Ultrasonic image processing method and device, ultrasonic equipment and storage medium

Abstract

The invention relates to the technical field of medical image processing, and discloses an ultrasonic image processing method and device, ultrasonic equipment and a storage medium. The ultrasonic image processing method comprises the steps that continuous ultrasonic images of a target scanning stage are obtained, the target scanning stage comprises the stage of conducting ultrasonic scanning on the abdomen of a pregnant woman according to a specified operation mode, and the specified operation mode is used for representing the scanning direction and position; ultrasonic images belonging to a no-load data category in the continuous ultrasonic images are identified, a plurality of first image sets are divided from the continuous ultrasonic images according to an identification result, and the ultrasonic images in the first image sets are ultrasonic images of a non-no-load data category; the ultrasonic images in the at least one first image set are analyzed, and the fetal position is determined according to the analysis result. The ultrasonic image reading process does not need to depend on user experience, and the efficiency and objectivity of ultrasonic image reading are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and particularly to an ultrasonic image processing method, apparatus, ultrasonic device, and storage medium. Background Art

[0002] At present, ultrasonic medical imaging technology has been widely applied. For example, an obstetric clinician can manually scan the abdomen of a pregnant woman with a conventional ultrasonic technique to obtain an ultrasonic image of the abdomen, and then analyze the situation reflected by the ultrasonic image based on experience, that is, the manual image reading process of the ultrasonic image; wherein, scanning refers to adjusting the position and orientation of an ultrasonic probe according to the operation experience of the ultrasonic device to obtain relevant ultrasonic images that meet the requirements. Obviously, to obtain the information represented in the ultrasonic image, the existing solutions require the user to have strong image reading ability of ultrasonic images, rely heavily on the user's experience, and have high difficulty and technical threshold. Summary of the Invention

[0003] In view of this, the present invention provides an ultrasonic image processing method, apparatus, ultrasonic device, and storage medium to solve the problems such as excessive dependence on user experience in the existing image reading process of ultrasonic images.

[0004] In a first aspect, the present invention provides an ultrasonic image processing method, which includes:

[0005] Obtain consecutive ultrasonic images in a target scanning stage, where the target scanning stage includes a stage of performing ultrasonic scanning on the abdomen of a pregnant woman according to a specified operation mode, and the specified operation mode is used to represent the scanning direction and position;

[0006] Identify the ultrasonic images belonging to the no-load data category in the consecutive ultrasonic images, and divide a plurality of first image sets from the consecutive ultrasonic images according to the identification result, where the ultrasonic images in the first image sets are ultrasonic images belonging to the non-no-load data category;

[0007] Analyze the ultrasonic images in at least one of the first image sets, and determine the fetal position according to the analysis result.

[0008] Based on the processing of the ultrasonic images of the abdomen of a pregnant woman obtained by the specified operation mode, the present invention can identify the ultrasonic images belonging to the non-no-load category, and determine the fetal position according to the analysis result of the ultrasonic images belonging to the non-no-load data category. It can be seen that the image reading process of the ultrasonic images of the present invention does not need to rely on the user's experience, greatly improves the efficiency and objectivity of ultrasonic image reading, and provides the user with the identification result of the fetal position without the need for the user to identify the image.

[0009] In an optional implementation manner, identifying the ultrasonic images belonging to the no-load data category in the consecutive ultrasonic images includes:

[0010] Classify each ultrasound image in a series of ultrasound images to screen out the ultrasound images belonging to the no-load data category.

[0011] The present invention can also distinguish and classify each ultrasound image in a series of ultrasound images by classification, so as to accurately screen out the ultrasound images belonging to the no-load data category.

[0012] In an alternative embodiment, divide a series of ultrasound images into multiple first image sets according to the recognition result, including:

[0013] Divide a series of ultrasound images into multiple first image sets and multiple second image sets according to the ultrasound images belonging to the no-load data category, and the images in the second image set are the ultrasound images of the no-load data category.

[0014] Determine multiple second image sets based on the ultrasound images belonging to the no-load data category, so as to determine multiple first image sets. The present invention can accurately determine whether the current ultrasound image belongs to the no-load data category or the non-no-load data category among all the ultrasound images in the target scanning stage.

[0015] In an alternative embodiment, classifying each ultrasound image in a series of ultrasound images includes:

[0016] Classify each ultrasound image in a series of ultrasound images through a trained first classification network to identify the ultrasound images belonging to the no-load data category; the input of the first classification network is the ultrasound image, and the output is the image category, and the image category includes the no-load data category.

[0017] Based on the trained first classification network, the present invention can classify multiple ultrasound images in the target scanning stage more accurately and quickly.

[0018] In an alternative embodiment, analyzing the ultrasound images in at least one first image set includes:

[0019] Classify the ultrasound images in at least one first image set through a trained second classification network to identify the classification results corresponding to the ultrasound images in the first image set; the input of the second classification network is the ultrasound image, and the output is the classification result, and the classification result includes the analysis result.

[0020] Based on the trained second classification network, the present invention can classify the ultrasound images in the first image set more accurately and quickly.

[0021] In an alternative embodiment, the multiple first image sets correspond one by one to the stages of performing ultrasound scanning on the pregnant woman's abdomen according to a specified operation method.

[0022] Based on the recognition of multiple first image sets, the present invention can specifically recognize and judge ultrasonic images in each scanning stage.

[0023] In an optional embodiment, the stages of ultrasonic scanning of a pregnant woman's abdomen include a first scanning stage; analyzing the ultrasonic images in at least one first image set, and determining the fetal position according to the analysis results, including:

[0024] If the classification result of the ultrasonic images in the first image set corresponding to the first scanning stage includes an orbital section, it is determined that the fetal position is the occipital posterior position.

[0025] By judging the classification result of the ultrasonic images in the first scanning stage, the present invention can also realize the recognition of the fetal position result of the occipital posterior position.

[0026] In an optional embodiment, the stages of ultrasonic scanning of a pregnant woman's abdomen further include a second scanning stage, and the second scanning stage is the stage after the first scanning stage; analyzing the ultrasonic images in at least one first image set, and determining the fetal position according to the analysis results, including:

[0027] If the classification result of the ultrasonic images in the first preset time period in the first image set corresponding to the second scanning stage includes an orbital section, it is determined that the fetal position is the occipital posterior position;

[0028] If the classification result of the ultrasonic images in the first preset time period in the first image set corresponding to the second scanning stage includes a spinal section, it is determined that the fetal position is the occipital anterior position;

[0029] If the classification result of the ultrasonic images in the first image set corresponding to the first scanning stage includes a cerebral midline section, and the classification result of the ultrasonic images in the second preset time period in the first image set corresponding to the second scanning stage includes a spinal section, it is determined that the fetal position is the right occipital transverse position;

[0030] If the classification result of the ultrasonic images in the first image set corresponding to the first scanning stage includes a cerebral midline section, and the classification result of the ultrasonic images in the second preset time period in the first image set corresponding to the second scanning stage includes an orbital section, it is determined that the fetal position is the left occipital transverse position.

[0031] By judging the classification result of the ultrasonic images in the second scanning stage, the present invention can realize the recognition of the fetal position results of the occipital posterior position, the occipital anterior position, the right occipital transverse position, and the left occipital transverse position.

[0032] In an optional embodiment, the stages of ultrasonic scanning of a pregnant woman's abdomen further include a third scanning stage, and the third scanning stage is the stage after the second scanning stage; analyzing the ultrasonic images in at least one first image set, and determining the fetal position according to the analysis results, including:

[0033] If the classification result of the ultrasound images in the third preset time period in the first image set corresponding to the third scanning stage includes an orbital section, determine that the fetal position is the occipital posterior position;

[0034] If the classification result of the ultrasound images in the third preset time period in the first image set corresponding to the third scanning stage includes a spinal section, determine that the fetal position is the occipital anterior position;

[0035] If the classification result of the ultrasound images in the first image set corresponding to the first scanning stage includes a cerebral midline section, and the classification result of the ultrasound images in the fourth preset time period in the first image set corresponding to the third scanning stage includes a spinal section, determine that the fetal position is the left occipital transverse position;

[0036] If the classification result of the ultrasound images in the first image set corresponding to the first scanning stage includes a cerebral midline section, and the classification result of the ultrasound images in the fourth preset time period in the first image set corresponding to the third scanning stage includes an orbital section, determine that the fetal position is the right occipital transverse position.

[0037] In a second aspect, the present invention provides an ultrasound image processing device, which includes:

[0038] An acquisition module, configured to acquire consecutive ultrasound images of a target scanning stage, where the target scanning stage includes a stage of performing an ultrasound scan on a pregnant woman's abdomen according to a specified operation method, and the specified operation method is used to characterize the scanning direction and position;

[0039] A segmentation module, configured to identify the ultrasound images belonging to the no-load data category in the consecutive ultrasound images, and to divide multiple first image sets from the consecutive ultrasound images according to the identification result, where the ultrasound images in the first image set are non-no-load data category ultrasound images;

[0040] An analysis module, configured to analyze the ultrasound images in at least one first image set, and to determine the fetal position according to the analysis result.

[0041] In a third aspect, the present invention provides an ultrasound device, including: a memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the ultrasound image processing method according to the first aspect or any corresponding embodiment thereof.

[0042] In a fourth aspect, the present invention provides a computer-readable storage medium, on which computer instructions are stored, and the computer instructions are used to cause a computer to execute the ultrasound image processing method according to the first aspect or any corresponding embodiment thereof. Description of the Drawings

[0043] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the related art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the related art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0044] Figure 1 It is a schematic flowchart of an ultrasonic image processing method according to an embodiment of the present invention;

[0045] Figure 2 It is a schematic flowchart of another ultrasonic image processing method according to an embodiment of the present invention;

[0046] Figure 3 It is a schematic flowchart of yet another ultrasonic image processing method according to an embodiment of the present invention;

[0047] Figure 4 It is a schematic diagram of the implementation principle for segmenting multiple ultrasonic images in the first image set corresponding to the target scanning stage according to an embodiment of the present invention;

[0048] Figure 5 It is a schematic diagram of the implementation principle for classifying multiple ultrasonic images in the first image set corresponding to the first scanning stage according to an embodiment of the present invention;

[0049] Figure 6 It is a schematic diagram of the implementation principle for classifying the ultrasonic images in the first preset time period in the first image set corresponding to the second scanning stage according to an embodiment of the present invention;

[0050] Figure 7 It is a schematic diagram of the implementation principle for classifying the ultrasonic images in the second preset time period in the first image set corresponding to the second scanning stage according to an embodiment of the present invention;

[0051] Figure 8 It is a schematic diagram of the implementation principle for classifying the ultrasonic images in the third preset time period in the first image set corresponding to the third scanning stage according to an embodiment of the present invention;

[0052] Figure 9 It is a schematic diagram of the implementation principle for classifying the ultrasonic images in the fourth preset time period in the first image set corresponding to the third scanning stage according to an embodiment of the present invention;

[0053] Figure 10 It is a schematic diagram of the implementation principle for obtaining the fetal position based on abdominal ultrasonic images according to an embodiment of the present invention;

[0054] Figure 11It is a structural block diagram of an ultrasonic image processing device according to an embodiment of the present invention;

[0055] Figure 12 It is a schematic diagram of the hardware structure of an ultrasonic device according to an embodiment of the present invention. Specific embodiments

[0056] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0057] In the related art, an ultrasonic doctor can manually draw images on the abdomen of a pregnant woman to find some sections and features, and analyze the ultrasonic images based on clinical experience to obtain the results of image analysis or recognition. Taking the process of determining the fetal position as an example, clinically, the fetal position is generally determined by means of conventional ultrasonic examinations and image analysis in the ultrasonic department, or by means of digital examination and subjective experience by clinicians during childbirth. Most of the techniques for determining the fetal position using conventional ultrasound by ultrasonic doctors are not automatic. It requires the ultrasonic doctor to manually draw images on the abdomen to find some specific sections and features, and then analyze the ultrasonic images based on their clinical experience to determine the fetal position; the above process is highly dependent on the user's own clinical experience and ultrasonic image reading ability. Therefore, if a delivery room doctor wants to use ordinary ultrasonic technology to determine the fetal position, due to their relatively little experience and ability in using ultrasonic technology, there are relatively large difficulties and technical thresholds. Although the fetal position can be judged by means of digital examination and subjective experience by clinicians during childbirth, during the digital examination process, it mainly relies on the clinician to insert a finger and judge the fetal condition by touch. However, the actual uterine cavity and birth canal environments are complex, and touch is always not intuitive and comprehensive. Therefore, it is highly dependent on the subjective judgment of the clinical experience of the delivery room doctor, so there will be subjective errors; moreover, the invasion will also bring relatively serious problems such as infection and discomfort to the pregnant woman.

[0058] According to an embodiment of the present invention, an embodiment of an ultrasonic image processing method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0059] In this embodiment, an ultrasonic image processing method is provided, which can be used for an ultrasonic device, Figure 1is a flowchart of an ultrasonic image processing method according to an embodiment of the present invention. As Figure 1 shown, the process includes the following steps:

[0060] Step S101, obtain continuous ultrasonic images of a target scanning phase, where the target scanning phase includes a phase of performing ultrasonic scanning on a pregnant woman's abdomen according to a specified operation mode, and the specified operation mode is used to characterize the scanning direction and position.

[0061] Among them, the target scanning phase may include multiple scanning phases, and the number of ultrasonic images in each scanning phase may be multiple.

[0062] In the embodiment of the present invention, specifically, according to the specified operation mode means performing continuous ultrasonic probe scanning according to a specified standard process to obtain continuous data; the continuous data includes ultrasonic data in the scanning phase and ultrasonic data in the non-scanning phase, and the continuous data is used to form continuous ultrasonic images.

[0063] For example, the phase of performing ultrasonic scanning on a pregnant woman's abdomen according to the specified operation mode may include, but is not limited to, a first scanning phase, a second scanning phase, and a third scanning phase. In the first scanning phase, from above the midline of the abdomen, the axial plane of the probe scans relatively evenly from top to bottom and stops when reaching the pubic symphysis. In the second scanning phase, after the first scanning phase ends, pick up the probe and rotate it 90 degrees, and the sagittal plane of the probe scans relatively uniformly from the vertical position in the middle of the abdomen to the operator's left (the right side of the pregnant woman) and stops when reaching the right boundary of the pregnant woman's abdomen. In the third scanning phase, after the second scanning phase ends, pick up the probe and return it to the midline position of the abdomen, and the sagittal plane of the probe scans relatively uniformly from the vertical position in the middle of the abdomen to the operator's right (the left side of the pregnant woman) and stops when reaching the left boundary of the pregnant woman's abdomen.

[0064] Step S102, identify the ultrasonic images belonging to the no-load data category in the continuous ultrasonic images, and divide multiple first image sets from the continuous ultrasonic images according to the identification result. The ultrasonic images in the first image sets are ultrasonic images of the non-no-load data category.

[0065] The multiple first image sets in this embodiment correspond one-to-one to the phases of performing ultrasonic scanning on a pregnant woman's abdomen according to the specified operation mode.

[0066] Among them, one first image set corresponds to one phase of performing ultrasonic scanning on a pregnant woman's abdomen. Based on the identification of the first image sets, for example, the identification of the first scanning phase, the second scanning phase, and the third scanning phase, this embodiment can accurately judge each phase of performing ultrasonic scanning on a pregnant woman's abdomen according to the specified operation mode, so as to identify and judge the ultrasonic images of the independent scanning phases targeted.

[0067] Specifically, the ultrasound images belonging to the no-load data category represent the ultrasound images in the non-actual scanning stage, and the ultrasound images in the non-actual scanning stage are used as the basis for segmentation.

[0068] As Figure 4 shown, a schematic diagram of the implementation principle of segmenting multiple ultrasound images in the target scanning stage in this embodiment is shown. For example, multiple ultrasound images between two sets of adjacent ultrasound images in the non-actual scanning stage are used as a first image set, and multiple ultrasound images between the ultrasound image in the pre-scanning stage and the adjacent ultrasound image belonging to the no-load data category are used as a first image set. Since the ultrasound images in the non-actual scanning stage contain fewer or no organ features, while the ultrasound images in the actual scanning stage contain more organ features, it is possible to effectively determine whether a certain frame of ultrasound image belongs to the ultrasound image of the no-load data category based on the organ feature recognition method of the ultrasound image. Among them, the ultrasound images in the actual scanning stage are useful ultrasound images.

[0069] Step S103: Analyze the ultrasound images in at least one first image set, and determine the fetal position according to the analysis result.

[0070] In this embodiment, the process of analyzing the ultrasound images in the first image set is a recognition process, and the corresponding fetal position can be determined according to the analysis result of the ultrasound images obtained from the recognition process.

[0071] For example, in this embodiment, the category of the ultrasound image can be determined by classifying the ultrasound image, and then the fetal position corresponding to the current ultrasound image can be recognized according to the determined category.

[0072] The ultrasound image processing method provided in this embodiment is based on the recognition of the ultrasound images of the pregnant woman's abdomen obtained according to the specified operation method, determines the ultrasound images belonging to the non-no-load data category, analyzes the ultrasound images belonging to the non-no-load data category, and determines the fetal position according to the analysis result. It can be seen that the present invention can automatically identify the fetal position, and the process of reading the ultrasound image does not need to rely on the user's experience, greatly improving the efficiency of the process of reading the ultrasound image and the objectivity of the result of reading the ultrasound image, and realizing the automatic determination of the fetal position without manual reading of the image. The ultrasound image processing method provided in this embodiment has the characteristic of visualization, and can obtain the image recognition result in the form of quantitative parameters to better solve the subjectivity problem of reading the ultrasound image. The ultrasound image processing method provided in this embodiment has the characteristics of intelligence and automation, making the recognition process of the ultrasound image intelligent and simple, and improving the efficiency of reading the ultrasound image; in addition, the solution provided in this embodiment does not require invasive examination compared with the finger examination method, so it will not cause problems such as infection and invasive pain.

[0073] In this embodiment, an ultrasound image processing method is provided, which can be used in an ultrasound device.Figure 2 is a flowchart of an ultrasonic image processing method according to an embodiment of the present invention. As Figure 2 shown, the process includes the following steps:

[0074] Step S201: Obtain consecutive ultrasonic images of a target scanning phase. The target scanning phase includes a phase of performing ultrasonic scanning on a pregnant woman's abdomen according to a specified operation mode, and the specified operation mode is used to characterize the scanning direction and position. For details, please refer to Figure 1 Step S101 of the embodiment shown herein, which will not be elaborated herein.

[0075] Step S202: Classify each ultrasonic image in the consecutive ultrasonic images to screen out ultrasonic images belonging to the no-load data category; according to the ultrasonic images belonging to the no-load data category, divide the consecutive ultrasonic images into multiple first image sets and multiple second image sets, and the images in the second image set are ultrasonic images of the no-load data category.

[0076] Among them, the ultrasonic images belonging to the no-load data category are ultrasonic images in the no-load phase, and the number of no-load phases can be multiple. For example Figure 4 the three no-load phases shown in

[0077] In this embodiment, a neural network model with an image classification function can be used to classify the ultrasonic images, and each image corresponds to an image classification result; in this embodiment, multiple image categories can be preset in advance, and each image category represents a classification result, and the specific content of the image category can be set according to the actual situation.

[0078] In this embodiment, the classification result may include the section features represented in the ultrasonic image, for example, may include but are not limited to orbital section, spinal section, cerebral midline section, no-load feature, and other features (i.e., features of non-determined regions), etc.

[0079] Combined with the foregoing embodiment, identifying the ultrasonic images belonging to the no-load data category in the consecutive ultrasonic images includes: classifying each ultrasonic image in the consecutive ultrasonic images to screen out the ultrasonic images belonging to the no-load data category.

[0080] Combined with the foregoing embodiment, dividing multiple first image sets from the consecutive ultrasonic images according to the identification result includes: dividing the consecutive ultrasonic images into multiple first image sets and multiple second image sets according to the ultrasonic images belonging to the no-load data category, and the images in the second image set are ultrasonic images of the no-load data category.

[0081] Step S203: Analyze the ultrasonic images in at least one first image set, and determine the fetal position according to the analysis result. For details, please refer to Figure 1Step S103 of the illustrated embodiment will not be elaborated herein.

[0082] In this embodiment, the ultrasonic images of the no-load data category are specifically used to classify the multiple ultrasonic images in the target scanning stage. This method can effectively divide the ultrasonic images that do not belong to the no-load data category (ultrasonic images of non-no-load data category), so as to specifically classify and identify the ultrasonic images that do not belong to the no-load data category, and obtain the classification results of useful ultrasonic images. This method significantly improves the pertinence and effectiveness of classifying the ultrasonic images in the target scanning stage, and further improves the accuracy of fetal position recognition.

[0083] In this embodiment, an ultrasonic image processing method is provided, which can be used in ultrasonic devices. Figure 3 It is a flowchart of the ultrasonic image processing method according to an embodiment of the present invention. As Figure 3 shown, the process includes the following steps:

[0084] Step S301, obtain consecutive ultrasonic images in the target scanning stage. The target scanning stage includes the stage of performing ultrasonic scanning on the pregnant woman's abdomen according to a specified operation method, and the specified operation method is used to represent the scanning direction and position. For details, please refer to Figure 1 Step S101 of the illustrated embodiment will not be elaborated herein.

[0085] Step S302, classify each ultrasonic image in the consecutive ultrasonic images through a trained first classification network to identify the ultrasonic images belonging to the no-load data category; the input of the first classification network is the ultrasonic image, and the output is the image category, and the image category includes the no-load data category; according to the ultrasonic images belonging to the no-load data category, divide the consecutive ultrasonic images into multiple first image sets and multiple second image sets, and the images in the second image set are ultrasonic images of the no-load data category.

[0086] Combined with the foregoing embodiments, classifying each ultrasonic image in the consecutive ultrasonic images includes: classifying each ultrasonic image in the consecutive ultrasonic images through a trained first classification network to identify the ultrasonic images belonging to the no-load data category; the input of the first classification network is the ultrasonic image, and the output is the image category, and the image category includes the no-load data category.

[0087] Specifically, in this embodiment, a trained first classification network is used to classify multiple ultrasonic images in the target scanning stage to identify the ultrasonic images belonging to the no-load data category; the input of the first classification network is the ultrasonic image, and the output is the image category, and the image category includes the no-load data category.

[0088] The first classification network in this embodiment is a feature classification network, which is used to classify each frame of ultrasound image to obtain a classification result. Specifically, the classification result may include: transabdominal - other, transabdominal - cerebral midline, transabdominal - spine, transabdominal - orbit, no load, these five categories. Of course, the present invention is not limited to these five categories. Under the guidance of this embodiment, more categories can also be set. Among them, transabdominal - other represents the classification result of the non - determined area. More specifically, the feature classification network in this embodiment is a section classification network, which is used to identify the above - mentioned various types of sections included in the ultrasound image.

[0089] Specifically, the first classification network is, for example, one or more of Alexnet (Alexandria network), VGG (Visual Geometry Group network), and Mobilenet (mobile network).

[0090] In the process of training the first classification network in this embodiment, a large number of section images containing spine, bilateral orbits, and cerebral midline features can be collected through an ultrasound device, and a data set can be made using the collected section images. For example, the proportion of various section images in the made data set is as follows: no load: spine: bilateral orbits: cerebral midline: other = 1:1:1:1:1; then the made data set can be respectively formed into a training set and a test set, and the first classification network can be trained based on deep learning frameworks such as PyTorch (an open - source Python machine learning library, based on Torch, used for applications such as natural language processing), TensorFlow (an end - to - end open - source machine learning platform), etc., and an excellent feature classification network model can be obtained by continuously optimizing.

[0091] This embodiment can effectively classify and identify the ultrasound images in the target scanning stage through the trained feature classification network (i.e., the feature classification network model), and can improve the classification accuracy and efficiency.

[0092] In this embodiment, the trained second classification network is used to classify the ultrasound images in the first image set to identify the classification results corresponding to the ultrasound images in the first image set; the input of the second classification network is the ultrasound image, and the output is the classification result, and the classification result includes the analysis result.

[0093] Among them, the second classification network is a section classification network, which is used to classify the ultrasound images in the first image set. Combining with the foregoing embodiments, the second classification network and the first classification network can be the same section classification network.

[0094] For the training process of the second classification network, it can be the same as the process of training the first classification network, and will not be elaborated here.

[0095] As Figures 4 to 9 shown, in this embodiment, the scanning phase between two adjacent no-load phases is named as the independent scanning phase, and the scanning phase between the pre-scanning phase and the adjacent no-load phase is determined as the independent scanning phase.

[0096] In this embodiment, the determined multiple independent scanning phases are respectively the first scanning phase, the second scanning phase, and the third scanning phase, and the number of ultrasonic images in each independent scanning phase can exceed thirteen hundred.

[0097] In this embodiment, the no-load phase corresponding one-to-one to the second image set is used as the basis for dividing the target scanning phase. This method is the specified rule followed by this embodiment, so as to analyze and segment the ultrasonic images, that is, to analyze and segment the acquired continuous data.

[0098] Among them, the multiple ultrasonic images in the first scanning phase form a first image set, the multiple ultrasonic images in the second scanning phase form a first image set, and the multiple ultrasonic images in the third scanning phase form a first image set.

[0099] In this embodiment, the ultrasonic images belonging to the no-load data category corresponding to each no-load phase form a second image set. Multiple such second image sets are determined among all the images in the entire target scanning phase, so that multiple independent scanning phases can be divided from the target scanning phase, and the division result is more accurate.

[0100] Step S303, classify the ultrasonic images in at least one first image set through the trained second classification network to identify the classification results corresponding to the ultrasonic images in the first image set; the input of the second classification network is the ultrasonic image and the output is the classification result, and the classification result includes the analysis result. Determine the fetal position according to the analysis result.

[0101] In this embodiment, the image recognition result is specifically the recognition result of the fetal position, such as including occipitoposterior position, occipitoposterior position, left occipitotransverse position, and right occipitotransverse position.

[0102] Combined with the foregoing embodiments, analyzing the ultrasonic images in at least one first image set includes: classifying the ultrasonic images in at least one first image set through the trained second classification network to identify the classification results corresponding to the ultrasonic images in the first image set; the input of the second classification network is the ultrasonic image and the output is the classification result, and the classification result includes the analysis result.

[0103] Compared with the related art, this embodiment can provide a method for automatically obtaining the fetal position based on abdominal ultrasound. The fetal position may include, but is not limited to, the fetal position during labor; and by analyzing each segment of the data after the aforementioned segmentation, the fetal position can be determined in combination with the types of features identified in each segment of the data.

[0104] Among them, the fetal position specifically describes the relationship between the indicating point of the fetal presenting part and the pelvic bone of the pregnant woman, mainly including cephalic presentation, breech presentation, and transverse presentation. Cephalic presentation can be further subdivided into fetal positions such as occipital anterior, occipital posterior, occipital left, and occipital right.

[0105] In particular, the monitoring of the fetal position during the labor stage is of great significance for the intervention of obstetricians. The general premise of the fetal position during the labor stage is cephalic presentation. For cephalic presentation, due to the different postures of the fetal head entering the pelvis in the late pregnancy, there are also fetal position classifications such as occipital anterior, occipital posterior, occipital left, and occipital right. Therefore, the determination of the fetal position during labor in this embodiment is mainly to determine which one of occipital anterior, occipital posterior, occipital left, occipital right, etc. the fetal position is in the case of cephalic presentation.

[0106] In some alternative embodiments, the stage of performing ultrasound scanning on the abdomen of the pregnant woman includes a first scanning stage. The specified operation method corresponding to the first scanning stage includes: starting from above the midline of the abdomen, the axial plane of the ultrasound probe scans relatively uniformly from top to bottom and stops when reaching the pubic symphysis. Among them, the ventral midline specifically refers to a vertical and thin line extending towards the xiphoid process at the midpoint of the umbilicus; it is relatively shallow for the average person compared to the pregnant woman, and is clearer for the pregnant woman. The axial plane is a human anatomical term, referring to the transverse section with the height direction as the normal line.

[0107] According to the classification result including the preset classification result, output the image recognition result corresponding to the preset classification result, including:

[0108] If the classification result of the ultrasound images in the first image set corresponding to the first scanning stage includes the orbital section, output that the fetal position corresponding to the orbital section is the occipital posterior position; among them, the preset classification result includes the orbital section, and the image recognition result includes that the fetal position is the occipital posterior position.

[0109] Such as Figure 5As shown, a schematic diagram of the implementation principle for classifying multiple ultrasound images in the first image set corresponding to the first scanning stage in this embodiment is shown. If the classification result of the ultrasound images in the first image set corresponding to the first scanning stage includes a midline brain section, it is determined that the fetal position corresponding to this midline brain section is the occipitotransverse position. Then, it is necessary to further classify the ultrasound images in the first image set corresponding to the second scanning stage or the third scanning stage to further determine whether this occipitotransverse position is specifically the left occipitotransverse position or the right occipitotransverse position. Among them, the preset classification result may also include the occipitotransverse position.

[0110] Based on the specified operation mode, in this embodiment, the recognition of the fetal position result as the occipitoposterior position is achieved by judging the classification result of the ultrasound images in the first image set corresponding to the first scanning stage.

[0111] In some optional implementation manners, the stage of performing ultrasound scanning on the pregnant woman's abdomen further includes a second scanning stage, which is the stage after the first scanning stage. The specified operation mode corresponding to the second scanning stage includes: lifting the ultrasound probe and rotating the probe 90 degrees, and the sagittal plane of the ultrasound probe scans relatively uniformly from the vertical position in the middle of the pregnant woman's abdomen to the right side of the abdomen until it reaches the right boundary of the pregnant woman's abdomen and stops. Among them, the sagittal section is a human anatomical term, specifically referring to the plane that divides the human body into left and right parts.

[0112] According to the classification result including the preset classification result, output the image recognition result corresponding to the preset classification result, including:

[0113] Step a1, if the classification result of the ultrasound images in the first preset time period in the first image set corresponding to the second scanning stage includes an orbital section, output that the fetal position corresponding to the orbital section is the occipitoposterior position.

[0114] In this embodiment, the first preset time period is the first 5 / 6 time period of the second scanning stage.

[0115] Step a2, if the classification result of the ultrasound images in the first preset time period in the first image set corresponding to the second scanning stage includes a spinal section, output that the fetal position corresponding to the spinal section is the occipitoanterior position.

[0116] As Figure 6 shown, a schematic diagram of the implementation principle for classifying the ultrasound images in the first preset time period in the first image set corresponding to the second scanning stage in this embodiment is shown. In the first 5 / 6 time period of the second scanning stage, if an orbital section appears, it indicates that the fetal position is the occipitoposterior position; if a spinal section appears, it indicates that the fetal position is the occipitoanterior position.

[0117] Step a3, if the classification result of the ultrasound images in the first image set corresponding to the first scanning stage includes the midline section of the brain and the classification result of the ultrasound images in the second preset time period in the first image set corresponding to the second scanning stage includes the spinal section, then output that the fetal position corresponding to the spinal section is right occipital transverse position.

[0118] In this embodiment, the second preset time period is the last 1 / 6 time period of the second scanning stage.

[0119] Step a4, if the classification result of the ultrasound images in the first image set corresponding to the first scanning stage includes the midline section of the brain and the classification result of the ultrasound images in the second preset time period in the second scanning stage includes the orbital section, then output that the fetal position corresponding to the orbital section is left occipital transverse position.

[0120] Among them, the preset classification result further includes the spinal section and the midline section of the brain, and the image recognition result further includes that the fetal position is occipital anterior position, the fetal position is right occipital transverse position, and the fetal position is left occipital transverse position.

[0121] As Figure 7 shown, it shows a schematic diagram of the implementation principle of classifying the ultrasound images in the first preset time period in the first image set corresponding to the second scanning stage in this embodiment. In the last 1 / 6 time period of the second scanning stage, if the spinal section appears, it indicates that the fetal position is right occipital transverse position, and if the orbital section appears, it indicates that the fetal position is left occipital transverse position.

[0122] Based on the specified operation mode, in this embodiment, by judging the classification result of the ultrasound images in the first image set corresponding to the second scanning stage and combining with the recognition of the classification result of the ultrasound images in the first image set corresponding to the first scanning stage, the recognition of the fetal position results of occipital posterior position, occipital anterior position, right occipital transverse position, and left occipital transverse position is realized.

[0123] In some optional implementation manners, the stage of performing ultrasound scanning on the pregnant woman's abdomen further includes a third scanning stage, the third scanning stage is the stage after the second scanning stage, and the specified operation mode corresponding to the third scanning stage includes: lifting the ultrasound probe back to the midline position of the abdomen, and the sagittal plane of the ultrasound probe scans relatively uniformly from the vertical position in the middle of the pregnant woman's abdomen to the left side of the abdomen until it reaches the left boundary of the pregnant woman's abdomen and stops.

[0124] According to the classification result including the preset classification result, output the image recognition result corresponding to the preset classification result, including:

[0125] Step b1, if the classification result of the ultrasound images in the third preset time period in the first image set corresponding to the third scanning stage includes the orbital section, then output that the fetal position corresponding to the orbital section is occipital posterior position.

[0126] In this embodiment, the third preset time period is the first 5 / 6 time period of the third scanning stage.

[0127] Step b2, if the classification result of the ultrasonic images in the third preset time period of the first image set corresponding to the third scanning stage contains a spinal cord section, then output the fetal position corresponding to the spinal cord section as the occipitoanterior position.

[0128] As Figure 8 shown, it shows a schematic diagram of the implementation principle of classifying the ultrasonic images in the third preset time period of the first image set corresponding to the third scanning stage of this embodiment. In the first 5 / 6 time period of the third scanning stage, if a spinal cord section appears, it indicates that the fetal position is the occipitoanterior position; if an orbital section appears, it indicates that the fetal position is the occipitoposterior position.

[0129] Step b3, if the classification result of the ultrasonic images in the first image set corresponding to the first scanning stage contains a cerebral midline section and the classification result of the ultrasonic images in the fourth preset time period of the first image set corresponding to the third scanning stage contains a spinal cord section, then output the fetal position corresponding to the spinal cord section as the left occipitotransverse position.

[0130] In this embodiment, the fourth preset time period is the last 1 / 6 time period of the third scanning stage.

[0131] Step b4, if the classification result of the ultrasonic images in the first image set corresponding to the first scanning stage contains a cerebral midline section and the classification result of the ultrasonic images in the fourth preset time period of the first image set corresponding to the third scanning stage contains an orbital section, then output the fetal position corresponding to the orbital section as the right occipitotransverse position.

[0132] Based on the specified operation mode, in this embodiment, by judging the classification result of the ultrasonic images in the third scanning stage and combining with the recognition of the classification result of the ultrasonic images in the first scanning stage, the recognition of the fetal position results of the occipitoposterior position, occipitoanterior position, right occipitotransverse position, and left occipitotransverse position is realized.

[0133] As Figure 9 shown, it shows a schematic diagram of the implementation principle of classifying the ultrasonic images in the fourth preset time period of the third scanning stage of this embodiment. In the last 1 / 6 time period of the third scanning stage, if a spinal cord section appears, it indicates that the fetal position is the left occipitotransverse position; if an orbital section appears, it indicates that the fetal position is the right occipitotransverse position.

[0134] As Figure 10As shown, this embodiment shows a schematic diagram of the implementation principle of obtaining the fetal position based on abdominal ultrasound images. Combining the foregoing embodiments, the principle of fetal position determination in the intrapartum stage will be taken as an example for illustration. The determination of fetal position in the intrapartum stage mainly relies on transabdominal ultrasound imaging in the "mid-sagittal plane (i.e., mid-longitudinal section)" and the "axial plane (i.e., transverse section)" for evaluation, which is the implementation basis of the present invention according to the specified operation method. The fetal position can be described as a circle like a clock, where ≥2:30h and ≤3:30h is left occipital transverse (LOT), ≥8:30h and ≤9:30h is right occipital transverse (ROT), >3:30h and <8:30h is occipital posterior (OP), and >9:30h and <2:30h is occipital anterior (OA). Combining Figure 10 As shown, in this embodiment, the circle used to describe the fetal position is simplified into a semi-circle. For example, when the probe is placed in the "abdominal sagittal position" of the parturient, if the fetal spine characteristics (i.e., spine section) can be observed in the ultrasound imaging, it is the occipital anterior position; then the ultrasound probe is moved downward and rotated to the "abdominal axial position". When two orbits of the fetus (i.e., orbit section) are shown in the ultrasound imaging, it indicates the occipital posterior position. If the brain midline characteristics (i.e., brain midline section) appear in the ultrasound image, it indicates the occipital transverse position; Figure 10 The detailed judgment process of the fetal position in the second scanning stage and the third scanning stage shown in [reference] has been described in detail in the foregoing embodiments and will not be elaborated here.

[0135] In summary, the embodiment of the present invention can provide a brand-new recognition algorithm for transabdominal fetal position in the intrapartum stage or the antenatal stage. This algorithm is scanned based on a specific scanning process, which can well solve the positioning problems of the probe scanning position and the scanning section direction; moreover, the embodiment of the present invention can combine the probe scanning position and the scanning section direction with the characteristic sections for analysis, and then can well judge the fetal position; based on this method, obstetric clinicians only need to scan the images according to the standardized process, and all the analysis of the data and the judgment of the results are automatically completed by the computer system. Such a judgment process does not need to rely on the clinical experience of obstetric clinicians and the ability to read ultrasound images, so it reduces the difficulty and technical threshold of using ultrasound to determine the fetal position, and greatly improves the efficiency and objectivity of the fetal position determination process.

[0136] In this embodiment, an ultrasound image processing device is also provided. This device is used to implement the above embodiments and preferred implementation manners, and those that have been described will not be elaborated. As used hereinafter, the term "module" can be a combination of software and / or hardware that can achieve a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0137] This embodiment provides an ultrasound image processing device, as Figure 11 shown, including:

[0138] An acquisition module 1101, configured to acquire consecutive ultrasound images in a target scanning phase, where the target scanning phase includes a phase of performing an ultrasound scan on a pregnant woman's abdomen according to a specified operation method, and the specified operation method is used to characterize the scanning direction and position.

[0139] A segmentation module 1102, configured to identify ultrasound images belonging to the no-load data category in the consecutive ultrasound images, and to divide a plurality of first image sets from the consecutive ultrasound images according to the identification result, where the ultrasound images in the first image sets are ultrasound images of the non-no-load data category.

[0140] An analysis module 1103, configured to analyze the ultrasound images in at least one first image set, and to determine the fetal position according to the analysis result.

[0141] In some alternative embodiments, the segmentation module 1102 includes:

[0142] An identification unit, configured to classify each ultrasound image in the consecutive ultrasound images to screen out the ultrasound images belonging to the no-load data category.

[0143] In some alternative embodiments, the segmentation module 1102 further includes:

[0144] A division unit, configured to divide the consecutive ultrasound images into a plurality of first image sets and a plurality of second image sets according to the ultrasound images belonging to the no-load data category, where the images in the second image sets are ultrasound images of the no-load data category.

[0145] In some alternative embodiments, the identification unit is specifically configured to classify each ultrasound image in the consecutive ultrasound images through a trained first classification network to identify the ultrasound images belonging to the no-load data category; the input of the first classification network is the ultrasound image, and the output is the image category, and the image category includes the no-load data category.

[0146] In some alternative embodiments, the analysis module 1103 is specifically configured to classify the ultrasound images in at least one first image set through a trained second classification network to identify the classification result corresponding to the ultrasound images in the first image set; the input of the second classification network is the ultrasound image, and the output is the classification result, and the classification result includes the analysis result.

[0147] In some alternative embodiments, the plurality of first image sets correspond one-to-one to the phases of performing an ultrasound scan on a pregnant woman's abdomen according to the specified operation method.

[0148] In some alternative embodiments, the phase of performing an ultrasound scan on a pregnant woman's abdomen includes a first scan phase.

[0149] The analysis module 1103 is specifically configured to determine that the fetal position is the occipital posterior position according to the classification result of the ultrasound images in the first image set corresponding to the first scanning stage, where the classification result includes an orbital section.

[0150] In some alternative embodiments, the stage of performing an ultrasound scan on the pregnant woman's abdomen further includes a second scanning stage, and the second scanning stage is a stage after the first scanning stage.

[0151] The analysis module 1103 is specifically configured to determine that the fetal position is the occipital posterior position according to the classification result of the ultrasound images in the first preset time period in the first image set corresponding to the second scanning stage, where the classification result includes an orbital section.

[0152] The analysis module 1103 is specifically configured to determine that the fetal position is the occipital anterior position according to the classification result of the ultrasound images in the first preset time period in the first image set corresponding to the second scanning stage, where the classification result includes a spinal column section.

[0153] The analysis module 1103 is specifically configured to determine that the fetal position is the right occipital transverse position according to the classification result of the ultrasound images in the first image set corresponding to the first scanning stage, where the classification result includes a cerebral midline section, and the classification result of the ultrasound images in the second preset time period in the first image set corresponding to the second scanning stage includes a spinal column section.

[0154] The analysis module 1103 is specifically configured to determine that the fetal position is the left occipital transverse position according to the classification result of the ultrasound images in the first image set corresponding to the first scanning stage, where the classification result includes a cerebral midline section, and the classification result of the ultrasound images in the second preset time period in the first image set corresponding to the second scanning stage includes an orbital section.

[0155] In some alternative embodiments, the stage of performing an ultrasound scan on the pregnant woman's abdomen further includes a third scanning stage, and the third scanning stage is a stage after the second scanning stage.

[0156] The analysis module 1103 is specifically configured to determine that the fetal position is the occipital posterior position according to the classification result of the ultrasound images in the third preset time period in the first image set corresponding to the third scanning stage, where the classification result includes an orbital section;

[0157] The analysis module 1103 is specifically configured to determine that the fetal position is the occipital anterior position according to the classification result of the ultrasound images in the third preset time period in the first image set corresponding to the third scanning stage, where the classification result includes a spinal column section;

[0158] The analysis module 1103 is specifically configured to determine that the fetal position is the left occipital transverse position according to the classification result of the ultrasound images in the first image set corresponding to the first scanning stage, where the classification result includes a cerebral midline section, and the classification result of the ultrasound images in the fourth preset time period in the first image set corresponding to the third scanning stage includes a spinal column section;

[0159] The analysis module 1103 is specifically configured to determine that the fetal position is the right occipital transverse position based on that the classification result of the ultrasonic images in the first image set corresponding to the first scanning stage includes the midline section of the brain, and the classification result of the ultrasonic images in the fourth preset time period in the first image set corresponding to the third scanning stage includes the orbital section.

[0160] The further function descriptions of the above-mentioned various modules and various units are the same as those in the corresponding embodiments above, and will not be repeated here.

[0161] The ultrasonic image processing device in this embodiment is presented in the form of functional units. Here, the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and a memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0162] The embodiment of the present invention also provides an ultrasonic device having the above-mentioned Figure 11 shown ultrasonic image processing device.

[0163] Please refer to Figure 12 , Figure 12 which is a schematic structural diagram of an ultrasonic device provided by an optional embodiment of the present invention. As Figure 12 shown, the ultrasonic device is communicatively connected to an ultrasonic probe to receive or acquire ultrasonic images collected by the ultrasonic probe. The ultrasonic device includes: one or more processors 10, a memory 20, a display 40, and an interface for connecting various components, including a high-speed interface and a low-speed interface. Each component is communicatively connected to each other using different buses and can be installed on a common main board or installed in other ways as needed. The processor can process instructions executed within the ultrasonic device, including instructions stored in the memory or on the memory to display graphic information of a GUI on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple ultrasonic devices can be connected, and each device provides some necessary operations (for example, as a server array, a set of blade servers, or a multi-processor system). Figure 12 Taking one processor 10 as an example in

[0164] The processor 10 may be a central processing unit, a network processor, or a combination thereof. Among them, the processor 10 may further include a hardware chip. The above-mentioned hardware chip may be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The above-mentioned programmable logic device may be a complex programmable logic device, a field programmable gate array, a generic array logic, or any combination thereof.

[0165] Among them, the memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiments.

[0166] The memory 20 may include a program storage area and a data storage area. Among them, the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created according to the use of the ultrasonic device, etc. In addition, the memory 20 may include a high-speed random access memory, and may further include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some alternative embodiments, the memory 20 may optionally include a memory remotely provided with respect to the processor 10, and these remote memories may be connected to the ultrasonic device through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0167] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk, or a solid-state drive; the memory 20 may further include a combination of the above-mentioned types of memories.

[0168] The ultrasonic device further includes a communication interface 30 for the ultrasonic device to communicate with other devices or communication networks.

[0169] An embodiment of the present invention further provides a computer-readable storage medium. The method according to the embodiment of the present invention may be implemented in hardware, firmware, or may be implemented as computer code recorded on a storage medium, or may be implemented as computer code originally stored in a remote storage medium or a non-temporary machine-readable storage medium and downloaded through a network and to be stored in a local storage medium, so that the method described herein may be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium may be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid-state drive, etc.; further, the storage medium may further include a combination of the above-mentioned types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component capable of storing or receiving software or computer code, and when the software or computer code is accessed and executed by the computer, the processor, or the hardware, the method shown in the above embodiments is implemented.

[0170] Although embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations fall within the scope defined by the appended claims.

Claims

1. An ultrasonic image processing method, characterized in that, The method includes: Obtaining consecutive ultrasound images of a target scanning stage, where the target scanning stage includes a stage of performing ultrasound scanning on a pregnant woman's abdomen in a specified operation mode, and the specified operation mode is used to characterize the scanning direction and position; Identifying the ultrasound images belonging to the no-load data category in the consecutive ultrasound images, and dividing the consecutive ultrasound images into multiple first image sets according to the identification result, where the ultrasound images in the first image set are non-no-load data category ultrasound images; Analyzing the ultrasound images in at least one of the first image sets, and determining the fetal position according to the analysis result.

2. The method according to claim 1, wherein The identifying the ultrasound images belonging to the no-load data category in the consecutive ultrasound images includes: Classifying each ultrasound image in the consecutive ultrasound images to screen out the ultrasound images belonging to the no-load data category.

3. The method according to claim 1, characterized in that, The dividing the consecutive ultrasound images into multiple first image sets according to the identification result includes: Dividing the consecutive ultrasound images into multiple first image sets and multiple second image sets according to the ultrasound images belonging to the no-load data category, where the images in the second image set are no-load data category ultrasound images.

4. The method according to claim 2, wherein The classifying each ultrasound image in the consecutive ultrasound images includes: Classifying each ultrasound image in the consecutive ultrasound images through a trained first classification network to identify the ultrasound images belonging to the no-load data category; the input of the first classification network is the ultrasound image, and the output is the image category, and the image category includes the no-load data category.

5. The method according to claim 1, characterized in that The analyzing the ultrasound images in at least one of the first image sets includes: Classifying the ultrasound images in at least one of the first image sets through a trained second classification network to identify the classification result corresponding to the ultrasound images in the first image set; the input of the second classification network is the ultrasound image, and the output is the classification result, and the classification result includes the analysis result.

6. The method according to claim 1, wherein The multiple first image sets correspond one-to-one to the stages of performing ultrasound scanning on the pregnant woman's abdomen in the specified operation mode.

7. The method according to any one of claims 1 to 6, characterized in that, The stage of performing ultrasound scanning on the pregnant woman's abdomen includes a first scanning stage; the analyzing the ultrasound images in at least one of the first image sets and determining the fetal position according to the analysis result includes: If the classification result of the ultrasound images in the first image set corresponding to the first scanning stage includes an orbital section, determining that the fetal position is the occipital posterior position.

8. The method according to claim 7, wherein The stage of performing ultrasound scanning on the pregnant woman's abdomen further includes a second scanning stage, and the second scanning stage is the stage after the first scanning stage; the analyzing the ultrasound images in at least one of the first image sets and determining the fetal position according to the analysis result includes: If the classification result of the ultrasound images in the first preset time period in the first image set corresponding to the second scanning stage includes an orbital section, determining that the fetal position is the occipital posterior position; If the classification result of the ultrasound images in the first preset time period in the first image set corresponding to the second scanning stage includes a spinal cord section, determine that the fetal position is the occipitoanterior position; If the classification result of the ultrasound images in the first image set corresponding to the first scanning stage includes a cerebral midline section, and the classification result of the ultrasound images in the second preset time period in the first image set corresponding to the second scanning stage includes a spinal cord section, determine that the fetal position is the right occipitotransverse position; If the classification result of the ultrasound images in the first image set corresponding to the first scanning stage includes a cerebral midline section, and the classification result of the ultrasound images in the second preset time period in the first image set corresponding to the second scanning stage includes an orbital section, determine that the fetal position is the left occipitotransverse position.

9. The method according to claim 8, wherein The stage of performing ultrasound scanning on the pregnant woman's abdomen further includes a third scanning stage, which is the stage after the second scanning stage; analyzing the ultrasound images in at least one of the first image sets and determining the fetal position according to the analysis result includes: If the classification result of the ultrasound images in the third preset time period in the first image set corresponding to the third scanning stage includes an orbital section, determine that the fetal position is the occipitoposterior position; If the classification result of the ultrasound images in the third preset time period in the first image set corresponding to the third scanning stage includes a spinal cord section, determine that the fetal position is the occipitoanterior position; If the classification result of the ultrasound images in the first image set corresponding to the first scanning stage includes a cerebral midline section, and the classification result of the ultrasound images in the fourth preset time period in the first image set corresponding to the third scanning stage includes a spinal cord section, determine that the fetal position is the left occipitotransverse position; If the classification result of the ultrasound images in the first image set corresponding to the first scanning stage includes a cerebral midline section, and the classification result of the ultrasound images in the fourth preset time period in the first image set corresponding to the third scanning stage includes an orbital section, determine that the fetal position is the right occipitotransverse position.

10. An ultrasonic image processing apparatus, characterized in that, The device includes: An acquisition module, configured to acquire consecutive ultrasound images of a target scanning stage, where the target scanning stage includes a stage of performing ultrasound scanning on the pregnant woman's abdomen according to a specified operation method, and the specified operation method is used to represent the scanning direction and position; A segmentation module, configured to identify the ultrasound images belonging to the no-load data category in the consecutive ultrasound images, and to divide a plurality of first image sets from the consecutive ultrasound images according to the identification result, where the ultrasound images in the first image sets are non-no-load data category ultrasound images; An analysis module, configured to analyze the ultrasound images in at least one of the first image sets, and to determine the fetal position according to the analysis result.

11. An ultrasonic device, characterized in that, Including: A memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the ultrasound image processing method according to any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that, Computer instructions are stored on the computer-readable storage medium, and the computer instructions are used to cause a computer to execute the ultrasonic image processing method according to any one of claims 1 to 9.

Images