Auscultation system based on Internet of Things, control method and equipment of auscultation system and storage medium

Abstract

The invention discloses an auscultation system based on the Internet of Things, a control method and device thereof and a storage medium, and relates to a general control system.The auscultation system based on the Internet of Things comprises an auscultation terminal, a transmission module, an Internet of Things box, a management and control platform and a remote server, and the auscultation terminal is used for collecting audio signals of a user; the transmission module is used for performing analog-to-digital conversion on the audio signal to obtain a digital audio signal and uploading the digital audio signal to the management and control platform and the Internet of Things box; the Internet of Things box is used for generating a first analysis result according to the received digital audio signal; the management and control platform is used for forwarding the digital audio signal to the remote server and updating a first model of the Internet of Things box; and the remote server is used for receiving the digital audio signal and generating a second analysis result according to the digital audio signal. According to the invention, the technical effect of remote high-fidelity diagnosis can be realized.

Description

Technical Field

[0001] This application relates to the field of general control system technology, and in particular to an Internet of Things-based auscultation system and its control method, device and storage medium. Background Technology

[0002] With the continuous advancement of modern medical technology, auscultation, as a fundamental and important clinical diagnostic method, plays a crucial role in the initial screening and diagnosis of diseases. Stethoscopes primarily rely on air conduction for sound, and doctors listen to physiological sounds such as heart and lung sounds and respiratory sounds through headphones, making judgments based on experience. However, users need to visit a medical institution for diagnosis, preventing them from having real-time access to information about their own condition. Summary of the Invention

[0003] The main purpose of this application is to provide an Internet of Things-based auscultation system and its control method, device and storage medium, which aims to solve the technical problem that users need to go to an institution for diagnosis, resulting in users not being able to know their own status in real time.

[0004] To achieve the above objectives, this application provides an Internet of Things (IoT)-based auscultation system, which includes an auscultation terminal, a transmission module, an IoT box, a control platform, and a remote server. The auscultation terminal is used to collect the user's audio signals. The transmission module is used to convert the audio signal from analog to digital and upload it to the control platform and the IoT box. The IoT box is used to generate a first analysis result based on the received digital audio signal; The control platform is used to forward the digital audio signal to the remote server and update the first model of the IoT box; The remote server is used to receive digital audio signals and generate a second analysis result based on the digital audio signals.

[0005] In one embodiment, the transmission module includes an auscultation sound acquisition and control submodule and an auscultation sound receiving and processing submodule. The auscultation sound acquisition and control submodule includes an audio acquisition submodule and a signal transmitting end. The auscultation sound receiving and processing submodule includes a signal receiving end and an adapter. The audio acquisition submodule is connected to the auscultation terminal.

[0006] In one embodiment, the IoT-based auscultation system further includes an analysis device. The adapter is connected to the analysis device, which is a mobile terminal, a mobile consultation vehicle, or a medical operation terminal. The analysis device calls the IoT box to analyze the digital audio signal and generate a first analysis result.

[0007] In one embodiment, the control platform is used to analyze historical digital audio signals to perform situational awareness. If an anomaly is detected, the anomaly is pushed to the IoT box to alert the user that an anomaly has occurred.

[0008] To achieve the above objectives, this application provides a control method for an Internet of Things (IoT)-based auscultation system, the control method comprising: In response to the trigger operation of the auscultation terminal, the acquired audio signal is converted from analog to digital audio signal; The digital audio signal, user identifier, and terminal identifier of the auscultation terminal are encrypted into a target signal file by the transmission module and sent to the control platform and the IoT box. The IoT box and the auscultation terminal are networked together. The control platform sends the target signal file to the remote server, and the IoT box processes the target signal file based on the first model to generate a first analysis result; In response to the second analysis result of the target signal file returned by the remote server, the first analysis result and the second analysis result are fused together to form a status report.

[0009] In one embodiment, the step of converting the acquired audio signal from analog to digital to a digital audio signal includes: The auscultation terminal collects the audio signal; The audio signal is uploaded to a preset link through the radio frequency fixed-frequency circuit of the signal transmitting end; The adapter receives the audio signal from the preset link through the signal receiving end and converts it from analog to digital into the digital audio signal.

[0010] In one embodiment, the step of uploading the audio signal to a preset link via the radio frequency fixed-frequency circuit of the signal transmitting end includes: The audio signal is amplified, filtered, operated, amplified, π-type filtered, and electrostatically processed by a radio frequency fixed-frequency circuit, and then uploaded to the preset link; The step of receiving the audio signal from the preset link and converting it from analog to digital into a digital audio signal includes: Based on the fixed-frequency receiving antenna of the signal receiving end, the audio signal is processed into the digital audio signal through operational amplifier, power amplifier, electrostatic processing and noise reduction.

[0011] In one embodiment, the step of processing the target signal file based on the first model to generate a first analysis result includes: The control platform sends the target signal file and the calling instruction of the first model to the IoT box; After receiving the target signal file, the IoT box performs preprocessing on the digital audio signal in the target signal file based on the first model, including noise reduction filtering, time-frequency domain conversion, and feature point extraction. The extracted audio features are matched with the pre-stored standard cardiopulmonary sounds and / or respiratory sounds feature library in the first model to identify the location and type of abnormal audio segments. Based on the location and type of the abnormal audio segment, and combined with the historical health data associated with the user identifier, a first analysis result is generated, including abnormal feature identifiers and preliminary lesion probabilities. Furthermore, to achieve the above objectives, this application also provides a control device for an IoT-based auscultation system. The IoT-based auscultation system control device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. The computer program is configured to implement the steps of the IoT-based auscultation system control method described above.

[0012] In addition, to achieve the above objectives, this application also provides a storage medium, which is a computer-readable storage medium, storing a program for implementing a control method for an Internet of Things-based auscultation system. The program for implementing the control method for an Internet of Things-based auscultation system is executed by a processor to implement the steps of the control method for an Internet of Things-based auscultation system as described above.

[0013] This application provides a control method for an IoT-based auscultation system. The auscultation system includes an auscultation terminal, a transmission module, an IoT box, a management platform, and a remote server. In response to a trigger operation by the auscultation terminal, the method converts the acquired audio signal from analog to digital audio signal. The digital audio signal, user identifier, and terminal identifier of the auscultation terminal are encrypted into a target signal file by the transmission module and sent to the management platform and the IoT box. The IoT box and the auscultation terminal are networked. The management platform sends the target signal file to the remote server, and the IoT box processes the target signal file based on a first model to generate a first analysis result. In response to a second analysis result returned by the remote server for the target signal file, the first and second analysis results are fused and processed into a status report. This solves the technical problem in related technologies where users need to go to an institution for diagnosis, preventing users from understanding their own status in real time, and achieves the technical effect of remote diagnosis. Attached Figure Description

[0014] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the IoT-based auscultation system of this application; Figure 2 This is a flowchart illustrating a first embodiment of the control method for an IoT-based auscultation system according to this application. Figure 3 This is a schematic diagram of the auscultation system provided in Embodiment 1 of the control method for the Internet of Things-based auscultation system of this application; Figure 4 This is a schematic diagram of the hardware structure of the control device for the IoT-based auscultation system of this application.

[0017] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0019] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0020] Currently, stethoscopes primarily rely on air conduction for sound. Doctors listen to physiological sounds such as heart and lung sounds and respiratory sounds through headphones and make judgments based on experience. However, users need to go to a medical institution for diagnosis, which prevents them from understanding their own condition in real time.

[0021] The main solution of this application is as follows: In response to the trigger operation of the auscultation terminal, the acquired audio signal is converted from analog to digital to a digital audio signal; the digital audio signal, the user identifier, and the terminal identifier of the auscultation terminal are encrypted into a target signal file by the transmission module and sent to the management and control platform; the management and control platform schedules the IoT boxes of the auscultation terminal network to process the target signal file based on the first model to generate a first analysis result; in response to the second analysis result of the target signal file returned by the remote server, the first analysis result and the second analysis result are fused and processed into a status report.

[0022] This application combines an auscultation terminal, a transmission module, an IoT box, a management and control platform, and a remote server to process the audio signals collected by the auscultation terminal. By processing the audio signals into digital audio signals and coordinating the parallel processing of the local IoT box and the remote server through the cloud, the results of both are combined to generate a status report, thus achieving the technical effect of remote diagnosis.

[0023] It should be noted that the executing entity in this embodiment can be a stethoscope system, or a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or a control device for an IoT-based stethoscope system capable of the above functions. This embodiment does not specifically limit the specific implementation. The following description uses a stethoscope system as the executing entity to illustrate this embodiment and the subsequent embodiments.

[0024] Based on this, refer to Figure 1 This application proposes an IoT-based auscultation system, which includes an auscultation terminal, a transmission module, an IoT box, a management platform, and a remote server. The auscultation terminal is used to collect the user's audio signal; the transmission module is used to convert the audio signal from analog to digital and upload it to the management platform and the IoT box; the IoT box is used to generate a first analysis result based on the received digital audio signal; the management platform is used to forward the digital audio signal to the remote server and update the first model of the IoT box; the remote server is used to receive the digital audio signal and generate a second analysis result based on the digital audio signal.

[0025] Optionally, the transmission module includes an auscultation sound acquisition and control submodule and an auscultation sound receiving and processing submodule. The auscultation sound acquisition and control submodule includes an audio acquisition submodule and a signal transmitting end. The auscultation sound receiving and processing submodule includes a signal receiving end and an adapter. The audio acquisition submodule is connected to the auscultation terminal.

[0026] In this embodiment, the adapter is a connector used to realize signal transmission and protocol conversion between different devices; that is, one end of the adapter is wired to devices such as computers and mobile treatment stations, and the other end is wirelessly connected to comfort devices or audio acquisition devices. The adapter is used to receive lossless carrier signals, which can be analog signals, i.e., lossless audio, and the mother's heart and lung sounds can be saved in high-fidelity WAV format at the local end, which can be a computer, mobile phone, mobile treatment vehicle, or other devices. In this embodiment, the audio acquired is stored in WAV format, which can achieve lossless audio quality and sampling frequencies of 48kHz, 96kHz and above.

[0027] Optionally, the IoT-based auscultation system further includes an analysis device. The adapter is connected to the analysis device, which is a mobile terminal, a mobile consultation vehicle, or a medical operation terminal. The analysis device calls the IoT box to analyze the digital audio signal and generate a first analysis result.

[0028] The analysis equipment is equipped with a software system that connects to the auscultation terminal via an audio signal receiving and processing submodule to acquire input data. This software system's functions include: activating the sound card and directly connecting to the network cloud service platform to achieve simultaneous on-site and cloud-based auscultation recording and broadcasting; interacting with cloud data and information: transmitting and storing auscultation data or information, communicating with the cloud server, extracting samples, sampling, cleaning, deep-learning, recognizing, and providing feedback, such as the extraction, recognition, and prompting of heart and lung sounds, and second heart and lung sounds; feature information extraction and interaction: machine information, user information, AI interaction information settings and timing prompts, natural speech, heart and lung auscultation, auscultation of other organs (vascular auscultation), etc.; optimal information transmission mode, low power consumption, low cost, easy recognition, easy operation, and high fidelity.

[0029] Optionally, the control platform is used to analyze historical digital audio signals to perform situational awareness. If an anomaly is detected, the anomaly is pushed to the IoT box to alert the user that an anomaly has occurred.

[0030] As an optional implementation, the auscultation terminal is equipped to users who require daily physiological sound monitoring. The auscultation terminal is designed as a portable device for convenient use anytime. Simultaneously, an IoT box and transmission module can be deployed in the user's home. The IoT box performs preliminary analysis of the received digital audio signals, and the transmission module ensures a stable connection between the auscultation terminal and the IoT box. The audio acquisition submodule within the auscultation sound acquisition and control submodule of the transmission module is connected to the auscultation terminal to acquire audio signals.

[0031] At the medical institution level, a management and control platform and a remote server are deployed. The management and control platform is responsible for data forwarding, model updates, and situational awareness, while the remote server provides more in-depth analysis and diagnostic capabilities. In addition, analytical devices, such as mobile terminals, mobile consultation vehicles, or medical operation terminals, are equipped with corresponding software systems. These devices are connected to the signal receiving end of the auscultation sound receiving and processing submodule of the transmission module via an adapter to achieve data interaction with the auscultation terminal.

[0032] Users use a stethoscope terminal to collect their own physiological audio signals, such as heart and lung sounds and breath sounds. The audio signals collected by the stethoscope terminal are transmitted to the stethoscope sound acquisition and control submodule of the transmission module. After the audio acquisition submodule acquires the signal, the signal transmitter sends the analog audio signal to the signal receiver of the stethoscope sound receiving and processing submodule. After receiving the analog audio signal, the signal receiver converts it into a digital audio signal through an adapter.

[0033] The converted digital audio signals are uploaded to the management platform and the IoT box deployed in the user's home via the network. At the same time, the software system on the analysis device activates the sound card and connects directly to the network cloud service platform to achieve simultaneous full-diagnosis recording and simultaneous external broadcasting at the on-site end (user end) and the cloud (remote server end), ensuring that doctors can obtain audio signals in real time.

[0034] After receiving digital audio signals, the IoT box deployed in the user's home analyzes the signals according to a preset first model and generates a first analysis result. The first model can be a basic diagnostic model trained on a large amount of clinical data, which can make preliminary judgments on common physiological sound abnormalities, such as simple abnormal heart and lung sounds, coarse breath sounds, etc.

[0035] The analysis device obtains initial analysis results by calling an IoT box. The software system on the analysis device can extract feature information and interact based on the initial analysis results. For example, it can provide users with preliminary analysis information through natural voice prompts, such as "A slight abnormality in heart and lung sounds has been detected, please pay attention." At the same time, it displays relevant machine information, user information, and AI interaction information on the device interface and provides time-series prompts.

[0036] The control platform forwards the received digital audio signals to a remote server. The remote server utilizes more powerful computing resources and more complex analysis algorithms, such as deep learning algorithms, to perform in-depth analysis of the digital audio signals and generate a second analysis result. This second analysis result provides more accurate and detailed diagnostic information, such as the specific disease type and severity.

[0037] During the analysis, the remote server also interacts with cloud data, transmitting and storing auscultation data or information, and performing sample extraction, sampling, cleaning, deep learning, recognition, and feedback from the AI ​​data on the cloud server. For example, for heart and lung sound signals, the focus is on extracting and recognizing the second heart and lung sound, and providing prompts based on the recognition results.

[0038] The control platform analyzes historical digital audio signals to perform situational awareness. By analyzing the trend of audio signal changes over a period of time, it determines whether the user's physiological state is abnormal. If an abnormality is detected, the control platform pushes the abnormality notification to the user's IoT box in their home. After receiving the abnormality notification, the IoT box uses analysis devices to inform the user of the abnormality through natural voice, pop-up windows, etc., and suggests that the user take further measures, such as going to the hospital for a detailed examination.

[0039] At the same time, the management platform will also synchronize abnormal situations to relevant medical staff in medical institutions, so that medical staff can understand the user's situation in a timely manner and provide the user with further medical advice and guidance.

[0040] The management platform is responsible for updating the initial model of the IoT boxes in users' homes. As clinical data accumulates and medical research deepens, the platform collects new case data and diagnostic results to optimize and update the initial model. By pushing the updated model to the IoT boxes, the accuracy and effectiveness of the IoT boxes' analysis of digital audio signals are improved.

[0041] In addition, the software system on the analysis device will continuously optimize and upgrade its functions based on actual usage and user feedback to ensure that the system always maintains the optimal information transmission mode, achieving the goals of low power consumption, low cost, easy identification, easy operation, and high fidelity, providing users and medical staff with a better user experience and diagnostic support.

[0042] Furthermore, by aggregating data from household IoT boxes within the region, disease risk assessment and population incidence trend prediction are achieved, providing support for public health decision-making. The management platform has added a "Regional Data Aggregation Module," which aggregates the first analysis results uploaded by each household IoT box according to administrative divisions, such as streets and communities. Results such as suspected respiratory abnormalities and irregular heart and lung sounds are included in the aggregation scope. Simultaneously, basic user information is linked, including age, gender, and past medical history, all of which have been anonymized. A spatiotemporal sequence analysis algorithm is used to perform multi-dimensional statistics on the aggregated data: Time dimension: calculating the frequency of abnormal signals at different times, such as the frequency difference between weekdays and weekends, and the frequency changes during morning and evening rush hours. Spatial dimension: generating a regional disease heat map, marking areas with dense abnormal signals. Population dimension: analyzing the incidence tendency of specific populations, such as people over 65 years old or specific occupational groups. Three levels of warning thresholds are set: general attention, moderate warning, and high warning. When the proportion of a certain type of abnormal signal in a certain area exceeds the threshold: warning information and heat map are pushed to the regional health service center. Health tips are pushed to users in high-risk areas via IoT boxes. For example, it may indicate a high incidence of respiratory diseases in the area and suggest reducing gatherings.

[0043] The management platform extracts seasonally relevant features from IoT device data, such as increased abnormal breathing sounds in winter and allergic cough signals in summer. It also combines regional meteorological data, including temperature, humidity, and pollen concentration, to build a correlation model. Using time-series prediction algorithms such as LSTM, and training the model on aggregated historical data from the past three years, it predicts disease trends for the next 1-3 months. For example, it can predict a 15% increase in the incidence of bronchitis among children in the region next month. Trend reports are provided to the disease control center to assist in developing vaccination plans or health education programs. The prediction results are also pushed to medical institutions in the region, allowing them to allocate medical resources in advance, such as increasing the number of pediatric outpatient staff.

[0044] Furthermore, the model is optimized using the massive amounts of data uploaded by the IoT boxes, and the updated model is then distributed to the terminals to improve local analysis accuracy. Each household IoT box uses its own accumulated digital audio signals, such as cardiopulmonary sound data from the past three months, to perform lightweight training based on the basic model distributed by the management platform. Only the model parameters are uploaded to the management platform, not the raw data, to avoid privacy leaks. The management platform's "model aggregation module" performs weighted aggregation of the parameters uploaded by all IoT boxes, with weights positively correlated with data quality and sample size, generating a globally optimized model, and verifying the model's accuracy through cross-validation. Cross-validation selects labeled data from some medical institutions. The management platform pushes incremental model parameters to the IoT boxes, updating only the differing parts to reduce transmission bandwidth. After receiving the parameters, the IoT boxes automatically merge them, complete the model update, and report the update results back to the platform.

[0045] The management platform categorizes and labels the digital audio signals uploaded by IoT boxes according to disease type (e.g., cardiovascular, respiratory) and user group (e.g., infants, the elderly), constructing segmented datasets. Labeling is performed by a team of professional physicians on a remote server. For high-frequency diseases, such as hypertensive heart disease, a "specialized recognition model" is trained, and feature extraction algorithms are optimized, for example, to improve the accuracy of recognizing the third heart and lung sound. For specific populations, such as obese users experiencing breath sound attenuation, a "group-adapted model" is trained, and signal noise reduction parameters are adjusted. Based on the historical data characteristics of the IoT boxes, such as if the household user is elderly, the management platform pushes suitable specialized models: for elderly user households, priority is given to updating the cardiovascular disease recognition model; for households with children, the focus is on updating the pneumonia-related breath sound analysis model.

[0046] In regional analysis scenarios, an "aggregation-analysis-early warning" chain transforms individual health data into a basis for public health decision-making. In model iteration scenarios, a "local training-global optimization-precise push" mechanism converts data into model capabilities while balancing privacy protection and terminal efficiency. These two expanded scenarios extend the system from "individual health monitoring" to "regional health management" and "intelligent model evolution," enhancing its overall ecosystem value.

[0047] Based on this, Embodiment 1 of this application proposes a control method for an Internet of Things-based auscultation system. Please refer to... Figure 2 The control method for the IoT-based auscultation system includes steps S10-S40: Step S10: In response to the triggering operation of the auscultation terminal, the acquired audio signal is converted from analog to digital audio signal.

[0048] In this embodiment, the auscultation terminal is a portable medical device integrating a high-sensitivity microphone, which can be triggered for data acquisition via physical buttons or a mobile app. The audio signal refers to the physiological acoustic signals generated by the activity of the human heart, blood vessels, or lungs, such as heart and lung sounds and breath sounds. Analog-to-digital conversion is performed by the chip built into the terminal, preserving high-frequency acoustic characteristics according to the set sampling rate, and using a 16-bit quantization bit depth to ensure signal detail.

[0049] Reference Figure 3 The auscultation system includes auscultation terminals, a transmission module, an IoT box, a management platform, and a remote server. Several auscultation terminals are used, and each terminal transmits digital audio signals to the management platform via the transmission module. The management platform connects to both the IoT box and the remote server, which can be a community health center or a medical institution.

[0050] As an optional implementation, upon detection of a trigger signal, a dual-microphone array synchronously acquires sound pressure signals with a 180° phase difference. The main microphone employs an electret condenser sensor, while an ambient microphone is used for noise cancellation. The analog signal is amplified to a 40dB gain by a low-noise preamplifier and then filtered by a second-order Butterworth low-pass filter to suppress high-frequency interference. Subsequently, it is oversampled by a 32-bit ADC at a sampling rate of 128kHz, reduced to an effective sampling rate of 16kHz by a digital decimation filter, generating a 32-bit floating-point PCM data stream. This stream is then encapsulated into data frames of 256 samples each to generate a digital audio signal.

[0051] Step S20: The digital audio signal, user identifier, and terminal identifier of the auscultation terminal are encrypted into a target signal file by the transmission module and sent to the management and control platform and the IoT box. The IoT box and the auscultation terminal are networked together.

[0052] In this embodiment, the user identifier is a unique number registered by the patient in the system (such as a mobile phone number hash), and the terminal identifier is a unique serial number written to the device at the factory (such as an IMEI code). The transmission module is used to encrypt and upload the digital audio signal to the cloud. The key is generated by a dynamic session key pre-negotiated between the management platform and the IoT box, and the target signal file is encapsulated in JSON format, containing audio data, user identifier, and terminal identifier.

[0053] As an optional implementation, the transmission module employs blockchain notarization technology to encrypt the digital audio signal, user identifier, and terminal identifier of the stethoscope terminal into a target signal file. Before encryption, a timestamp and hash value are added to the target signal file, and it is uploaded to the management platform while being registered on the consortium blockchain node to ensure data immutability. Simultaneously, the transmission module sends the target signal file to the IoT box.

[0054] In step S30, the control platform sends the target signal file to the remote server, and the IoT box processes the target signal file based on the first model to generate a first analysis result.

[0055] In this embodiment, the IoT box is an edge computing gateway, deployed with a lightweight CNN model, i.e., the first model, for classifying heart and lung sounds. The first analysis result is the result parsed by the IoT box from the digital audio signal.

[0056] As an optional implementation, the control platform schedules the nearest edge node based on geographic location information, or schedules edge nodes networked with the auscultation terminal. The IoT box runs an optimized audio classification model using TensorFlowLite to extract time- and frequency-domain features from the signal, identifying features such as S1 / S2 heart and lung sound splitting and pathological murmurs, generating a structured analysis report containing heart rate variability and murmur type, i.e., the first analysis result. Simultaneously, the control platform sends the target signal file to a remote server.

[0057] Step S40: In response to the second analysis result of the target signal file returned by the remote server, the first analysis result and the second analysis result are fused and processed into a status report.

[0058] In this embodiment, the remote server is deployed with an LSTM-based time-series analysis model and a medical expert rule engine, or it may be a device with experts, such as a community health center or medical institution. The second analysis result is the result of expert analysis of the digital audio signal, or the result of analysis of the digital audio signal by the time-series analysis model and the medical expert rule engine. The status report is a result that combines the first and second analysis results to generate a result that is more personalized to the user.

[0059] As an optional implementation, the management platform uses evidence theory to fuse the results, weighting and fusing the probability distribution output by the edge model with the diagnostic suggestions of the expert system to generate a multi-dimensional status report that includes heart rate and confidence index, which is then pushed to the user terminal in real time.

[0060] For example, a user uses a smart stethoscope to collect heart and lung sound data; the device automatically triggers the collection through bioelectrical signal detection. After the data is encrypted and uploaded to the cloud, the management platform schedules edge nodes at the community health service center to perform preliminary analysis, identifying the S3 gallop rhythm characteristic. Simultaneously, the deep learning model of the telemedicine platform, combined with the patient's electronic medical record, indicates a risk of left ventricular dysfunction. The system integrates the dual analysis results, generates a red alert status report, and pushes it to the patient's mobile phone and the family doctor's terminal.

[0061] In this embodiment, the auscultation system includes an auscultation terminal, a transmission module, an IoT box, a management and control platform, and a remote server. In response to a trigger operation by the auscultation terminal, the system converts the collected audio signal from analog to digital audio signal. The digital audio signal, user identifier, and terminal identifier of the auscultation terminal are encrypted into a target signal file by the transmission module and sent to the management and control platform. The management and control platform schedules the IoT boxes connected to the auscultation terminal network to process the target signal file based on a first model and generate a first analysis result. In response to a second analysis result returned by the remote server for the target signal file, the first and second analysis results are fused and processed into a status report. This solves the technical problem in related technologies where users need to go to an institution for diagnosis, preventing users from understanding their own status in real time, and achieves the technical effect of remote diagnosis.

[0062] Based on any of the above embodiments, in Embodiment 2 of this application, the auscultation system further includes an adapter, the adapter being provided with a signal receiving end, and the auscultation terminal being provided with a signal transmitting end. Step S10 includes: Step S11: The auscultation terminal collects the audio signal.

[0063] In this embodiment, the auscultation terminal is an intelligent module modified from a traditional stethoscope, with a built-in MEMS microphone array and adaptive beamforming technology to focus and collect heart and lung sound signals.

[0064] As an optional implementation, the auscultation terminal detects the contact pressure through a triaxial accelerometer. When the pressure value is in the range of 2 to 3 N, it triggers the acquisition. The dual microphone array synchronously acquires the sound pressure signal with a 180° phase difference. The main microphone uses an electret condenser sensor with a sensitivity of -38dB, and the ambient microphone is used to acquire background noise.

[0065] Step S12: The audio signal is uploaded to the preset link through the radio frequency fixed frequency circuit of the signal transmitting end.

[0066] As an optional implementation, the signal transmitter integrates a 2.4GHz radio frequency chip, adopts GFSK modulation, and the preset link is a private wireless channel based on the IEEE 802.15.4 protocol.

[0067] For example, the audio signal is initially digitized by a 12-bit ADC at a sampling rate of 48kHz, and then modulated into a 433MHz carrier signal by a radio frequency fixed-frequency circuit. Direct sequence spread spectrum (DS-SS) technology is used to improve anti-interference capability, and the data transmission rate is set to 250kbps to ensure audio quality while reducing power consumption.

[0068] As another alternative implementation, a proprietary protocol is used at both the signal transmitter and receiver to achieve lossless data transmission.

[0069] In step S13, the adapter receives the audio signal from the preset link through the signal receiving end and converts it from analog to digital into the digital audio signal.

[0070] In this embodiment, the adapter is a rechargeable smart gateway device, the signal receiving end integrates a superheterodyne receiver circuit, and the analog-to-digital conversion uses a 24-bit Σ-Δ ADC.

[0071] As an optional implementation, the signal receiver locks a 433MHz carrier signal through a phase-locked loop (PLL) circuit. After mixing, intermediate frequency amplification, and demodulation, the baseband audio signal is recovered. A 24-bit Σ-Δ ADC then performs high-precision analog-to-digital conversion at a sampling rate of 192kHz. The signal is then reduced to an effective sampling rate of 44.1kHz by a digital filter, generating a 32-bit floating-point PCM data stream, which is finally uploaded to the management platform via USB or Wi-Fi.

[0072] For example, a cardiologist uses a smart stethoscope to collect a patient's heart and lung sounds. The stethoscope module transmits the audio signal to a clinic adapter via 433MHz radio frequency. The adapter receives the signal, performs high-precision analog-to-digital conversion, generates a digital audio file, and uploads it to the hospital information system. The doctor can then view the real-time heart and lung sound waveforms and AI analysis results via a remote server to quickly determine if the patient has mitral stenosis.

[0073] This embodiment uses a distributed acquisition architecture to transfer the analog-to-digital conversion function to the adapter, thereby reducing the size of the auscultation terminal and achieving lossless signal conversion.

[0074] Optionally, the step of uploading the audio signal to a preset link via the radio frequency fixed-frequency circuit of the signal transmitting end includes: Step S121: The audio signal is amplified, filtered, operated, amplified, π-type filtered, and electrostatically processed by the radio frequency fixed-frequency circuit, and then uploaded to the preset link.

[0075] In this embodiment, the radio frequency fixed-frequency circuit refers to the radio frequency processing circuit integrated within the signal transmitting end of the stethoscope terminal, used for power amplification and signal optimization of the audio signal. The power amplifier amplifies the signal transmission strength; the filter removes noise interference; the operational amplifier performs signal amplification; the π-type filter further filters out high-frequency noise through a π-type LC filter circuit; and electrostatic discharge (ESD) protection devices prevent damage to the circuit from static electricity. The preset link is a wireless transmission channel established based on a specific frequency band.

[0076] As an optional implementation, the audio signal first enters the power amplifier of the RF fixed-frequency circuit to boost the signal power to a level suitable for long-distance transmission; then it passes through a bandpass filter to filter out interference signals outside the frequency band; subsequently, the signal is adjusted by an operational amplifier to ensure stable signal amplitude; it is amplified again to enhance signal strength; then it passes through a π-type LC filter circuit to filter out residual high-frequency noise; finally, it undergoes electrostatic discharge treatment through an ESD electrostatic protection device and is transmitted to the preset link via the 2.4GHz frequency band using GFSK modulation.

[0077] As another optional implementation method for uploading the audio signal to a preset link through the radio frequency fixed-frequency circuit of the signal transmitter, the signal transmitter achieves multi-directional transmission through broadcast 84MHz.

[0078] Optionally, the step of receiving the audio signal from the preset link and converting it from analog to digital is divided into the following steps: Step S131: Based on the fixed-frequency receiving antenna of the signal receiving end, the audio signal is processed into the digital audio signal through operational amplifier, power amplifier, electrostatic processing and noise reduction.

[0079] In this embodiment, the fixed-frequency receiving antenna is an antenna used by the adapter signal receiver to receive signals in a specific frequency band. The operational amplifier amplifies the received weak signal; the power amplifier further increases the signal power; electrostatic discharge protection is achieved through ESD protection circuitry; and noise reduction is achieved by removing environmental interference through a digital noise reduction algorithm.

[0080] As an optional implementation, after the fixed-frequency receiving antenna receives the audio signal from the preset link, it first sends the signal to the operational amplifier for amplification to enhance the signal strength; then it further enhances the signal power through the power amplifier; then it is protected against electrostatic discharge (ESD) by the ESD electrostatic protection device; finally, it uses an adaptive digital noise reduction algorithm to dynamically adjust the parameters according to the environmental noise characteristics to remove background noise interference, and then the built-in 24-bit ADC performs analog-to-digital conversion to generate a digital audio signal.

[0081] For example, a user uses a smart auscultation terminal at home to collect heart and lung sounds. The radio frequency fixed-frequency circuit inside the auscultation terminal performs a series of processes on the collected audio signal, including power amplification and filtering, before sending it out through a preset link. An adapter placed in the living room receives the signal through a fixed-frequency receiving antenna, and after processing such as operational amplifier and noise reduction, it is converted into a digital audio signal and uploaded to a management platform for subsequent analysis and diagnosis. Finally, the user receives a status report containing health status analysis on their mobile phone.

[0082] This embodiment uses a radio frequency fixed-frequency circuit to perform multiple processing steps on the audio signal before transmission, and an adapter to optimize the conversion of the received signal. This reduces interference and loss during signal transmission, improves the transmission quality and stability of the audio signal, and provides a reliable data foundation for subsequent accurate analysis and diagnosis.

[0083] Based on any of the above embodiments, in Embodiment 3 of this application, step S20 includes: Step S21: Obtain the terminal identifier of the auscultation terminal and the user identifier of the login account of the auscultation terminal.

[0084] In this embodiment, the terminal identifier is a globally unique identification code (such as UUID or IMEI) that is fixed when the stethoscope terminal leaves the factory, used for device identity authentication; the user identifier is a unique account code (such as a mobile phone number hash value) generated when the user registers, which is associated with the user's personal health record.

[0085] As an optional implementation, when the auscultation terminal is started, the system automatically reads the terminal identifier stored in the secure element (SE); after the user logs in to the APP via fingerprint or facial recognition, the system retrieves the user identifier from the local encrypted database to ensure that the data reading process complies with GDPR security standards.

[0086] Step S22: Obtain the data key of the IoT box in the auscultation terminal network.

[0087] In this embodiment, the data key is a symmetric encryption key pre-negotiated between the IoT box and the management platform, and it adopts a dynamic update mechanism (such as changing it every 24 hours).

[0088] As an optional implementation, after the stethoscope terminal establishes a connection with the IoT box via Bluetooth or Wi-Fi, it sends a key request command. After verifying the legitimacy of the terminal identifier, the IoT box retrieves the current data key from the secure storage module (TPM), encrypts it using the terminal's public key, and sends it back, ensuring that the key transmission process is protected against theft and tampering.

[0089] Step S23: Encrypt the digital audio signal, the user identifier, and the terminal identifier into a first data packet based on the data key.

[0090] In this embodiment, the AES-256-GCM algorithm is used for data encryption. The GCM mode provides authentication encryption function to ensure data integrity.

[0091] As an optional implementation, the system encapsulates the digital audio signal, user identifier, and terminal identifier into raw data in JSON format, encrypts it using the data key obtained from the IoT box using AES-256-GCM, generates a first data packet containing ciphertext, authentication tag, and random number, and calculates the data hash value (SHA-256) for integrity verification.

[0092] In step S24, the transmission module transmits the first data packet to the management and control platform through the encrypted channel corresponding to the terminal identifier.

[0093] In this embodiment, the encrypted channel is established based on the TLS 1.3 protocol, and each terminal identifier corresponds to a unique session key.

[0094] As an optional implementation, after parsing the terminal identifier, the transmission module obtains the corresponding session key from the key management system and establishes a secure connection using the TLS 1.3 protocol. The first data packet, after being processed into frames, is transmitted to the management platform through this encrypted channel. Forward error correction (FEC) technology is enabled during transmission to ensure reliable data transmission in weak network environments.

[0095] For example, a diabetic patient uses a smart stethoscope at home to monitor their heart health. Upon startup, the device automatically acquires a terminal identifier, and the user obtains a user identifier by logging in via facial recognition on an app. The stethoscope connects to a home IoT box, obtains a data key after authentication, and encrypts the collected digital audio signals to generate the first data packet. The transmission module securely uploads the data packet to a management platform through a pre-configured encrypted channel. Subsequently, an AI model and a doctor conduct a joint diagnosis and provide feedback on the results.

[0096] This embodiment achieves end-to-end encryption of data from the terminal to the cloud through multi-layer encryption mechanisms and dynamic key management, effectively protecting the privacy and security of users' medical data. Based on the dedicated encryption channel of the terminal identifier and transmission optimization technology, it enhances the security and reliability of data transmission, providing a solid data security foundation for remote medical diagnosis.

[0097] Based on any of the above embodiments, in Embodiment 4 of this application, the step of generating a first analysis result by processing the target signal file based on the first model includes: Step S31: The control platform sends the target signal file and the calling instruction of the first model to the IoT box.

[0098] In this embodiment, the control platform is a distributed cluster with high-performance computing and storage capabilities, responsible for managing and scheduling system resources. The target signal file is an encrypted data packet containing user audio data, identifiers, and other information. The first model is a lightweight machine learning model deployed on the IoT box for audio signal analysis. The invocation command is a command sent by the control platform to the IoT box to start the model processing task.

[0099] As an optional implementation, after receiving the target signal file, the management platform selects a suitable IoT box for task allocation based on the IoT box's load status and geographical location. The target signal file and invocation instructions are then sent to the selected IoT box via a message queue (such as RabbitMQ) to ensure reliable and asynchronous data transmission.

[0100] Step S32: After receiving the target signal file, the IoT box preprocesses the digital audio signal in the target signal file based on the first model, including noise reduction filtering, time-frequency domain conversion, and feature point extraction.

[0101] In this embodiment, the IoT box is an edge computing device with certain computing and storage capabilities, used for preliminary processing of audio signals locally. Noise denoising filtering removes noise interference from the audio signal using digital filters, improving signal quality. Time-frequency domain conversion converts the time-domain audio signal to the frequency domain for better analysis of its frequency characteristics. Feature point extraction extracts representative feature points from the audio signal for subsequent matching and analysis.

[0102] As an optional implementation, after receiving the target signal file, the IoT box first decrypts and verifies it. Then, wavelet transform is used for denoising filtering to remove high-frequency noise and low-frequency interference. Next, the audio signal is converted into a time-frequency domain representation using short-time Fourier transform (STFT) to obtain a spectrogram. Finally, a peak detection algorithm is used to extract feature points from the spectrogram, such as the S1 and S2 peaks of heart and lung sounds.

[0103] Step S33: Match the extracted audio features with the pre-stored standard cardiopulmonary sound and / or respiratory sound feature library in the first model to identify the location and type of abnormal audio segments.

[0104] In this embodiment, the standard heart and lung sounds and / or breath sounds feature library is a pre-stored set of normal audio features in the first model, used for comparison with the extracted audio features. Matching determines the presence of an anomaly by calculating the similarity between audio features. The location of the abnormal audio segment refers to the specific time position in the audio signal where the anomaly occurs. The anomaly type refers to the classification of the anomaly determined based on the matching results, such as arrhythmia, shortness of breath, etc.

[0105] As an optional implementation, the IoT box compares the extracted audio features one by one with each feature in a standard feature library, using the Dynamic Time Warping (DTW) algorithm to calculate the similarity. When the similarity is below a certain threshold, it is determined to be an abnormal audio segment, and the type of abnormality is determined based on the feature matching results. For example, if the frequency and intensity of heart and lung sounds match the arrhythmia features in the standard feature library within a certain time period, it is identified as an arrhythmia abnormality.

[0106] Step S34: Based on the location of the abnormal audio segment and the abnormal type, and combined with the historical health data associated with the user identifier, generate a first analysis result containing abnormal feature identifiers and preliminary lesion probability.

[0107] In this embodiment, the historical health data associated with the user identifier refers to past health information related to the user, such as medical records and examination reports. The abnormal feature identifier is a specific description and label of the abnormal situation to facilitate subsequent analysis and processing. The preliminary lesion probability is an estimate of the likelihood of the lesion calculated based on the location of the abnormal audio segment, the type of abnormality, and historical health data.

[0108] Based on any of the above embodiments, in Embodiment 5 of this application, after the step of encrypting the digital audio signal, user identifier, and terminal identifier of the auscultation terminal into a target signal file via the transmission module and sending it to the management and control platform, the following steps are included: Step A10: Remove personal information contained in the digital audio signal in the target signal file to obtain the target digital signal.

[0109] In this embodiment, personal information includes, but is not limited to, user biometric data, geographic location information, and timestamps that can identify an individual. The removal operation is implemented through data desensitization algorithms, using irreversible hash desensitization (such as SHA-256) to process the user identifier, and using voice activity detection (VAD) technology to remove any environmental dialogue sounds that may be contained in the audio signal.

[0110] As an optional implementation, the system scans the time-domain waveform of the digital audio signal using a natural language processing (NLP) model to identify and filter speech segments containing keywords such as names and addresses; at the same time, it performs format-preserving desensitization on the user identification field (e.g., converting "1381234" to "1381234" format but with irreversible obfuscation of the actual value) to ensure that the target digital signal meets the requirements of the "Guideline for Health and Medical Data Security" (GB / T39725-2020).

[0111] Step A20: Use the target digital signal as training data to train the second model.

[0112] In this embodiment, the second model is a deep learning model deployed in the cloud, employing a hybrid architecture of convolutional neural networks (CNN) and recurrent neural networks (RNN) to improve the generalization ability of cardiopulmonary sound classification. The training process uses a federated learning framework to ensure that the data does not leave the domain.

[0113] As an optional implementation, the target digital signal is segmented into 10-second segments, and a 40-dimensional feature vector is extracted using Mel-frequency cepstral coefficients (MFCC) to construct a training dataset. End-to-end training is performed using the Adam optimizer, with the loss function being a weighted sum of cross-entropy and focal loss. The iteration period is set to 100 rounds, the batch size to 32, and the validation set ratio to 20%. During training, the model's transfer performance on public datasets (such as the PhysioNet cardiopulmonary sound database) is monitored in real time to prevent overfitting.

[0114] Step A30: In response to the data synchronization command, update the first model of the IoT box based on the second model.

[0115] In this embodiment, data synchronization instructions are generated periodically by the management platform (e.g., at 2 AM daily) or triggered based on model performance metrics (e.g., a ≥1.5% improvement in validation set accuracy). The update process employs model differential compression technology, transmitting only the changes in weight parameters.

[0116] As an optional implementation, the management platform generates an incremental update package (DeltaPackage) by comparing the weight parameters of the second model and the first model locally located on the IoT box. The update package is pushed to the target IoT box using the gRPC protocol. Upon receiving the update package, the box first performs an integrity check (SHA-512 hash matching), and then replaces the model file using a hot update mechanism, ensuring uninterrupted service throughout the process. After the update is complete, the IoT box automatically runs test cases (such as pre-stored normal / abnormal heart and lung sound samples) to verify the correctness of the model's functionality and sends the results back to the cloud.

[0117] For example, a regional medical center collected 1,000 anonymized abnormal cardiopulmonary sounds to train a second cloud-based model. After three days of iterative training, the model's accuracy in identifying mitral regurgitation increased from 88% to 93%. Upon detecting this significant performance improvement, the management platform sent data synchronization instructions to the IoT boxes in all community health service centers within the region. Each box received the model update package and upgraded during non-clinic hours, enabling the new model to be used for cardiopulmonary sound analysis the following day, thus increasing the detection rate of early valvular lesions in primary healthcare institutions by approximately 5%.

[0118] This embodiment constructs a closed-loop mechanism of "data anonymization - cloud training - edge update" to achieve continuous model evolution while protecting user privacy, solving the problem of traditional edge computing models being difficult to dynamically optimize. Incremental update technology reduces the traffic consumption of model upgrades, and the hot update mechanism ensures uninterrupted primary healthcare services, effectively improving the system's practicality and scalability.

[0119] Based on any of the above embodiments, in Embodiment Six of this application, the step of fusing the first analysis result and the second analysis result into a status report in response to the second analysis result for the target signal file returned by the remote server includes: Step S41: Obtain the first metadata of the first analysis result and the second metadata of the second analysis result.

[0120] In this embodiment, the first metadata is data used to describe information related to the first analysis result. The generation timestamp records the specific time when the first analysis result was generated. The IoT box identifier is used to identify which IoT box generated the analysis result, and the model version number indicates the version of the first model used at that time. The second metadata is descriptive information related to the second analysis result. The remote server identifier indicates which remote server the second analysis result came from, the expert qualification level reflects the professional level of the experts participating in the analysis, and the analysis completion time records the moment when the second analysis result was completed.

[0121] As an optional implementation, the system can directly extract the corresponding metadata information from the database storing the first and second analysis results. The first metadata includes the generation timestamp, IoT box identifier, and model version number, while the second metadata includes the remote server identifier, expert qualification level, and analysis completion time.

[0122] Step S42: If the first metadata and the second metadata pass the verification, determine the intersection interval between the abnormal feature identifier in the first analysis result and the manually annotated abnormal segment in the second analysis result.

[0123] In this embodiment, passing the verification means that the first and second metadata meet certain rules and conditions in terms of time and source. The anomaly feature identifier is a specific identifier for anomalies in the first analysis result, while the manually annotated anomaly segment is an anomaly audio segment manually marked by experts in the second analysis result. The intersection interval is the range of the anomaly audio segment jointly determined by these two methods.

[0124] As an optional implementation, the system can verify whether the metadata passes the test by comparing time information, device identification, etc., in the first and second analysis results. For anomaly identifiers and manually marked anomaly segments, the intersection range can be determined by comparing them on the timeline.

[0125] Step S43: If the intersection interval meets the confidence requirement, assign fusion weights to the first analysis result and the second analysis result.

[0126] In this embodiment, the confidence level requirement refers to the reliability and credibility of the intersection interval reaching a certain standard. The fusion weight is a numerical value used to determine the proportion of the first analysis result and the second analysis result during fusion.

[0127] As an alternative implementation, the system can determine whether the confidence level requirement is met based on factors such as the length of the intersection interval and the degree of conformity with known standards. The allocation of fusion weights can be determined based on factors such as the historical accuracy records of the first and second analysis results and the authority of experts.

[0128] Step S44: Based on the fusion weight, the first analysis result and the second analysis result are fused to form the status report.

[0129] In this embodiment, the fusion process involves combining the first analysis result and the second analysis result according to the fusion weight to generate the final status report.

[0130] As an optional implementation, the system can perform a weighted average of the indicators in the first and second analysis results according to the fusion weights to obtain the corresponding content in the status report.

[0131] For example, suppose the first analysis indicates the possible presence of a certain heart disease, with the abnormal feature identified as "A". The second analysis shows that an abnormal segment manually annotated by experts is also related to this disease, and the intersection interval meets the confidence requirement. Based on the fusion weights, the first and second analysis results are fused to generate a status report. The report indicates the likelihood of the heart disease and provides corresponding recommendations.

[0132] This embodiment obtains and verifies metadata, determines the intersection range, assigns fusion weights, and performs fusion processing. It can comprehensively consider the information from the first analysis result and the second analysis result to generate a more accurate and reliable status report, providing users with more valuable health diagnoses and suggestions.

[0133] Optionally, the step of fusing the first analysis result and the second analysis result into the status report based on the fusion weights includes: Step S441: Cross-validate the location and type of the abnormal audio segment, and determine the abnormal segment by combining the intersection interval.

[0134] In this embodiment, cross-validation refers to ensuring consistency and accuracy by comparing the location and type of abnormal audio segments in the first and second analysis results. The intersection interval refers to the time range of abnormal audio segments jointly identified in the first and second analysis results.

[0135] As an optional implementation, the system first compares the location and type of abnormal audio segments in the first and second analysis results. If there is a high degree of consistency between the two in terms of time range and abnormal type, the abnormal segment within the intersection interval is determined as the final abnormal segment. If there are differences, the reasons for the differences are further analyzed, which may be due to different analysis methods or data processing methods. In this case, the system can combine expert opinions or other relevant information to determine the final abnormal segment.

[0136] Step S442: Based on the lesion probability value corresponding to the abnormal segment, the fusion probability value is obtained by weighted calculation using the fusion weight.

[0137] In this embodiment, the lesion probability value refers to the estimated probability of the lesion corresponding to the abnormal segment in the first analysis result and the second analysis result. The fusion weight is a weight value determined based on the reliability and authority of the first analysis result and the second analysis result.

[0138] As an optional implementation, the system performs a weighted calculation based on the lesion probability values ​​corresponding to the abnormal segments in the first and second analysis results, and a pre-determined fusion weight. Specifically, the lesion probability value in the first analysis result is multiplied by its corresponding fusion weight, and the lesion probability value in the second analysis result is multiplied by its corresponding fusion weight. The two products are then added together to obtain the fusion probability value.

[0139] Step S443: Combine the historical health data associated with the user identifier to correct the fusion probability value to obtain the final probability.

[0140] In this embodiment, historical health data refers to past health information associated with a user's identifier, such as medical records and examination reports. Adjusting the fusion probability value refers to adjusting the fusion probability value based on historical health data to improve diagnostic accuracy.

[0141] As an optional implementation, the system combines the fusion probability value with historical health data associated with the user's identifier. If the user has a relevant medical history or other health problems, the fusion probability value is adjusted based on this information. For example, if the user has a history of heart disease, the fusion probability value is appropriately increased; if the user does not have a relevant medical history, the fusion probability value is appropriately decreased. In this way, the user's health status can be assessed more accurately.

[0142] Step S444: Based on the final probability, basic information, location of abnormal audio segments, and abnormal type, the status report is generated in conjunction with the suggested scheme.

[0143] In this embodiment, the final probability refers to the corrected fusion probability value, reflecting the likelihood that a user has a certain disease. Basic information refers to the user's basic information, such as age, gender, height, and weight. Abnormal audio segment location and abnormal type refer to the time range and abnormal type of the finally determined abnormal segment. The recommended solution is a treatment or health management suggestion formulated based on the final probability and other relevant information.

[0144] As an optional implementation, the system generates a status report based on the final probability, basic information, location of the abnormal audio segment, and anomaly type, combined with suggested solutions. The status report includes the user's basic information, the location and type of the abnormal audio segment, the final probability, and suggested solutions. Suggested solutions can be customized based on the magnitude of the final probability and other relevant information. For example, if the final probability is high, further examination or treatment is recommended; if the final probability is low, health management measures such as dietary control and exercise are suggested.

[0145] For example, a patient uses a smart stethoscope for examination. Both the first and second analysis results identify an abnormal audio segment and both suggest it may be related to a heart condition. The system cross-validates the location and type of the abnormal audio segment, determining the final abnormal segment by combining the intersection interval. Then, based on the disease probability values ​​corresponding to the abnormal segment in the first and second analysis results, and a pre-determined fusion weight, a weighted calculation is performed to obtain a fusion probability value. Next, the fusion probability value is corrected by incorporating historical health data associated with the user's identifier, resulting in a final probability. Finally, based on the final probability, basic information, location of the abnormal audio segment, and type of abnormality, a status report is generated, along with suggested treatment options. The status report shows that the patient's final probability is 70%, recommending further examination and treatment.

[0146] This embodiment improves diagnostic accuracy by cross-validating the location and type of abnormal audio segments and identifying the abnormal segments based on their intersection intervals. Furthermore, it enhances diagnostic accuracy by performing weighted calculations based on fusion weights to obtain a fusion probability value, which is then corrected using historical health data. Finally, it generates a status report, providing users with detailed health information and suggestions to help them better manage their health.

[0147] This application provides a control device for an IoT-based auscultation system. The control device for the IoT-based auscultation system includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the control method for the IoT-based auscultation system in the above embodiment 1.

[0148] The following is for reference. Figure 4 The diagram illustrates a structural schematic of a control device suitable for implementing an IoT-based auscultation system according to embodiments of this application. The control device for the IoT-based auscultation system in these embodiments may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, personal digital assistants (PDAs), tablets, and in-vehicle terminals, as well as fixed terminals such as digital TVs and desktop computers. Figure 4 The control device for the IoT-based auscultation system shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0149] like Figure 4As shown, the control device of the IoT-based auscultation system may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 1002 or a program loaded from storage device 1003 into random access memory (RAM) 1004. The random access memory 1004 also stores various programs and data required for the operation of the control device of the IoT-based auscultation system. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the control device of the IoT-based auscultation system to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows a control device for an IoT-based auscultation system with various systems, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented alternatively.

[0150] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0151] The IoT-based auscultation system control device provided in this application, employing the IoT-based auscultation system control method described in the above embodiments, solves the technical problem that users need to go to an institution for diagnosis, resulting in users not being able to understand their own status in real time. Compared with the prior art, the beneficial effects of the IoT-based auscultation system control device provided in this application are the same as those of the IoT-based auscultation system control device provided in the above embodiments, and other technical features in this IoT-based auscultation system control device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0152] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0153] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0154] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the control method of the Internet of Things-based auscultation system in the above embodiments.

[0155] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, radio frequency (RF), or any suitable combination thereof.

[0156] The aforementioned computer-readable storage medium may be included in the control device of the Internet of Things-based auscultation system; or it may exist independently and not be assembled into the control device of the Internet of Things-based auscultation system.

[0157] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by the control device of the IoT-based auscultation system, the control device of the IoT-based auscultation system: in response to a trigger operation of the auscultation terminal, converts the acquired audio signal from analog to digital to a digital audio signal; encrypts the digital audio signal, the user identifier, and the terminal identifier of the auscultation terminal through the transmission module into a target signal file, and sends it to the management and control platform; the management and control platform schedules the IoT boxes of the auscultation terminal network to process the target signal file based on a first model to generate a first analysis result; and in response to a second analysis result for the target signal file returned by the remote server, merges the first analysis result and the second analysis result into a status report.

[0158] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0159] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0160] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0161] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the control method of the above-described IoT-based auscultation system. This solves the technical problem that users need to go to an institution for diagnosis, preventing them from understanding their own status in real time. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the control method of the IoT-based auscultation system provided in the above embodiments, and will not be repeated here.

[0162] This application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the control method for an Internet of Things-based auscultation system as described above.

[0163] The computer program product provided in this application can solve the technical problem that users need to go to an institution for diagnosis, resulting in users not being able to understand their own status in real time. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as the beneficial effects of the control method of the Internet of Things-based auscultation system provided in the above embodiments, and will not be repeated here.

[0164] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent scope of this application.

Claims

1. A stethoscope system based on the Internet of Things, characterized in that, The IoT-based auscultation system includes an auscultation terminal, a transmission module, an IoT box, a management and control platform, and a remote server. The auscultation terminal is used to collect the user's audio signals. The transmission module is used to convert the audio signal from analog to digital and upload it to the control platform and the IoT box. The IoT box is used to generate a first analysis result based on the received digital audio signal; The control platform is used to forward the digital audio signal to the remote server and update the first model of the IoT box; The remote server is used to receive digital audio signals and generate a second analysis result based on the digital audio signals.

2. The IoT-based auscultation system as described in claim 1, characterized in that, The transmission module includes an auscultation sound acquisition and control submodule and an auscultation sound receiving and processing submodule. The auscultation sound acquisition and control submodule includes an audio acquisition submodule and a signal transmitting end. The auscultation sound receiving and processing submodule includes a signal receiving end and an adapter. The audio acquisition submodule is connected to the auscultation terminal.

3. The IoT-based auscultation system as described in claim 2, characterized in that, The IoT-based auscultation system also includes an analysis device. The adapter is connected to the analysis device, which can be a mobile terminal, a mobile consultation vehicle, or a medical operation terminal. The analysis device calls the IoT box to analyze the digital audio signal and generate a first analysis result.

4. The IoT-based auscultation system as described in claim 1, characterized in that, The control platform is used to analyze historical digital audio signals to perform situational awareness. If an anomaly is detected, it will push the anomaly to the IoT box to alert the user.

5. A control method for an Internet of Things-based auscultation system, characterized in that, Applied to the IoT-based auscultation system as described in any one of claims 1-4, the control method of the IoT-based auscultation system includes: In response to the trigger operation of the auscultation terminal, the acquired audio signal is converted from analog to digital audio signal; The digital audio signal, user identifier, and terminal identifier of the auscultation terminal are encrypted into a target signal file by the transmission module and sent to the control platform and the IoT box. The IoT box and the auscultation terminal are networked together. The control platform sends the target signal file to the remote server, and the IoT box processes the target signal file based on the first model to generate a first analysis result; In response to the second analysis result of the target signal file returned by the remote server, the first analysis result and the second analysis result are fused together to form a status report.

6. The control method for the Internet of Things-based auscultation system as described in claim 5, characterized in that, The step of converting the acquired audio signal from analog to digital to a digital audio signal includes: The auscultation terminal collects the audio signal; The audio signal is uploaded to a preset link through the radio frequency fixed-frequency circuit of the signal transmitting end; The adapter receives the audio signal from the preset link through the signal receiving end and converts it from analog to digital into the digital audio signal.

7. The control method for the Internet of Things-based auscultation system as described in claim 6, characterized in that, The step of uploading the audio signal to a preset link via the radio frequency fixed-frequency circuit of the signal transmitting end includes: The audio signal is amplified, filtered, operated, amplified, π-type filtered, and electrostatically processed by a radio frequency fixed-frequency circuit, and then uploaded to the preset link; The step of receiving the audio signal from the preset link and converting it from analog to digital into a digital audio signal includes: Based on the fixed-frequency receiving antenna of the signal receiving end, the audio signal is processed into the digital audio signal through operational amplifier, power amplifier, electrostatic processing and noise reduction.

8. The control method for the Internet of Things-based auscultation system as described in claim 5, characterized in that, The step of processing the target signal file based on the first model to generate the first analysis result includes: The control platform sends the target signal file and the calling instruction of the first model to the IoT box; After receiving the target signal file, the IoT box performs preprocessing on the digital audio signal in the target signal file based on the first model, including noise reduction filtering, time-frequency domain conversion, and feature point extraction. The extracted audio features are matched with the pre-stored standard cardiopulmonary sounds and / or respiratory sounds feature library in the first model to identify the location and type of abnormal audio segments. Based on the location of the abnormal audio segment and the type of abnormality, combined with the historical health data associated with the user identifier, a first analysis result is generated, which includes abnormal feature identifiers and preliminary lesion probability.

9. A control device for an Internet of Things-based auscultation system, characterized in that, The control device of the IoT-based auscultation system includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the control method for the IoT-based auscultation system as described in any one of claims 5 to 8.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the control method for the Internet of Things-based auscultation system as described in any one of claims 5 to 8.

Images