Wearable ultrasound-based long-period ultrasonic image monitoring method and system

Abstract

The invention belongs to the technical field of image processing. According to the long-period ultrasonic image monitoring method and system based on wearable ultrasound, feature extraction is carried out on an obtained ultrasonic image sequence at each sampling moment, and a preliminary feature extraction result is obtained; performing linear projection processing on the preliminary feature extraction result to obtain an image feature sequence at each sampling moment; carrying out self-attention calculation on the image feature sequence of each sampling moment and the image feature sequence of the adjacent sampling moment to obtain self-adaptive features of each sampling moment, carrying out self-adaptive gating calculation on the self-adaptive features of each adjacent sampling moment, and obtaining an attention feature sequence after passing through a feedforward neural network; and obtaining an ultrasonic image monitoring result according to the attention feature sequence. According to the method, the dynamic change of the ultrasonic sequence in the long-period image sequence can be adaptively concerned, and the long-period ultrasonic image monitoring precision of the wearing individual is improved.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a long-period ultrasound imaging monitoring method and system based on wearable ultrasound. Background Art

[0002] The statements in this section merely provide background art related to the present invention and do not necessarily constitute prior art.

[0003] Wearable devices, capable of continuously and noninvasively monitoring the physiological state of internal human tissues, represent a research trend in precision digital medicine. Traditional clinical ultrasound examinations require three key elements: an ultrasound physician, a patient, and ultrasound equipment. This limitation places the patient in a resting position (prone or supine). Wearable ultrasound devices, with their wearable and conformable design, enable real-time, assisted monitoring of human ultrasound images around the clock, over extended periods, without interfering with normal living conditions. Furthermore, wearable ultrasound imaging enables continuous sampling of deep tissues and organs over days to months, helping clinicians monitor health, observe disease progression, and assess disease risks.

[0004] Although wearable ultrasound devices have many advantages, they have the following problems: although existing wearable ultrasound devices can realize long-period ultrasound image acquisition, they cannot perform long-period ultrasound image monitoring, do not focus on the specific physiological structures in the ultrasound image sequence and their dynamic pattern changes, and cannot effectively process long-period image sequences in real time. Summary of the Invention

[0005] In order to address the shortcomings of the existing technology, the present invention provides a long-period ultrasound imaging monitoring method and system based on wearable ultrasound, which can adaptively focus on the dynamic changes of the ultrasound sequence in the long-period image sequence, thereby improving the accuracy of long-period ultrasound imaging monitoring of the wearer.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a long-period ultrasound imaging monitoring method based on wearable ultrasound.

[0007] A long-period ultrasound imaging monitoring method based on wearable ultrasound includes the following processes: Perform feature extraction on the ultrasonic image sequence acquired at each sampling moment to obtain preliminary feature extraction results; Performing linear projection processing on the preliminary feature extraction results to obtain an image feature sequence at each sampling moment; Perform self-attention calculation on the image feature sequence at the current sampling moment and the image feature sequence at adjacent sampling moments to obtain the adaptive features at the current sampling moment; Perform self-attention calculation on the image feature sequence at the next sampling moment and the image feature sequence at the adjacent sampling moments to obtain the adaptive features at the next sampling moment; Adaptive gating calculation is performed on the adaptive features of the current sampling moment and the next adjacent sampling moment, and the attention feature sequence is obtained after passing through the feedforward neural network; According to the attention feature sequence, the ultrasonic image monitoring result at the current sampling moment is obtained.

[0008] As a further limitation of the first aspect of the present invention, the adaptive feature of the current sampling moment And the adaptive features of the next sampling moment ,include: ; ; in, For the current moment The weight coefficient of For the next moment The weight coefficient of For the current moment The output of the gating mechanism, For the next moment The output of the gating mechanism, is the weight coefficient, and For the current moment The input of the gating mechanism, and For the next moment The input of the gating mechanism.

[0009] As a further limitation of the first aspect of the present invention, the current sampling time and the next adjacent sampling time Adaptive gating calculation is performed based on the adaptive features of , including: ,in, , For and The output after fusion using the gating mechanism, is the training weight matrix in the adaptive gating mechanism, is the activation function.

[0010] As a further limitation of the first aspect of the present invention, the monitoring model is fine-tuned by continuously collecting ultrasound data of the wearing individual, and the weight parameters of the monitoring model are updated online.

[0011] As a further limitation of the first aspect of the present invention, the attention feature sequence is subjected to Transformer calculation, and the result of the Transformer calculation is subjected to multiple rounds of self-attention calculation, adaptive gating calculation and Transformer calculation to obtain an ultrasound imaging monitoring result; Based on the ultrasound imaging monitoring results, abnormal changes in the wearer are monitored, and when the change is greater than the set threshold, an alarm message is generated.

[0012] As a further limitation of the first aspect of the present invention, based on the collected ultrasound data of the wearing individual, set thresholds under exercise conditions and normal life conditions are established to automatically adapt to the volatility of the wearing individual data.

[0013] In a second aspect, the present invention provides a long-period ultrasound imaging monitoring system based on wearable ultrasound.

[0014] A long-period ultrasound imaging monitoring system based on wearable ultrasound, comprising: The preliminary feature extraction unit is configured to: perform feature extraction on the acquired ultrasound image sequence at each sampling moment to obtain a preliminary feature extraction result; A linear projection processing unit is configured to: perform linear projection processing on the preliminary feature extraction result to obtain an image feature sequence at each sampling moment; The first feature extraction unit is configured to: perform self-attention calculation on the image feature sequence at the current sampling moment and the image feature sequence at the adjacent sampling moments to obtain the adaptive feature at the current sampling moment; The second feature extraction unit is configured to: perform self-attention calculation on the image feature sequence at the next sampling moment and the image feature sequence at the adjacent sampling moments to obtain the adaptive feature at the next sampling moment; An attention calculation unit is configured to perform adaptive gating calculation on the adaptive features of the current sampling moment and the next adjacent sampling moment, and obtain an attention feature sequence after passing through a feedforward neural network; The monitoring result generating unit is configured to obtain the ultrasound image monitoring result at the current sampling moment according to the attention feature sequence.

[0015] In a third aspect, the present invention provides a wearable ultrasound device comprising: a processor and a computer-readable storage medium; a processor adapted to execute a computer program; A computer-readable storage medium having a computer program stored therein, wherein when the computer program is executed by the processor, the long-period ultrasound imaging monitoring method based on wearable ultrasound is implemented as described in the first aspect of the present invention.

[0016] In a fourth aspect, the present invention provides a computer-readable storage medium storing a computer program, wherein the computer program is suitable for being loaded by a processor and executing the long-period ultrasound imaging monitoring method based on wearable ultrasound as described in the first aspect of the present invention.

[0017] In a fifth aspect, the present invention provides a computer program product, comprising a computer program. When the computer program is executed by a processor, the computer program implements the long-period ultrasound imaging monitoring method based on wearable ultrasound as described in the first aspect of the present invention.

[0018] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention innovatively proposes a long-period ultrasound imaging monitoring method based on wearable ultrasound. It is specifically used to perform long-period modeling of the characteristics of long-period wearable ultrasound image sequences. It focuses more on specific physiological structures in ultrasound image sequences and their dynamic pattern changes, thereby improving the accuracy of long-period ultrasound imaging monitoring of the wearer.

[0019] 2. The present invention calculates the attention of adjacent key image sequences, takes the ultrasound image features at time t to calculate the Q vector (i.e., query vector), and simultaneously takes the ultrasound image features of its adjacent image sequences to calculate the K vector (i.e., key vector) and V vector (i.e., value vector), and calculates the attention between ultrasound sequences. In this way, the calculation of the sequence attention mechanism of the entire ultrasound image sequence is completed in the form of a sliding window. This attention mechanism helps to improve the model's understanding of long-period ultrasound image sequence data. The model can pay attention to the relationship between a certain time step and the previous and next time steps, thereby capturing the pattern of dynamic changes in ultrasound images.

[0020] 3. This invention introduces an adaptive gating mechanism to adjust the sequence attention The degree of transmission of moment series information to the next layer. In ultrasound image sequence processing, the importance of features at different time steps in health monitoring is inconsistent. The adaptive gating mechanism can help the model selectively focus on the features of some specific time steps while suppressing unimportant information. Through a dynamic weight mechanism, the feature information of the ultrasound image sequence is screened and weighted. This adaptive gating mechanism is particularly important because it can ignore unimportant details and amplify potential anomalies or changing trends, thereby improving the accuracy of health risk prediction.

[0021] Advantages of additional aspects of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0023] Figure 1 A schematic diagram of the process of the long-period ultrasound imaging monitoring method based on wearable ultrasound provided in Example 1 of the present invention; Figure 2 Schematic diagram of a long-period image sequence self-attention module provided in Example 1 of the present invention; Figure 3 A schematic diagram of a long-period ultrasound imaging monitoring system based on wearable ultrasound provided in Example 2 of the present invention; Figure 4 A schematic diagram of a computer device provided in Example 3 of the present invention. DETAILED DESCRIPTION

[0024] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0025] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0026] In the absence of conflict, the embodiments of the present invention and the features thereof may be combined with each other.

[0027] Example 1: This implementation proposes a long-term ultrasound imaging monitoring method based on wearable ultrasound, which can be widely used in scenarios such as elderly health monitoring, chronic disease management, and postoperative recovery monitoring. Through continuous analysis of ultrasound images, it helps medical institutions better understand the health status of patients. Specifically, it includes the following processes: S1: Data queue construction and preprocessing.

[0028] Wearable ultrasound devices can continuously collect large amounts of real-time data. During data queue construction, ultrasound images of deep tissues and organs are captured every 30 minutes (other intervals are possible, such as 1 hour or 20 minutes, but are not specifically limited here). These images are not limited to the heart, blood vessels, muscles, and abdomen. Due to hardware limitations, wearable ultrasound imaging devices often have poor image quality, requiring image enhancement such as denoising to improve data quality and minimize the impact of sensor noise or motion artifacts.

[0029] S2: Implementation of long-period image sequence ultrasound imaging monitoring algorithm based on wearable ultrasound, such as Figure 1 Specifically, the process includes the following: S2.1: The original long-period ultrasound image sequence is processed by the Visual Encoder (the visual encoder is responsible for extracting preliminary features of the image and converting the two-dimensional image data into feature representation for subsequent module processing; S2.2: The features output by the visual encoder are linearly projected to further reduce the dimension and generate a compact feature vector; S2.3: WUS Sequence Encoder (Wearable Ultrasound Long-Period Ultrasound Image Sequence Encoder) The sequence encoder is responsible for extracting deeper temporal and spatial features from the image sequence. The WUS Sequence Encoder contains multiple WUS Attention Transformer Blocks (Long-Period Image Sequence Self-Attention Modules) to adaptively focus on the dynamic changes of the ultrasound sequence in the long-period image sequence.

[0030] In this implementation, a wearable ultrasound long-period image sequence self-attention mechanism (WUS Attention) is proposed in the WUS Attention Transformer Block (Long-period Image Sequence Self-Attention Module). This mechanism is specifically used to perform long-period modeling of the features of long-period wearable ultrasound image sequences, focusing more on specific physiological structures in ultrasound image sequences and their dynamic pattern changes.

[0031] The core of WUS Attention Transformer Block is to calculate the attention of adjacent key image sequences, for example, The Query vector is calculated based on the ultrasound image features of the time, and the Key vector and Value vector are calculated based on the ultrasound image features of the adjacent image sequences. The attention between the ultrasound sequences is calculated, and the sequence attention mechanism of the entire ultrasound image sequence is calculated in the form of a sliding window. This attention mechanism helps to improve the model's understanding of long-period ultrasound image sequence data. The model can pay attention to the relationship between a certain time step and the previous and next time steps, thereby capturing the dynamic change pattern of the ultrasound image.

[0032] The present invention also introduces an adaptive gating mechanism for sequence attention to adjust the degree of transmission of sequence information at time t to the next layer. In ultrasound image sequence processing, the importance of features at different time steps in health monitoring is inconsistent. The adaptive gating mechanism can help the model selectively focus on the features of some specific time steps while suppressing unimportant information. Through a dynamic weight mechanism, the feature information of the ultrasound image sequence is screened and weighted. This adaptive gating mechanism is particularly important because it can ignore unimportant details and amplify potential abnormalities or changing trends, thereby improving the accuracy of health risk prediction. Figure 2 As shown in Figure 2, the specific formula of WUS Attention Transformer Block (long-period image sequence self-attention module) is as follows: Step 1: Perform layer standardization, specifically including: (1); (2); (3); (4); in, for Ultrasound image features at the moment, for Ultrasound image features at the moment, for Ultrasound image features at the moment, for Ultrasound image features at the moment, for Normalized ultrasound image features at time, for Normalized ultrasound image features at time, for Normalized ultrasound image features at the moment, for Normalized ultrasound image features at the time.

[0033] Step 2: For long-period image sequences, use the self-attention operator to perform self-attention calculation, specifically including: (5); (6); (7); (8); (9); (10); (11); (12); (13); (14); (15); (16); (17); (18); in, for The key weight matrix at time t, for The value weight matrix at time , for The query weight matrix at time t, for The key weight matrix at time t, for The value weight matrix at time , and are all the results of self-attention calculation. is the activation function, represents the dimension of the bond matrix, for The key vector at time, for The value vector at time instant, for The key vector at time, for The value vector at time instant, for The value vector at time instant, for The key vector at time, and For the current moment The input of the gating mechanism, and For the next moment The input of the gating mechanism.

[0034] Step 3: Adopt an adaptive gating mechanism to calculate the self-attention. Specifically, it includes: (19); (20); (twenty one); (twenty two); (twenty three); (twenty four); (25).

[0035] in, represents a feedforward neural network, , , is the training weight matrix in the adaptive gating mechanism, For the current moment The weight coefficient of For the next moment The weight coefficient of For the current moment The output of the gating mechanism, For the next moment The output of the gating mechanism, For and The output after fusion using the gating mechanism, is the final output after being processed by the feedforward neural network, is the activation function, is the weight coefficient.

[0036] This implementation method includes multiple WUS Attention Transformer Blocks and Transformer modules connected in sequence, and obtains the final monitoring result output through multiple iterative calculations.

[0037] In this implementation, in long-period ultrasound imaging monitoring, the features at different moments in the sequence contain ultrasound image feature information that changes continuously over time. Through the WUS Attention Transformer Block attention mechanism, the model can focus on the relationship between a certain time step and the previous and next time steps, while adaptively enhancing or suppressing the features of different time steps. This enables the model to retain key features when processing long time series, thereby capturing the dynamic change pattern of the ultrasound structure.

[0038] Existing wearable ultrasound devices are unable to adaptively modify monitoring models for specific individuals and are unable to adapt to the physical conditions of different wearers. In view of this, this implementation uses personalized model training for the wearable ultrasound's long-period ultrasound imaging monitoring algorithm. Each person's health status and physical characteristics vary, and personalized model training is performed by continuously collecting individual ultrasound data. The basic model is fine-tuned using individual health data (such as ultrasound data from recent weeks or months), and the model's weight parameters are continuously updated online to better adapt the model to the individual's specific health trends and characteristics.

[0039] In this implementation, anomaly detection can be performed based on the long-term monitoring algorithm model. Once certain abnormal changes are detected, such as abnormal enlargement of organs or irregular morphological changes, abnormal warnings can be issued in a timely manner. At the same time, based on the learning of historical ultrasound imaging data by the long-term monitoring algorithm model, a dynamically changing threshold range (under exercise and normal living conditions) is established to automatically adapt to the volatility of individual health data.

[0040] Example 2: like Figure 3 As shown, this implementation provides a long-period ultrasound imaging monitoring system based on wearable ultrasound, including: The preliminary feature extraction unit is configured to: perform feature extraction on the acquired ultrasound image sequence at each sampling moment to obtain a preliminary feature extraction result; A linear projection processing unit is configured to: perform linear projection processing on the preliminary feature extraction result to obtain an image feature sequence at each sampling moment; The first feature extraction unit is configured to: perform self-attention calculation on the image feature sequence at the current sampling moment and the image feature sequence at the adjacent sampling moments to obtain the adaptive feature at the current sampling moment; The second feature extraction unit is configured to: perform self-attention calculation on the image feature sequence at the next sampling moment and the image feature sequence at the adjacent sampling moments to obtain the adaptive feature at the next sampling moment; An attention calculation unit is configured to perform adaptive gating calculation on the adaptive features of the current sampling moment and the next adjacent sampling moment, and obtain an attention feature sequence after passing through a feedforward neural network; The monitoring result generating unit is configured to obtain the ultrasound image monitoring result at the current sampling moment according to the attention feature sequence.

[0041] The specific working process of each of the above units is described in Example 1 and will not be repeated here.

[0042] It is understandable that each of the above-mentioned units can be separately or completely combined into one or several other units to form a unit, or one (or some) of the units can be further divided into multiple functionally smaller units to form a unit, which can achieve the same operation without affecting the realization of the technical effects of the embodiments of the present application. The above-mentioned units are divided based on logical functions. In actual applications, the functions of one unit can also be implemented by multiple units, or the functions of multiple units can be implemented by one unit. In other embodiments of the present application, the system may also include other units. In actual applications, these functions can also be implemented with the assistance of other units and can be implemented by the collaboration of multiple units.

[0043] According to another embodiment of the present application, the system described in this embodiment can be constructed and the method of Example 1 of the present application can be implemented by running a computer program (including program code) capable of executing the steps involved in the corresponding method described in Example 1 on a general-purpose computing device such as a computer, which includes processing elements and storage elements such as a central processing unit (CPU), random access memory (RAM), and read-only memory (ROM). The computer program can be recorded on, for example, a computer-readable recording medium, and loaded into the above-mentioned computing device through the computer-readable recording medium and run therein.

[0044] Example 3: like Figure 4 As shown, this implementation provides an electronic device, which includes a processor 1001, a communication interface 1002, and a computer-readable storage medium 1003. The processor 1001, the communication interface 1002, and the computer-readable storage medium 1003 may be connected via a bus or other means.

[0045] Among them, the communication interface 1002 is used to receive and send data, the computer-readable storage medium 1003 can be stored in the memory of the electronic device, the computer-readable storage medium 1003 is used to store a computer program, the computer program includes program instructions, and the processor 1001 is used to execute the program instructions stored in the computer-readable storage medium 1003.

[0046] The processor 1001 (also called CPU (Central Processing Unit)) is the computing core and control core of the electronic device, which is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to implement corresponding method processes or corresponding functions.

[0047] The processor 1001 is configured to execute the following process: Perform feature extraction on the ultrasonic image sequence acquired at each sampling moment to obtain preliminary feature extraction results; Performing linear projection processing on the preliminary feature extraction results to obtain an image feature sequence at each sampling moment; Perform self-attention calculation on the image feature sequence at the current sampling moment and the image feature sequence at adjacent sampling moments to obtain the adaptive features at the current sampling moment; Perform self-attention calculation on the image feature sequence at the next sampling moment and the image feature sequence at the adjacent sampling moments to obtain the adaptive features at the next sampling moment; Adaptive gating calculation is performed on the adaptive features of the current sampling moment and the next adjacent sampling moment, and the attention feature sequence is obtained after passing through the feedforward neural network; According to the attention feature sequence, the ultrasonic image monitoring result at the current sampling moment is obtained.

[0048] The specific working process is described in Example 1 and will not be repeated here.

[0049] Example 4: This implementation provides a computer-readable storage medium (Memory). This computer-readable storage medium is a memory device within an electronic device that stores programs and data. It should be understood that the computer-readable storage medium herein may include both built-in storage media within the electronic device and, of course, extended storage media supported by the electronic device. The computer-readable storage medium provides storage space that stores the processing system of the electronic device.

[0050] Furthermore, the storage space also stores one or more instructions suitable for being loaded and executed by the processor. These instructions may be one or more computer programs (including program code). It should be noted that the computer-readable storage medium herein may be a high-speed RAM memory or a non-volatile memory, such as at least one disk storage device; alternatively, it may be at least one computer-readable storage medium located remotely from the processor.

[0051] In one embodiment, the computer-readable storage medium stores one or more instructions; the processor loads and executes the one or more instructions stored in the computer-readable storage medium to implement the following process: Perform feature extraction on the ultrasonic image sequence acquired at each sampling moment to obtain preliminary feature extraction results; Performing linear projection processing on the preliminary feature extraction results to obtain an image feature sequence at each sampling moment; Perform self-attention calculation on the image feature sequence at the current sampling moment and the image feature sequence at adjacent sampling moments to obtain the adaptive features at the current sampling moment; Perform self-attention calculation on the image feature sequence at the next sampling moment and the image feature sequence at the adjacent sampling moments to obtain the adaptive features at the next sampling moment; Adaptive gating calculation is performed on the adaptive features of the current sampling moment and the next adjacent sampling moment, and the attention feature sequence is obtained after passing through the feedforward neural network; According to the attention feature sequence, the ultrasonic image monitoring result at the current sampling moment is obtained.

[0052] The specific working process is described in Example 1 and will not be repeated here.

[0053] Example 5: This implementation provides a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform the following process: Perform feature extraction on the ultrasonic image sequence acquired at each sampling moment to obtain preliminary feature extraction results; Performing linear projection processing on the preliminary feature extraction results to obtain an image feature sequence at each sampling moment; Perform self-attention calculation on the image feature sequence at the current sampling moment and the image feature sequence at adjacent sampling moments to obtain the adaptive features at the current sampling moment; Perform self-attention calculation on the image feature sequence at the next sampling moment and the image feature sequence at the adjacent sampling moments to obtain the adaptive features at the next sampling moment; Adaptive gating calculation is performed on the adaptive features of the current sampling moment and the next adjacent sampling moment, and the attention feature sequence is obtained after passing through the feedforward neural network; According to the attention feature sequence, the ultrasonic image monitoring result at the current sampling moment is obtained.

[0054] The specific working process is described in Example 1 and will not be repeated here.

[0055] Those skilled in the art can be aware that units and algorithm steps of each example described in combination with the embodiments disclosed in the application can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the application.

[0056] In the above embodiments, all or part of the embodiments can be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions according to the embodiments of the application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. Computer instructions can be stored in or transmitted by a computer-readable storage medium. Computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through wired (for example, coaxial cable, optical fiber, digital line (DSL)) or wireless (for example, infrared, wireless, microwave, etc.) mode. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data processing device such as a server, data center, etc. integrated with one or more available media. The available media can be magnetic media (for example, floppy disk, hard disk, magnetic tape), optical media (for example, DVD), or semiconductor media (for example, solid state disk (SSD)) and the like.

[0057] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A long-period ultrasound imaging monitoring method based on wearable ultrasound, characterized in that: The following processes are included: Perform feature extraction on the ultrasonic image sequence acquired at each sampling moment to obtain preliminary feature extraction results; Performing linear projection processing on the preliminary feature extraction results to obtain an image feature sequence at each sampling moment; Perform self-attention calculation on the image feature sequence at the current sampling moment and the image feature sequence at adjacent sampling moments to obtain the adaptive features at the current sampling moment; Perform self-attention calculation on the image feature sequence at the next sampling moment and the image feature sequence at the adjacent sampling moments to obtain the adaptive features at the next sampling moment; Adaptive gating calculation is performed on the adaptive features of the current sampling moment and the next adjacent sampling moment, and the attention feature sequence is obtained after passing through the feedforward neural network; According to the attention feature sequence, the ultrasonic image monitoring result at the current sampling moment is obtained.

2. The wearable ultrasound-based long-period ultrasound imaging monitoring method according to claim 1, wherein: Adaptive features at the current sampling moment And the adaptive features at the next sampling moment ,include: ； ； in, For the current moment The weight coefficient of For the next moment The weight coefficient of For the current moment The output of the gating mechanism, For the next moment The output of the gating mechanism, is the weight coefficient, and For the current moment The input of the gating mechanism, and For the next moment The input of the gating mechanism.

3. The long-period ultrasound imaging monitoring method based on wearable ultrasound according to claim 2, characterized in that: For the current sampling time and the next adjacent sampling time Adaptive gating calculation is performed based on the adaptive features of , including: ； in, , For and The output after fusion using the gating mechanism, is the training weight matrix in the adaptive gating mechanism, is the activation function.

4. The long-period ultrasound imaging monitoring method based on wearable ultrasound according to any one of claims 1 to 3, characterized in that: By continuously collecting ultrasound data from the wearer, the monitoring model is fine-tuned and the weight parameters of the monitoring model are updated online.

5. The long-period ultrasound imaging monitoring method based on wearable ultrasound according to any one of claims 1 to 3, characterized in that: Performing Transformer calculation on the attention feature sequence, and then performing multiple rounds of self-attention calculation, adaptive gating calculation, and Transformer calculation on the result of the Transformer calculation to obtain an ultrasound imaging monitoring result; Based on the ultrasound imaging monitoring results, abnormal changes in the wearer are monitored, and when the change is greater than the set threshold, an alarm message is generated.

6. The wearable ultrasound-based long-period ultrasound imaging monitoring method according to claim 5, wherein: Based on the collected ultrasonic data of the wearer, threshold values ​​are established under exercise conditions and normal living conditions to automatically adapt to the volatility of the wearer's data.

7. A long-period ultrasound imaging monitoring system based on wearable ultrasound, characterized in that: include: The preliminary feature extraction unit is configured to: perform feature extraction on the acquired ultrasound image sequence at each sampling moment to obtain a preliminary feature extraction result; A linear projection processing unit is configured to: perform linear projection processing on the preliminary feature extraction result to obtain an image feature sequence at each sampling moment; The first feature extraction unit is configured to: perform self-attention calculation on the image feature sequence at the current sampling moment and the image feature sequence at the adjacent sampling moments to obtain the adaptive feature at the current sampling moment; The second feature extraction unit is configured to: perform self-attention calculation on the image feature sequence at the next sampling moment and the image feature sequence at the adjacent sampling moments to obtain the adaptive feature at the next sampling moment; An attention calculation unit is configured to perform adaptive gating calculation on the adaptive features of the current sampling moment and the next adjacent sampling moment, and obtain an attention feature sequence after passing through a feedforward neural network; The monitoring result generating unit is configured to obtain the ultrasound image monitoring result at the current sampling moment according to the attention feature sequence.

8. A wearable ultrasound device, characterized in that: include: a processor and a computer-readable storage medium; a processor adapted to execute a computer program; A computer-readable storage medium having a computer program stored therein, wherein when the computer program is executed by the processor, the long-period ultrasound imaging monitoring method based on wearable ultrasound is implemented according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and the computer program is suitable for being loaded by a processor and executing the long-period ultrasound imaging monitoring method based on wearable ultrasound according to any one of claims 1 to 6.

10. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, it implements the long-period ultrasound imaging monitoring method based on wearable ultrasound according to any one of claims 1 to 6.

Images