Intelligent wearable heart sound detection system

Abstract

The invention relates to an intelligent wearable heart sound detection system, and belongs to the technical field of sensors. The intelligent wearable heart sound detection system comprises a wearable high-sensitivity sensor, a signal processing system, a machine learning system and an AI data processing system. The wearable high-sensitivity sensor collects heart sound of a user to generate a sensor signal; the signal processing system processes the sensor signal to generate a heart sound signal; the machine learning system establishes and updates a heart sound computer model through a machine learning algorithm; and the AI data processing system draws a phonocardiogram reflecting heart sound characteristics based on the heart sound computer model and the heart sound signals. Therefore, cardiac murmur can be visually reflected on the phonocardiogram to serve as a reference basis for a user and a doctor to judge the heart health condition, meanwhile, the wearable high-sensitivity sensor is adopted so that cardiovascular examination can be conducted through daily wearing, the use method is convenient, and the application range is quite wide.

Description

Technical Field

[0001] The present invention relates to the field of sensor technology, in particular to the field of heart sound detection technology, and specifically refers to an intelligent wearable heart sound detection system. Background Art

[0002] Cardiovascular disease ranks first in the mortality rate of diseases in my country, and the number of cardiovascular disease patients in China has reached 110 million. One in every five adults suffers from cardiovascular disease, and one in every 13 cardiovascular disease patients suffers from acute myocardial infarction. Daily examination and early screening of cardiovascular diseases are of great significance.

[0003] At present, routine cardiovascular examinations include auscultation, phonocardiogram, electrocardiogram, and cardiac CT in large hospitals. Auscultation and phonocardiogram can judge the health of the heart based on the doctor's experience and ability by analyzing heart murmurs. Statistics from the international medical community show that experienced cardiologists can make judgments with an accuracy rate of 80% through auscultation and analysis of phonocardiograms. In non-central cities and remote areas, the accuracy rate is much lower than this ratio.

[0004] The electrocardiogram, which has a high penetration rate, is very effective in diagnosing cardiovascular diseases, but is difficult to apply to routine cardiovascular examinations.

[0005] Therefore, how to provide a heart sound detection system that can be used in daily cardiovascular examinations and reflect heart murmurs has become an urgent problem to be solved in this field. Summary of the invention

[0006] The purpose of the present invention is to overcome the shortcomings of the above-mentioned prior art and provide an intelligent wearable heart sound detection system that is easy to use for daily cardiovascular examinations, reflects heart murmurs, and has a convenient method of use.

[0007] In order to achieve the above-mentioned purpose, the intelligent wearable heart sound detection system of the present invention has the following structure:

[0008] The intelligent wearable heart sound detection system includes: a wearable high-sensitivity sensor, a signal processing system, a machine learning system and an AI data processing system.

[0009] Among them, the wearable high-sensitivity sensor collects the user's heart sounds and generates sensor signals;

[0010] The signal processing system filters and amplifies the sensor signal to generate a heart sound signal;

[0011] The machine learning system is based on the international heart sound library standard data and the historical data of heart sound signals collected, and uses machine learning algorithms to establish and update the heart sound computer model;

[0012] The AI ​​data processing system draws a phonocardiogram reflecting the characteristics of the heart sounds based on the heart sound computer model and the heart sound signal.

[0013] In the intelligent wearable heart sound detection system, the wearable high-sensitivity sensor is a piezoelectric sensor, which is arranged on the clothing near the heart.

[0014] In the intelligent wearable heart sound detection system, the piezoelectric sensor is any one or more of a piezoelectric fiber sensor, a PVDF piezoelectric sensor and a piezoelectric film electret sensor.

[0015] In the intelligent wearable heart sound detection system, the piezoelectric fiber sensor is a cable-shaped sensor, which includes a coaxial structure consisting of a core electrode, a dielectric layer, a shielding layer and a self-shielding sheath, and at least a portion of the clothing is woven by the cable-shaped sensor.

[0016] In the intelligent wearable heart sound detection system, the clothing is a vest, a bra or a neck cover.

[0017] In the intelligent wearable heart sound detection system, the signal processing system acquires the sensor signal through multi-point matrix signal acquisition.

[0018] In the intelligent wearable heart sound detection system, the machine learning system is specifically used to classify the collected historical data of heart sound signals according to their heart sound murmur types based on the international heart sound library standard data, and to establish and update the classified heart sound computer model associated with the murmur type through a machine learning algorithm.

[0019] In the intelligent wearable heart sound detection system, the AI ​​data processing system is specifically used to select a classified heart sound computer model corresponding to the current heart sound signal; adopt the murmur standard graph of the classified heart sound computer model, and in the phonocardiogram of the current heart sound signal, draw a part that meets the murmur standard graph according to the murmur standard graph.

[0020] In the intelligent wearable heart sound detection system, when the current heart sound signal conforms to more than one classified heart sound computer model, the machine learning system is used to establish a new comprehensive type of heart sound murmur based on the international heart sound library standard data, and the comprehensive type of heart sound murmur is associated with the original heart sound murmur type, and a murmur standard graph of a new comprehensive type of heart sound murmur is formed based on the associated heart sound murmur type.

[0021] The intelligent wearable heart sound detection system also includes: a personal terminal, which is connected to the wearable high-sensitivity sensor for data signals, and the signal processing system, machine learning system and AI data processing system are all arranged in the personal terminal.

[0022] The intelligent wearable heart sound detection system also includes: a network server, which is connected to the personal terminal via a network, and is used to store the international heart sound library standard data, heart sound signal historical data, heart sound computer model, heart sound murmur type and heart sound map.

[0023] The intelligent wearable heart sound detection system using this invention includes: a wearable high-sensitivity sensor, a signal processing system, a machine learning system and an AI data processing system. The wearable high-sensitivity sensor collects the user's heart sounds to generate a sensor signal; the sensor signal is processed by the signal processing system to generate a heart sound signal; the machine learning system establishes and updates the heart sound computer model through a machine learning algorithm; the AI ​​data processing system draws a heart sound graph that reflects the characteristics of the heart sound based on the heart sound computer model and the heart sound signal. In this way, the heart murmur can be intuitively reflected on the heart sound graph, which can serve as a reference for users and doctors to judge the health of the heart. At the same time, the use of a wearable high-sensitivity sensor allows for daily cardiovascular examinations, which is easy to use and has a wide range of applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 Schematic diagram of the structural modules of the intelligent wearable heart sound detection system of the present invention. DETAILED DESCRIPTION

[0025] In order to more clearly understand the technical content of the present invention, the following embodiments are given in detail.

[0026] See also Figure 1 FIG. 1 is a schematic diagram of a structure module of the intelligent wearable heart sound detection system of the present invention.

[0027] In one embodiment, the intelligent wearable heart sound detection system includes: a wearable high-sensitivity sensor, a signal processing system, a machine learning system, and an AI data processing system.

[0028] The wearable high-sensitivity sensor collects the user's heart sounds and generates sensor signals. The wearable high-sensitivity sensor is a piezoelectric sensor, which is arranged on the clothing near the heart.

[0029] The signal processing system acquires the sensor signal through multi-point matrix signal acquisition, and filters and amplifies the sensor signal to generate a heart sound signal;

[0030] The machine learning system is based on the international heart sound library standard data and the historical data of heart sound signals collected, and uses machine learning algorithms to establish and update the heart sound computer model;

[0031] The AI ​​data processing system draws a phonocardiogram reflecting the characteristics of the heart sounds based on the heart sound computer model and the heart sound signal.

[0032] In a preferred embodiment, the piezoelectric sensor is any one or more of a piezoelectric fiber sensor, a PVDF piezoelectric sensor, and a piezoelectric film electret sensor. The piezoelectric fiber sensor is a cable-shaped sensor, which includes a coaxial structure consisting of a core electrode, a dielectric layer, a shielding layer, and a self-shielding sheath, and at least a portion of the clothing is woven from the cable-shaped sensor. The woven clothing can be a vest, a bra, or a neck cover, etc.

[0033] The signal processing system filters the sensor signal, performs Butterworth bandpass filtering on the heart sound signal to reduce noise, sets the passband of the heart sound signal to 30-150 Hz, and performs bidirectional zero phase shift filtering on the heart sound signal to avoid the influence of phase shift on the propagation relationship.

[0034] In a further preferred embodiment, the machine learning system is specifically used to classify the collected heart sound signal history data according to its heart sound murmur type based on the international heart sound library standard data, and establish and update the classified heart sound computer model associated with the murmur type through the machine learning algorithm. The heart sound murmur type includes systolic murmur, diastolic murmur and continuous murmur at various positions of the heart.

[0035] In another further preferred embodiment, the AI ​​data processing system is specifically used to select a classified heart sound computer model corresponding to the current heart sound signal; use the murmur standard graph of the classified heart sound computer model, and in the phonocardiogram of the current heart sound signal, draw a portion that meets the murmur standard graph according to the murmur standard graph. Specifically, a portion of the graph in the phonocardiogram of the current heart sound signal is matched with the murmur standard graph, and the portion that is identical or similar to the murmur standard graph is highlighted in a distinguishing manner.

[0036] In the intelligent wearable heart sound detection system, when the current heart sound signal conforms to more than one classified heart sound computer model, the machine learning system is used to establish a new comprehensive type of heart sound murmur based on the international heart sound library standard data, and the comprehensive type of heart sound murmur is associated with the original heart sound murmur type, and a murmur standard graph of a new comprehensive type of heart sound murmur is formed based on the associated heart sound murmur type.

[0037] In a more preferred embodiment, the intelligent wearable heart sound detection system also includes: a personal terminal and a network server. The personal terminal can be a mobile phone, a tablet, a PC, etc., or it can be other wearable devices such as a bracelet watch. The personal terminal is connected to the wearable high-sensitivity sensor to achieve data signal connection, and the signal processing system, the machine learning system and the AI ​​data processing system are all arranged in the personal terminal. The network server is connected to the personal terminal via a network, and the network server is used to store the international heart sound library standard data, heart sound signal historical data, heart sound computer model, heart sound murmur type and heart sound map.

[0038] In practical applications, the high-sensitivity sensors in the present invention can be respectively used in the following three types:

[0039] 1. Piezoelectric fiber sensor

[0040] The piezoelectric fiber sensor is a coaxial cable-shaped sensor consisting of a core electrode (signal electrode), a dielectric layer, a shielding layer, and a sheath self-shielding. When the piezoelectric fiber sensor is squeezed or bent by an external force, the amount of charge or voltage generated by the sensor is proportional to the applied pressure.

[0041] The piezoelectric fiber sensor can work for a long time at a temperature of -79℃ to 120℃, and its service life at room temperature is more than 30 years; the self-shielding structure enables the sensor to adapt to complex electromagnetic environments; the waterproofness ensures that the sensor can withstand washing or underwater environments. According to actual needs, the piezoelectric fiber sensor can be cut into any length, and a soft, breathable, washable smart fabric can be obtained through a weaving process.

[0042] Based on the use of this sensor, a multi-point matrix signal acquisition system is used to detect the noises in various parts of the heart at multiple points, effectively extract weak heart sound signals, and output the heart sound signals to the machine learning algorithm end through a self-developed algorithm.

[0043] Fabrics made of piezoelectric fibers can be made into vests, bras, and neck covers to achieve the purpose of multi-point signal collection.

[0044] 2. PVDF piezoelectric sensor, combined with self-developed signal acquisition, filtering, and amplification circuit systems, compares the standard data of the international heart sound library, develops an algorithm to automatically compare and effectively collect heart murmur signals, and outputs them to the machine learning algorithm end.

[0045] 3. Piezoelectric film electret sensor, the application method is the same as PVDF sensor.

[0046] The signal acquisition and processing system develops a signal acquisition and processing system based on different sensors. Through complex algorithms, it compares the heart sound signals in the standard database, collects weak heart sound signals, shields and filters noise, and outputs valid heart sound signals to the machine learning data end.

[0047] The AI ​​algorithm heart sound signal processing and diagnosis system, based on providing effective heart sound signals, establishes a computer model and draws a heart sound graph. Through a machine learning system supported by big data, AI intelligently analyzes the murmur type corresponding to the heart murmur and reflects it on the heart sound graph through different colors or distinctive labels.

[0048] The backend data center is used to establish a backend database, accumulate heart sound data, and analyze the heart sound data of different populations and regions through big data models, providing users and doctors with an analytical basis for cardiovascular health status.

[0049] The user's personal terminal can be a mobile phone, tablet, PC, etc. running a specific APP application, or other wearable devices such as a bracelet and a watch. The personal terminal is connected to the wearable high-sensitivity sensor to achieve data signal connection. The signal processing system, machine learning system and AI data processing system can all be set in the personal terminal, or they can be run by the background data center. The user's personal terminal also displays the phonocardiogram generated by the intelligent wearable phonocardiogram detection system of the present invention, especially on the phonocardiogram, different colors or different markings are used to reflect the noise and display the corresponding noise type for user reference.

[0050] The intelligent wearable heart sound detection system using this invention includes: a wearable high-sensitivity sensor, a signal processing system, a machine learning system and an AI data processing system. The wearable high-sensitivity sensor collects the user's heart sounds to generate a sensor signal; the sensor signal is processed by the signal processing system to generate a heart sound signal; the machine learning system establishes and updates the heart sound computer model through a machine learning algorithm; the AI ​​data processing system draws a heart sound graph that reflects the characteristics of the heart sound based on the heart sound computer model and the heart sound signal. In this way, the heart murmur can be intuitively reflected on the heart sound graph, which can serve as a reference for users and doctors to judge the health of the heart. At the same time, the use of a wearable high-sensitivity sensor allows for daily cardiovascular examinations, which is easy to use and has a wide range of applications.

[0051] In this specification, the present invention has been described with reference to specific embodiments thereof. However, it is apparent that various modifications and variations may be made without departing from the spirit and scope of the present invention. Therefore, the specification and drawings should be regarded as illustrative rather than restrictive.

Claims

1. An intelligent wearable heart sound detection system, characterized in that: include: Wearable high-sensitivity sensor collects user's heart sounds and generates sensor signals; A signal processing system, which filters and amplifies the sensor signal to generate a heart sound signal; The machine learning system builds and updates the heart sound computer model based on the international heart sound library standard data and the historical data of the collected heart sound signals, and through machine learning algorithms; The AI ​​data processing system draws a phonocardiogram reflecting the characteristics of the heart sounds based on the heart sound computer model and the heart sound signal.

2. The intelligent wearable heart sound detection system according to claim 1, characterized in that: The wearable high-sensitivity sensor is a piezoelectric sensor, which is arranged on the clothing near the heart.

3. The intelligent wearable heart sound detection system according to claim 2, characterized in that: The piezoelectric sensor is any one or more of a piezoelectric fiber sensor, a PVDF piezoelectric sensor and a piezoelectric film electret sensor.

4. The intelligent wearable heart sound detection system according to claim 3, characterized in that: The piezoelectric fiber sensor is a cable-shaped sensor, which includes a coaxial structure consisting of a core electrode, a dielectric layer, a shielding layer and a self-shielding sheath. At least a portion of the clothing is woven from the cable-shaped sensor, and the clothing is a vest, a bra or a neck cover.

5. The intelligent wearable heart sound detection system according to claim 1, characterized in that: The signal processing system acquires the sensor signal through multi-point matrix signal acquisition.

6. The intelligent wearable heart sound detection system according to claim 1, characterized in that: The machine learning system is specifically used to classify the collected historical data of heart sound signals according to their heart sound murmur types based on the standard data of the International Heart Sound Library, and to establish and update a classified heart sound computer model associated with the murmur type through a machine learning algorithm.

7. The intelligent wearable heart sound detection system according to claim 6, characterized in that: The AI ​​data processing system is specifically used to select a classified heart sound computer model corresponding to the current heart sound signal; adopt the murmur standard graph of the classified heart sound computer model, and in the heart sound graph of the current heart sound signal, draw the part that meets the murmur standard graph according to the murmur standard graph.

8. The intelligent wearable heart sound detection system according to claim 7, characterized in that: When the current heart sound signal conforms to more than one classified heart sound computer model, the machine learning system is used to establish a new comprehensive heart sound murmur type based on the international heart sound library standard data, and the comprehensive heart sound murmur type is associated with the original heart sound murmur type, and a murmur standard graph of the new comprehensive heart sound murmur type is formed based on the associated heart sound murmur type.

9. The intelligent wearable heart sound detection system according to claim 8, characterized in that: Also includes: A personal terminal, which realizes data signal connection with the wearable high-sensitivity sensor, and the signal processing system, machine learning system and AI data processing system are all arranged in the personal terminal.

10. The intelligent wearable heart sound detection system according to claim 9, characterized in that: Also includes: A network server is connected to the personal terminal via a network, and is used to store the international heart sound library standard data, heart sound signal history data, heart sound computer model, heart sound murmur type and heart sound map.

Images