CN211300047U

CN211300047U - Intelligent cardiopulmonary auscultation system based on neural network

Info

Publication number: CN211300047U
Application number: CN201921761066.7U
Authority: CN
Inventors: 李海霞; 吴云飞; 朱慧博
Original assignee: Suqian College
Current assignee: Suqian College
Priority date: 2019-10-21
Filing date: 2019-10-21
Publication date: 2020-08-21
Anticipated expiration: 2029-10-21

Abstract

The utility model relates to an intelligent heart-lung auscultation system based on a neural network, which comprises a micro-sound acquisition module, a power supply module, a signal processing module, a PC analysis module and a server; the micro sound acquisition module acquires heart sound and breath sound signals of a human body; the power supply module converts commercial power into stable positive and negative direct current and outputs the stable positive and negative direct current; the signal processing module processes the acquired data and transmits the processed data to a computer (PC); the PC analysis module performs software analysis on the signals and predicts the signals by using a neural network algorithm; the server creates a database, stores and screens the uploaded data, and continuously trains and releases a new neural network by using the new data for downloading by a PC. The utility model supports the collection of heart sound and breath sound, and can update data under the condition of networking, thus ensuring the accuracy of data analysis; and data analysis and processing under the situation of no network are supported, and disease condition prediction is made.

Description

Intelligent cardiopulmonary auscultation system based on neural network

Technical Field

The utility model relates to a domestic little electrical apparatus technical field, in particular to intelligence cardiopulmonary auscultation system based on neural network.

Background

China is a developing country, 9 hundred million people live in rural areas, and the condition of shortage of medical services and medicines in the rural areas is not fundamentally changed, so that diagnosis and treatment effect evaluation of many cardiopulmonary diseases still depend on a stethoscope to a great extent. Even the initial impression diagnosis and targeted further instrumental examinations by first-visit physicians in large hospitals, even in cardiology-specific hospitals, in cities, are determined by medical history and physical examinations, including those seen in cardiopulmonary auscultation.

Heart and breath sounds are important physiological signals of the human body: the heart sound is a general name of sound generated in the mechanical movement of the heart and the cardiovascular system, comprises physiological and pathological information of each part of the heart and interaction of each part of the heart, has important value in cardiovascular diseases, is an important method for noninvasive detection of the cardiovascular diseases, and has the irreplaceable advantages of electrocardiogram and ultrasonic electrocardiogram; the respiratory sound is a general term for the sound generated by the respiratory system during ventilation movement, and includes physiological and pathological information of the lung. The traditional stethoscope only relies on sound recognition, has low resolution, can be qualitatively analyzed and judged only by personal experience of doctors, and cannot be effectively used by people without relevant medical knowledge; the existing digital electronic stethoscope does not depend on medical knowledge, and can support the functions of recording, storing and playing the auscultated cardiopulmonary sound data, but the price is too expensive to popularize. In addition, auscultation products on the market at present mostly need upload data to data center through the network, rely on remote doctor to its analysis, and the limitation is more, the timeliness is relatively poor.

The patent with publication number CN102697520B discloses an electronic stethoscope based on intelligent recognition function, which comprises a processor unit, and a signal acquisition unit, a peripheral driving unit, a storage unit and a data bus interface unit connected with the processor unit, wherein the signal acquisition unit acquires heart sound and breath sound signals and performs preprocessing on the heart sound and breath sound signals, the processor unit specifically realizes a mode recognition algorithm for the heart sound and the breath sound and separates the heart sound and the breath sound, and completes intelligent recognition and classification of the heart sound and breath sound signals, and manages other hardware units; the storage unit is used for storing the program and the extension program thereof, storing the heart sound and breathing sound data, the standard heart sound and breathing sound and the listening patterns of each typical case of diseases and outputting and playing the listening patterns; the peripheral driving unit and the data bus interface unit are used for realizing peripheral operation function driving and data communication. The electronic stethoscope passes through

The portable function is realized by additionally arranging the central processing chip, but the price is high due to the additionally arranged central processing chip, and the portable function is not beneficial to common household.

SUMMERY OF THE UTILITY MODEL

The utility model discloses to not enough in the background art, provided one kind and be applicable to domestic, low price based on neural network intelligence cardiopulmonary auscultation system, increased the accuracy that the analysis result judged simultaneously.

In order to realize the purpose, the utility model discloses a technical scheme as follows:

an intelligent cardiopulmonary auscultation system based on a neural network is characterized by comprising a micro-sound acquisition module, a power supply module, a signal processing module, a PC analysis module and a remote server; the micro sound collecting module comprises a stethoscope and a high-sensitivity sound pick-up which are sequentially connected and is used for collecting heart sound and breath sound signals of a human body; the signal processing module is connected with the micro-sound acquisition module, comprises a pre-amplification circuit, a band-pass filter circuit and an amplification lifting circuit which are sequentially connected and is used for amplifying, filtering and the like the acquired signals; the PC analysis module is provided with a sound card, and the sound card is connected with the signal analysis module and used for converting the analog signals into digital signals; the server is connected with the PC analysis module and is used for analyzing and predicting the signals output by the signal analysis module by using a neural network algorithm, creating a database, storing and screening the uploaded data, and continuously training a new neural network by using new data under the condition of network existence; the power supply module is connected with the signal processing module and is used for converting commercial power into stable positive and negative direct current to be output.

Preferably, the high-sensitivity sound pick-up in the micro-sound acquisition module is WM-034 CZ.

Preferably, the pre-amplification circuit in the signal processing module is NE5542 AP.

Preferably, the band-pass filtering circuit in the signal processing module comprises two TLC04 and one LM358 AP.

Preferably, the amplifying and lifting circuit in the signal processing module comprises an OP07CP and an LM 386.

Preferably, the power supply module comprises a transformer, a rectifying circuit, a filtering circuit and a voltage stabilizing and reducing circuit.

Further, the power supply module includes two 3N246, two LM7909CT, and two LM7905 CT.

Compared with the prior art, the utility model discloses the beneficial effect who has:

(1) the utility model discloses increased the enlarged lifting circuit at the signal processing part, the lifting circuit signal can be discerned by PC's sound card and is converted into digital signal by analog signal through enlarging.

(2) The utility model discloses directly be connected with the PC, need not central processing chip, simple easy-to-use, low cost, small in size.

(3) The method can detect the abnormalities of heart and lung noise, extra sound and the like in real time by utilizing local application algorithm analysis without means of network communication, remote assistance and the like, judge the abnormality types and predict the causes.

(4) In order to increase the accuracy of judgment, the server continuously obtains new samples from the user, improves the reliability of the neural network by enlarging the number of the samples, and updates the neural network sample library in real time when the user obtains the networking; in addition, the trained neural network is only updated to the PC application module, and the storage space and the operation resource of the PC of the user are not occupied.

Drawings

Fig. 1 is a structural diagram of the present invention.

Fig. 2 is a schematic block diagram of the present invention.

Fig. 3 is a circuit diagram of the power supply module of the present invention.

Fig. 4 is a front-end amplification circuit diagram of the present invention.

Fig. 5 is a circuit diagram of the band-pass filter of the present invention.

Fig. 6 is a circuit diagram of the level lifting circuit of the present invention.

In the figure: 1. the stethoscope probe, 2, high sensitivity adapter, 3, signal processing module, 4, preamplification circuit, 5, band-pass filter circuit, 6, the lifting circuit that amplifies, 7, sound card, 8, PC analysis module, 9, remote server, 10, power supply module.

Detailed Description

The present invention will be further described with reference to the following examples and accompanying drawings.

As shown in fig. 1, an intelligent cardiopulmonary auscultation system based on a neural network comprises a micro sound collecting module, a power supply module 10, a signal processing module 3, a PC analyzing module 8 and a remote server 9. The micro sound collecting module is formed by connecting a stethoscope probe 1 with a high-sensitivity sound pickup 2 and is connected with the signal processing module 3; the signal processing module 3 consists of a preamplifier circuit 4, a band-pass filter circuit 5 and an amplifying and lifting circuit 6 which are connected in sequence; the signal processing module 3 is connected with a sound card 7 in the PC analysis module 8; the server 9 is connected with the PC analysis module 8 and is used for algorithm updating and data interaction; the power supply module 10 is connected with the signal processing module 3.

The specific process comprises the following steps:

as shown in fig. 2, the heart sound signal or the breath sound signal of a human body is collected by a micro sound collecting module consisting of a stethoscope and a high-sensitivity sound pickup device, the heart sound signal or the breath sound signal collected by the micro sound collecting module is weak, a pre-amplification circuit (shown in fig. 4) is required to amplify the weak heart sound signal or the breath sound signal, in order to remove the interference of other noises and obtain the heart sound signal and the breath sound signal which are low-frequency and audible to human ears, a band-pass filter circuit (shown in fig. 5) is arranged to keep the signal frequency at 20-700Hz, and in order to enable the signal to meet the requirement of a/D conversion, an amplification lifting circuit (shown in fig. 6) is arranged to enable the signal to be identified by a sound card of a PC and converted into a digital. The PC analysis module obtains the digital signal transmitted by the sound card, then stores the data, locally analyzes the signal through an application algorithm, extracts the characteristic value in a single period of the heart sound or the breath sound, judges whether the heart sound or the breath sound is abnormal or not by utilizing a neural network, obtains information such as abnormal types and the like, and can realize updating under the condition that the algorithm is in a network state. And finally, displaying the analyzed signal waveform and the obtained result to achieve the effect of observing heart sound and breath sound for a long time and remind the user whether to seek medical advice in time.

As further shown in fig. 3, the power supply module includes a transformer, a rectifying circuit, a filtering circuit, and a voltage stabilizing and dropping circuit.

The transformer firstly reduces the voltage of a 220V power grid into 12V alternating voltage, then the alternating current is changed into positive and negative 12V direct current through the rectifying circuit, power frequency interference is eliminated through the filter circuit, a voltage curve is smooth and stable, finally the positive and negative 12V direct current is changed into stable and reliable positive and negative 9V and positive and negative 5V through the voltage reduction and voltage stabilization circuit, stable positive and negative direct current is provided for a chip of the signal processing module, and the acquired signals are more accurate.

Finally, it should be noted that: the above embodiments are only used for illustrating the present invention and do not limit the technical solution described in the present invention; thus, while the present invention has been described in detail with reference to the various embodiments thereof, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted; all such modifications and variations are intended to be included herein within the scope of this disclosure and the present invention and within the scope and spirit of the following claims.

Claims

1. An intelligent cardiopulmonary auscultation system based on a neural network is characterized by comprising a micro-sound acquisition module, a power supply module, a signal processing module, a PC analysis module and a remote server; the micro sound collecting module comprises a stethoscope and a high-sensitivity sound pick-up which are sequentially connected and is used for collecting heart sound and breath sound signals of a human body; the signal processing module is connected with the micro-sound acquisition module, comprises a pre-amplification circuit, a band-pass filter circuit and an amplification lifting circuit which are sequentially connected and is used for amplifying and filtering acquired signals; the PC analysis module is provided with a sound card, and the sound card is connected with the signal analysis module and used for converting the analog signals into digital signals; the server is connected with the PC analysis module and is used for analyzing and predicting the signals output by the signal analysis module by using a neural network algorithm, creating a database, storing and screening the uploaded data, and continuously training a new neural network by using new data under the condition of network existence; the power supply module is connected with the signal processing module and is used for converting commercial power into stable positive and negative direct current to be output.

2. The intelligent neural network-based cardiopulmonary auscultation system of claim 1, wherein: and a high-sensitivity sound pickup in the micro sound acquisition module is WM-034 CZ.

3. The intelligent neural network-based cardiopulmonary auscultation system of claim 1, wherein: the pre-amplifying circuit in the signal processing module is NE5542 AP.

4. The intelligent neural network-based cardiopulmonary auscultation system of claim 1, wherein: the bandpass filtering circuit in the signal processing module includes two TLC04 and one LM358 AP.

5. The intelligent neural network-based cardiopulmonary auscultation system of claim 1, wherein: the amplifying and lifting circuit in the signal processing module comprises an OP07CP and an LM 386.

6. The intelligent neural network-based cardiopulmonary auscultation system of claim 1, wherein: the power supply module comprises a transformer, a rectifying circuit, a filter circuit and a voltage stabilizing and reducing circuit.

7. The intelligent neural network-based cardiopulmonary auscultation system of claim 6, wherein: the power supply module includes two 3N246, two LM7909CT, and two LM7905 CT.