Ultrasonic image data management method and system based on artificial intelligence

Abstract

The invention discloses an artificial intelligence-based ultrasonic image data management method and system, and the method comprises the steps: collecting ultrasonic data of a target object through an ultrasonic detection device, and obtaining n pieces of ultrasonic data; determining a corresponding ultrasonic image in each piece of ultrasonic data in the n pieces of ultrasonic data to obtain n ultrasonic images; segmenting each ultrasonic image in the n ultrasonic images to obtain m segmented ultrasonic images; annotating the m segmented ultrasonic images through a preset image annotation model to obtain m annotation data; transmitting the m annotation data and the m segmented ultrasonic images to a storage module; ultrasonic detection parts corresponding to the m segmented ultrasonic images are determined, and i ultrasonic detection parts are obtained; and the m annotation data and the m segmented ultrasonic images are stored through a storage module according to the i ultrasonic detection parts, and the stored data are classified and indexed. By adopting the embodiment of the invention, the refinement degree of ultrasonic image data management is improved.

Description

Technical Field

[0001] The present application relates to the technical field of data management, and particularly to an artificial intelligence-based ultrasonic image data management method and system. Background Art

[0002] Ultrasonic data is the data generated by ultrasonic detection equipment, also known as ultrasonic image data. It is a type of non-invasive image data. Using the physical principle of ultrasonic waves, ultrasonic waves are emitted by an ultrasonic probe, and then the reflected acoustic wave signals are recorded. The signals are converted into image data to display the internal part structure and morphology of the object under test. Ultrasonic data is widely used in the medical field for detecting diseases, diagnosing conditions, and observing treatment effects, etc.

[0003] Currently, ultrasonic image data is usually classified according to simple information such as inspection date and patient number, and an ultrasonic image is stored as a whole, which cannot meet complex clinical needs and data analysis requirements, and the degree of refinement is low. For example, when studying the ultrasonic imaging characteristics of thyroid diseases, it is difficult to quickly screen out thyroid ultrasonic images with the same pathological characteristics, different ages, genders, and disease stages from a large amount of data; therefore, how to improve the refinement degree of ultrasonic image data management has become an urgent problem to be solved. Summary of the Invention

[0004] The embodiments of the present application provide an artificial intelligence-based ultrasonic image data management method and system, which improve the refinement degree of ultrasonic image data management.

[0005] In a first aspect, the embodiments of the present application provide an artificial intelligence-based ultrasonic image data management method, which is applied to a control module of an ultrasonic image data management device. The ultrasonic image data management device further includes: an ultrasonic detection device and a storage module. The method includes:

[0006] Collect ultrasonic data of a target object through the ultrasonic detection device to obtain n ultrasonic data; n is a positive integer;

[0007] Determine the corresponding ultrasonic image in each of the n ultrasonic data to obtain n ultrasonic images;

[0008] Segment each of the n ultrasonic images to obtain m segmented ultrasonic images; m is an integer greater than or equal to n;

[0009] Label the m segmented ultrasonic images through a preset image annotation model to obtain m annotation data;

[0010] Transmit the m annotation data and the m segmented ultrasonic images to the storage module;

[0011] Determine the ultrasonic detection parts corresponding to the m segmented ultrasonic images to obtain i ultrasonic detection parts; i is a positive integer less than or equal to m;

[0012] The storage module stores the m labeled data and the m segmented ultrasonic images according to the i ultrasonic detection parts, and classifies and indexes the stored data.

[0013] In a second aspect, an embodiment of the present application provides an artificial intelligence-based ultrasonic image data management system, which is applied to a control module of an ultrasonic image data management device. The ultrasonic image data management device further includes: an ultrasonic detection device and a storage module. The system includes: an acquisition module, a determination module, and a management module, where:

[0014] The acquisition module is configured to collect ultrasonic data of a target object through the ultrasonic detection device to obtain n ultrasonic data; n is a positive integer;

[0015] The determination module is configured to determine the ultrasonic images corresponding to each of the n ultrasonic data to obtain n ultrasonic images;

[0016] The management module is configured to segment each of the n ultrasonic images to obtain m segmented ultrasonic images; m is an integer greater than or equal to n; label the m segmented ultrasonic images through a preset image annotation model to obtain m labeled data; and transmit the m labeled data and the m segmented ultrasonic images to the storage module;

[0017] The determination module is further configured to determine the ultrasonic detection parts corresponding to the m segmented ultrasonic images to obtain i ultrasonic detection parts; i is a positive integer less than or equal to m;

[0018] The management module is further configured to store the m labeled data and the m segmented ultrasonic images according to the i ultrasonic detection parts through the storage module, and classify and index the stored data.

[0019] In a third aspect, the present application provides an electronic device, including: a processor and a memory. The memory is used to store one or more programs. Among them, the above one or more programs are stored in the above memory and are configured to be executed by the above processor. The above programs include instructions for executing the steps in the first aspect of the present application.

[0020] In a fourth aspect, the present application provides a computer-readable storage medium. Among them, the above computer-readable storage medium stores a computer program for electronic data exchange. Among them, the above computer program enables a computer to execute some or all of the steps described in the first aspect of the present application.

[0021] Fifth aspect, the present application provides a computer program product, wherein the computer program product includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to cause a computer to execute some or all of the steps described in the first aspect of the present application. The computer program product can be a software installation package.

[0022] Implementing the present application has the following beneficial effects:

[0023] It can be seen that the artificial intelligence-based ultrasonic image data management method described in the present application includes: collecting ultrasonic data of a target object through an ultrasonic detection device to obtain n ultrasonic data; determining the corresponding ultrasonic images in each of the n ultrasonic data to obtain n ultrasonic images; segmenting each of the n ultrasonic images to obtain m segmented ultrasonic images; labeling the m segmented ultrasonic images through a preset image labeling model to obtain m labeled data; transmitting the m labeled data and the m segmented ultrasonic images to a storage module; determining the ultrasonic detection parts corresponding to the m segmented ultrasonic images to obtain i ultrasonic detection parts; storing the m labeled data and the m segmented ultrasonic images by the storage module according to the i ultrasonic detection parts, and classifying and indexing the stored data. By segmenting each ultrasonic image, different tissue structures or regions of interest in the ultrasonic image are separately extracted to obtain multiple segmented ultrasonic images. In this way, an ultrasonic image containing multiple complex structures is refined into multiple segmented ultrasonic images with specific targets, making the subsequent management and analysis of each specific structure more targeted, avoiding information confusion and omission that may occur during overall processing, and thus improving the refinement degree of ultrasonic image data management. Description of the Drawings

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the background art, the following will describe the drawings required to be used in the embodiments of the present application or the background art.

[0025] Figure 1 is a schematic diagram of an ultrasonic image data management device provided by an embodiment of the present application;

[0026] Figure 2 is a flowchart of an artificial intelligence-based ultrasonic image data management method provided by an embodiment of the present application;

[0027] Figure 3 is a block diagram of the functional units of an artificial intelligence-based ultrasonic image data management system provided by an embodiment of the present application;

[0028] Figure 4 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. Detailed implementation manners

[0029] To enable those skilled in the art to better understand the solution of this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this application without creative efforts shall fall within the protection scope of this application.

[0030] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned accompanying drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally further includes steps or units not listed, or optionally further includes other steps or units inherent to these processes, methods, products or devices.

[0031] Referring to "embodiments" herein means that the specific features, structures or characteristics described in connection with the embodiments can be included in at least one embodiment of this application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0032] The electronic devices described in the embodiments of this application may include smart phones (such as Android phones, iOS phones, Windows Phone phones, etc.), tablet computers, handheld computers, laptop computers, video matrices, monitoring platforms, mobile internet devices (MID) or wearable devices, etc. The above are only examples, not an exhaustive list, including but not limited to the above devices. Of course, the above electronic devices may also be servers, for example, cloud servers.

[0033] Please refer to Figure 1 , Figure 1 which is a schematic diagram of an ultrasonic image data management device provided by an embodiment of this application. The ultrasonic image data management device includes: an ultrasonic detection device, a storage module, and a control module, where:

[0034] An ultrasonic detection device may include an ultrasonic probe and a main unit. The ultrasonic probe is a component used to transmit and receive ultrasonic waves, and there are various types according to different examination purposes and sites. For example, high-frequency probes (7.5 - 10 MHz) are used to examine superficial tissues such as the thyroid gland, mammary gland, etc.; low-frequency probes (2 - 5 MHz) are suitable for deep tissues such as the liver, kidney, etc. The main unit is responsible for generating electrical signals to drive the probe to emit ultrasonic waves and processing the ultrasonic echo signals received by the probe, including operations such as amplification, filtering, and digitization. The core function of the ultrasonic detection device is to collect ultrasonic data of the target object. During the examination, the ultrasonic probe contacts the patient's body and emits ultrasonic waves into the human tissue. When the ultrasonic waves propagate in the tissue, phenomena such as reflection, refraction, and scattering will occur due to the acoustic impedance differences of different tissues. The probe receives these echo signals and converts them into electrical signals for transmission to the main unit. After the main unit processes these electrical signals, ultrasonic data is formed, which is the basis for generating ultrasonic images subsequently and is also the starting point of the entire ultrasonic image data management process.

[0035] The storage module can be a combination of various storage media and architectures. Common storage media include hard disk drives (HDDs), solid-state drives (SSDs), tape libraries, and cloud storage. Hard disk drives have the characteristics of large capacity and relatively low cost, and are suitable for storing a large amount of ultrasonic image data; solid-state drives have fast read and write speeds and can be used to store frequently accessed data to improve data access efficiency; tape libraries are suitable for long-term archiving and backup of data; cloud storage provides high scalability and flexibility, facilitating data sharing and remote access between multiple institutions. The storage architecture can be centralized storage for convenient unified management; or it can be distributed storage to improve the reliability, availability, and scalability of the storage system. The main role of the storage module is to store ultrasonic image data and its related information. It can receive ultrasonic images, annotation data, etc. transmitted from the ultrasonic detection device or the control module and store them according to certain rules. For example, classify and store the data according to the ultrasonic examination site (such as the heart, liver, etc.), and at the same time establish an index for convenient subsequent retrieval and query. The storage module is also responsible for data backup and maintenance. By backing up data regularly (such as a combination of full backups and incremental backups), data loss is prevented; and the stored data is regularly cleaned and sorted to optimize the utilization of storage resources.

[0036] The control module can be a unit containing a processor, memory, and related control software. The processor is responsible for executing data processing and control instructions, and the memory is used for temporarily storing data and intermediate results during program operation. The control module can be used to receive ultrasonic data collected by an ultrasonic detection device, process it to determine the corresponding ultrasonic image therein. For example, by processing the ultrasonic data with a reconstruction algorithm, it is converted into a visual ultrasonic image format. Additionally, it can also perform operations such as segmenting and annotating the ultrasonic image. Using a built-in image segmentation algorithm, the ultrasonic image is segmented into multiple parts with specific meanings (such as segmenting out different organ images), and then annotation information (such as organ names, lesion characteristics, etc.) is added through a preset image annotation model. After that, the control module transmits the processed annotation data and segmented ultrasonic image to the storage module for storage, and controls the storage module to classify and index the data to ensure the orderly progress of the entire data management process.

[0037] Please refer to Figure 2 , Figure 2 is a flowchart of an artificial intelligence-based ultrasonic image data management method provided by an embodiment of the present application. This method can be applied to the control module of an ultrasonic image data management device, and the ultrasonic image data management device further includes: an ultrasonic detection device, a storage module. The method includes:

[0038] S201. Collect ultrasonic data of a target object through the ultrasonic detection device to obtain n ultrasonic data; n is a positive integer.

[0039] In an embodiment of the present application, the ultrasonic detection device may include at least one of the following: an ultrasonic diagnostic instrument, an ultrasonic endoscope device, an intravascular ultrasonic device, etc., which is not limited herein.

[0040] In a specific embodiment, an ultrasonic detection plan formulated by a target doctor for a target object can be obtained, the target human area to be detected is determined according to the ultrasonic detection plan, and the ultrasonic detection device is started to perform ultrasonic detection on the target human area to obtain n ultrasonic data; wherein, the target doctor is the doctor who diagnoses the target object.

[0041] S202. Determine the corresponding ultrasonic image in each of the n ultrasonic data to obtain n ultrasonic images.

[0042] Optionally, in step S202, the determining the corresponding ultrasonic image in each of the n ultrasonic data to obtain n ultrasonic images may include the following steps:

[0043] A1. Determine the corresponding first ultrasonic image in each of the n ultrasonic data to obtain n first ultrasonic images;

[0044] A2. Detect whether each of the n first ultrasonic images meets a preset qualified condition;

[0045] A3. If it meets the preset qualified condition, determine the n ultrasonic images according to the n first ultrasonic images;

[0046] A4. If it does not meet the preset qualified condition, determine the first ultrasonic images that do not meet the preset qualified condition among the n first ultrasonic images, obtaining a first ultrasonic images, and n - a first ultrasonic images among the n first ultrasonic images other than the a first ultrasonic images; a is a natural number less than or equal to n;

[0047] A5. Determine the unqualified reasons why each of the a first ultrasonic images does not meet the preset qualified condition, obtaining a unqualified reasons;

[0048] A6. Perform repair processing on the a first ultrasonic images according to the a unqualified reasons to obtain a second ultrasonic images;

[0049] A7. Determine the n ultrasonic images according to the n - a first ultrasonic images and the a second ultrasonic images.

[0050] In the embodiments of the present application, the preset qualified condition can be preset in advance or by default; the unqualified reasons may include one of the following: device-related reasons (such as probe problems, device parameter setting problems), target object-related reasons (such as target object body movement, interference from special physiological states in the body), environment-related reasons (such as electromagnetic interference, temperature and humidity effects), and imaging algorithm-related reasons (such as beamforming algorithm errors, improper gray mapping), etc., which are not limited herein.

[0051] In a specific embodiment, the first ultrasonic image corresponding to each of the n ultrasonic data can be determined to obtain n first ultrasonic images. Specifically, each ultrasonic data includes multiple ultrasonic signals, and the ultrasonic signals in the n ultrasonic data can be processed by using the method of gray mapping to obtain n first ultrasonic images. For example, the intensity range of the ultrasonic signals can be mapped to the gray value range (generally 0 - 255), the part with lower signal intensity is mapped to a darker gray (close to 0), and the part with higher signal intensity is mapped to a brighter gray (close to 255). In this way, ultrasonic signals with different intensities can be represented by different grayscales, thereby forming an ultrasonic image with light and dark contrast. Among them, the bright area usually represents strongly reflecting tissues (such as bones), and the dark area represents weakly reflecting tissues (such as liquids); then, it can be detected whether each of the n first ultrasonic images meets the preset qualified condition;

[0052] If all of the n first ultrasound images meet the preset qualified conditions, the n first ultrasound images can be directly determined as the n ultrasound images;

[0053] If there are first ultrasound images among the n first ultrasound images that do not meet the preset qualified conditions, then the first ultrasound images that do not meet the preset qualified conditions among the n first ultrasound images can be determined, obtaining a first ultrasound images, and n - a first ultrasound images among the n first ultrasound images excluding the a first ultrasound images; then, the unqualified reasons for each of the a first ultrasound images not meeting the preset qualified conditions can be determined, obtaining a unqualified reasons. Specifically, an image analysis software (or an ultrasound doctor) can be used to check each first ultrasound image, identify the places in the image that clearly do not meet the qualified conditions. For example, it is found that there are large blurred areas in the image, the tissue boundaries cannot be distinguished, or there are obvious artifacts (such as ghosting, abnormal comet-tail artifacts, etc.). At the same time, these preliminarily judged unqualified features are recorded. Then, the unqualified reasons can be determined based on these unqualified features, thereby obtaining a unqualified reasons. For example, assuming that the recorded unqualified feature is "the image is blurred because the target object moves, resulting in artifacts", then the unqualified reason is: reasons related to the target object; then, the a first ultrasound images can be repaired according to the a unqualified reasons to obtain a second ultrasound images; finally, the n - a first ultrasound images and the a second ultrasound images can be used as the n ultrasound images.

[0054] In this way, by detecting whether the first ultrasound images meet the preset qualified conditions, images with poor quality can be screened out, which helps to ensure that the ultrasound images finally used for diagnosis or other purposes have high quality and reduce diagnostic errors caused by image quality problems. For example, in some scenarios with high requirements for detecting subtle lesions (such as early tumor screening), clear and standard-compliant images can more accurately display the lesion characteristics and provide reliable diagnostic basis for doctors.

[0055] Optionally, in step A6, the repairing the a first ultrasound images according to the a unqualified reasons to obtain a second ultrasound images may include the following steps:

[0056] B1. Determine the repair operations corresponding to the a unqualified reasons to obtain a repair operations;

[0057] B2. Repair the corresponding first ultrasound images among the a first ultrasound images according to the a repair operations to obtain a third ultrasound images;

[0058] B3. Detect whether each of the a third ultrasound images meets the preset qualified conditions;

[0059] B4. If the preset qualification conditions are met, determine the a second ultrasonic images according to the a third ultrasonic images;

[0060] B5. If the preset qualification conditions are not met, determine the third ultrasonic images among the a third ultrasonic images that do not meet the preset qualification conditions, obtaining b third ultrasonic images, and the a - b third ultrasonic images among the a third ultrasonic images other than the b third ultrasonic images; b is a natural number less than or equal to a;

[0061] B6. Determine the acquisition parameters corresponding to each of the b third ultrasonic images, obtaining b acquisition parameters;

[0062] B7. Determine the unqualified reasons corresponding to the b third ultrasonic images among the a unqualified reasons, obtaining b unqualified reasons;

[0063] B8. Determine the fine - tuning factors corresponding to the b unqualified reasons, obtaining b fine - tuning factors;

[0064] B9. Adjust the corresponding acquisition parameters among the b acquisition parameters according to the b fine - tuning factors, obtaining b target acquisition parameters;

[0065] B10. Use the ultrasonic detection device to acquire ultrasonic images of the target object with the b target acquisition parameters respectively, obtaining b fourth ultrasonic images;

[0066] B11. Determine the a second ultrasonic images according to the a - b third ultrasonic images and the b fourth ultrasonic images.

[0067] In the embodiments of the present application, the repair operation may include at least one of the following: filtering and denoising operation, image sharpening operation, gray - level stretching operation, histogram equalization operation, artifact repair operation, etc., which are not limited herein; the acquisition parameters may include at least one of the following: frequency parameter, gain parameter, transmit power parameter, focusing parameter, etc., which are not limited herein.

[0068] In a specific embodiment, the repair operations corresponding to the a unqualified reasons may be determined, obtaining a repair operations. Specifically, the mapping relationship between the preset unqualified reasons and the repair operations may be stored in advance, and based on this mapping relationship, the a repair operations corresponding to the a unqualified reasons are determined; then, the corresponding first ultrasonic images among the a first ultrasonic images may be repaired according to the a repair operations, obtaining a third ultrasonic images; detect whether each of the a third ultrasonic images meets the preset qualification conditions; if all of the a third ultrasonic images meet the preset qualification conditions, the a third ultrasonic images may be determined as the a second ultrasonic images.

[0069] If there are third ultrasonic images that do not meet the preset qualified conditions among a third ultrasonic images, then the third ultrasonic images that do not meet the preset qualified conditions among the a third ultrasonic images can be found, obtaining b third ultrasonic images, and a - b third ultrasonic images among the a third ultrasonic images except for the b third ultrasonic images; the acquisition parameters corresponding to each of the b third ultrasonic images can be determined, obtaining b acquisition parameters. Specifically, the acquisition time of each of the b third ultrasonic images can be obtained, obtaining b acquisition times, and then, the acquisition record of the ultrasonic detection device can be obtained. The acquisition record can include the acquisition parameters and the acquisition time. Thus, b acquisition parameters can be obtained from the acquisition record according to the b acquisition times; then, the unqualified reasons corresponding to each of the b third ultrasonic images among the b third ultrasonic images can be found, obtaining b unqualified reasons; further, the fine-tuning factors corresponding to the b unqualified reasons can be determined, obtaining b fine-tuning factors. For example, the mapping relationship between the preset unqualified reasons and the fine-tuning factors can be stored in advance, and based on this mapping relationship, the b fine-tuning factors corresponding to the b unqualified reasons can be determined. The value range of the fine-tuning factors can be -0.3 to 0.3; then, the corresponding acquisition parameters among the b acquisition parameters can be adjusted according to the b fine-tuning factors. The specific calculation formula is as follows:

[0070] The first target acquisition parameter = the first acquisition parameter * (1 + the first fine-tuning factor);

[0071] Wherein, the first acquisition parameter is any one of the b acquisition parameters, the first fine-tuning factor is the fine-tuning factor corresponding to the first acquisition parameter among the b fine-tuning factors, and the first target acquisition parameter is the target acquisition parameter corresponding to the first acquisition parameter among the b target acquisition parameters. By calculating according to the above formula b times, b target acquisition parameters can be obtained; then, the ultrasonic images of the target object can be acquired by the ultrasonic detection device respectively with the b target acquisition parameters, obtaining b fourth ultrasonic images; finally, these a - b third ultrasonic images and the b fourth ultrasonic images can be determined as a second ultrasonic images.

[0072] Thus, by performing a repair operation on the unqualified first ultrasonic image to obtain a third ultrasonic image, and then detecting again whether it meets the preset qualified conditions, a multi-level quality optimization process is formed, improving the quality of the ultrasonic image, and thereby, increasing the number of qualified images that meet the diagnostic requirements. For example, for an image that is unqualified due to contrast problems, a contrast repair operation is first performed. If it is still unqualified, the acquisition parameters are further analyzed and fine-tuned for re-acquisition, so that the finally obtained ultrasonic image can meet the quality standards required for clinical diagnosis or research.

[0073] S203. Segment each of the n ultrasonic images to obtain m segmented ultrasonic images; m is an integer greater than or equal to n.

[0074] In the embodiments of the present application, the control module can perform segmentation processing on the n ultrasonic images, and each ultrasonic image is segmented into one or more images, thereby obtaining m segmented ultrasonic images.

[0075] Optionally, in step S203, the step of segmenting each of the n ultrasonic images to obtain m segmented ultrasonic images may include the following steps:

[0076] C1. Segment the target ultrasonic image based on a preset image segmentation method to obtain c segmented ultrasonic images; the target ultrasonic image is any one of the n ultrasonic images; c is a positive integer less than or equal to m;

[0077] C2. Determine the clarity corresponding to each of the c segmented ultrasonic images to obtain c clarities;

[0078] C3. Obtain the device clarity corresponding to the ultrasonic detection device;

[0079] C4. Determine the clarities among the c clarities that are less than the device clarity to obtain d clarities; d is a natural number less than or equal to c;

[0080] C5. Determine the segmented ultrasonic images corresponding to the d clarities among the c segmented ultrasonic images to obtain d segmented ultrasonic images;

[0081] C6. Determine the difference between each of the d clarities and the device clarity to obtain d clarity differences;

[0082] C7. Process the d segmented ultrasonic images based on the d clarity differences to obtain d target segmented ultrasonic images;

[0083] C8. Replace the d segmented ultrasonic images among the c segmented ultrasonic images with the d target segmented ultrasonic images to obtain the segmented ultrasonic images corresponding to the target ultrasonic image.

[0084] In the embodiments of the present application, the preset image segmentation method can be preset in advance or by default.

[0085] In specific embodiments, the target ultrasound image can be segmented based on a preset image segmentation method to obtain c segmented ultrasound images. For example, the preset image segmentation method can be the region growing method. Assuming the target ultrasound image is a breast ultrasound image, the central pixel of a breast nodule can be used as a seed point, and based on the gray-scale similarity and texture similarity of the pixels inside the nodule, the surrounding pixels can be continuously merged into the nodule region, thereby achieving the segmentation of the nodule and obtaining c segmented ultrasound images. Then, the clarity corresponding to each of the c segmented ultrasound images can be determined to obtain c clarity values. Specifically, an image analysis software can be used to analyze the c segmented ultrasound images to obtain c clarity values. Then, the device clarity corresponding to the ultrasound detection device can be obtained. Specifically, the target device type of the ultrasound detection device can be obtained, and the device clarity can be determined according to the target device type. For example, the mapping relationship between the preset device type and the clarity can be pre-stored, and the device clarity corresponding to the target device type can be determined based on this mapping relationship. Then, the clarity values among the c clarity values that are less than the device clarity can be found to obtain d clarity values.

[0086] Further, the segmented ultrasound images corresponding to the d clarity values among the c segmented ultrasound images can be determined to obtain d segmented ultrasound images. Then, the difference between each of the d clarity values and the device clarity can be calculated to obtain d clarity differences. Then, the d segmented ultrasound images can be processed based on the d clarity differences to obtain d target segmented ultrasound images. Finally, the d target segmented ultrasound images can be used to replace the d segmented ultrasound images among the c segmented ultrasound images to obtain the updated c segmented ultrasound images, that is, the segmented ultrasound images corresponding to the target ultrasound image.

[0087] In this way, by determining the clarity of each segmented ultrasound image and comparing it with the device clarity corresponding to the ultrasound detection device, the segmented ultrasound images with clarity lower than the device standard can be screened out. This step is based on an objective clarity measurement standard, ensuring that the subsequent processing is targeted at those parts of the images that really need to improve the quality, thereby improving the resource utilization rate. For example, in some high-precision ultrasound examinations, such as eye ultrasound or small blood vessel ultrasound, high requirements are placed on the device clarity. Through this comparison, the segmented images with poor quality can be accurately located.

[0088] Optionally, in step C7, the processing the d segmented ultrasound images based on the d clarity differences to obtain d target segmented ultrasound images may include the following steps:

[0089] D1. Determine the target clarity enhancement method corresponding to the target clarity difference; the target clarity enhancement method includes one of the following: image filtering processing, gray-scale transformation enhancement, histogram equalization; the target clarity difference is any one of the d clarity differences; the target clarity difference corresponds to the first segmented ultrasound image among the d segmented ultrasound images;

[0090] D2. Process the first segmented ultrasound image based on the target clarity enhancement method to obtain a reference segmented ultrasound image;

[0091] D3. Determine the target segmented ultrasound image corresponding to the first segmented ultrasound image according to the reference segmented ultrasound image.

[0092] In the embodiments of the present application, the target clarity enhancement method corresponding to the target clarity difference can be determined. Specifically, the mapping relationship between the preset clarity difference and the clarity enhancement method can be pre-stored, and the target clarity enhancement method corresponding to the target clarity difference can be determined based on this mapping relationship. Then, the target clarity enhancement method can be used to process the first segmented ultrasound image to obtain a reference segmented ultrasound image. Further, the target segmented ultrasound image corresponding to the first segmented ultrasound image can be determined according to the reference segmented ultrasound image. Specifically, the reference clarity of the reference segmented ultrasound image can be obtained. When the reference clarity is greater than or equal to the device clarity, the reference segmented ultrasound image can be directly determined as the target segmented ultrasound image corresponding to the first segmented ultrasound image.

[0093] When the reference clarity is less than the device clarity, it indicates that the clarity enhancement degree of the target clarity enhancement method for the first segmented ultrasound image is insufficient or the enhancement fails. At this time, the first ultrasound detection part corresponding to the first segmented ultrasound image can be determined, and the first ultrasound detection part of the target object can be re-collected by the ultrasound detection device to obtain a new ultrasound image, and this new ultrasound image can be determined as the target segmented ultrasound image corresponding to the first segmented ultrasound image.

[0094] In this way, by determining the target clarity enhancement method corresponding to the target clarity difference, the most suitable enhancement method can be selected according to the specific clarity deficiency degree of each segmented ultrasound image. Different clarity differences reflect different severity degrees and natures of image quality problems. For example, a smaller clarity difference may only require simple gray-scale transformation enhancement to fine-tune the contrast and brightness of the image to make the image details more obvious; while a larger clarity difference may require more complex image filtering processing to remove noise and blur and restore the clarity of the image. This targeted processing method can effectively improve the quality of the segmented ultrasound image.

[0095] S204. Use a preset image annotation model to annotate the m segmented ultrasound images, obtaining m annotation data.

[0096] In the embodiments of this application, the preset image annotation model can be preset in advance or by default.

[0097] In a specific embodiment, the m segmented ultrasound images can be sequentially input into the preset image annotation model, and the preset image annotation model annotates the m segmented ultrasound images to obtain m annotation data.

[0098] S205. Transmit the m annotation data and the m segmented ultrasound images to the storage module.

[0099] In the embodiments of this application, the control module can transmit the m annotation data and the m segmented ultrasound images to the storage module. Specifically, a suitable transmission protocol can be selected according to the nature of the transmitted data, the transmission distance, security requirements, etc., to obtain the target transmission protocol. For example, for local transmission, a simple file copy protocol (such as the SMB (Server Message Block) protocol for file sharing transmission within a local area network) is sufficient. If it is a remote transmission and data security needs to be ensured, HTTPS (Hypertext Transfer Protocol Secure) or FTPS (File Transfer Protocol based on SSL / TLS) can be used. When transmitting a large number of ultrasound images and annotation data, a dedicated data transmission protocol can also be considered, such as the Storage Service-Class protocol of DICOM, which is designed specifically for the storage and transmission of medical image data, can effectively handle the transmission of a large amount of image data, and supports the management and error recovery of the transmission process. Then, the control module can use the target transmission protocol to transmit the m annotation data and the m segmented ultrasound images, and transmit these data to the storage module.

[0100] S206. Determine the ultrasonic detection parts corresponding to the m segmented ultrasound images, obtaining i ultrasonic detection parts; i is a positive integer less than or equal to m.

[0101] In the embodiments of this application, the ultrasonic detection parts corresponding to the m segmented ultrasound images can be determined, obtaining i ultrasonic detection parts. Specifically, the detection parts can be determined by identifying the anatomical features in each segmented ultrasound image. For example, if the ultrasound image shows typical liver shape, blood vessel distribution, and texture features, it can be determined that the detection part is the liver. Thus, m ultrasonic detection parts are obtained. Then, the m ultrasonic detection parts can be de-duplicated, removing the repeated ultrasonic detection parts to obtain i ultrasonic detection parts.

[0102] S207. Store the m labeled data and the m segmented ultrasound images according to the i ultrasonic detection parts through the storage module, and classify and index the stored data.

[0103] In the embodiment of the present application, the m labeled data and the m segmented ultrasound images can be stored according to the i ultrasonic detection parts through the storage module, and the stored data can be classified and indexed. Specifically, in the storage module, a hierarchical storage structure can be established according to the ultrasonic detection parts. For example, taking the file system as an example, a top-level folder named after the ultrasonic detection part can be created first, such as "liver examination", "heart examination", etc. Sub-folders are then created under each top-level folder according to the specific examination date, examination number or other relevant information to store the corresponding m labeled data and m segmented ultrasound images, and an index is created for each data.

[0104] Optionally, in step S207, the storing the m labeled data and the m segmented ultrasound images according to the i ultrasonic detection parts and classifying and indexing the stored data may include the following steps:

[0105] E1. Classify the m labeled data and the m segmented ultrasound images according to the i ultrasonic detection parts to obtain i data sets; each data set corresponds to an ultrasonic detection part; each data set includes at least one labeled data and at least one segmented ultrasound image;

[0106] E2. Compress the i data sets to obtain i compressed data sets;

[0107] E3. Store the i compressed data sets into the storage module, and generate corresponding indexes for each of the i compressed data sets to obtain i indexes.

[0108] In the embodiments of the present application, m annotation data and m segmented ultrasound images can be classified according to i ultrasonic detection parts, and the annotation data and the segmented ultrasound images corresponding to the same ultrasonic detection part are divided into a data set. Since there are i ultrasonic detection parts, m annotation data and m segmented ultrasound images can be divided into i data sets. Then, the i data sets can be compressed to obtain i compressed data sets. Finally, the i compressed data sets can be stored in a storage module, and a corresponding index is generated for each compressed data set in the i compressed data sets to obtain i indexes. Specifically, in the storage module, a folder dedicated to storing these compressed data sets can be created, such as named "compressed ultrasound data set", and the i compressed data sets are placed in this newly created "compressed ultrasound data set" folder through operations such as file copying and moving, and a corresponding index is generated for each compressed data set to obtain i indexes. For example, indexes can be created by combining three fields of "ultrasonic detection part", "examination date", and "name of the target object", so that when querying according to these three conditions simultaneously (such as finding the examination data of a certain part of a certain object on a specific date), relevant data can be located more quickly.

[0109] In this way, by compressing the i data sets, the storage space occupied by the data can be significantly reduced, and the storage space utilization rate can be improved.

[0110] Optionally, in step E2, the compressing the i data sets to obtain i compressed data sets may include the following steps:

[0111] F1. Determine the first data type corresponding to the i data sets;

[0112] F2. Determine the second data type corresponding to the storage module;

[0113] F3. Convert the i data sets from the first data type to the second data type to obtain i first data sets;

[0114] F4. Obtain the target device performance parameters of the ultrasonic image data management device;

[0115] F5. Select a compression method corresponding to the target device performance parameters from a preset compression method library to obtain p compression methods; p is a positive integer;

[0116] F6. Determine the data quality loss parameters of each compression method in the p compression methods to obtain p data quality loss parameters;

[0117] F7. Obtain the target data importance level corresponding to the i first data sets;

[0118] F8. Determine the target loss parameter range corresponding to the importance level of the target data;

[0119] F9. Determine the data quality loss parameters among the p data quality loss parameters that fall within the target loss parameter range, obtaining q data quality loss parameters; q is a positive integer less than or equal to p;

[0120] F10. Determine the q compression methods corresponding to the q data quality loss parameters;

[0121] F11. Determine the compression ratio of each compression method among the q compression methods, obtaining q compression ratios;

[0122] F12. Determine the maximum compression ratio among the q compression ratios, and determine the compression method corresponding to the maximum compression ratio as the target compression method;

[0123] F13. Compress the i first data sets based on the target compression method to obtain the i compressed data sets.

[0124] In the embodiments of the present application, the data type includes one of the following: string type, JSON type, XML type, etc., which are not limited herein; the target device performance parameters may include one of the following: storage read and write speed, floating-point operations per second, etc., which are not limited herein.

[0125] In a specific embodiment, the first data type corresponding to the i data sets may be determined first. Specifically, the data storage format of the i data sets may be checked, and then the first data type may be determined according to this data storage format; then, the second data type corresponding to the storage module may be determined. Specifically, the usage document of the storage module may be obtained, and the data type that the storage module can store may be determined according to this usage document to obtain the second data type; then, the i data sets may be converted from the first data type to the second data type to obtain the i first data sets. Specifically, the conversion method for converting the first data type to the second data type may be obtained to obtain the first conversion method. For example, if the first data type is the string type and the second data type is the JSON type, the first conversion method may be a string splitting operation; then, the i data sets may be processed using this first conversion method to obtain the i first data sets; then, the target device performance parameters of the ultrasonic image data management device may be obtained. Specifically, the user manual of the ultrasonic image data management device may be obtained, and the target device performance parameters may be found from this user manual.

[0126] Next, a compression method corresponding to the performance parameters of the target device can be selected from the preset compression method library to obtain p compression methods. Specifically, the device performance parameters required for each compression method in the preset compression method library can be obtained to get x device performance parameters, where x is an integer greater than or equal to p. Then, the device performance parameters less than the performance parameters of the target device can be found from the x device performance parameters to obtain p device performance parameters, and the p compression methods corresponding to these p device performance parameters in the preset compression method library can be determined. Then, the data quality loss parameter of each compression method among the p compression methods can be determined to obtain p data quality loss parameters. Specifically, the mapping relationship between the preset compression method and the data quality loss parameter can be stored in advance, and the p data quality loss parameters corresponding to the p compression methods can be determined based on this mapping relationship.

[0127] It should be noted that the data quality loss parameter is used to represent the loss of image details after data compression.

[0128] Furthermore, the target data importance level corresponding to the i first data sets can be obtained. Specifically, the target data usage of the i first data sets can be obtained first, and the target data importance level can be determined according to the target data usage. For example, the mapping relationship between the preset data usage and the data importance level can be stored in advance, and the target data importance level corresponding to the target data usage can be determined based on this mapping relationship. Then, the target loss parameter range corresponding to the target data importance level can be determined. Similarly, the mapping relationship between the preset data importance level and the loss parameter range can be stored in advance, and the target loss parameter range corresponding to the target data importance level can be determined based on this mapping relationship. Next, the data quality loss parameters within the target loss parameter range can be found from the p data quality loss parameters to obtain q data quality loss parameters. Then, the q compression methods corresponding to these q data quality loss parameters among the p compression methods can be determined. The compression ratio of each of these q compression methods can be obtained to get q compression ratios. Specifically, the mapping relationship between the preset compression method and the compression ratio can be stored in advance, and the q compression ratios corresponding to the q compression methods can be determined based on this mapping relationship. Then, the maximum compression ratio among the q compression ratios can be found, and the compression method corresponding to this maximum compression ratio is the target compression method. The i first data sets are compressed by the target compression method to obtain i compressed data sets.

[0129] Thus, based on q compression methods that meet the data quality requirements, further determining their compression ratios and selecting the target compression method corresponding to the maximum compression ratio to compress the i first data sets can minimize the storage space occupied by the data. The amount of ultrasonic image data is usually large, and effective compression can significantly save storage resources and reduce storage costs. Especially for management devices that have accumulated a large amount of ultrasonic examination data over a long period, reasonable compression can enable them to store more data within a limited storage capacity, extend the usage cycle of the storage device, and at the same time facilitate operations such as data backup and migration, improving the utilization efficiency of the storage space.

[0130] In summary, implementing this application has the following beneficial effects:

[0131] It can be seen that the artificial intelligence-based ultrasonic image data management method described in this application includes: collecting ultrasonic data of a target object through an ultrasonic detection device to obtain n ultrasonic data; determining the corresponding ultrasonic images in each of the n ultrasonic data to obtain n ultrasonic images; segmenting each of the n ultrasonic images to obtain m segmented ultrasonic images; labeling the m segmented ultrasonic images through a preset image annotation model to obtain m labeled data; transmitting the m labeled data and the m segmented ultrasonic images to a storage module; determining the ultrasonic detection parts corresponding to the m segmented ultrasonic images to obtain i ultrasonic detection parts; storing the m labeled data and the m segmented ultrasonic images by the storage module according to the i ultrasonic detection parts, and classifying and indexing the stored data. By segmenting each ultrasonic image, different tissue structures or regions of interest in the ultrasonic image are separately extracted to obtain multiple segmented ultrasonic images. In this way, an ultrasonic image containing multiple complex structures is refined into multiple segmented ultrasonic images with specific targets, making the subsequent management and analysis of each specific structure more targeted, avoiding information confusion and omission that may occur during overall processing, and thus improving the refinement degree of ultrasonic image data management.

[0132] Please refer to Figure 3 , Figure 3 FIG. is a functional unit composition block diagram of an artificial intelligence-based ultrasonic image data management system 300 provided by an embodiment of this application, which is applied to a control module of an ultrasonic image data management device. The ultrasonic image data management device further includes: an ultrasonic detection device, a storage module. The artificial intelligence-based ultrasonic image data management system 300 includes: an acquisition module 301, a determination module 302, a management module 303, where:

[0133] The acquisition module 301 is configured to collect ultrasonic data of a target object through the ultrasonic detection device to obtain n ultrasonic data; n is a positive integer;

[0134] The determining module 302 is configured to determine the corresponding ultrasound image in each of the n ultrasound data, obtaining n ultrasound images;

[0135] The management module 303 is configured to segment each of the n ultrasound images, obtaining m segmented ultrasound images; m is an integer greater than or equal to n; label the m segmented ultrasound images through a preset image annotation model, obtaining m labeled data; and transmit the m labeled data and the m segmented ultrasound images to the storage module;

[0136] The determining module 302 is further configured to determine the corresponding ultrasound detection parts of the m segmented ultrasound images, obtaining i ultrasound detection parts; i is a positive integer less than or equal to m;

[0137] The management module 303 is further configured to store the m labeled data and the m segmented ultrasound images according to the i ultrasound detection parts through the storage module, and classify and index the stored data.

[0138] Optionally, in terms of determining the corresponding ultrasound image in each of the n ultrasound data to obtain n ultrasound images, the determining module 302 is specifically configured to:

[0139] Determine the corresponding first ultrasound image in each of the n ultrasound data, obtaining n first ultrasound images;

[0140] Detect whether each of the n first ultrasound images meets a preset qualified condition;

[0141] If it meets the preset qualified condition, determine the n ultrasound images according to the n first ultrasound images;

[0142] If it does not meet the preset qualified condition, determine the first ultrasound images that do not meet the preset qualified condition among the n first ultrasound images, obtaining a first ultrasound images, and the n - a first ultrasound images among the n first ultrasound images except the a first ultrasound images; a is a natural number less than or equal to n;

[0143] Determine the unqualified reasons why each of the a first ultrasound images does not meet the preset qualified condition, obtaining a unqualified reasons;

[0144] Perform repair processing on the a first ultrasound images according to the a unqualified reasons, obtaining a second ultrasound images;

[0145] Determine the n ultrasound images according to the n - a first ultrasound images and the a second ultrasound images.

[0146] Optionally, in the aspect of repairing the a first ultrasonic images according to the a unqualified reasons to obtain a second ultrasonic images, the determining module 302 is specifically configured to:

[0147] Determine the repair operations corresponding to the a unqualified reasons to obtain a repair operations;

[0148] Repair the corresponding first ultrasonic images in the a first ultrasonic images according to the a repair operations to obtain a third ultrasonic images;

[0149] Detect whether each of the a third ultrasonic images meets the preset qualified conditions;

[0150] If it meets the preset qualified conditions, determine the a second ultrasonic images according to the a third ultrasonic images;

[0151] If it does not meet the preset qualified conditions, determine the third ultrasonic images that do not meet the preset qualified conditions in the a third ultrasonic images to obtain b third ultrasonic images, and a - b third ultrasonic images among the a third ultrasonic images except the b third ultrasonic images; b is a natural number less than or equal to a;

[0152] Determine the acquisition parameters corresponding to each of the b third ultrasonic images to obtain b acquisition parameters;

[0153] Determine the unqualified reasons corresponding to the b third ultrasonic images among the a unqualified reasons to obtain b unqualified reasons;

[0154] Determine the fine-tuning factors corresponding to the b unqualified reasons to obtain b fine-tuning factors;

[0155] Adjust the corresponding acquisition parameters in the b acquisition parameters according to the b fine-tuning factors to obtain b target acquisition parameters;

[0156] Collect ultrasonic images of the target object by the ultrasonic detection device respectively with the b target acquisition parameters to obtain b fourth ultrasonic images;

[0157] Determine the a second ultrasonic images according to the a - b third ultrasonic images and the b fourth ultrasonic images.

[0158] Optionally, in the aspect of segmenting each of the n ultrasonic images to obtain m segmented ultrasonic images, the management module 303 is specifically configured to:

[0159] Segment the target ultrasound image based on a preset image segmentation method to obtain c segmented ultrasound images; the target ultrasound image is any one of the n ultrasound images; c is a positive integer less than or equal to m;

[0160] Determine the clarity corresponding to each of the c segmented ultrasound images to obtain c clarities;

[0161] Obtain the device clarity corresponding to the ultrasound detection device;

[0162] Determine the clarities among the c clarities that are less than the device clarity to obtain d clarities; d is a natural number less than or equal to c;

[0163] Determine the segmented ultrasound images corresponding to the d clarities among the c segmented ultrasound images to obtain d segmented ultrasound images;

[0164] Determine the difference between each of the d clarities and the device clarity to obtain d clarity differences;

[0165] Process the d segmented ultrasound images based on the d clarity differences to obtain d target segmented ultrasound images;

[0166] Replace the d segmented ultrasound images among the c segmented ultrasound images with the d target segmented ultrasound images to obtain the segmented ultrasound images corresponding to the target ultrasound image.

[0167] Optionally, in the aspect of processing the d segmented ultrasound images based on the d clarity differences to obtain d target segmented ultrasound images, the management module 303 is specifically configured to:

[0168] Determine a target clarity enhancement method corresponding to the target clarity difference; the target clarity enhancement method includes one of the following: image filtering processing, gray-scale transformation enhancement, histogram equalization; the target clarity difference is any one of the d clarity differences; the target clarity difference corresponds to a first segmented ultrasound image among the d segmented ultrasound images;

[0169] Process the first segmented ultrasound image based on the target clarity enhancement method to obtain a reference segmented ultrasound image;

[0170] Determine the target segmented ultrasound image corresponding to the first segmented ultrasound image according to the reference segmented ultrasound image.

[0171] Optionally, in the aspect of storing the m labeled data and the m segmented ultrasound images according to the i ultrasound detection parts, and classifying and indexing the stored data, the management module 303 is specifically configured to:

[0172] Classify the m labeled data and the m segmented ultrasound images according to the i ultrasonic detection parts to obtain i data sets; each data set corresponds to one ultrasonic detection part; each data set includes at least one labeled data and at least one segmented ultrasound image;

[0173] Compress the i data sets to obtain i compressed data sets;

[0174] Store the i compressed data sets in the storage module, and generate a corresponding index for each compressed data set in the i compressed data sets to obtain i indexes.

[0175] Optionally, in terms of compressing the i data sets to obtain i compressed data sets, the management module 303 is specifically configured to:

[0176] Determine the first data type corresponding to the i data sets;

[0177] Determine the second data type corresponding to the storage module;

[0178] Convert the i data sets from the first data type to the second data type to obtain i first data sets;

[0179] Obtain the target device performance parameters of the ultrasonic image data management device;

[0180] Select a compression method corresponding to the target device performance parameters from a preset compression method library to obtain p compression methods; p is a positive integer;

[0181] Determine the data quality loss parameters of each compression method in the p compression methods to obtain p data quality loss parameters;

[0182] Obtain the target data importance level corresponding to the i first data sets;

[0183] Determine the target loss parameter range corresponding to the target data importance level;

[0184] Determine the data quality loss parameters among the p data quality loss parameters that are within the target loss parameter range to obtain q data quality loss parameters; q is a positive integer less than or equal to p;

[0185] Determine the q compression methods corresponding to the q data quality loss parameters;

[0186] Determine the compression ratios of each compression method in the q compression methods to obtain q compression ratios;

[0187] Determine the maximum compression ratio among the q compression ratios, and determine the compression method corresponding to the maximum compression ratio as the target compression method;

[0188] Compress the i first data sets based on the target compression method to obtain the i compressed data sets.

[0189] In specific implementation, the artificial intelligence-based ultrasonic image data management system 300 described in the embodiments of the present invention can also execute other implementation manners described in the artificial intelligence-based ultrasonic image data management method provided in the embodiments of the present invention, which will not be elaborated here.

[0190] Please refer to Figure 4 , Figure 4 which is a schematic structural diagram of an electronic device provided in an embodiment of the present application. The electronic device includes a processor, a memory, a communication interface, and one or more programs. The processor, the memory, and the communication interface are interconnected through a bus. The above one or more programs are stored in the above memory and are configured to be executed by the above processor. The above programs include parts or all of the steps for executing the artificial intelligence-based ultrasonic image data management method in the embodiments of the present application.

[0191] The embodiments of the present application also provide a computer storage medium. The computer storage medium stores a computer program for electronic data exchange, and the computer program enables a computer to execute part or all of the steps of any method described in the method embodiments above. The above computer includes an electronic device.

[0192] The embodiments of the present application also provide a computer program product. The above computer program product includes a non-transitory computer-readable storage medium storing a computer program. The above computer program is operable to enable a computer to execute part or all of the steps of any method described in the method embodiments above. The above computer program product can be a software installation package, and the above computer includes an electronic device.

[0193] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present application is not limited by the described action sequence, because according to the present application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the present application.

[0194] In the above embodiments, the descriptions of the respective embodiments have their own emphases. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0195] In several embodiments provided by this application, it should be understood that the disclosed device can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the above division of units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling, direct coupling, or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of devices or units can be in electrical or other forms.

[0196] The units described above as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or they can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0197] In addition, each functional unit in various embodiments of this application can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0198] If the above integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on such an understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the above methods in various embodiments of this application. And the aforementioned memory includes: USB flash drives, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), mobile hard disks, magnetic disks, or optical disks and other various media that can store program codes.

[0199] The above has introduced the embodiments of this application in detail. Specific examples are used in this article to elaborate on the principle and implementation manner of this application. The description of the above embodiments is only used to help understand the method and its core idea of this application; at the same time, for those of ordinary skill in the art, according to the idea of this application, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to this application.

Claims

1. An ultrasonic image data management method based on artificial intelligence, characterized in that: A control module applied to an ultrasonic image data management device, wherein the ultrasonic image data management device further comprises: an ultrasonic detection device and a storage module, and the method comprises: The ultrasonic detection device collects ultrasonic data of the target object to obtain n ultrasonic data; n is a positive integer; Determine an ultrasonic image corresponding to each ultrasonic data in the n ultrasonic data to obtain n ultrasonic images; Segmenting each of the n ultrasonic images to obtain m segmented ultrasonic images, where m is an integer greater than or equal to n; Annotating the m segmented ultrasound images by using a preset image annotation model to obtain m annotation data; Transmitting the m labeled data and the m segmented ultrasound images to the storage module; Determine the ultrasonic detection parts corresponding to the m segmented ultrasonic images to obtain i ultrasonic detection parts, where i is a positive integer less than or equal to m; The storage module stores the m labeled data and the m segmented ultrasound images according to the i ultrasound detection locations, and classifies and indexes the stored data.

2. The method according to claim 1, characterized in that The step of determining the ultrasonic image corresponding to each ultrasonic data in the n ultrasonic data to obtain n ultrasonic images includes: Determine a first ultrasonic image corresponding to each ultrasonic data in the n ultrasonic data to obtain n first ultrasonic images; Detecting whether each of the n first ultrasonic images meets a preset qualification condition; If the preset qualified condition is met, determining the n ultrasonic images according to the n first ultrasonic images; If the preset qualified condition is not met, determining the first ultrasonic images that do not meet the preset qualified condition among the n first ultrasonic images, obtaining a first ultrasonic images, and na first ultrasonic images other than the a first ultrasonic images among the n first ultrasonic images; a is a natural number less than or equal to n; Determine a failure reason for each of the a first ultrasonic images not meeting the preset qualification condition, and obtain a failure reasons; Performing repair processing on the a first ultrasonic images according to the a reasons for failure to meet the requirements, to obtain a second ultrasonic images; The n ultrasonic images are determined according to the na first ultrasonic images and the a second ultrasonic images.

3. The method according to claim 2, characterized in that The repairing process of the a first ultrasonic images according to the a reasons for failure to meet the requirements to obtain a second ultrasonic images includes: Determine the repair operations corresponding to the a reasons for non-conformity, and obtain a repair operations; Performing repair processing on corresponding first ultrasonic images among the a first ultrasonic images according to the a repair operations to obtain a third ultrasonic images; Detecting whether each of the a third ultrasonic images meets the preset qualified condition; If the preset qualified condition is met, determining the a second ultrasonic images according to the a third ultrasonic images; If the preset qualified condition is not met, the third ultrasonic images that do not meet the preset qualified condition are determined among the a third ultrasonic images, b third ultrasonic images are obtained, and ab third ultrasonic images among the a third ultrasonic images except the b third ultrasonic images; b is a natural number less than or equal to a; Determine an acquisition parameter corresponding to each of the b third ultrasonic images to obtain b acquisition parameters; Determine the unqualified reasons corresponding to the b third ultrasound images among the a unqualified reasons, and obtain b unqualified reasons; Determine the fine-tuning factors corresponding to the b reasons for non-conformity to obtain b fine-tuning factors; Adjust corresponding acquisition parameters among the b acquisition parameters according to the b fine-tuning factors to obtain b target acquisition parameters; Acquire ultrasonic images of the target object using the ultrasonic detection equipment with the b target acquisition parameters to obtain b fourth ultrasonic images; The a second ultrasonic images are determined according to the ab third ultrasonic images and the b fourth ultrasonic images.

4. The method according to any one of claims 1 to 3, characterized in that: The step of segmenting each of the n ultrasonic images to obtain m segmented ultrasonic images includes: Segmenting the target ultrasonic image based on a preset image segmentation method to obtain c segmented ultrasonic images; the target ultrasonic image is any one of the n ultrasonic images; c is a positive integer less than or equal to m; Determine the definition corresponding to each of the c segmented ultrasonic images to obtain c definitions; Obtaining the device definition corresponding to the ultrasonic detection device; Determine the definition that is less than the definition of the device among the c definitions, and obtain d definitions; d is a natural number less than or equal to c; Determine the segmented ultrasonic images corresponding to the d sharpnesses in the c segmented ultrasonic images, and obtain d segmented ultrasonic images; Determine the difference between each of the d definition and the device definition to obtain d definition difference values; Processing the d segmented ultrasonic images based on the d clarity differences to obtain d target segmented ultrasonic images; The d segmented ultrasonic images among the c segmented ultrasonic images are replaced by the d target segmented ultrasonic images to obtain a segmented ultrasonic image corresponding to the target ultrasonic image.

5. The method according to claim 4, characterized in that The step of processing the d segmented ultrasound images based on the d definition differences to obtain d target segmented ultrasound images includes: Determine a target definition enhancement method corresponding to the target definition difference; the target definition enhancement method includes one of the following: image filtering processing, grayscale transformation enhancement, and histogram equalization; the target definition difference is any one of the d definition differences; the target definition difference corresponds to a first segmented ultrasonic image in the d segmented ultrasonic images; Processing the first segmented ultrasonic image based on the target definition enhancement method to obtain a reference segmented ultrasonic image; A target segmented ultrasonic image corresponding to the first segmented ultrasonic image is determined according to the reference segmented ultrasonic image.

6. The method according to any one of claims 1 to 3, characterized in that: The storing the m labeled data and the m segmented ultrasound images according to the i ultrasound detection parts, and classifying and indexing the stored data, includes: Classifying the m annotated data and the m segmented ultrasound images according to the i ultrasound detection parts to obtain i data sets; each data set corresponds to an ultrasound detection part; and each data set includes at least one annotated data and at least one segmented ultrasound image; Compressing the i data sets to obtain i compressed data sets; The i compressed data sets are stored in the storage module, and a corresponding index is generated for each of the i compressed data sets to obtain i indexes.

7. The method according to claim 6, characterized in that The compressing the i data sets to obtain i compressed data sets includes: Determine a first data type corresponding to the i data sets; Determine a second data type corresponding to the storage module; Convert the i data sets from the first data type to the second data type to obtain i first data sets; Acquiring target device performance parameters of the ultrasonic image data management device; Selecting a compression method corresponding to the performance parameter of the target device from a preset compression method library to obtain p compression methods; p is a positive integer; Determine a data quality loss parameter of each compression method in the p compression methods to obtain p data quality loss parameters; Obtaining target data importance levels corresponding to the i first data sets; Determine a target loss parameter range corresponding to the target data importance level; Determine the data quality loss parameters in the target loss parameter range among the p data quality loss parameters to obtain q data quality loss parameters; q is a positive integer less than or equal to p; Determine q compression methods corresponding to the q data quality loss parameters; Determine the compression ratio of each of the q compression methods to obtain q compression ratios; Determine a maximum compression ratio among the q compression ratios, and determine a compression method corresponding to the maximum compression ratio as a target compression method; The i first data sets are compressed based on the target compression method to obtain the i compressed data sets.

8. An ultrasonic image data management system based on artificial intelligence, characterized in that: A control module applied to an ultrasonic image data management device, wherein the ultrasonic image data management device further comprises: an ultrasonic detection device, a storage module, and the system comprises: an acquisition module, a determination module, and a management module, wherein: The acquisition module is used to acquire ultrasonic data of the target object through the ultrasonic detection device to obtain n ultrasonic data; n is a positive integer; The determination module is used to determine the ultrasonic image corresponding to each ultrasonic data in the n ultrasonic data to obtain n ultrasonic images; The management module is used to segment each of the n ultrasonic images to obtain m segmented ultrasonic images, where m is an integer greater than or equal to n; annotate the m segmented ultrasonic images using a preset image annotation model to obtain m annotation data; and transmit the m annotation data and the m segmented ultrasonic images to the storage module; The determination module is further used to determine the ultrasonic detection parts corresponding to the m segmented ultrasonic images to obtain i ultrasonic detection parts; i is a positive integer less than or equal to m; The management module is also used to store the m labeled data and the m segmented ultrasound images according to the i ultrasound detection parts through the storage module, and classify and index the stored data.

9. An electronic device, characterized in that: include: A processor and a memory, wherein the memory is used to store one or more programs and is configured to be executed by the processor, wherein the program includes instructions for executing the steps in the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: A computer program for electronic data exchange is stored, wherein the computer program enables a computer to execute the method according to any one of claims 1 to 7.

Images