Wearable cardiac ultrasound patch, cardiovascular disease identification system, and method thereof

Abstract

The present application provides a wearable heart ultrasound patch, a cardiovascular disease identification system and a method thereof. A silicon photon integrated circuit layer and a micro-electro-mechanical integrated circuit layer are integrated into a fiber type micro sensor, and the micro sensor is packaged with a computing chip to realize the detection and processing of heart ultrasound in a wearable and convenient way without time or geographical restrictions. Each silicon photon component in the silicon photon integrated circuit layer and each piezoelectric component in the micro-electro-mechanical integrated circuit layer are respectively arranged in a two-dimensional array to enable the wearable heart ultrasound patch to scan more angles. Furthermore, the cardiovascular disease identification system can process the heart ultrasound, electrocardio signals and heart sound signals of the subject, and the data of the three signals are verified with each other to improve the detection rate of comprehensive cardiovascular diseases and reduce the time and misjudgment of doctors in diagnosing cardiovascular diseases.

Description

Technical Field

[0001] This invention relates to a wearable cardiac ultrasound patch, a cardiovascular disease identification system and method, and more specifically, to a wearable cardiac ultrasound patch, a cardiovascular disease identification system and method that integrates a silicon photonic integrated circuit layer and a microelectromechanical integrated circuit layer into a fiber optic microsensor and packages it with a computing chip to realize cardiac ultrasound detection and processing. Background Technology

[0002] With advancements in medicine and increased life expectancy, population aging has become an inevitable trend. As people age, physiological changes occur in the cardiovascular system, leading to a higher risk of cardiovascular disease in the elderly. Furthermore, the rapid pace of modern life, driven by global economic progress and high working hours, has drastically increased the stress of modern life, resulting in irregular work and rest habits. Advances in food technology and transportation have also led to the availability of many processed foods and heavily seasoned, flavorful foods. All these lifestyle factors contribute significantly to the increased incidence of modern cardiovascular disease. Therefore, echocardiography is a crucial method for detecting cardiovascular disease. Traditional echocardiography is a highly technical and time-consuming procedure. The equipment is expensive, scarce, and in high demand. Patients applying for an echocardiography scan must wait one to two months, and after the scan, they must wait another one to two weeks for the diagnostic results. This prolonged diagnostic process can easily cause patients to miss the optimal treatment window for cardiovascular disease.

[0003] Current cardiac ultrasound scanning methods involve highly technical measurement and processing, making talent training difficult and resulting in a shortage of qualified personnel. Furthermore, current cardiac ultrasound scanning techniques are performed manually. Technicians need to place the ultrasound probe at different positions within the patient's chest cavity to avoid structures such as the ribs and lungs, finding the optimal scanning angle to clearly display different parts of the heart. This typically involves multiple probe angle adjustments to acquire images from different cross-sections. If the technician lacks experience in positioning and angle adjustment, re-examinations are frequently necessary. Because the heart is constantly beating, technicians need to simultaneously observe and capture key moments of cardiac activity, such as images during systole and diastole. Therefore, technicians must possess excellent technical and anatomical knowledge during the scanning process to ensure the clinical accuracy of the images. After the scan, technicians also need to perform preliminary image labeling and measurements, including the size of cardiac structures, wall thickness, and chamber diameter, to provide reference for physician diagnosis.

[0004] As can be seen from the above, the accuracy of current cardiac ultrasound scanning largely depends on the experience and operating skills of technicians, and the production of cardiovascular disease diagnostic reports requires a lot of professional manpower and time. Therefore, how to reduce the need for manual operation in cardiac ultrasound scanning and improve the efficiency of cardiac ultrasound detection has become a very important issue. Summary of the Invention

[0005] In view of the various problems of the prior art, the main objective of the present invention is to provide a wearable cardiac ultrasound patch, a cardiovascular disease identification system and method thereof, wherein the wearable cardiac ultrasound patch is used to be installed on the skin in the heart region to obtain cardiac images. The wearable cardiac ultrasound patch includes: a package having an adhesive layer on one side for adhering to the skin in the heart region; and a fiber optic microsensor housed in the package, the fiber optic microsensor including a silicon photonic integrated circuit layer of a plurality of silicon photonic components arranged in a two-dimensional array and a microelectromechanical integrated circuit layer of a plurality of piezoelectric components arranged in a two-dimensional array. The circuit layer and the microelectromechanical integrated circuit (MEMS) layer are vertically stacked, with each silicon photonic component corresponding to each piezoelectric component. The silicon photonic integrated circuit layer is disposed facing the adhesive layer. The MEMS layer is used to emit ultrasound waves and propagate them to the heart region. When the ultrasound echo acts on each piezoelectric component and causes material deformation, the material deformation is captured by the silicon photonic component and converted into a phase change of the optical signal. The silicon photonic integrated circuit layer outputs the electrical signal converted from the optical signal. A computing chip is also included in the package. The computing chip is vertically stacked with the fiber optic microsensor. The computing chip uses 3D beamforming technology to drive each piezoelectric component to emit ultrasound waves and performs digital processing and artificial intelligence data analysis on the electrical signal output by the silicon photonic integrated circuit layer. This allows the ultrasound beam to be focused on different angles and depths of the heart, acquiring ultrasound data from multiple angles and generating raw data for cardiac ultrasound images.

[0006] Preferably, in the aforementioned wearable cardiac ultrasound patch, the computing chip includes: an ultrasound transducer driver for providing the required voltage and current to multiple piezoelectric components of the microelectromechanical integrated circuit layer to control each piezoelectric component to emit the ultrasound waves; an analog front-end circuit for amplifying the electrical signal output by the silicon photonic integrated circuit layer, performing noise suppression and filtering to reduce noise, and quantizing the electrical signal into a digital signal through analog-to-digital conversion; and a digital control system for receiving the digital signal processed by the analog front-end circuit, performing digital signal processing (DSP) and data parsing to generate the raw data of the cardiac ultrasound image, and storing the raw data of the cardiac ultrasound image for subsequent reconstruction processing of the cardiac ultrasound image.

[0007] Preferably, in the aforementioned wearable cardiac ultrasound patch, the computing chip further includes: a computing unit for performing artificial intelligence calculations, interpretation, and identification on the raw data of the cardiac ultrasound images generated by the digital control system; when it is identified that the raw data of the cardiac ultrasound images does not contain valid data of the heart structure, feedback information is given to the digital control system to control the ultrasonic transducer driver to adjust the parameters of the ultrasonic waves emitted by each of the piezoelectric components, so that the raw data of the cardiac ultrasound images generated by the digital control system contains valid data of the heart structure, thereby reflecting the image of the heart region and achieving the accuracy of subsequent cardiac ultrasound image reconstruction processing.

[0008] Preferably, in the aforementioned wearable cardiac ultrasound patch, the computing chip also has data transmission and storage functions to store the raw data of the generated cardiac ultrasound image for subsequent cardiac ultrasound image reconstruction processing, or to transmit the raw data of the cardiac ultrasound image to a cardiovascular disease identification system. The cardiovascular disease identification system has a multi-view identification network module and a cardiovascular disease detection network module. The multi-view identification network module stores multiple view identification models to perform cardiac ultrasound image reconstruction processing on the raw data of the cardiac ultrasound image to generate a cardiac ultrasound image to be identified. Furthermore, it identifies the corresponding view identification model based on the cardiac ultrasound image to be identified to obtain the cardiac ultrasound view image and cardiac ultrasound image quality required for clinical diagnosis. When the cardiovascular disease identification system determines that the image to be identified… When a cardiac ultrasound image does not conform to the cardiac ultrasound viewing angle or quality, the computing chip drives each piezoelectric component to adjust the ultrasound parameters of the three-dimensional beamforming to transmit ultrasound waves, thereby obtaining raw data of cardiac ultrasound images at different focal positions. This process continues until the cardiac ultrasound image to be identified, obtained based on subsequently generated raw data, conforms to the cardiac ultrasound viewing angle and quality. The cardiovascular disease detection network module has a cardiac ultrasound signal database for analyzing cardiac ultrasound images that conform to the cardiac ultrasound viewing angle and quality. Feature parameters are extracted from the analyzed cardiac ultrasound images, and cardiac ultrasound images are classified with cardiovascular diseases using the cardiac ultrasound signal database and cardiac ultrasound algorithm to provide a cardiovascular disease identification report.

[0009] The cardiovascular disease identification system of the present invention can perform data transmission processing with the aforementioned wearable cardiac ultrasound patch. The computing chip of the wearable cardiac ultrasound patch has a data transmission function to transmit the raw data of the cardiac ultrasound image generated by the computing chip to the cardiovascular disease identification system. The cardiovascular disease identification system includes: a multi-view identification network module for performing cardiac ultrasound image reconstruction processing on the raw data of the cardiac ultrasound image to generate a cardiac ultrasound image to be identified. The multi-view identification network module stores multiple view identification models, each view identification model having at least one cardiac ultrasound view image and its corresponding cardiac ultrasound image quality. Based on the cardiac ultrasound image to be identified, the corresponding view identification model is found to identify the cardiac ultrasound view image and cardiac ultrasound image quality required for clinical diagnosis. When it is determined that the cardiac ultrasound image to be identified does not conform to the cardiac ultrasound viewing angle image or cardiac ultrasound image quality, the computing chip drives each of the piezoelectric components to adjust the ultrasound parameters of the three-dimensional beamforming to transmit ultrasound, so as to obtain the original data of cardiac ultrasound images at different focusing positions, until the cardiac ultrasound image to be identified obtained based on the subsequently generated original data conforms to the cardiac ultrasound viewing angle image and cardiac ultrasound image quality; and a cardiovascular disease detection network module, which has a cardiac ultrasound signal database, is used to analyze the cardiac ultrasound images to be identified that conform to the cardiac ultrasound viewing angle image and cardiac ultrasound image quality, to extract feature parameters from the analyzed cardiac ultrasound images to be identified, and to classify cardiac ultrasound images and cardiovascular diseases according to the cardiac ultrasound signal database and cardiac ultrasound algorithm, so as to provide a cardiovascular disease identification report.

[0010] Preferably, in the above-mentioned cardiovascular disease identification system, the cardiovascular disease detection network module also has a cardiac audio signal database and receives sound signals generated by cardiac activity captured by a phonocardiography (PCG) detection device, so as to analyze and diagnose the function and health status of the heart based on the cardiac audio signal database.

[0011] Preferably, in the above-mentioned cardiovascular disease identification system, the cardiovascular disease detection network module further includes an electrocardiogram (ECG) signal database and receives ECG signals detected by an electrocardiogram (ECG) detection device to monitor heart rhythm and diagnose heart-related diseases based on the ECG signal database.

[0012] The cardiovascular disease identification method of the present invention utilizes the above-mentioned wearable cardiac ultrasound patch and a cardiovascular disease identification system for data transmission processing to achieve cardiovascular disease identification. The cardiovascular disease identification method includes the following steps: (1) The computing chip of the wearable cardiac ultrasound patch transmits the raw data of the generated cardiac ultrasound image to the cardiovascular disease identification system; (2) The cardiovascular disease identification system receives the raw data of the cardiac ultrasound image transmitted by the computing chip and performs cardiac ultrasound image reconstruction processing to generate a cardiac ultrasound image to be identified; (3) The cardiovascular disease identification system determines whether the cardiac ultrasound image to be identified meets the cardiac ultrasound perspective image and cardiac ultrasound image quality required for clinical diagnosis. When it is determined that the cardiac ultrasound image to be identified does not meet the cardiac ultrasound perspective image or When the quality of the cardiac ultrasound image is affected, the computing chip drives each of the piezoelectric components to adjust the ultrasound parameters of the three-dimensional beamforming to transmit ultrasound, so as to generate raw data of cardiac ultrasound images at different focal positions, until the cardiac ultrasound image to be identified generated by the cardiovascular disease identification system based on the raw data subsequently transmitted from the computing chip meets the cardiac ultrasound perspective image and cardiac ultrasound image quality; and (4) the cardiovascular disease identification system analyzes the cardiac ultrasound image to be identified that meets the cardiac ultrasound perspective image and cardiac ultrasound image quality, extracts feature parameters from the analyzed cardiac ultrasound image to be identified, and classifies cardiac ultrasound images and cardiovascular diseases according to cardiac ultrasound signal database and cardiac ultrasound algorithm to provide a cardiovascular disease identification report.

[0013] Preferably, in the above-mentioned cardiovascular disease identification method, before performing step (1), the following steps are included: (1-1) The computing chip of the wearable cardiac ultrasound patch interprets the raw data of the generated cardiac ultrasound image; (1-2) After interpretation, the computing chip analyzes whether the raw data of the cardiac ultrasound image has valid data of the heart structure. If it does not have valid data of the heart structure, proceed to step (1-3); if it has valid data of the heart structure, proceed to step (1-4); (1-3) When the computing chip analyzes that the raw data of the cardiac ultrasound image does not have valid data of the heart structure, it adjusts the parameters of the ultrasound emitted by each piezoelectric component so that the raw data of the cardiac ultrasound image subsequently generated by the computing chip contains valid data of the heart structure, and then performs step (1); and (1-4) When the computing chip analyzes that the raw data of the cardiac ultrasound image has valid data of the heart structure, it performs step (1).

[0014] In summary, the wearable cardiac ultrasound patch, cardiovascular disease identification system, and method of the present invention integrate a silicon photonics integrated circuit layer and a microelectromechanical integrated circuit layer into a fiber optic microsensor, which is packaged with a computing chip. This allows the patient to conveniently receive cardiac ultrasound detection and processing in a wearable manner as needed. Furthermore, the silicon photonic components in the silicon photonics integrated circuit layer and the piezoelectric components in the microelectromechanical integrated circuit layer are stacked and arranged in a two-dimensional array, enabling the wearable cardiac ultrasound patch to scan more angles. For piezoelectric micromachined ultrasound transducers (pMUTs), ultrasound reception also originates from the deformation of the piezoelectric layer caused by thin film vibration. The wearable cardiac ultrasound patch of the present invention utilizes silicon photonics technology to sense thin film vibration, thus improving the sensitivity issues of piezoelectric materials. Moreover, the advantages of silicon photonics technology—low noise, high speed, and high sensitivity—can reduce the driving voltage. Therefore, the wearable cardiac ultrasound patch of the present invention can operate with lower power consumption. Furthermore, the cardiovascular disease identification system of the present invention can detect the patient's cardiac ultrasound, electrocardiogram and heart sound signals, and obtain cardiovascular disease-related parameters after signal processing. These parameters are then fed into an artificial intelligence cardiovascular disease identification and diagnosis algorithm to classify cardiovascular diseases, such as valvular heart disease, heart failure, coronary artery disease and arrhythmia, thereby reducing the time and misdiagnosis of cardiovascular diseases for physicians. Attached Figure Description

[0015] Figure 1 The diagram shown is an exploded view illustrating the structure of the wearable cardiac ultrasound patch of the present invention.

[0016] Figure 2 The diagram shown illustrates the structure of the fiber optic microsensor in the wearable cardiac ultrasound patch of the present invention.

[0017] Figure 3 The diagram shown is a basic architecture block diagram of an embodiment of the computing chip of the wearable cardiac ultrasound patch of the present invention.

[0018] Figure 4 This is shown as an application example of the wearable cardiac ultrasound patch of the present invention.

[0019] Figure 5 The diagram shown is a block diagram illustrating the basic architecture of the cardiovascular disease identification system and the wearable cardiac ultrasound patch for data transmission processing according to the present invention.

[0020] Figure 6 The diagram shown is a basic system architecture block diagram illustrating another embodiment of the cardiovascular disease identification system of the present invention.

[0021] Figure 7 The diagram shown illustrates the processing flow of the cardiovascular disease identification method of the present invention.

[0022] Figure 8 The flowchart shown illustrates the processing of raw data from cardiac ultrasound images in the cardiovascular disease identification method of the present invention.

[0023] Component designation explanation

[0024] 1.1' Wearable cardiac ultrasound patch

[0025] 10 Package

[0026] 100 adhesive layers

[0027] 11 Fiber Optic Miniature Sensors

[0028] 111 Silicon Photonics Module

[0029] 110 Silicon Photonics Integrated Circuit Layer

[0030] 115 Microelectromechanical Integrated Circuit Layer

[0031] 116 Piezoelectric Components

[0032] 12 computing chips

[0033] 120 computing chip

[0034] 121 Ultrasonic transducer driver

[0035] 122 Analog Front-End Circuit

[0036] 123 Digital Control System

[0037] 124 computing units

[0038] 2,2' Cardiovascular Disease Identification System

[0039] 20 multi-view recognition network modules

[0040] 21 Cardiovascular Disease Detection Network Module

[0041] 23 Information Transmission and Processing Module

[0042] 24 User Interface Module

[0043] 25. Cloud Database Construction Module

[0044] 26 AI information combined with physiological information processing

[0045] Module

[0046] 30 Parasternal region

[0047] 31. Apical region

[0048] 32. Subcostal region Detailed Implementation

[0049] The following description, accompanied by illustrations, illustrates the technical content of the present invention through specific embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments. Various details in this specification can also be modified and changed based on different viewpoints and applications without departing from the spirit of the present invention.

[0050] The wearable cardiac ultrasound patch of this invention reduces the cumbersome process of traditional cardiac ultrasound measurement by providing cardiac ultrasound image detection and processing. Moreover, the detection and processing is not limited by environmental and time constraints. Through the cardiovascular disease identification system of this invention, cardiovascular diseases can be identified and predicted in real time to assist doctors in diagnosis. Doctors can obtain measurement reports in a shorter time, improve the quality of medical care, and it can also be applied to telemedicine, providing a more convenient and efficient way to measure cardiovascular diseases.

[0051] Firstly, as Figure 1 The diagram shown is an exploded view of the wearable cardiac ultrasound patch 1 of the present invention. The wearable cardiac ultrasound patch 1 is installed on the skin in the heart region, specifically on the left side of the chest, to capture cardiac images. The wearable cardiac ultrasound patch 1 includes: a package 10, a fiber optic microsensor 11, and a computing chip 12. The package 1 encapsulates the fiber optic microsensor 11 and the computing chip 12. Please refer to the accompanying documentation. Figure 2 The diagram shows the structure of the fiber optic microsensor 11. The fiber optic microsensor 11 includes a silicon photonic integrated circuit layer 110 with a plurality of silicon photonic components 111 arranged in a two-dimensional array and a microelectromechanical integrated circuit layer 115 with a plurality of piezoelectric components 116 arranged in a two-dimensional array. The silicon photonic integrated circuit layer 110 and the microelectromechanical integrated circuit layer 115 are vertically stacked, and each of the plurality of silicon photonic components 111 corresponds to each of the plurality of piezoelectric components 116. The computing chip 12 is vertically stacked with the fiber optic microsensor 11 and electrically connected to the plurality of silicon photonic components 111 and the plurality of piezoelectric components 116.

[0052] An adhesive layer 100 is provided on one side of the package 10 for adhering to the skin of the heart region, and the silicon photonic integrated circuit layer 110 faces the adhesive layer 100. Each of the piezoelectric components 116 of the microelectromechanical integrated circuit layer 115 is used to emit ultrasound waves and propagate to the heart region. When the ultrasound echo acts on each of the piezoelectric components 116 and causes material deformation, the material deformation is captured by the silicon photonic component 111 and converted into a phase change of the optical signal. In other words, the phase change of the ultrasound wave is ultimately mapped into the phase modulation of the light wave through the piezoelectric effect and the sensitivity of the optical components, so as to convert the deformation into an optical signal to reflect the change caused by the ultrasound wave.

[0053] The computing chip 12 uses 3D beamforming technology to control the piezoelectric components 116 on the microelectromechanical integrated circuit layer 115 to emit ultrasound waves, and performs digital processing and artificial intelligence data analysis on the electrical signals output by the silicon photonic integrated circuit layer 110. Specifically, based on the intensity and echo time of the reflected waves, different tissue densities and their distance from the fiber optic microsensor 11 can be analyzed. In cardiac ultrasound examinations, there is a close relationship between the frequency of the ultrasound waves and their penetration depth and image resolution. Low-frequency ultrasound waves (approximately 2-5 MHz) have strong penetrating power and are suitable for observing deeper structures, such as the deep structures of the adult heart. The longer wavelength of low-frequency waves allows them to penetrate tissues more easily without being rapidly absorbed, thus obtaining clear images at deeper locations. High-frequency ultrasound waves (5-10 MHz) have shallower penetration depth and are suitable for observing superficial tissues or smaller structures near the body surface, such as the heart or superficial tissues in children. While the shorter wavelength of high-frequency waves provides higher resolution, the energy is more easily absorbed by human tissue, resulting in a reduced penetration depth. As can be seen from the foregoing, since there is a close relationship between the frequency of ultrasound and its penetration depth and image resolution, the wearable cardiac ultrasound patch 1 of the present invention automatically adjusts the frequency of ultrasound emitted by each of the plurality of piezoelectric components 116 through the computing chip 12, so that the ultrasound beam can be focused on different angles and depths of the heart, obtain ultrasound data from multiple angles, and generate raw data of cardiac ultrasound images.

[0054] The microelectromechanical integrated circuit layer 115 is based on the piezoelectric effect technology of, for example, a piezoelectric micromachined ultrasonic transducer (pMUT). It utilizes the properties of piezoelectric materials to convert electrical energy into mechanical vibrations to emit ultrasonic waves, and receives ultrasonic echoes reflected back from tissues or structures of different densities (such as heart muscle, heart chambers, and blood) as the ultrasonic waves propagate. The silicon photonic integrated circuit layer 110 detects minute material deformations (not illustrated) in the structure of the silicon photonic integrated circuit layer 110, such as optical resonators or optical waveguides, caused by the ultrasonic echo, based on the optical mechanical vibration detection principle of an optical micro-machined ultrasonic sensor (OMUS). These deformations are converted into changes in optical signals. These deformations change the transmission path or phase of the optical signals, causing interference or frequency changes. Such optical changes represent the characteristics of the received acoustic waves. This optical modulation technique, combined with silicon photonics technology, transmits optical signals, allowing the ultrasonic echoes to undergo optical processing within the optical micro-machined ultrasonic sensor to become high-precision, low-noise data, facilitating further analysis. Because the silicon photonic component 111 can precisely modulate the phase or intensity of the optical signal, it helps to identify information such as the relative strength and time difference in the echo process of ultrasound, thereby generating high-resolution signal output. This technology can save echo information at different levels as optical phase information, and after being converted into electrical signals, it is used by the computing chip 12 for digitization and to generate raw data for cardiac ultrasound images required for subsequent cardiac image reconstruction. In addition, the optical signal processing through the silicon photonic component 111 can improve the sensitivity and accuracy of ultrasound sensing, and make it less sensitive to external electromagnetic interference (EMI), thus significantly improving the signal-to-noise ratio and eliminating the problem that electromagnetic interference in the medical environment may affect the performance of traditional electronic sensors.

[0055] Therefore, the wearable cardiac ultrasound patch 1 of the present invention is an ultrasound sensing technology that combines silicon photonics and microelectromechanical systems (MEMS), mainly utilizing optical resonance and mechanical vibration to detect ultrasound signals. The working principle of this sensor is based on the correlation between wavelength shift of the optical resonant cavity and mechanical deformation. When an ultrasound signal strikes the sensor's thin film or structure, it causes changes in the optical path, which are reflected in the change in the optical signal, thereby achieving accurate ultrasound measurement. It is particularly noteworthy that the silicon photonic integrated circuit layer 110 also has a photodetector (e.g., a photodiode), which is typically located adjacent to or below the silicon photonic component 111 to detect the modulated optical signal inside the silicon photonic component 111 in real time. The optical signal is finally detected by the photodetector and converted into an electrical signal, which is transmitted to the computing chip 12 for digitization and generation of raw data for subsequent cardiac ultrasound image reconstruction.

[0056] To facilitate the use of the wearable cardiac ultrasound patch of the present invention by the examinee, that is, to eliminate the need for the examinee to pay special attention to the specific position of the wearable cardiac ultrasound patch on the skin of the heart region, the computing chip 12 of the wearable cardiac ultrasound patch 1 of the present invention performs a position calibration operation. That is, the computing chip 12 performs an initialization operation when it first drives each of the piezoelectric components 116 to emit ultrasound waves, and calibrates the position of the package 10 and the heart region. It should be specifically noted that the computing chip 12 pre-stores identification data for multiple cardiac detection zones and corresponding ultrasound control parameters. The cardiac detection zones include at least the parasternal, apical, and subcostal regions. Therefore, the computing chip 12 can control the ultrasound emission parameters of each piezoelectric component 116 based on different cardiac detection zones. The determination of the cardiac detection zone can be based on the raw data of the first cardiac ultrasound image generated by the computing chip 12. This raw data mainly comes from the ultrasound reflected sound wave signal. The signal contains echo characteristics of tissues at different depths (such as reflection intensity, echo delay, etc.), thus providing a digital reflection of the cardiac region structure and tissue density. Based on the determined cardiac region structure and tissue density, the application position of the wearable cardiac ultrasound patch 1 is determined. The cardiac detection area is determined by the application position, and the corresponding identification data and ultrasound control parameters corresponding to each identification data are found. That is, the cardiac detection area confirmed by the computing chip 12 after position calibration, and the piezoelectric components 116 are adjusted according to the identification data corresponding to the cardiac detection area to generate ultrasound waves with corresponding waveforms and wavelengths, so that the ultrasound beam is focused on the angle and depth of the heart.

[0057] Next, please refer to Figure 3This diagram illustrates the basic architecture of a computing chip in an embodiment of the wearable cardiac ultrasound patch of the present invention. The computing chip 120 in this embodiment includes: an ultrasound transducer driver 121, an analog front-end (AFE) circuit 122, a digital control system 123, and a computing unit 124. Please refer to the accompanying documentation. Figure 1 and Figure 2 The ultrasonic transducer driver 121 provides the required voltage and current to the multiple piezoelectric components 116 of the microelectromechanical integrated circuit layer 115 of the fiber optic microsensor 11, thereby driving the multiple piezoelectric components 116 of the microelectromechanical integrated circuit layer 115 of the fiber optic microsensor 11 located on the skin in the heart region. This causes the multiple piezoelectric components 116 to generate corresponding ultrasound waves according to the driving parameters. These ultrasound waves encounter tissues or structures of different densities (such as heart muscle, heart chambers, and blood) and are reflected back when they encounter tissues or structures of different densities. The ultrasonic transducer driver 121 adjusts the ultrasound parameters, such as frequency, according to the depth and resolution requirements of the heart detection. Lower frequency signals can penetrate deeper tissues but have lower resolution, while higher frequencies are suitable for shallow tissue detection and have higher resolution. When the reflected wave arrives at the plurality of piezoelectric components 116, the plurality of piezoelectric components 116 will undergo slight deformation due to the effect of the wave. At this time, the plurality of silicon photonic components 111 on the silicon photonic integrated circuit layer 110 of the fiber optic micro-sensor 11 will be affected by the piezoelectric deformation of each of the plurality of piezoelectric components 116, which will further affect the optical structure (such as optical resonator or optical waveguide) in the silicon photonic component 111, so that the optical signal is modulated according to the deformation to output an electrical signal.

[0058] The analog front-end circuit 122 is used to amplify the electrical signals converted by the multiple silicon photonic components 111 of the silicon photonic integrated circuit layer 110, and to perform noise suppression and filtering to reduce noise, ensure signal quality, and make the signals supplied to the digital control system 123 for image reconstruction processing more accurate. The circuit also quantizes the electrical signals into digital signals through analog-to-digital conversion.

[0059] The digital control system 123 receives the digital signal processed by the analog front-end circuit 122, performs digital signal processing (DSP) and data parsing to generate the raw data of the cardiac ultrasound image, and stores the raw data of the cardiac ultrasound image for subsequent cardiac ultrasound image reconstruction processing. The wearable cardiac ultrasound patch 1' of the present invention first performs preprocessing to generate the raw data of the cardiac ultrasound image. For high-resolution cardiac ultrasound images, the reconstruction processing is performed by a back-end device (e.g., a remote server) (which will be described in detail later). Figure 4 Therefore, the digital control system 123 also has a data transmission function, which is used to transmit the raw data of the generated cardiac ultrasound image to the outside world, so as to remotely transmit it to the aforementioned back-end device or to information processing devices such as mobile phones or computers. The cardiac ultrasound image can be reconstructed in the back-end device or information processing device, and the cardiac ultrasound image can be classified with cardiovascular diseases, so that medical staff can detect abnormalities or potential risks in the patient's heart in real time based on the cardiac ultrasound image, and take corresponding measures for emergency situations.

[0060] To ensure the validity of the raw cardiac ultrasound image data required for reconstruction processing by backend devices or information processing equipment, the computing unit 124 of the computing chip 120 of the wearable cardiac ultrasound patch 1' of this invention can perform artificial intelligence calculations, interpretation, and identification on the raw cardiac ultrasound image data generated by the digital control system 123. When it is identified that the raw cardiac ultrasound image data does not contain valid data on the heart structure, feedback information is sent to the digital control system 123 to control the ultrasound transducer driver 121 to adjust the frequency and focusing mode of the ultrasound waves emitted by each piezoelectric component 116, so that the raw cardiac ultrasound image data generated by the digital control system 123 contains valid data on the heart structure, ensuring the accuracy of cardiac ultrasound image reconstruction processing by the backend device. Therefore, the function provided by the computing unit 124 is similar to that of an edge computing device, pre-analyzing the validity of the raw data to be provided to the backend device or information processing equipment for cardiac ultrasound image reconstruction, thereby reducing the computational load on the backend device or information processing equipment. In addition, compared with the digital control system 123, the computing unit 124 provides auxiliary calculation and management functions. According to the current detection needs, it can adjust the power, frequency and focusing mode of the ultrasound transducer driver 121, and dynamically adjust the ultrasound parameters according to the patient's cardiac characteristics to improve the accuracy of diagnosis.

[0061] To further optimize the adjustment of ultrasound parameters to meet the personalized testing needs of different patients with different cardiac characteristics while ensuring diagnostic accuracy, similarly to the above, since the wearable cardiac ultrasound patch of this invention has a data transmission function, in addition to transmitting the raw data of the cardiac ultrasound image generated by the computing chip 120 to the information processing device for cardiac ultrasound image reconstruction processing, the computing chip 120 can also receive control information from the information processing device. The control information includes adjusting the power, frequency, and focusing mode of the ultrasound transducer driver 121. Correspondingly, the information processing device has a built-in cardiac... A cardiovascular disease identification application provides a user interface for users (e.g., medical personnel) located on the information processing device to view reconstructed cardiac ultrasound images. If the user is not satisfied with the quality of the cardiac ultrasound image, they can input the desired viewing angle through the user interface. The cardiovascular disease identification application obtains the ultrasound parameters of the desired viewing angle and transmits them to the wearable cardiac ultrasound patch, causing the computing unit 124 to adjust the power, frequency, and focusing mode of the ultrasound transducer driver 121 according to the ultrasound parameters. In short, the interaction architecture between the wearable cardiac ultrasound patch of the present invention and the aforementioned information processing device or backend device is based on the design principle of the Internet of Things (IoT), enabling smooth data exchange and achieving functions such as remote control or status monitoring. Therefore, the wearable cardiac ultrasound patch of the present invention can be applied to telemedicine, providing a more convenient and efficient method for measuring cardiovascular diseases.

[0062] It should be noted that the wearable cardiac ultrasound patch of this invention can be used individually, and the number of patches can be increased or decreased for extended applications. The more patches there are, the more comprehensive and complete the captured cardiac ultrasound images will be; therefore, more patches can improve the detection range and accuracy. This embodiment uses three wearable cardiac ultrasound patches for detection, such as... Figure 4 As shown, each wearable cardiac ultrasound patch is respectively attached to the parasternal region 30, the apex region 31, and the subcostal region 32, enabling each wearable cardiac ultrasound patch to cover a 120-degree emission range to obtain 360-degree cardiac ultrasound imaging. The position of the wearable cardiac ultrasound patch and the ultrasound emission angle of each piezoelectric material inside it determine the viewing angle and structural visibility of the image. The three wearable cardiac ultrasound patches are arranged in a... Figure 4When performing cardiac ultrasound examination using the setup shown, the following images can be obtained: Parasternal Long Axis View (PLAX), Parasternal Short Axis View (PSAX), Apical Four-Chamber View (A4CH) from the apex of the heart, Subcostal Inferior Vena Cava View (SIVC) from below the sternum (below the xiphoid process), and Subcostal Long Axis View (SLAX) from below the sternum (long axis) and below the xiphoid process, showing the overall cardiac structure. By controlling the ultrasound parameters of each piezoelectric component 116 in the microelectromechanical integrated circuit layer 115, the ultrasound beam can be focused on different angles and depths of the heart, generating multi-angle ultrasound data and obtaining multiple cardiac ultrasound images, which can then be used by physicians to diagnose various cardiovascular diseases.

[0063] Next, please refer to Figure 5 This is a block diagram illustrating the basic architecture of the cardiovascular disease identification system of the present invention and the wearable cardiac ultrasound patch for data transmission processing. As described above, the computing chip of the wearable cardiac ultrasound patch of the present invention has a data transmission function to transmit the raw data of the cardiac ultrasound image generated by the computing chip to the cardiovascular disease identification system 2. The cardiovascular disease identification system 2 includes: a multi-view identification network module 20 and a cardiovascular disease detection network module 21. The multi-view identification network module 20 is used to perform cardiac ultrasound image reconstruction processing on the raw data of the cardiac ultrasound image to generate a cardiac ultrasound image to be identified. The multi-view identification network module stores multiple view identification models. Each view identification model has at least one cardiac ultrasound view image and its corresponding cardiac ultrasound image quality. Based on the cardiac ultrasound image to be identified, the corresponding view identification model is found to find the cardiac ultrasound view image and cardiac ultrasound image quality required for clinical diagnosis.

[0064] The aforementioned multi-view identification model provides diverse information, enabling physicians to quickly and accurately examine different structures and functions of the heart, avoiding the blind spots of a single viewpoint. Combining multi-view images can aid in the diagnosis of heart diseases such as valvular heart disease, cardiac defects, or cardiomyopathy, and also provides an assessment of cardiac function, facilitating a comprehensive analysis of cardiac systolic and diastolic function and hemodynamics. The quality of cardiac ultrasound images refers to the clarity, accuracy, and signal integrity of the cardiac images, helping physicians accurately diagnose the health status of the heart and blood vessels. The quality of cardiac ultrasound images is affected by several factors, primarily including spatial resolution, contrast resolution, and depth penetration.

[0065] Therefore, the multi-view recognition network module 20 performs the aforementioned cardiac ultrasound perspective image or cardiac ultrasound image quality judgment on the cardiac ultrasound image to be identified generated after image reconstruction. When it is determined that the cardiac ultrasound image to be identified does not meet the required cardiac ultrasound perspective image or cardiac ultrasound image quality, the computing chip drives each of the piezoelectric components to adjust the parameters of the three-dimensional beamforming to transmit ultrasound waves, so as to obtain the original data of cardiac ultrasound images at different focusing positions, until the cardiac ultrasound image to be identified obtained based on the subsequently generated original data meets the cardiac ultrasound perspective image and cardiac ultrasound image quality. The aforementioned adjustment of the three-dimensional beamforming parameters includes: adjusting the phase difference and amplitude difference between adjacent piezoelectric components, and changing the emission delay time of the piezoelectric components, etc. Since the ultrasound beam can be controlled to different angles as needed to achieve three-dimensional imaging, by adjusting the phase difference and amplitude difference between adjacent piezoelectric components, the beam can be pointed in a specific direction; in addition, the focusing position affects the imaging depth and resolution, and changing the emission delay time of the piezoelectric components can focus the beam at different depths. Because the phase of a wave changes over time, and the phase describes a specific state (position) of a wave within its period, phase difference and time delay are essentially interconvertible concepts in beamforming. The phase difference caused by time delay directly determines the direction and focusing ability of the beam. By adjusting these parameters, the angle and depth of beamforming can be precisely controlled, thereby improving the accuracy and resolution of cardiac ultrasound imaging.

[0066] The cardiovascular disease detection network module 21 has a cardiac ultrasound signal database for analyzing cardiac ultrasound images that conform to the cardiac ultrasound perspective and image quality. It extracts feature parameters from the analyzed cardiac ultrasound images and, based on the cardiac ultrasound signal database and cardiac ultrasound algorithms, classifies the cardiac ultrasound images and cardiovascular diseases to provide a cardiovascular disease identification report. The feature parameters are numerical values ​​extracted from cardiac ultrasound images or data. These data have specific diagnostic significance and can be used to describe the anatomical structure and function of the heart. Common feature parameters include: the size of the heart chambers (e.g., left ventricular diameter, right atrial area), the thickness of the heart wall (e.g., degree of myocardial hypertrophy), the morphology and motion characteristics of the heart valves (e.g., mitral valve orifice area), the motion velocity and pattern of the ventricular wall (regional myocardial motion disorder assessment), blood flow velocity and volume, etc. These parameters quantitatively describe the health status or pathological changes of the heart and become important bases for disease diagnosis and prediction. The cardiac ultrasound algorithm is a computational method based on data and image processing technology. Its purpose is to extract useful features from ultrasound images and analyze them, enabling the classification of cardiac ultrasound images and cardiovascular diseases. The cardiac ultrasound algorithm includes the following processes: (1) improving image contrast and edge clarity to facilitate the segmentation of cardiac structures; (2) feature extraction, based on machine learning or deep learning methods, automatically identifying and quantifying cardiac structural and functional parameters from images; (3) data matching: comparing the extracted features with a cardiac ultrasound signal database, which contains a large number of labeled normal and pathological samples, which can help the cardiac ultrasound algorithm identify cardiovascular diseases; (4) disease diagnosis: classifying cardiac ultrasound images and cardiovascular diseases based on abnormal patterns of feature parameters, and providing diagnostic reports generated by the cardiac ultrasound algorithm, such as determining whether there is valvular stenosis or myocardial ischemia. Therefore, the cardiovascular disease detection network module 21, by combining feature parameters with the cardiac ultrasound algorithm, can significantly reduce errors in human diagnosis. On the other hand, the cardiac ultrasound algorithm can generate personalized reports based on the examinee's characteristics, providing personalized disease assessments and treatment recommendations. Through continuous data collection and analysis, the cardiac ultrasound algorithm can monitor disease progression and evaluate treatment effectiveness. This combination of technologies not only improves diagnostic efficiency but also enhances the clinical application value of medical imaging.

[0067] Echocardiography is primarily used for assessing cardiac structure and dynamic function, such as chamber size, myocardial motion, and valvular function. However, it cannot provide direct information about the details of cardiac electrical activity. Therefore, the cardiovascular disease identification system of this invention can also receive PCG and ECG signals from a phonocardiography (PCG) detection device and an electrocardiography (ECG) detection device, respectively. The ECG signal clearly reflects the electrophysiological activity of the heart and is highly sensitive for diagnosing arrhythmias, myocardial ischemia, and infarction. The PCG signal, by recording heart sounds, can detect abnormal sounds in the early stages of certain valvular diseases (such as stenosis or regurgitation). The data from echocardiography, PCG, and ECG mutually verify each other, improving the detection rate of comprehensive cardiovascular diseases and helping doctors make more accurate diagnoses. Figure 6 As shown, it is a block diagram of the basic system architecture of another embodiment of the cardiovascular disease identification system of the present invention. The cardiovascular disease identification system 2' of this embodiment includes: an information transmission and processing module 23, a user interface module 24, a cloud database construction module 25, and an AI information combined with physiological information processing module 26.

[0068] The information transmission and processing module 23 is used to process the received raw echocardiogram data, PCG signals, and ECG signals and perform signal processing, such as feature extraction. Regarding the raw echocardiogram data, the information transmission and processing module 23 includes, for example, the above-described features. Figure 5 The processing of the multi-view identification network module 20 shown is used for cardiac ultrasound image reconstruction. Since the PCG and ECG signals are time-series signals rather than image data, the information transmission processing module 23 ensures accurate synchronization between the PCG and ECG signals. Especially in multimodal diagnosis, in certain situations, these signals can be converted into image form to be combined with cardiac ultrasound images to improve diagnostic results. It is worth noting that, in this embodiment, in addition to the wearable cardiac ultrasound patch of the present invention, the patient in the cardiovascular disease identification system 2' also has PCG and ECG patches attached for sensing, and the sensed PCG and ECG signals are transmitted to the cardiovascular disease identification system 2'.

[0069] The user interface module 24 is used to display the results processed by the information transmission and processing module 23, and to generate relevant parameters and data. It can also mark abnormal signals for medical staff to interpret.

[0070] The cloud database construction module 25 constructs a cardiac ultrasound image database, a PCG database, and an ECG database. It stores the cardiac ultrasound images, PCG signals, and ECG signals detected by the examinee and input into the cardiovascular disease identification system 2' for subsequent processing by physicians. The processing may include verifying abnormal markings through the user interface module 24, such as marking four diseases: valvular heart disease, congestive heart failure, coronary artery disease, and the previously developed arrhythmia. The data and markings stored in the cardiac ultrasound image database, PCG database, and ECG database can be used to develop cardiovascular disease identification models.

[0071] The AI ​​information, combined with the physiological information processing module 26, can automatically analyze data and identify common cardiac abnormalities such as arrhythmia, atrial fibrillation, tachycardia, valvular heart disease, heart failure, and coronary artery disease, thus assisting in medical diagnosis. Since the physiological information includes ECG and PCG signals, it can be used to establish an AI-based cardiovascular disease identification model. This model combines relevant disease parameters with various neural networks (CNNs) and artificial intelligence technologies to classify cardiovascular diseases, identifying four types: valvular heart disease, heart failure, coronary artery disease, and arrhythmia. Computer-assisted disease prediction can then be performed, reducing the time and accuracy of physician diagnosis. This identification model assists in the correlation analysis of echocardiogram, ECG, and PCG signals, and it is expected that the four cardiovascular diseases can be predicted and identified using only two relatively easy-to-detect physiological signals: ECG and PCG. This will significantly improve the diagnostic efficiency of related diseases. Furthermore, the algorithm automates the medical grading of symptom severity, transforming the cumbersome judgment process for a disease by technicians and doctors into an automatic calculation by the AI ​​information combined with the physiological information processing module 26. This process performs cardiac ultrasound imaging and cardiovascular disease classification, generates relevant parameters and data, marks abnormal signals, and compiles all analysis results into a preliminary identification report for doctors' reference, aiming to improve the efficiency and accuracy of diagnosis.

[0072] Next, please refer to Figure 7This is a flowchart illustrating the cardiovascular disease identification method of the present invention. The cardiovascular disease identification method of the present invention uses the above-mentioned wearable cardiac ultrasound patch for cardiovascular disease identification processing. The wearable cardiac ultrasound patch and a cardiovascular disease identification system perform data transmission processing to realize cardiovascular disease identification. The cardiovascular disease identification method first performs step S1, in which the computing chip of the wearable cardiac ultrasound patch transmits the raw data of the generated cardiac ultrasound image to the cardiovascular disease identification system, and then proceeds to step S2.

[0073] In step S2, the cardiovascular disease identification system receives the raw data of the cardiac ultrasound image transmitted by the computing chip, performs cardiac ultrasound image reconstruction processing to generate a cardiac ultrasound image to be identified, and then proceeds to step S3.

[0074] In step S3, the cardiovascular disease identification system determines whether the cardiac ultrasound image to be identified meets the requirements of the cardiac ultrasound viewing angle and quality for clinical diagnosis. If the cardiac ultrasound image to be identified does not meet the requirements of the cardiac ultrasound viewing angle or quality, the system proceeds to step S4; conversely, if the cardiac ultrasound image to be identified meets the requirements of the cardiac ultrasound viewing angle and quality, the system proceeds to step S5. The aforementioned multi-view identification model provides diverse information to enable doctors to quickly and accurately examine different structures and functions of the heart, avoiding blind spots from a single viewpoint. Combining multi-view images can also help diagnose heart disease. The aforementioned cardiac ultrasound image quality refers to the clarity, accuracy, and signal integrity of the cardiac image, which helps doctors accurately diagnose the health status of the heart and blood vessels.

[0075] In step S4, the cardiovascular disease identification system determines that the cardiac ultrasound image to be identified does not meet the cardiac ultrasound viewing angle or cardiac ultrasound image quality required for clinical diagnosis. It then instructs the computing chip to drive each of the piezoelectric components to adjust the ultrasound parameters of the three-dimensional beamforming to transmit ultrasound waves, thereby generating raw data of cardiac ultrasound images at different focal positions. The system then returns to step S2. In short, this process continues until the cardiac ultrasound image to be identified generated by the cardiovascular disease identification system through image reconstruction processing based on the raw data subsequently transmitted from the computing chip meets the cardiac ultrasound viewing angle and cardiac ultrasound image quality requirements.

[0076] In step S5, the cardiovascular disease identification system analyzes the cardiac ultrasound images that conform to the cardiac ultrasound perspective and image quality, extracts feature parameters from the analyzed cardiac ultrasound images, and classifies the cardiac ultrasound images and cardiovascular diseases according to the cardiac ultrasound signal database and cardiac ultrasound algorithm, providing a cardiovascular disease identification report for physicians' reference to reduce the time and misdiagnosis of cardiovascular diseases. On the other hand, the cardiovascular disease identification report can also provide personalized condition assessment and treatment suggestions based on the characteristics of the examinee, with dynamic monitoring, in order to improve the efficiency of diagnosis.

[0077] The cardiovascular disease identification method of the present invention further includes processing the raw data of the cardiac ultrasound images generated by the computing chip of the wearable cardiac ultrasound patch, so that the raw data of the cardiac ultrasound images processed by the cardiovascular disease identification system is useful information, thereby avoiding the occupation of the cardiovascular disease identification system's resources and increasing the computational load, such as... Figure 8 As shown, step S10 is performed first, in which the computing chip of the wearable cardiac ultrasound patch interprets the raw data of the generated cardiac ultrasound image, and then proceeds to step S11.

[0078] In step S11, the computing chip interprets and analyzes whether the raw data of the cardiac ultrasound image contains valid data on the cardiac structure. If it does not contain valid data on the cardiac structure, the process proceeds to step S12; if it does contain valid data on the cardiac structure, the process proceeds to step S13. The aforementioned valid data on the cardiac structure refers to ultrasound echo information that clearly characterizes and reflects the cardiac structure (e.g., heart wall, valves, chambers, blood flow direction and velocity) and related functions. This data forms the basis for the subsequent cardiac ultrasound image reconstruction by the cardiovascular disease identification system.

[0079] In step S12, when the computing chip analyzes that the original data of the cardiac ultrasound image does not contain valid data of the heart structure, it adjusts the parameters of the ultrasound emitted by each piezoelectric component, such as frequency, focusing depth or direction, and returns to step S10 until the original data of the cardiac ultrasound image subsequently generated by the computing chip contains valid data of the heart structure, so as to ensure the accuracy of the cardiac ultrasound image reconstruction processing performed by the cardiovascular disease identification system in the subsequent process.

[0080] In step S13, when the computing chip analyzes and finds that the raw data of the cardiac ultrasound image contains valid data of the heart structure, the raw data of the cardiac ultrasound image is transmitted to the cardiovascular disease identification system for reconstruction processing of the cardiac ultrasound image.

[0081] In summary, the wearable cardiac ultrasound patch, cardiovascular disease identification system, and method of the present invention integrate a silicon photonics integrated circuit layer and a microelectromechanical integrated circuit layer into a fiber optic miniature sensor, which is packaged with a computing chip. This allows the examinee to conveniently perform cardiac ultrasound detection and processing in a wearable manner, regardless of time or location. Furthermore, the silicon photonic components in the silicon photonics integrated circuit layer and the piezoelectric components in the microelectromechanical integrated circuit layer are stacked and arranged in a two-dimensional array, enabling the wearable cardiac ultrasound patch to scan from more angles. Moreover, the cardiovascular disease identification system of the present invention can detect the examinee's cardiac ultrasound, electrocardiogram (ECG), and heart sound signals. After signal processing, cardiovascular disease-related parameters are obtained and incorporated into an artificial intelligence cardiovascular disease identification and diagnosis algorithm for classification of cardiovascular diseases, such as valvular heart disease, heart failure, coronary artery disease, and arrhythmia. Computer-aided disease prediction is also performed to reduce the time and accuracy of physicians in diagnosing cardiovascular diseases.

[0082] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can make modifications and changes to the above embodiments without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention should be as set forth in the claims.

Claims

1. A wearable cardiac ultrasound patch, characterized in that, The wearable cardiac ultrasound patch, designed to be attached to the skin in the heart region to obtain cardiac images, comprises: An encapsulation body having an adhesive layer on one side for attaching to the skin in the heart region; A fiber optic microsensor is housed within the package. The fiber optic microsensor includes a silicon photonic integrated circuit layer comprising a plurality of silicon photonic components arranged in a two-dimensional array, and a microelectromechanical integrated circuit layer comprising a plurality of piezoelectric components arranged in a two-dimensional array. The silicon photonic integrated circuit layer and the microelectromechanical integrated circuit layer are vertically stacked, with each silicon photonic component corresponding to each piezoelectric component. The silicon photonic integrated circuit layer faces the adhesive layer. The microelectromechanical integrated circuit layer is used to emit ultrasound waves that propagate to the heart region. When the ultrasound echo acts on each piezoelectric component and induces material deformation, the material deformation is captured by the silicon photonic components and converted into a phase change of an optical signal. The silicon photonic integrated circuit layer outputs the electrical signal converted from the optical signal. A computing chip is housed in the package and is perpendicularly stacked with the fiber optic microsensor. The computing chip uses 3D beamforming technology to drive each piezoelectric component to emit ultrasound waves and performs digital processing and artificial intelligence data analysis on the electrical signals output by the silicon photonic integrated circuit layer, so that the ultrasound beam is focused on different angles and depths of the heart, obtains ultrasound data from multiple angles, and generates raw data of cardiac ultrasound images.

2. The wearable cardiac ultrasound patch according to claim 1, characterized in that, The computing chip includes: An ultrasonic transducer driver is provided to supply the required voltage and current to multiple piezoelectric components of the microelectromechanical integrated circuit layer to control each piezoelectric component to emit the ultrasonic waves; An analog front-end circuit is used to amplify the electrical signal output from the silicon photonic integrated circuit layer, perform noise suppression and filtering to reduce noise, and quantize the electrical signal into a digital signal through analog-to-digital conversion; and A digital control system is used to receive the digital signal processed by the analog front-end circuit, perform digital signal processing and data parsing to generate the raw data of the cardiac ultrasound image, and store the raw data of the cardiac ultrasound image for subsequent reconstruction processing of the cardiac ultrasound image.

3. The wearable cardiac ultrasound patch according to claim 2, characterized in that, The computing chip further includes: a computing unit for performing artificial intelligence calculations, interpretation, and identification on the raw data of cardiac ultrasound images generated by the digital control system. When it is identified that the raw data of the cardiac ultrasound images does not contain valid data of the heart structure, feedback information is sent to the digital control system to control the ultrasonic transducer driver to adjust the parameters of the ultrasonic waves emitted by each piezoelectric component, so that the raw data of the cardiac ultrasound images generated by the digital control system contains valid data of the heart structure, thereby reflecting the image of the heart region and achieving the accuracy of subsequent cardiac ultrasound image reconstruction processing.

4. The wearable cardiac ultrasound patch according to claim 2, characterized in that, The digital control system also has a data transmission function, which is used to transmit the raw data of the generated cardiac ultrasound images.

5. The wearable cardiac ultrasound patch according to claim 4, characterized in that, It can also transmit data with an information processing device. The digital control system is used to transmit the raw data of the generated cardiac ultrasound image to the information processing device, or the digital control system receives control information from the information processing device. The control information includes adjusting the power, frequency and focusing mode of the ultrasound transducer driver to optimize the ultrasound parameter adjustment.

6. The wearable cardiac ultrasound patch according to claim 1, characterized in that, The computing chip performs an initialization operation when it first drives each of the piezoelectric components to emit ultrasound waves, in order to perform position calibration on the package and the heart region. The ultrasound reflection data of the heart region contained in the original data of the cardiac ultrasound image is used as the basis for position calibration, so as to adjust the angle and depth at which the ultrasound beam emitted by each of the piezoelectric components is focused on the heart.

7. The wearable cardiac ultrasound patch according to claim 1, characterized in that, The computing chip also has data transmission and storage functions to store the raw data of the generated echocardiogram for subsequent echocardiogram reconstruction processing, or to transmit the raw data of the echocardiogram to a cardiovascular disease identification system. The cardiovascular disease identification system has a multi-view identification network module and a cardiovascular disease detection network module. The multi-view identification network module stores multiple view identification models to perform echocardiogram reconstruction processing on the raw data of the echocardiogram to generate an echocardiogram to be identified. Based on the echocardiogram to be identified, the system finds the corresponding view identification model to obtain the echocardiogram view image and quality required for clinical diagnosis. The system can detect when the echocardiogram to be identified does not meet the required standards. When the quality of the cardiac ultrasound image or cardiac ultrasound image is consistent with the cardiac ultrasound perspective image, the computing chip drives each of the piezoelectric components to adjust the ultrasound parameters of the three-dimensional beamforming to transmit ultrasound waves, so as to obtain the original data of cardiac ultrasound images at different focal positions, until the cardiac ultrasound image to be identified obtained based on the subsequently generated original data conforms to the cardiac ultrasound perspective image and cardiac ultrasound image quality; the cardiovascular disease detection network module has a cardiac ultrasound signal database, which is used to analyze the cardiac ultrasound image to be identified that conforms to the cardiac ultrasound perspective image and cardiac ultrasound image quality, to extract feature parameters from the analyzed cardiac ultrasound image to be identified, and to classify cardiac ultrasound images and cardiovascular diseases through the cardiac ultrasound signal database and cardiac ultrasound algorithm, so as to provide a cardiovascular disease identification report.

8. A cardiovascular disease identification system, characterized in that, It can perform data transmission processing with the wearable cardiac ultrasound patch as described in claim 1, and the computing chip of the wearable cardiac ultrasound patch has a data transmission function to transmit the raw data of the cardiac ultrasound image generated by the computing chip to the cardiovascular disease identification system, the cardiovascular disease identification system comprising: A multi-view recognition network module is used to perform cardiac ultrasound image reconstruction processing on the raw data of the cardiac ultrasound images to generate a cardiac ultrasound image to be identified. The multi-view recognition network module stores multiple view recognition models, each of which has at least one cardiac ultrasound view image and its corresponding cardiac ultrasound image quality. Based on the cardiac ultrasound image to be identified, the corresponding view recognition model is found to determine the cardiac ultrasound view image and cardiac ultrasound image quality required for clinical diagnosis. When it is determined that the cardiac ultrasound image to be identified does not conform to the cardiac ultrasound view image or cardiac ultrasound image quality, the computing chip drives each piezoelectric component to adjust the ultrasound parameters of the three-dimensional beamforming to transmit ultrasound waves, thereby obtaining raw data of cardiac ultrasound images at different focal positions. This process continues until the cardiac ultrasound image to be identified obtained based on subsequently generated raw data conforms to the cardiac ultrasound view image and cardiac ultrasound image quality. A cardiovascular disease detection network module has a cardiac ultrasound signal database for analyzing cardiac ultrasound images that conform to the cardiac ultrasound perspective and image quality. It extracts feature parameters from the analyzed cardiac ultrasound images and classifies cardiac ultrasound images and cardiovascular diseases according to the cardiac ultrasound signal database and cardiac ultrasound algorithm to provide a cardiovascular disease identification report.

9. The cardiovascular disease identification system according to claim 8, characterized in that, The cardiovascular disease detection network module also has a cardiac audio signal database and receives sound signals generated by heart activity captured by a phonocardiography (PCG) detection device, so as to analyze and diagnose the function and health status of the heart based on the cardiac audio signal database.

10. The cardiovascular disease identification system according to claim 8, characterized in that, The cardiovascular disease detection network module also includes an electrocardiogram (ECG) signal database and receives ECG signals detected by an electrocardiogram (ECG) detection device to monitor heart rhythm and diagnose heart-related diseases based on the ECG signal database.

11. The cardiovascular disease identification system according to claim 8, characterized in that, When the number of wearable cardiac ultrasound patches is three, each wearable cardiac ultrasound patch can cover a 120-degree emission range to obtain 360-degree cardiac ultrasound imaging. Through the multi-view recognition network module, the computing chip controls the angle and depth of cardiac ultrasound emission based on the multi-view recognition model.

12. A method for identifying cardiovascular diseases using the wearable cardiac ultrasound patch of claim 1, based on the electrical connector according to claim 11, characterized in that... The wearable cardiac ultrasound patch transmits data to a cardiovascular disease identification system to achieve cardiovascular disease identification. The cardiovascular disease identification method includes the following steps: (1) The computing chip of the wearable cardiac ultrasound patch transmits the raw data of the generated cardiac ultrasound image to the cardiovascular disease identification system. (2) The cardiovascular disease identification system receives the raw data of the cardiac ultrasound image transmitted by the computing chip and performs cardiac ultrasound image reconstruction processing to generate a cardiac ultrasound image to be identified. (3) The cardiovascular disease identification system determines whether the cardiac ultrasound image to be identified meets the cardiac ultrasound viewing angle and cardiac ultrasound image quality required for clinical diagnosis. When it is determined that the cardiac ultrasound image to be identified does not meet the cardiac ultrasound viewing angle or cardiac ultrasound image quality, the computing chip drives each of the piezoelectric components to adjust the ultrasound parameters of the three-dimensional beamforming to transmit ultrasound waves, thereby generating raw data of cardiac ultrasound images at different focal positions. This continues until the cardiac ultrasound image to be identified generated by the cardiovascular disease identification system through image reconstruction processing based on the raw data subsequently transmitted from the computing chip meets the cardiac ultrasound viewing angle and cardiac ultrasound image quality. (4) The cardiovascular disease identification system analyzes the cardiac ultrasound images that conform to the cardiac ultrasound perspective image and cardiac ultrasound image quality, and extracts feature parameters from the analyzed cardiac ultrasound images. Based on the cardiac ultrasound signal database and cardiac ultrasound algorithm, cardiac ultrasound images are classified with cardiovascular diseases to provide a cardiovascular disease identification report.

13. The cardiovascular disease identification method according to claim 12, characterized in that, When three wearable cardiac ultrasound patches are placed on the skin in the heart region, they are respectively placed beside the sternum, at the apex of the heart, and below the ribs, so that each wearable cardiac ultrasound patch can cover a 120-degree emission range to obtain 360-degree cardiac ultrasound imaging.

14. The cardiovascular disease identification method according to claim 12, characterized in that, The ultrasonic parameters include: adjusting the phase difference and amplitude difference between adjacent piezoelectric components, and changing the emission delay time of the piezoelectric components.

15. The cardiovascular disease identification method according to claim 12, characterized in that, Before performing step (1), the following steps are included: (1-1) The computing chip of the wearable cardiac ultrasound patch interprets the raw data of the generated cardiac ultrasound image; (1-2) After the computing chip interprets the data, it analyzes whether the original data of the cardiac ultrasound image has valid data of the cardiac structure. If it does not have valid data of the cardiac structure, it proceeds to step (1-3); if it has valid data of the cardiac structure, it proceeds to step (1-4). (1-3) When the computing chip determines that the raw data of the cardiac ultrasound image does not contain valid data on the heart structure, it adjusts the parameters of the ultrasound emitted by each piezoelectric component so that the raw data of the cardiac ultrasound image subsequently generated by the computing chip contains valid data on the heart structure, and then performs the above step (1); and (1-4) When the computing chip analyzes the raw data of the cardiac ultrasound image and finds that it contains valid data of the cardiac structure, then the above step (1) is performed.

Images