的特征解剖结构拟合为平面,将所述拟合的所述平面确定为正中矢状面;The characteristic anatomical structure is detected in the three-dimensional image by the target detection algorithm, the detected characteristic anatomical structure is fitted as a plane, and the fitted plane is determined as the midsagittal plane;通过预训练的机器学习模型从所述三维图像中确定正中矢状面。The midsagittal plane is determined from the three-dimensional image by a pretrained machine learning model.
- 根据权利要求23-32中任意一项所述的方法,其特征在于,所述解剖结构包括以下至少一项:小脑、侧脑室、脉络丛、大脑外侧裂和丘脑。The method of any one of claims 23-32, wherein the anatomical structure comprises at least one of the following: cerebellum, lateral ventricle, choroid plexus, sylvian fissure, and thalamus.
- 一种超声成像***,其特征在于,包括:An ultrasound imaging system, comprising:超声探头,用于向受测胎儿颅脑发射超声波并接收超声回波,得到超声回波信号;The ultrasonic probe is used to transmit ultrasonic waves to the brain of the tested fetus and receive ultrasonic echoes to obtain ultrasonic echo signals;处理器,用于对超声回波信号进行处理,得到所述胎儿颅脑的超声图像;a processor, configured to process the ultrasonic echo signal to obtain the ultrasonic image of the fetal skull;显示器,用于显示所述超声图像;a display for displaying the ultrasound image;存储器,用于存储可执行的程序指令;memory for storing executable program instructions;所述处理器,还用于基于所述可执行的程序指令执行如权利要求1-33任意一项所述的方法。The processor is further configured to execute the method according to any one of claims 1-33 based on the executable program instructions.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2020/140985 WO2022141085A1 (en) | 2020-12-29 | 2020-12-29 | Ultrasonic detection method and ultrasonic imaging system |
CN202080107963.5A CN116762093A (en) | 2020-12-29 | 2020-12-29 | Ultrasonic detection method and ultrasonic imaging system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2020/140985 WO2022141085A1 (en) | 2020-12-29 | 2020-12-29 | Ultrasonic detection method and ultrasonic imaging system |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022141085A1 true WO2022141085A1 (en) | 2022-07-07 |
Family
ID=82258774
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2020/140985 WO2022141085A1 (en) | 2020-12-29 | 2020-12-29 | Ultrasonic detection method and ultrasonic imaging system |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN116762093A (en) |
WO (1) | WO2022141085A1 (en) |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130202174A1 (en) * | 2012-02-06 | 2013-08-08 | Samsung Medison Co., Ltd. | Method and apparatus for obtaining symmetry information of objects |
CN104414680A (en) * | 2013-08-21 | 2015-03-18 | 深圳迈瑞生物医疗电子股份有限公司 | Three-dimensional ultrasonic imaging method and system |
CN104657979A (en) * | 2014-12-24 | 2015-05-27 | 中国科学院深圳先进技术研究院 | Ultrasonic image feature detection method and system |
CN105426808A (en) * | 2014-09-23 | 2016-03-23 | 深圳先进技术研究院 | Intra-brain sagittal line measurement method and system |
CN108852287A (en) * | 2018-05-04 | 2018-11-23 | 北京雅森科技发展有限公司 | A method of the selection symmetrical region of interest of brain |
CN109044400A (en) * | 2018-08-31 | 2018-12-21 | 上海联影医疗科技有限公司 | Ultrasound image mask method, device, processor and readable storage medium storing program for executing |
CN111368586A (en) * | 2018-12-25 | 2020-07-03 | 深圳迈瑞生物医疗电子股份有限公司 | Ultrasonic imaging method and system |
-
2020
- 2020-12-29 WO PCT/CN2020/140985 patent/WO2022141085A1/en active Application Filing
- 2020-12-29 CN CN202080107963.5A patent/CN116762093A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130202174A1 (en) * | 2012-02-06 | 2013-08-08 | Samsung Medison Co., Ltd. | Method and apparatus for obtaining symmetry information of objects |
CN104414680A (en) * | 2013-08-21 | 2015-03-18 | 深圳迈瑞生物医疗电子股份有限公司 | Three-dimensional ultrasonic imaging method and system |
CN105426808A (en) * | 2014-09-23 | 2016-03-23 | 深圳先进技术研究院 | Intra-brain sagittal line measurement method and system |
CN104657979A (en) * | 2014-12-24 | 2015-05-27 | 中国科学院深圳先进技术研究院 | Ultrasonic image feature detection method and system |
CN108852287A (en) * | 2018-05-04 | 2018-11-23 | 北京雅森科技发展有限公司 | A method of the selection symmetrical region of interest of brain |
CN109044400A (en) * | 2018-08-31 | 2018-12-21 | 上海联影医疗科技有限公司 | Ultrasound image mask method, device, processor and readable storage medium storing program for executing |
CN111368586A (en) * | 2018-12-25 | 2020-07-03 | 深圳迈瑞生物医疗电子股份有限公司 | Ultrasonic imaging method and system |
Also Published As
Publication number | Publication date |
---|---|
CN116762093A (en) | 2023-09-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110338840B (en) | Three-dimensional imaging data display processing method and three-dimensional ultrasonic imaging method and system | |
CN111179227B (en) | Mammary gland ultrasonic image quality evaluation method based on auxiliary diagnosis and subjective aesthetics | |
ES2914387T3 (en) | immediate study | |
CN111368586B (en) | Ultrasonic imaging method and system | |
WO2020027228A1 (en) | Diagnostic support system and diagnostic support method | |
JP2018501005A (en) | Ultrasound imaging apparatus and method for segmenting anatomical objects | |
CN115429325A (en) | Ultrasonic imaging method and ultrasonic imaging equipment | |
JP2022546344A (en) | Image processing for stroke feature acquisition | |
Yaqub et al. | Volumetric segmentation of key fetal brain structures in 3D ultrasound | |
CN112137693B (en) | Imaging method and device for four-dimensional ultrasonic guided puncture | |
WO2021120065A1 (en) | Automatic measurement method and ultrasonic imaging system for anatomical structure | |
WO2022141085A1 (en) | Ultrasonic detection method and ultrasonic imaging system | |
US20220249060A1 (en) | Method for processing 3d image data and 3d ultrasonic imaging method and system | |
WO2023133929A1 (en) | Ultrasound-based human tissue symmetry detection and analysis method | |
CN115035086A (en) | Intelligent tuberculosis skin test screening and analyzing method and device based on deep learning | |
US20220087645A1 (en) | Guided lung coverage and automated detection using ultrasound devices | |
Al et al. | Reinforcement learning-based automatic diagnosis of acute appendicitis in abdominal ct | |
Bohil et al. | Psychophysical methods and signal detection: Recent advances in theory | |
CN111862014A (en) | ALVI automatic measurement method and device based on left and right ventricle segmentation | |
US20200315569A1 (en) | System and method for determining condition of fetal nervous system | |
CN117575999B (en) | Focus prediction system based on fluorescent marking technology | |
WO2022134049A1 (en) | Ultrasonic imaging method and ultrasonic imaging system for fetal skull | |
CN114391878B (en) | Ultrasonic imaging equipment | |
CN111603199B (en) | Three-dimensional reconstruction ultrasonic diagnosis system based on body surface positioning measuring instrument | |
Khazendar | Computer-aided diagnosis of gynaecological abnormality using B-mode ultrasound images |
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: 20967433 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 202080107963.5 Country of ref document: CN |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 20967433 Country of ref document: EP Kind code of ref document: A1 |