Medical patrol bracelet based on multi-mode AI and Internet of Things

Abstract

The invention discloses a medical patrol bracelet based on multi-mode AI and Internet of Things. The medical patrol bracelet is suitable for medical help and daily health management scenes of old people and special groups. The bracelet integrates a hardware layer, a software layer and a communication layer, wherein the hardware layer comprises a main control module, a voice interaction module, a positioning module, a health monitoring module and the like which take an ESP32-S3 development board as a core; the software layer adopts a multi-mode AI dynamic adaptation technology, a real-time path optimization algorithm based on hospital node weight and the like to realize voice interaction, intelligent navigation and health data analysis functions; and the communication layer realizes data interaction between the bracelet and a cloud server and a mobile phone APP through the Internet of Things technology. According to the method, personalized voice interaction between six local accent and three foreign languages is supported through a multi-mode AI dynamic adaptation technology; a real-time path optimization algorithm based on hospital node weight is combined with a GCN model to realize dynamic navigation; a dual recording mechanism of voiceprint recognition is adopted, and doctor-patient communication records are synchronized to an applet for family members to trace back after being subjected to noise reduction processing.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical health monitoring and auxiliary equipment, in particular to a medical tour diagnosis bracelet based on multi-modal AI and Internet of Things. BACKGROUND

[0002] With the deepening of the aging of society, the demand for medical services and health management of the elderly group is increasing.

[0003] In real life, 90% of the elderly think that the medical process is complex and needs to be accompanied by diagnosis, and 70% of the respondents feedback that the hospital guidance is not clear. However, due to the current social work rhythm, children often cannot accompany the elderly to see a doctor on weekdays, and cannot grasp the physical health status of the elderly in real time, leading to the deterioration of some diseases due to the lack of timely intervention in the early stage.

[0004] Under this background, intelligent medical treatment becomes an important direction to solve the above problems, and the present application is based on big data processing, positioning system, wireless communication, Internet of Things and artificial intelligence technologies, aiming to provide comprehensive medical assistance and health management services for the elderly and special groups. SUMMARY

[0005] To solve the above problems, the present application aims to provide a medical tour diagnosis bracelet based on multi-modal AI and Internet of Things, which sets various hardware modules and functional modules in the bracelet body, and each module works together to realize comprehensive medical help and health monitoring functions.

[0006] To achieve the above purpose, the technical scheme of the present application is as follows: A medical tour diagnosis bracelet based on multi-modal AI and Internet of Things, comprising hardware modules arranged in the bracelet body and functional modules installed in a cloud AI analysis platform; the hardware modules include a core control module, a voice interaction module, a positioning and navigation module, a health monitoring module, a display module, a power module and a switch module; the functional modules include a GPS positioning and navigation function module, a data recording and analysis function module, an accompanying diagnosis function module, a daily health monitoring function module and an emergency communication function module; each hardware module and functional module is used for medical help and daily health management of the elderly and special groups.

[0007] Further, the voice interaction module includes an INMP441 microphone, a MAX98357I2 audio amplification module, and an 8-ohm 0.25-watt small horn speaker; the INMP441 microphone (107) is used to collect user voice signals, the MAX98357I2 audio amplification module performs noise reduction and amplification processing on the voice signals, and the 8-ohm 0.25-watt small horn speaker is used to output voice replies; the voice interaction function module realizes real-time interaction through voice recognition and natural language processing technology, supports six local accents and three foreign languages, and locally stores high-frequency medical questions and answers to adapt to network-free scenarios.

[0008] Further, the core control module uses an ESP32-S3 development board, supports Wi-Fi and Bluetooth connections, is used to transmit the collected audio data and health data to the background server in real time, and cooperates with the INMP441 microphone, the MAX98357I2 audio amplification module, and the 8-ohm 0.25-watt small horn speaker to realize conversion of voice signals and text and real-time interaction.

[0009] Further, the positioning and navigation module includes a GPS positioning GNSS module and a vibration feedback unit; the GPS positioning GNSS module is used to obtain the real-time position of the user, and the vibration feedback unit indicates a left turn by rapid vibration and a straight ahead by slow vibration; the GPS positioning and navigation function module combines hospital 3D map nodalization and dynamic path optimization algorithms, calculates a navigation path based on a GCN model, and the path weight formula is W=α•D+β•C+γ•M, where D is the distance, C is the congestion degree, M is the medical history matching degree, α+β+γ=1, and navigation services are provided through voice prompts and vibration feedback.

[0010] Further, the GCN model can update the node weights of the hospital 3D map in real time to adapt to changes in the hospital layout, and when the user inputs a target location, the system calculates an optimal path based on the updated node weights and provides navigation services.

[0011] Further, the health monitoring module includes a photoelectric volume sensor for collecting user heart rate and blood pressure health data; the daily health monitoring function module monitors user health data in real time, and when it detects that the blood sugar is too high, the heart rate is too fast, or the blood pressure is abnormal, it reminds the user through voice and vibration, and supports the user to set or modify up to three groups of medicine alarms through voice.

[0012] Further, the emergency communication function module automatically sends an alarm ring and position information to the pre-stored emergency contact mobile phone APP when monitoring the user's dangerous state, and supports emergency signal transmission in a network-free environment, ensuring that family members can learn about the user's situation in a timely manner; the data recording and analysis function uses big data processing technology and deep learning and neural network technology to store, clean and preprocess the collected user health data and doctor-patient communication recordings, detects and predicts trends through training of an AI model, and generates a report containing disease analysis, doctor's advice, medication time and daily precautions; wherein the doctor-patient communication recording is synchronized to the family member's mobile phone APP after noise reduction processing by an endpoint detection algorithm.

[0013] Further, the power module is a 3.7V lithium battery, and the 3.7V lithium battery is provided with a 5V2A charging and discharging integrated module for providing charging and discharging functions.

[0014] Further, the accompanying medical treatment function module allows the user to inquire at any time when visiting a hospital through voice interaction, and obtain the location, registration information, floor explanation and medication information of the hospital; at the same time, the user's entire diagnosis process is recorded, and the diagnosis information is synchronized to the patient's family member's mobile phone APP, facilitating follow-up treatment.

[0015] Further, the voice interaction function module adopts a multi-modal AI dynamic adaptation technology, adjusts according to the language and accent of each user through voice signal preprocessing and feature extraction technology, and automatically switches models according to the network status and battery capacity by combining the multi-model AI hot switching technology of mainstream AI platforms, to realize a personalized AI voice medical assistant.

[0016] Beneficial effects: The present application solves the problems of communication barriers, delayed health monitoring reminders and slow emergency response of the elderly and special groups when seeking medical treatment, and can provide voice interaction, health monitoring (heart rate, blood pressure, etc.), emergency communication, accompanying medical treatment, GPS positioning and navigation functions, etc. Health data is uploaded to the cloud through the Internet of Things to realize remote monitoring, which facilitates remote care by family members, improves user convenience, disease prevention ability and quality of life, and promotes the convenience and intelligence of medical and health services. BRIEF DESCRIPTION OF DRAWINGS

[0017] The accompanying drawings, which form a part of this application, are included to provide a further understanding of the application, and are incorporated in and constitute a part of this application. The embodiments of the present application and their descriptions are used to explain the present application, and do not constitute an improper limitation on the present application. In the drawings: Figure 1 The structure of the medical patrol bracelet based on multi-modal AI and Internet of Things described in the embodiments of the present application is shown in the structure diagram; Figure 2 The function diagram of the medical patrol bracelet based on multi-modal AI and Internet of Things described in the embodiments of the present application is shown in the function diagram. Figure 3 The core architecture schematic diagram of the medical tour diagnosis bracelet based on multi-modal AI and Internet of Things according to the embodiment of the present application is shown in the figure. Figure 4 The AI voice interaction function application example process diagram of the medical tour diagnosis bracelet based on multi-modal AI and Internet of Things according to the embodiment of the present application is shown in the figure. Figure 5 The home care application example process diagram of the medical tour diagnosis bracelet based on multi-modal AI and Internet of Things according to the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0018] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0019] The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0020] Embodiment 1 Reference Figures 1-5 A medical tour diagnosis bracelet based on multi-modal AI and Internet of Things, comprising a hardware module arranged in a bracelet body 1 and a function module installed in a cloud AI analysis platform; the hardware module comprises a core control module 101, a voice interaction module, a positioning and navigation module 102, a health monitoring module 103, a display module 104, a power module 105 and a switch module 106; the function module comprises a GPS positioning and navigation function module, a data recording and analysis function module, an accompanying medical function module, a daily health monitoring function module and an emergency communication function module; each hardware module and function module cooperates to help the elderly and special groups in medical treatment and daily health management.

[0021] This embodiment realizes the systematic design of the medical tour diagnosis bracelet through the cooperative architecture of the hardware module and the function module, ensures the orderly linkage of each component (such as core control, voice interaction, positioning and navigation, etc.), avoids functional fragmentation, provides an integrated solution of “medical help + health management” for the elderly and special groups, and solves the problem of single function of traditional devices that cannot meet the comprehensive needs.

[0022] Targeted adaptation advantage: specially designed for the use scene of the elderly and special groups, the configuration of hardware and function module is carried out around the core needs (such as simplifying operation, strengthening monitoring, emergency response, etc.), ensuring the practicality and ease of use of the device, improving the acceptance and wearing rate of the target users.

[0023] In a specific implementation, the bracelet body of the embodiment includes a detachable shell and a sleeve-shaped bracelet shell, which are spliced by using a mortise and tenon structure; each hardware module is integrated and installed on a PCB expansion board; the bracelet terminal interacts with a mobile phone APP through a cloud AI analysis platform; the display module uses an OLED 0.91-inch screen, and the switch module uses a small wave switch with a side corner support.

[0024] In a specific example, the voice interaction module includes an INMP441 microphone 107, a MAX98357I2 audio amplification module 108, and an 8-ohm 0.25-watt small horn speaker 109; the INMP441 microphone 107 is used to collect user voice signals, the MAX98357I2 audio amplification module 108 performs noise reduction and amplification processing on the voice signals, and the 8-ohm 0.25-watt small horn speaker 109 is used to output voice replies; the voice interaction function module realizes real-time interaction through voice recognition and natural language processing technology, supports 6 local accents and 3 foreign languages, and locally stores high-frequency medical questions and answers to adapt to network-free scenarios.

[0025] Efficient voice interaction advantage: the combination of microphone, audio amplification module, and speaker, together with noise reduction processing technology, ensures clear voice signal collection, accurate processing, and easy-to-understand output, solving the interaction barriers caused by hearing or accent problems of the elderly.

[0026] Multi-scene adaptation advantage: supporting 6 local accents and 3 foreign languages, taking into account the needs of users with different language backgrounds; local storage of high-frequency medical questions and answers can still provide basic services in network-free scenarios, avoiding function failure due to network problems and improving the reliability of the device.

[0027] In a specific example, the core control module 101 uses an ESP32-S3 development board, supports Wi-Fi and Bluetooth connection, is used to transmit the collected audio data and health data to the background server in real time, and cooperates with the INMP441 microphone 107, the MAX98357I2 audio amplification module 108, and the 8-ohm 0.25-watt small horn speaker 109 to realize the conversion of voice signals and text and real-time interaction.

[0028] Data transmission efficiency: the ESP32-S3 development board supports dual connection of Wi-Fi and Bluetooth, ensuring real-time transmission of audio, health, and other data, providing a high-speed channel for the whole process of "collection-processing-reply" of voice interaction, achieving millisecond-level response, and avoiding delay affecting user experience.

[0029] Function integration advantage: As the core control module, it cooperates with the voice interaction component to realize seamless conversion of "voice-text-voice", which can complete the interaction without complex operation, meet the use habits of the elderly "no touch screen dependence", and reduce the operation threshold.

[0030] In a specific example, the positioning and navigation module 102 includes a GPS positioning GNSS module and a vibration feedback unit; the GPS positioning GNSS module is used to obtain the real-time position of the user, and the vibration feedback unit indicates left turn by rapid vibration and straight ahead by slow vibration; the GPS positioning and navigation function module combines hospital 3D map nodalization and dynamic path optimization algorithm, calculates the navigation path based on the GCN model, and the path weight formula is W=α•D+β•C+γ•M, where D is the distance, C is the congestion degree, M is the disease history matching degree, and α+β+γ=1, and provides navigation service through voice prompt and vibration feedback.

[0031] Precise navigation advantage: GPS and GNSS modules cooperate to improve positioning accuracy, combine hospital 3D map nodalization data and dynamic path optimization algorithm, calculate the optimal path through the weight formula (comprehensive distance, congestion degree, and disease history matching degree), solve the problem of complex hospital environment and inaccurate traditional navigation, and reduce the user's time to find the way.

[0032] Multi-dimensional feedback advantage: The dual-channel design of voice prompt + vibration feedback (rapid = left turn, slow = straight ahead) takes into account hearing and touch, adapts to the possible hearing loss of the elderly, and ensures effective transmission of navigation information.

[0033] In a specific example, the GCN model can update the node weight of the hospital 3D map in real time to adapt to the changes of the hospital layout. When the user inputs the target location, the system calculates the optimal path based on the updated node weight and provides navigation service.

[0034] Dynamic adaptation advantage: The GCN model updates the node weight in real time, which can quickly adapt to the changes of the hospital layout (such as relocation of examination rooms, temporary closure of passages, etc.), avoid navigation errors caused by lagging map information, ensure that users always get the latest and most accurate path guidance, and improve the efficiency of medical treatment.

[0035] In a specific example, the health monitoring module 103 includes a photoelectric volume sensor for collecting user heart rate and blood pressure health data; the daily health monitoring function module monitors user health data in real time, and when detecting high blood sugar, rapid heart rate, and abnormal blood pressure, it reminds the user through voice and vibration, and supports the user to set or modify up to three groups of medicine alarms through voice.

[0036] Health monitoring real-time: The photoelectric plethysmography sensor continuously collects data such as heart rate and blood pressure, and abnormal conditions (such as rapid heart rate and abnormal blood pressure) are reminded in real time through voice and vibration, achieving early detection and early intervention of diseases and reducing sudden health risks.

[0037] Drug management humanization: Support voice setting / modification of up to three groups of medicine alarm clocks, in line with the habit of "voice operation first" of the elderly, solve the problem of forgetting to take medicine and taking wrong medicine, and improve the compliance of taking medicine.

[0038] In a specific example, the emergency communication function module automatically sends warning ringtones and location information to the pre-stored emergency contact mobile APP when monitoring the user's dangerous state, and supports emergency signal transmission in a network-free environment, ensuring that family members can learn about the user's situation in a timely manner; The data recording and analysis function uses big data processing technology and deep learning and neural network technology to store, clean and preprocess the collected user health data and doctor-patient communication recordings, detect and predict trends through training AI models, and generate reports containing disease analysis, doctor's advice, medicine taking time and daily precautions; Among them, the doctor-patient communication recording is synchronized to the family mobile APP after noise reduction processing by the endpoint detection algorithm (VAD).

[0039] Emergency response timeliness: Automatically send warnings and location information to emergency contacts in dangerous situations, and support network-free transmission, solving the pain point that no one knows about the emergency situation of the elderly living alone and gaining time for timely treatment.

[0040] Data value deepening: Through big data and deep learning technology to process health data and doctor-patient recordings, generate professional analysis reports, and recordings can be traced back after noise reduction processing, making it easy for family members to understand the diagnosis and treatment process, realizing "remote follow-up treatment", and strengthening the effectiveness of family care.

[0041] In a specific implementation: The implementation steps of the emergency communication function module of the embodiment include: Dangerous state triggering mechanism: The health monitoring module (including photoelectric plethysmography sensor, etc.) collects user heart rate, blood pressure and other data in real time, and generates a monitoring value every 5 seconds. When any of the following abnormal states is monitored for 3 times in a row, it is determined to be a "dangerous state": Blood pressure ≥ 160 / 95 mmHg or ≤ 90 / 60 mmHg; Heart rate ≥ 120 beats / minute or ≤ 50 beats / minute.

[0042] Emergency signal generation and transmission With network environment: the core control module (ESP32-S3 development board) immediately integrates the current location information (provided by the GPS positioning GNSS module), generates early warning information containing "dangerous state type + real-time positioning + monitoring data", and transmits it to the cloud server through Wi-Fi or Bluetooth, and then pushes it to the pre-stored emergency contact person's mobile phone APP, and triggers the continuous ringing (volume is the maximum volume of the mobile phone) and pop-up reminder of the APP end.

[0043] Without network environment: the bracelet locally starts the standby communication protocol, broadcasts emergency signals through Bluetooth (if there are family members' Bluetooth devices such as mobile phones nearby), or switches to the SMS channel (relying on the built-in low-power communication chip), and sends the simplified early warning information (including the approximate location area and the danger type) to the emergency contact person's mobile phone number, ensuring uninterrupted signal.

[0044] Contact confirmation and feedback: After the emergency contact person's mobile phone APP receives the early warning, it needs to click the "confirmed" button within 10 minutes and feedback to the bracelet end; if it is not confirmed within the time limit, the bracelet automatically dials the emergency contact person's phone number (plays the early warning content through voice synthesis technology) in the pre-stored order until it responds.

[0045] Implementation steps of data recording and analysis function: Data acquisition and storage Health data: the health monitoring module (photoelectric plethysmograph sensor, etc.) collects heart rate and blood pressure data every 30 seconds, which is transmitted to the ESP32-S3 development board through the PCB expansion board, and the original data (encrypted format) is stored locally for nearly 72 hours, and at the same time, it is synchronized to the cloud database through Internet of Things technology.

[0046] Doctor-patient communication recording: the INMP441 microphone automatically starts recording when the user is treated (distinguishes doctor-patient conversation through voiceprint recognition to avoid irrelevant noise), and the audio data is processed by the MAX98357I2 audio amplification module and temporarily stored in the bracelet local memory (supports continuous recording for 8 hours).

[0047] Data preprocessing Health data: adopt big data cleaning technology to eliminate outliers (such as jump data caused by sensor false touch), and use deep learning model (LSTM neural network) to smooth the continuous data and extract daily and weekly health index trends.

[0048] Recording processing: use endpoint detection algorithm (VAD) to remove environmental noise (such as hospital corridor noise) in the recording, and keep the clear doctor-patient conversation content, and then separate the doctor's and patient's voice segments through voiceprint recognition technology to generate timestamped structured recording files.

[0049] AI analysis and report generation The cloud server calls the trained AI model (such as Qwen 2.5, deep seek 3V), performs trend prediction (such as the risk of blood pressure rising in the next 7 days) on the preprocessed health data, and generates disease analysis combined with the user's medical history (pre-existing in the cloud).

[0050] For doctor-patient recordings, key information (such as diagnosis conclusion, medication dosage, and review time) is extracted through natural language processing (NLP) technology, and automatically organized into "doctor's advice", "medication time", "daily precautions", etc. Form a structured report.

[0051] Data synchronization and backtracking The processed health report and noise reduction recording are synchronized to the family member's mobile phone APP through an encrypted channel, supporting time axis review (such as "2024 October 15 Heart Department Recording" and "Blood Pressure Fluctuation Report in the Past Week").

[0052] Family members can manually download or share reports to the doctor's end through the APP to assist remote diagnosis and treatment; the bracelet locally retains nearly 30 days of report cache for easy offline viewing by the user.

[0053] Through the above implementation, the emergency communication function module ensures rapid response in dangerous situations, and the data recording and analysis function realizes the whole process management of health data, providing accurate and traceable health management basis for users and family members.

[0054] In a specific example, the power module 105 is a 3.7V lithium battery, which is equipped with a 5V2A charging and discharging integrated module 1051 for providing charging and discharging functions.

[0055] Endurance and charging convenience: 3.7V lithium battery with 5V2A charging and discharging integrated module ensures long-term endurance of the device, while simplifying the charging operation (no need for complex wiring), meeting the needs of the elderly for "long endurance and easy charging" of the device, avoiding function interruption due to insufficient power.

[0056] In a specific example, the accompanying medical function module allows users to inquire at any time during hospital visits through voice interaction, obtaining the location, registration information, floor commentary, and medication information of the hospital; at the same time, it records the entire diagnosis process and synchronizes the diagnosis information to the patient's family member's mobile phone APP, facilitating follow-up treatment.

[0057] Medical assistance comprehensiveness: real-time provision of registration, floor, and medication information through voice interaction, equivalent to a "personal accompanying doctor", solving the problem of the elderly's unfamiliarity with the medical process and lack of accompanying personnel, and improving the self-reliance of medical treatment.

[0058] Information synchronization advantage: diagnostic process record and information synchronization to family APP, realizing the whole process understanding of family to the treatment process, facilitating the follow-up cooperation of doctors for rehabilitation management, forming the collaborative closed loop of "doctor-patient-user-family".

[0059] In a specific example, the voice interaction function module adopts a multi-modal AI dynamic adaptation technology, adjusts according to the language and accent of each user through voice signal preprocessing and feature extraction technology, combines the multi-model AI hot switching technology of mainstream AI platforms (Qwen 2.5, deep seek 3V, bean bag 1.5Pro, Qwen Max 2.5, etc.), automatically switches the model according to the network state and battery capacity, and realizes the personalized AI voice medical assistant.

[0060] Personalized service advantage: multi-modal AI dynamic adaptation technology adjusts the interaction mode according to the language and accent of the user, realizes the "thousand faces" voice assistant experience, solves the problem of "standardized interaction not fitting individual needs" of traditional devices, and improves the use comfort.

[0061] Resource optimization advantage: multi-model AI hot switching technology automatically switches the model according to the network state and battery capacity, reduces energy consumption and network occupation while ensuring function, prolongs the endurance and avoids the influence of network fluctuation, and improves the operation efficiency of the device.

[0062] The above only describes the preferred embodiments of the present application and does not limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A multi-modal AI and Internet of Things based medical patrol wristband, characterized in that, The application relates to a health management bracelet for the elderly and special groups, which comprises a hardware module arranged in a bracelet body (1) and a functional module installed in a cloud AI analysis platform; the hardware module comprises a core control module (101), a voice interaction module, a positioning and navigation module (102), a health monitoring module (103), a display module (104), a power module (105) and a switch module (106); the functional module comprises a GPS positioning and navigation function module, a data recording and analysis function module, an accompanying medical treatment function module, a daily health monitoring function module and an emergency communication function module; the hardware module and the functional module are cooperatively used for medical treatment help and daily health management of the elderly and special groups.

2. The bracelet of claim 1, wherein, The voice interaction module comprises an INMP441 microphone (107), a MAX98357I2 audio amplification module (108) and an 8-ohm 0.25-watt small horn loudspeaker (109); the INMP441 microphone (107) is used for collecting user voice signals, the MAX98357I2 audio amplification module (108) is used for carrying out noise reduction and amplification processing on the voice signals, and the 8-ohm 0.25-watt small horn loudspeaker (109) is used for outputting voice replies; the voice interaction function module realizes real-time interaction through voice recognition and natural language processing technology, supports six local accents and three foreign languages, and locally stores high-frequency medical questions and answers to adapt to a network-free scene.

3. The bracelet of claim 2, wherein, The core control module (101) adopts an ESP32-S3 development board, supports Wi-Fi and Bluetooth connection, is used for real-time transmission of collected audio data and health data to a background server, and is matched with the INMP441 microphone (107), the MAX98357I2 audio amplification module (108) and the 8-ohm 0.25-watt small horn loudspeaker (109) to realize conversion of voice signals and texts and real-time interaction.

4. The bracelet of claim 1, wherein, The positioning and navigation module (102) comprises a GPS positioning GNSS module and a vibration feedback unit; the GPS positioning GNSS module is used for acquiring a real-time position of a user, and the vibration feedback unit indicates left turning through rapid vibration and straight going through slow vibration; the GPS positioning and navigation function module is combined with hospital 3D map nodalization and a dynamic path optimization algorithm, a navigation path is calculated based on a GCN model, a path weight formula is W=alpha•D+beta•C+gamma•M (wherein D is distance, C is congestion degree, M is disease history matching degree, alpha+beta+gamma=1), and navigation service is provided through voice prompt and vibration feedback.

5. The bracelet of claim 4, wherein, The GCN model can update node weights of the hospital 3D map in real time to adapt to hospital layout changes, when a user inputs a target position, the system calculates an optimal path based on the updated node weights and provides navigation service.

6. The bracelet of claim 1, wherein, The health monitoring module (103) comprises a photoelectric volume sensor and is used for collecting user heart rate and blood pressure health data; the daily health monitoring function module monitors user health data in real time, when high blood sugar, rapid heart rate and abnormal blood pressure are detected, the user is reminded through voice and vibration, and the user can set or modify up to three groups of medicine alarms through voice.

7. The bracelet of claim 1, wherein, The emergency communication function module automatically sends warning ringtones and location information to the pre-stored emergency contact mobile phone APP when monitoring the user's dangerous state, and supports emergency signal transmission in a network-free environment, ensuring that family members can learn about the user's situation in a timely manner; The data recording and analysis function uses big data processing technology and deep learning and neural network technology to store, clean and preprocess the collected user health data and doctor-patient communication recordings, detects and predicts trends through training AI models, and generates reports containing disease analysis, doctor's advice, medication time and daily precautions; Among them, the doctor-patient communication recording is synchronized to the family member's mobile phone APP after noise reduction processing by the endpoint detection algorithm.

8. The bracelet of claim 1, wherein, The power module (105) is a 3.7V lithium battery, and the 3.7V lithium battery is provided with a 5V2A charging and discharging integrated module (1051) for providing charging and discharging functions.

9. The bracelet of claim 1, wherein, The accompanying medical function module allows users to inquire at any time when they are in the hospital for treatment, obtain the location, registration information, floor commentary, and medication information of the hospital for treatment; at the same time, records the entire diagnosis process of the user, and synchronously uploads the diagnosis information to the patient's family member's mobile phone APP, facilitating follow-up treatment.

10. The bracelet of claim 1, wherein, The voice interaction function module adopts a multi-modal AI dynamic adaptation technology, adjusts according to the language and accent of each user through voice signal preprocessing and feature extraction technology, and automatically switches models according to network status and battery capacity by combining the multi-model AI hot switching technology of mainstream AI platforms, to realize personalized AI voice medical assistants.

Images