Acoustic stimulation nerve regulation service management system combining AI and big data analysis

Abstract

The invention discloses an acoustic stimulation nerve regulation and control service management system combining AI and big data analysis, and the system comprises a cloud center layer, a region edge layer, a doctor management layer and a patient terminal layer which are connected with one another, the cloud is provided with a client management module, and an exclusive account is established for each patient; and an individualized sound stimulation scheme is generated and dynamically optimized by using an AI algorithm to realize remote closed-loop regulation and control. The platform supports remote demonstration and teaching of experts, autonomous operation and maintenance of clients and a mutual assistance network of the experts, and a cloud-side-end collaborative AIoT intelligent system is formed. The system is provided with a national-level data supervision interface, so that safety and controllability are ensured. Through the above mode, the AI and big data analysis combined acoustic stimulation nerve regulation and control service management system improves the accuracy and efficiency of personalized scheme formulation and optimization by constructing a collection-analysis-decision-execution-feedback-optimization closed-loop process; therefore, cloud-edge-end-human collaborative closed-loop intelligent regulation and control management with individualized nerve regulation and control and compliance safety considered is formed.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence and digital medical technology, and in particular to an acoustic stimulation neuromodulation service management system that combines AI and big data analysis. Background Technology

[0002] In recent years, acoustic stimulation neuromodulation (ASN) has been applied and clinically validated in the treatment of tinnitus, sudden deafness and tinnitus, sleep disorders, anxiety, depression, and otogenic vertigo. However, existing modulation systems still have the following problems in their widespread application:

[0003] 1. The development of personalized treatment plans relies solely on expert experience, lacking parameter generation and optimization mechanisms supported by AI algorithms and big data analysis;

[0004] 2. The remote intervention and follow-up system is weak, making it impossible to dynamically adjust the treatment plan according to the patient's real-time condition;

[0005] 3. The lack of a cross-institutional expert collaboration platform makes it difficult for experts to share experience and help each other when encountering difficult cases;

[0006] 4. The lack of a structured management and relearning mechanism for clinical big data hinders algorithm training and scientific research translation;

[0007] 5. Significant gaps in data security and regulation limit the large-scale application of AI in the medical field.

[0008] Therefore, there is an urgent need for an innovative neuromodulation platform that integrates AI intelligent decision-making, big data learning, cloud-edge-device collaborative control, remote expert teaching, personalized treatment plan formulation, and national-level safety supervision. Summary of the Invention

[0009] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0010] This invention provides an acoustic stimulation neuromodulation service management system that combines AI and big data analytics. It integrates an AIoT (Artificial Intelligence of Things) architecture, edge computing technology, a cloud-based multi-expert collaborative system, and a medical data security supervision mechanism. The system comprises interconnected cloud-based central layers, regional edge layers, physician management layers, and patient terminal layers. Data transmission is achieved through bidirectional secure encrypted communication, constructing a complete digital closed-loop process of data acquisition, analysis, decision-making, execution, feedback, and optimization. This enables closed-loop intelligent and precise neuromodulation through cloud-edge-device-human collaboration.

[0011] The cloud-based central layer includes an AI algorithm and big data analysis module, a cloud storage and customer management module, an expert collaboration and task scheduling module, an AI model and computing power management module, and a national medical data supervision interface module. The AI ​​algorithm and big data analysis module, based on deep learning and reinforcement learning models, performs big data analysis on patients' multimodal data to formulate and distribute individualized acoustic stimulation neuromodulation treatment plans to the patient terminal layer. It dynamically adjusts and optimizes these plans based on real-time feedback. The individualized acoustic stimulation neuromodulation treatment plans consist of digital stimulation signal parameters that can be transmitted over a network. This unique physical intervention advantage allows the plans to be transmitted from the cloud-based central layer to the regional edge layer, the physician management layer, and the patient terminal layer, thereby achieving smooth execution of the plans and accurate output of stimulation signals. The expert collaboration and task scheduling module is used to build a multi-expert horizontal communication network for remote collaboration. The national medical data supervision interface module is used to achieve real-time synchronization and compliance supervision of clinical data, algorithm logs, and model call records.

[0012] The edge layer of the region includes AI edge expert terminals with AI edge computing capabilities. These AI edge expert terminals connect upwards to the cloud center layer and downwards to the doctor management layer and the patient terminal layer, and are horizontally interconnected. The AI ​​edge expert terminals include:

[0013] Data preprocessing module: Filters, removes artifacts, and compresses features on collected EEG, MEG, fNIRS, ERP-P300, HRV, hearing threshold, and psychoacoustic test data to reduce upload bandwidth.

[0014] Edge computing module: Calls the model of the cloud central layer to independently execute AI inference and generate and optimize individualized solutions locally, achieving millisecond-level response;

[0015] Distributed encrypted data caching module: Automatically saves local feedback data when the network is interrupted, and automatically synchronizes it with the cloud model and database after the network connection is restored; performs distributed encryption, key classification and local authentication to ensure regional privacy compliance;

[0016] Remote Doctor / Patient Guidance Module: Used to enable remote guidance and demonstration of operations by platform / edge experts;

[0017] The physician management layer is used for intelligent control of information monitoring, individualized treatment plan editing, cloud platform / regional edge expert teaching operations, and hospital patient data storage.

[0018] The patient terminal layer is used for the execution of the protocol and the collection of feedback information.

[0019] In a preferred embodiment of the present invention, the cloud center layer further includes a cloud storage and customer management module, which is used to establish independent exclusive accounts or sub-modules for end customers and store all records and parameter information in the database for model retraining and knowledge base updates.

[0020] In a preferred embodiment of the present invention, the cloud center layer cloud computing support module further includes an AI model and computing power management submodule for accessing an external AI cloud service platform.

[0021] In a preferred embodiment of the present invention, the regional edge layer further includes: a data preprocessing module, an edge computing cloud interface module, and a doctor management input interface module.

[0022] In a preferred embodiment of the present invention, the physician management layer includes a patient view and efficacy monitoring interface module, an individualized treatment plan adjustment module, an expert mentoring and skills transfer module, and a task scheduling and intelligent reminder module.

[0023] In a preferred embodiment of the present invention, the physician management layer further includes a patient consultation / scale completion module, a test data input port module, a regional edge layer expert monitoring access port module (including hospital patient data storage), and an expert teaching and clinical demonstration module.

[0024] In a preferred embodiment of the present invention, the patient terminal layer includes a protocol execution module, a physiological signal acquisition module, an intelligent feedback module, and an AI follow-up assistant module.

[0025] In a preferred embodiment of the present invention, the scheme execution module includes a wearable mobile sound stimulation therapy device, a medical institution personalized sound stimulation therapy device, and a wearable hearing aid.

[0026] In a preferred embodiment of the present invention, the cloud central layer, the regional edge layer, the doctor management layer, and the patient terminal layer constitute an AI intelligent agent to achieve bidirectional interconnection between the physical world and the information world through the Internet of Things, comprising:

[0027] Digital twin patient model: Based on the patient's multimodal information data, an individualized virtual-reality mapping model is established to realize a digital twin patient profile;

[0028] Real-time dynamic mapping: The AI ​​learning process is linked with the actual feedback from patients so that when patient information data changes, it is updated and fed back to the AI ​​algorithm and big data analysis module;

[0029] AI self-learning closed loop: Utilizing group data and individual feedback to continuously train the model, achieving vertical closed-loop management and horizontal multi-cloud platform / regional edge expert collaboration.

[0030] A control method for an acoustic stimulation neuromodulation service management system combining AI and big data analytics, comprising the following steps:

[0031] (1) Data collection and uploading: Real-time collection and uploading of patients' multimodal information data, including but not limited to electroencephalogram (EEG), magnetoencephalogram (MEG), functional near-infrared spectroscopy (fNIRS), event-related potentials (ERP-P300), heart rate variability (HRV), hearing threshold, psychoacoustics, and questionnaire data;

[0032] (2) AI analysis and solution generation: Cloud-based AI models and big data analysis of multimodal data generate initial individualized acoustic stimulation neuromodulation treatment plans; among them, experts at the platform / regional edge layer can participate in the formulation or modification of modulation treatment plans through network mutual assistance and collaboration;

[0033] (3) Review and distribution of the plan: After the regional edge layer experts and / or physician management physicians review or adjust the individualized sound stimulation neuromodulation treatment plan, it is distributed to the patient terminal layer;

[0034] (4) Protocol execution feedback: During the protocol execution process, patient feedback information data is collected and uploaded in real time;

[0035] (5) Closed-loop optimization: The AI ​​algorithm and big data analysis module updates or optimizes the individualized acoustic stimulation neuromodulation treatment plan based on feedback information data to achieve dynamic iterative optimization;

[0036] (6) Synchronized supervision: All process data is synchronized to the medical data center through the national medical data supervision interface module for supervision, realizing safe, intelligent and closed-loop remote neuromodulation therapy.

[0037] The beneficial effects of this invention are as follows: By connecting the cloud central layer, regional edge layer, doctor management layer, and patient terminal layer, a closed-loop process of data collection, analysis, decision-making, execution, feedback, and optimization is constructed. This forms a cloud-edge-terminal-human collaborative, personalized, and compliant intelligent control and management system, which effectively improves the efficiency, accuracy, effectiveness, and security of solution formulation and optimization. Through the system network connecting one or more cloud central layers, hundreds of regional edge layers, experts, tens of thousands of clinicians, and millions of patients, it provides safe and efficient personalized and precise acoustic stimulation neuromodulation therapy, rehabilitation, and chronic disease management for a large-scale population of neurological and cognitive dysfunction patients distributed throughout the country. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein:

[0039] Figure 1 This is the overall architecture diagram of the acoustic stimulation neuromodulation service management system that combines AI and big data analysis according to the present invention;

[0040] Figure 2 This is a schematic diagram of the platform cloud-edge-device-human collaborative structure and functional module distribution in the acoustic stimulation neuromodulation service management system that combines AI and big data analysis according to the present invention.

[0041] Figure 3 This is a block diagram of the AI ​​and big data analysis module in the acoustic stimulation neuromodulation service management system that combines AI and big data analysis according to the present invention.

[0042] Figure 4 This is a flowchart of the patient-doctor-cloud three-party interaction process in the acoustic stimulation neuromodulation service management system that combines AI and big data analysis according to the present invention.

[0043] Figure 5 This is a closed-loop control logic diagram of reinforcement learning in the acoustic stimulation neuromodulation service management system that combines AI and big data analysis according to the present invention.

[0044] Figure 6 This is a diagram of the acoustic stimulation signal generation and transmission path in the acoustic stimulation neuromodulation service management system that combines AI and big data analysis according to the present invention.

[0045] Figure 7 This is a schematic diagram of the closed loop of expert collaborative teaching and patient efficacy management in the acoustic stimulation neuromodulation service management system that combines AI and big data analysis according to the present invention.

[0046] Figure 8 This invention relates to an AIoT intelligent agent and a national regulatory framework diagram within an acoustic stimulation neuromodulation service management system that combines AI and big data analysis. Detailed Implementation

[0047] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] Please see Figure 1-8 The embodiments of the present invention include:

[0049] An acoustic stimulation neuromodulation service management system that combines AI and big data analytics, possessing the following capabilities:

[0050] • Artificial intelligence and big data analysis and decision-making—automatically collect and fuse multimodal physiological and psychological information data, including electroencephalogram (EEG), magnetoencephalogram (MEG), functional near-infrared spectroscopy (fNIRS), event-related potentials (ERP-P300), heart rate variability (HRV), hearing threshold, body temperature, psychoacoustics, and questionnaire scale information, for modeling and efficacy prediction. Based on deep learning and reinforcement learning algorithms, it achieves personalized acoustic stimulation neuromodulation parameter optimization, and utilizes AI algorithms and big data analysis to generate, iteratively adjust, and control personalized acoustic stimulation programs.

[0051] • Intelligent Internet of Things (AIoT) and Edge Computing technologies enable cloud-edge-device collaborative closed-loop control, establish a four-level collaborative system of "expert-doctor-patient-cloud", build a cloud-based expert mutual assistance network, realize real-time collaboration and sharing of difficult cases, ensure personalized diagnosis and treatment effects and service quality, support remote training, demonstration and capability transfer, and help primary hospitals establish multidisciplinary team (MDT) disease centers;

[0052] • Medical information security and national-level data supervision – Ensuring the compliance and security of medical data through encrypted transmission and authorized cloud supervision mechanisms.

[0053] This will lead to the formation of a medical intelligent system based on AIoT and digital twins, realizing the integration of the physical world and the information world.

[0054] I. System Overall Structure

[0055] The AcousticStimulation AI-Cloud Neuromodulation Platform (AS-Cloud) system, which combines AI and big data analysis, adopts a four-layer collaborative architecture consisting of a cloud central layer, a regional edge layer, a physician management layer, and a patient terminal layer. This architecture forms a real-time data interaction and closed-loop feedback relationship between the cloud, edge, terminal, and human, with cloud computing nodes supporting the network.

[0056] The system achieves bidirectional data transmission through secure encrypted communication (TLS / SSL), constructing a complete digital closed-loop process of "collection—analysis—decision-execution—feedback—optimization". The platform supports concurrent remote precision diagnosis and treatment operations for thousands of experts, tens of thousands of doctors, and hundreds of thousands of patients, and is compatible with medical devices, mobile apps, and wearable devices.

[0057] II. Cloud Layer

[0058] The cloud-based central layer is the system's intelligent control and core data processing unit, including the cloud server system, which undertakes functions such as big data processing, AI model training, and customer account management. It includes the following modules:

[0059] (1) AI Algorithm and Big Data Analysis Module

[0060] The AI ​​algorithm and big data analysis module constructs a three-dimensional mapping model of "acoustic stimulation parameters – neural feedback information – therapeutic effect" based on deep learning and reinforcement learning algorithms (DNN + RL). The model's input includes offline collected data such as EEG, MEG, fNIRS, ERP-P300, HRV, hearing threshold, psychoacoustic and behavioral tests, body temperature and emotion questionnaires, etc. Through the policy network model π(a|S), it obtains individualized acoustic stimulation parameters (frequency, amplitude, rhythm, vocal tract balance, duration, etc.) and formulates individualized acoustic stimulation neuromodulation treatment plans. It performs self-learning based on the obtained patient feedback information to dynamically adjust and optimize the individualized acoustic stimulation neuromodulation treatment plans.

[0061] The internal logical structure and data flow path of the AI ​​algorithm and big data analysis module include:

[0062] Input data layer: EEG, MEG, fNIRS, ERP-P300, HRV, hearing threshold (HTD), psychoacoustic (TAT), body temperature and self-assessment questionnaire information, etc.

[0063] • Feature fusion layer: performs feature extraction and weighted fusion of multimodal physiological and behavioral signals;

[0064] AI Decision Core Layer:

[0065] Deep neural networks (DNNs) are used for multidimensional feature mapping;

[0066] The reinforcement learning policy network π(a|S) is responsible for generating the optimal acoustic stimulus parameters in real time;

[0067] The self-learning function is used for model retraining and dynamic updates.

[0068] • Output layer: Individualized acoustic stimulus parameter set (frequency, rhythm, amplitude, tract balance, duration, etc.).

[0069] • Data feedback layer: Physiological and subjective feedback signals uploaded in real time by the patient are fed back to the AI ​​model for correction and re-optimization.

[0070] like Figure 5 The reinforcement learning closed-loop control logic in the dynamic optimization of individualized acoustic stimulation schemes described in this application includes:

[0071] a. Input signals: EEG power spectrum, phase synchronization index (PSI), HRV standard deviation (SDNN), subjective emotion score and other information data;

[0072] b. State Evaluation: Analyzing the current neurological functional status and treatment trends;

[0073] c. Policy Decision (π(a|S)): Generate new stimulus parameters based on the real-time state;

[0074] d. Execution Unit: Drives the acoustic stimulation device to output a new scheme;

[0075] e. Feedback: Collect signals of changes in treatment efficacy and return them to the AI ​​model to achieve self-learning;

[0076] f. Safety Gate: Immediately pauses or adjusts the output when abnormal physiological responses (such as sudden changes in heart rate, sleep awakening, etc.) are detected.

[0077] (2) Cloud storage and customer management module

[0078] The cloud storage customer management module establishes a dedicated account or sub-module for each end-user (such as hospitals, rehabilitation centers, and individual doctor accounts). With the assistance of offline doctor clients, cloud platform experts can remotely create patient case files, collect examination data, conduct video consultations, and develop individualized acoustic stimulation neuromodulation treatment plans based on the patient's physiological and psychological information and questionnaire data. After the patient experiences the treatment, parameters are iteratively adjusted and optimized based on feedback until a stable, accurate, and effective treatment plan is obtained for long-term patient treatment and rehabilitation. All collaboration records and parameter modification results are stored in the cloud database.

[0079] This module allows customers to manage their patients and treatments independently after logging into the cloud, forming a progressive management mechanism of "expert guidance - customer independence".

[0080] (3) Expert Collaboration and Task Scheduling Module

[0081] Expert Collaboration and Task Scheduling Module: This module establishes a multi-expert horizontal communication network, supporting multiple experts to simultaneously guide multiple offline or online doctors. It facilitates assistance with difficult cases, experience sharing, and horizontal collaboration among experts, ensuring treatment effectiveness and service quality. When an expert encounters a difficult case, they can initiate a "collaboration request" with a single click through this module. The system automatically schedules suitable experts to join, forming a remote joint consultation, collaborative solution creation, and experience sharing.

[0082] The system can save collaboration records and parameter change trajectories to a cloud database for model retraining and knowledge base updates.

[0083] (4) AI Model and Computing Power Management Module

[0084] AI Model and Computing Power Management Module: By connecting to external commercial AI cloud service platforms (such as Alibaba Cloud Service Platform), users can pay to rent their large AI models and AI computing power.

[0085] The cloud center layer can rent large-scale AI models, AI computing power resources, and data storage services from nationally recognized commercial cloud platforms (such as Alibaba Cloud, Huawei Cloud, Baidu Cloud, etc.) to perform model training and multi-center task distribution suitable for the platform of this invention.

[0086] The system adopts a distributed training architecture and Kubernetes scheduling, supporting cross-center (regional edge layer) model synchronization and load balancing to ensure high availability and computational security.

[0087] (5) National Medical Data Supervision Interface Module

[0088] National Medical Data Supervision Interface Module: A medical data supervision terminal is set up in the cloud and directly interconnected with the nationally authorized medical data center. All clinical data, logs and model call records are synchronized to the supervision terminal in real time, realizing traceability and regulatory compliance of the entire process of medical data transmission, storage and retrieval.

[0089] This module complies with the Personal Information Protection Law, the Cybersecurity Law, and the medical device software lifecycle standard YY / T0664-2020.

[0090] (6) AI Intelligent Agent Module (AI Big Data Internet of Things AIoT)

[0091] like Figure 2 and Figure 8 As shown, the system of this application acts as an intelligent agent, realizing bidirectional interconnection and seamless integration between the physical world (experts, doctors, patients, devices) and the information world (AI algorithms, data, solutions, learning, education). This AIoT structure enables the system to not only provide treatment services, but also to perform dynamic prediction, risk warning and long-term rehabilitation management.

[0092] 1. Digital Twin Patient Model: Based on the patient's physiological and psychological test information, historical treatment information, and environmental factors, an individualized virtual-reality mapping model is established to realize a "digital twin patient profile".

[0093] 2. Real-time dynamic mapping: The AI ​​learning process and the patient's actual feedback empower each other to form an intelligent collaborative ecosystem. When the patient's physiological and / or psychological state changes, the cloud-based digital twin patient model is updated in real time and fed back to the AI ​​algorithm and big data analysis module.

[0094] 3. AI self-learning closed loop: Utilizing patient group data and individual feedback to continuously train the model, it achieves vertical integrated intelligent closed-loop management of "perception-decision-control-optimization", and can also realize horizontal multi-cloud platform / regional edge expert collaboration.

[0095] The cloud-based central layer of this application introduces AI decision-making algorithms, customer management modules, expert mutual assistance networks, AIoT intelligent connections, and national-level data supervision interfaces into the cloud platform. This enables a fully digital closed-loop system that covers the entire process from patient data collection, AI algorithms and big data analysis, individualized parameter generation, remote treatment plan formulation to efficacy feedback and re-optimization, forming a disease-specific intelligent neuromodulation system with self-learning, self-management, and multi-center collaboration.

[0096] III. Edge Layer

[0097] The regional edge layer is deployed at regional medical centers or service agent nodes, undertaking functions such as local computing, data preprocessing, low-latency AI inference, real-time response tasks and caching, and remote expert guidance for doctors in diagnosing and treating patients. Specifically, it includes:

[0098] • Local execution of AI model inference and rapid response reduces cloud load;

[0099] • Automatically cache treatment feedback data in case of network latency or interruption;

[0100] • Supports regional data encryption and distributed secure storage;

[0101] • Provide a real-time visualization interface for expert terminals;

[0102] • Provide real-time, personalized clinical diagnosis and treatment support, mentorship, training, and learning for primary care physicians.

[0103] The region edge layer includes the following modules:

[0104] (1) Data preprocessing module: Filters, removes artifacts and compresses features on the collected EEG, MEG, fNIRS, ERP-P300, HRV, hearing threshold, psychoacoustic test and other information data to reduce upload traffic.

[0105] (2) Edge computing module: calls the model of the cloud center layer, performs AI inference and individualized parameter generation locally, and achieves millisecond-level response.

[0106] (3) Distributed encrypted data caching module: Automatically saves local feedback data when the network is interrupted and uploads it synchronously after the connection is restored; performs distributed encryption, key classification and local authentication to ensure regional privacy compliance.

[0107] (4) Edge computing cloud interface module: used to realize the interconnection between the edge computing module and the data preprocessing module, the distributed encrypted data caching module, etc.;

[0108] (5) Remote guidance module for doctors / patients: Platform / edge experts can use this module to provide remote guidance and demonstrations;

[0109] (6) Doctor Management Input Interface Module: Used to connect to and receive information or requests from the doctor management team;

[0110] (7) AI Edge Expert Terminal

[0111] Each cloud platform / regional edge expert is equipped with an offline terminal (Expert EdgeNode) with AI edge computing capabilities, supporting local model inference and small-to-medium-sized data analysis. It can independently complete patient data processing and treatment plan generation without cloud connection and maintain periodic synchronization with the cloud.

[0112] AI edge expert terminals have the following characteristics:

[0113] • Built-in small and medium-sized AI analysis module, capable of independent model reasoning and scheme adjustment;

[0114] • Preliminary analysis of patient data and prediction of treatment efficacy can be completed without a network connection;

[0115] • Automatically and periodically synchronize with the cloud-based central layer to ensure data and model consistency;

[0116] • Supports secure remote login and multi-expert collaborative debugging.

[0117] The AI ​​edge expert terminal can connect upwards to the cloud central layer, downwards to the doctor management layer and patient terminal layer, and horizontally to other AI edge expert terminals. This design improves expert work efficiency and system reliability, while reducing cloud dependence and response latency.

[0118] IV. Doctor Layer

[0119] The physician management system serves as an intelligent control console for the operation and collaborative interaction between regional edge experts and clinicians. It provides functions such as a visual monitoring interface, individualized treatment plan editing, cloud platform / regional edge expert mentoring, and hospital-wide patient data storage. Specifically, it includes:

[0120] (1) Patient view and efficacy monitoring interface module

[0121] It displays in real time the patient's hearing curve, psychoacoustic test, EEG / MEG / fNIRS / ERP-P300 monitoring and analysis, HRV, hearing threshold, changes in THI (Tinnitus Severity Scale) / PSQI (Pittsburgh Sleep Scale) / HAMA (Hamilton Anxiety Scale) / HAMD (Hamilton Depression Scale) and other scales, changes in body temperature and mood questionnaire scores, sound stimulation records, and efficacy evaluation (characteristic value) trend charts, etc.

[0122] (2) Individualized Solution Adjustment Module

[0123] Doctors can fine-tune parameters (such as sound stimulation frequency, amplitude, rhythm, and duration) based on AI recommendations, and then send the results to the patient's device after confirmation.

[0124] (3) Expert mentoring and competency transfer module

[0125] Cloud-based or regional edge layer experts can remotely demonstrate treatment plan formulation, parameter adjustment, and efficacy evaluation in real time, while doctors can simultaneously view the entire process through the expert teaching and skills transfer module. Through long-term clinical teaching (6–24 months), doctors gradually acquire the ability to independently handle common and difficult clinical cases, achieving a phased shift from "expert-led" to "customer-driven operation and maintenance."

[0126] (4) Task scheduling and intelligent reminder module

[0127] Tasks are automatically prioritized based on the severity of the patient's condition, the stage of treatment, and feedback status. When stagnation or abnormal fluctuations in treatment efficacy are detected, the system automatically reminds the doctor to conduct a follow-up examination and suggests intervention strategies.

[0128] (5) Patient interview / scale completion module

[0129] Doctors can use this module to take patient consultations and fill in questionnaire information.

[0130] (6) Detection data input port module

[0131] The detection data input port module can connect to various offline detection devices via wired or wireless means to acquire monitoring data in real time.

[0132] (7) Regional edge layer expert monitoring access port module

[0133] The regional edge layer expert monitoring access port module enables connection between the doctor's and expert's terminals, allowing experts to provide real-time guidance and monitoring.

[0134] V. Client Layer

[0135] The patient terminal layer serves as the entry point for protocol implementation and feedback collection, used for individualized chronic disease treatment and rehabilitation feedback and follow-up, specifically including:

[0136] • Receive treatment plans from the cloud, regional terminal, or doctor's terminal and implement treatment;

[0137] • Collect test data, including hearing, EEG, MEG, fNIRS, ERP-P300, HRV, psychoacoustics, and subjective scores;

[0138] • Automatically upload treatment feedback to the cloud;

[0139] • Supports AI-powered follow-up reminders and self-rehabilitation guidance;

[0140] It features local storage and offline functionality, allowing the system to run independently in environments without a network connection and delay data uploads.

[0141] The patient terminal layer mainly includes the following functional modules:

[0142] (1) Protocol execution module: Receives individualized acoustic stimulation neuromodulation treatment protocol and drives the acoustic output module (amplitude / frequency / wideband noise or bi-beat / envelope modulation noise, AM / FM / BB / EN modulation mode, etc.) to output signals according to the parameter information in the protocol; mainly includes wearable mobile (home) acoustic stimulation therapy instrument (module), medical institution personalized acoustic stimulation therapy instrument (module), mobile APP, wearable hearing aid instrument (module) or its integrated device.

[0143] (2) Physiological signal acquisition module: Acquires information such as EEG, MEG, fNIRS, ERP-P300, HRV, body temperature, sleep activity, and subjective scale scores.

[0144] (3) Intelligent feedback module: Automatically records changes in treatment efficacy, compliance data and subjective experience information, and uploads them to the cloud.

[0145] (4) AI follow-up assistant module: Combines automatic push reminders during the treatment cycle, education and rehabilitation guidance content and questionnaire feedback, follow-up doctors to conduct tests and assessments, etc., to enhance patient compliance and improve the quality of chronic disease management.

[0146] like Figure 4 As shown, the three-way information interaction and closed-loop control logic of the system in this application during remote diagnosis and treatment—the patient's end, the doctor's end, and the cloud—is as follows:

[0147] Patient-side: Collect physiological and psychological data → Upload to the cloud;

[0148] Cloud Center: AI models generate or update personalized sound stimulation plans → automatically distribute them to doctors;

[0149] Doctor's side: Cloud platform / regional edge experts review and fine-tune the plan → distribute it to the patient's side for execution;

[0150] Feedback phase: Patients complete the treatment plan and upload feedback on treatment effectiveness → Cloud algorithm updates;

[0151] Expert mentoring system: Cloud-based experts can provide real-time demonstrations and remote guidance to doctors, gradually cultivating their independent operating abilities;

[0152] Expert Collaboration Link: When encountering difficult cases, the expert collaboration module can be triggered to call on other experts to participate in the diagnosis and treatment.

[0153] like Figure 6 As shown, the acoustic stimulation signal generation and transmission path includes (the diagram indicates security protection and encrypted transmission mechanisms to ensure the accuracy of treatment signal transmission and patient safety):

[0154] 1. AI parameter generation unit: Outputs individualized acoustic stimulus parameters (frequency, modulation method, waveform, rhythm, duration, and stimulus energy, etc.) from a cloud-based AI model.

[0155] 2. Digital Signal Synthesizer: Generates acoustic signals such as AM (amplitude modulation), FM (frequency modulation), BB (broadband noise or binaural beat), or EN (envelope noise);

[0156] 3. Digital-to-Analog Converter (DAC): Converts digital signals into analog waveforms;

[0157] 4. Acoustic amplification and output module: After being amplified and filtered for safety, the output is sent to the patient's headphones or hearing aid;

[0158] 5. Transmission and feedback channels: Signal parameters and execution status are fed back to the cloud via BLE / Wi-Fi to achieve real-time monitoring and closed-loop control.

[0159] VI. System Workflow

[0160] 1. Registration and record keeping: Doctors or clients create patient records in the cloud and upload examination data and questionnaire information;

[0161] 2. Data Acquisition and Upload: The patient terminal collects and uploads signals such as EEG, MEG, fNIRS, ERP-P300, HRV, hearing threshold, psychoacoustics, and questionnaire scales in real time;

[0162] 3. AI Analysis and Solution Generation: Cloud-based AI and big data analysis of multimodal data to generate initial individualized acoustic stimulation solutions;

[0163] 4. Expert review and issuance: Experts or doctors confirm and issue the treatment plan via the cloud;

[0164] 5. Treatment Implementation and Feedback: The patient executes the treatment plan and uploads physiological / psychological feedback information in real time;

[0165] 6. Closed-loop optimization: The AI ​​model updates parameters based on feedback, achieving dynamic iterative optimization;

[0166] 7. Synchronized with regulatory authorities: Data is automatically synchronized to national-level regulatory interfaces to ensure security and compliance throughout the entire process.

[0167] VII. Security and Compliance Mechanisms

[0168] To ensure medical data security and system stability, the system in this application adopts the following measures:

[0169] • End-to-end encrypted bidirectional transmission (TLS / SSL);

[0170] Multi-level user authentication and hierarchical access control;

[0171] • Model version is traceable, and operation logs are automatically recorded;

[0172] • Off-site disaster recovery backup and high availability redundancy;

[0173] • Medical device software management system that conforms to YY / T 0664-2020 and GB 9706.1-2020 standards.

[0174] VIII. Implementation Results and Performance Verification

[0175] 1. The effects achieved by this application include:

[0176] (1) Added customer management module and individual account system

[0177] Each client (hospital) has its own dedicated cloud sub-module and an independent account on the cloud platform, which can independently manage patient files, treatment plans and feedback records, enabling personalized services and independent operation and maintenance.

[0178] (2) Expert remote teaching and capacity transfer mechanism

[0179] By demonstrating treatments in the cloud and using AI-assisted decision-making, we can gradually train primary care physicians to become therapists with independent diagnostic and treatment capabilities and the ability to solve difficult cases, thus achieving a sustainable knowledge transfer system.

[0180] like Figure 7 As shown, the collaborative relationship and knowledge transfer mechanism among the three-tiered structure of experts, doctors, and patients include:

[0181] The Experts layer, located at the top, comprises multiple cloud / edge layer experts who form an interconnected expert collaboration network for collaborative assistance on complex cases, experience sharing, and protocol monitoring. The Doctors layer, in the middle, connects to the Experts layer via a two-way channel, receiving remote mentoring and supervisory support, and undertaking patient protocol review and clinical execution. The Patients layer (Patient Device / App), at the bottom, receives treatment plans and uploads efficacy feedback for physician assessment and cloud-based AI model optimization.

[0182] The diagram shows the knowledge flow channels on the left and right sides: experience sharing among experts (Experience Sharing Path), knowledge transfer from experts to doctors (Knowledge Transfer Path), and patient feedback flowing back to the AI ​​model through doctors and experts to form a closed-loop optimization (Treatment Feedback Path).

[0183] Figure 7 Overall, this invention reflects a multi-layered collaborative mechanism within the platform that enables the large-scale expansion of expert capabilities, the enhancement of physician capabilities, and the closed-loop management of patient treatment outcomes.

[0184] (3) Expert Horizontal Collaboration Network

[0185] When a single expert encounters a difficult case, they can initiate a request for help or collaboration with a single click. The platform will automatically dispatch other experts to participate, enabling real-time mutual assistance and experience sharing.

[0186] (4) AIoT intelligent agents and digital twin architecture

[0187] Centered on a cloud platform, it integrates AI algorithms, IoT devices, and big data analytics to deeply couple the physical and information worlds, forming a specialized neuromodulation system that features intelligent feedback and a symbiotic relationship between the virtual and real worlds.

[0188] (5) National-level regulatory interface and security compliance mechanism

[0189] The cloud platform is equipped with a monitoring terminal that is directly connected to the National Medical Data Center, ensuring that all data transmission, analysis, and storage are carried out under a secure monitoring system.

[0190] (6) Edge AI expert terminal and distributed computing power system

[0191] Each cloud platform / regional edge expert has an independently operating edge AI terminal, enabling intelligent collaboration and rapid response between the cloud, edge, and terminal, thereby improving system stability and processing efficiency.

[0192] (7) Multimodal data-driven personalized acoustic stimulation calculation

[0193] By integrating multi-source signal features such as EEG, MEG, fNIRS, ERP-P300, HRV, hearing, and psychoacoustics, the frequency, amplitude, rhythm, and duration of multimodal acoustic stimulation are optimized in real time through a reinforcement learning strategy network π(a|S) to achieve dynamic closed-loop therapy.

[0194] 2. Through simulation and validation in a multi-center clinical environment, this system can achieve the following:

[0195] • Single-center concurrent users ≥ 100,000;

[0196] • The response time of the acoustic stimulus parameters is ≤ 1 second;

[0197] • The accuracy rate of efficacy prediction is ≥ 90%;

[0198] • Patient compliance improved by approximately 35%;

[0199] • Cloud-edge-device collaboration latency ≤ 200 milliseconds;

[0200] • Data regulatory compliance rate is 100%.

[0201] The results show that the system significantly improves the intelligence level, remote processing efficiency and safety and reliability of the acoustic neuromodulation system, providing a sustainable technological foundation for digital rehabilitation management.

[0202] IX. Practical Applications and Social Value

[0203] The system proposed in this application can build a unified intelligent neuromodulation service network nationwide, and has the following value:

[0204] • Supports thousands of cloud platform / regional edge experts to provide concurrent remote guidance to tens of thousands of doctors and millions of patients for personalized, precise diagnosis and treatment of specific diseases.

[0205] • Enhance the treatment capabilities of primary hospitals, establish multidisciplinary team (MDT) disease centers, and realize AI-enabled precision chronic disease rehabilitation management services;

[0206] • Establish a national-level big data infrastructure for neural regulation to promote the integration of scientific research and industrialization;

[0207] • Achieve full-lifecycle chronic disease management and intelligent efficacy tracking to improve patient adherence;

[0208] Promoting the integration of traditional Chinese and Western medicine and the digital transformation of neurorehabilitation has broad social and economic benefits.

[0209] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. An acoustic stimulation neuromodulation service management system combining AI and big data analysis, characterized in that, This system comprises an interconnected cloud-based central layer, regional edge layer, physician management layer, and patient terminal layer. Data transmission is achieved through bidirectional secure encrypted communication, constructing a complete digital closed-loop process of data acquisition, analysis, decision-making, execution, feedback, and optimization. This enables closed-loop intelligent and precise neuromodulation through cloud-edge-device-human collaboration. The cloud-based central layer includes an AI algorithm and big data analysis module, an expert collaboration and task scheduling module, an AI agent module, a cloud storage and customer management module, a cloud computing support module, and a national medical data supervision interface module. The AI ​​algorithm and big data analysis module, based on deep learning and reinforcement learning models, performs big data analysis on patients' multimodal data to formulate and distribute individualized acoustic stimulation neuromodulation diagnoses. The treatment plan is transmitted to the patient terminal layer, and the individualized acoustic stimulation neuromodulation treatment plan is dynamically adjusted and optimized based on real-time feedback information. The individualized acoustic stimulation neuromodulation treatment plan consists of digital signal parameters that can be transmitted over a network, enabling the plan to be transmitted from the cloud central layer to the regional edge layer, the doctor management layer, and the patient terminal layer. Thus, the execution of the plan and the clinical output of the stimulation signal are effectively realized through the system network. The expert collaboration and task scheduling module is used to build a multi-expert horizontal communication network for remote collaboration. The cloud computing support module is used for cloud expert operation and maintenance of the cloud central layer. The national medical data supervision interface module is used to realize real-time synchronization and compliance supervision of clinical data, algorithm logs, and model call records. The edge layer of the region includes AI edge expert terminals with AI edge computing capabilities. These AI edge expert terminals connect upwards to the cloud center layer and downwards to the doctor management layer and the patient terminal layer, and are horizontally interconnected. The AI ​​edge expert terminals include: Edge computing module: This includes calling the model in the cloud central layer to independently execute AI inference and generate and optimize personalized solutions locally; Distributed encrypted data caching module: Automatically saves local feedback data when the network is interrupted, and automatically synchronizes it with the cloud model and database after the network connection is restored; Remote Doctor / Patient Guidance Module: Used to enable remote guidance and demonstration of operations by platform / edge experts; The physician management input interface module is used for two-way communication between marginal experts and clinicians. The physician management layer is responsible for: intelligent control of information monitoring, individualized treatment plan editing, cloud platform / regional edge expert teaching operations, input of objective test data and subjective consultation scales, implementation guidance of personalized neuromodulation plans for patients, and local storage of patient data on the hospital client side. The patient terminal layer is used for the execution of the protocol and the collection of feedback information.

2. The acoustic stimulation neuromodulation service management system combining AI and big data analysis according to claim 1, characterized in that, The cloud storage and customer management module is used to establish independent exclusive accounts or sub-modules for end customers and store all records and parameter information in the database for model retraining and knowledge base updates.

3. The acoustic stimulation neuromodulation service management system combining AI and big data analysis according to claim 1, characterized in that, The cloud computing support module also includes an AI model and computing power management submodule for accessing external AI cloud service platforms.

4. The acoustic stimulation neuromodulation service management system combining AI and big data analysis according to claim 1, characterized in that, The regional edge layer also includes a data preprocessing module and an edge computing cloud interface module.

5. The acoustic stimulation neuromodulation service management system combining AI and big data analysis according to claim 1, characterized in that, The physician management system includes a patient view and efficacy monitoring interface module, an individualized treatment plan adjustment module, an expert mentoring and skills transfer module, and a task scheduling and intelligent reminder module.

6. The acoustic stimulation neuromodulation service management system combining AI and big data analysis according to claim 5, characterized in that, The physician management system also includes a patient consultation / scale completion module, a test data input port module, a regional edge layer expert monitoring access port module, and an expert teaching and clinical demonstration module.

7. The acoustic stimulation neuromodulation service management system combining AI and big data analysis according to claim 1, characterized in that, The patient terminal layer includes a protocol execution module, a physiological signal acquisition module, an intelligent feedback module, and an AI follow-up assistant module.

8. The acoustic stimulation neuromodulation service management system combining AI and big data analysis according to claim 7, characterized in that, The implementation module of the solution includes a wearable mobile sound stimulation therapy device, a medical institution sound stimulation therapy device, and a wearable hearing aid.

9. The acoustic stimulation neuromodulation service management system combining AI and big data analysis according to any one of claims 1-9, characterized in that, The cloud central layer, the regional edge layer, the doctor management layer, and the patient terminal layer constitute an AI intelligent agent (AIoT) to achieve bidirectional interconnection between the physical and information worlds through the Internet of Things. This includes: Digital twin patient model: Based on the patient's multimodal information data, an individualized virtual-reality mapping model is established to realize a digital twin patient profile; Real-time dynamic mapping: The AI ​​learning process is linked with the actual feedback from patients so that when patient information data changes, it is updated and fed back to the AI ​​algorithm and big data analysis module. AI self-learning closed loop: Utilizing group data and individual feedback to continuously train the model, achieving vertical closed-loop management and horizontal multi-cloud platform / regional edge expert collaboration.

10. A control method for an acoustic stimulation neuromodulation service management system combining AI and big data analysis, comprising the following steps: (1) Data collection and uploading: Real-time collection and uploading of patients' multimodal information data, including EEG, MEG, fNIRS, ERP-P300, HRV, hearing threshold, psychoacoustics and questionnaire scale data; (2) AI analysis and solution generation: Cloud-based AI models and big data analysis of multimodal data generate initial individualized acoustic stimulation neuromodulation treatment plans; among them, experts at the platform / regional edge layer can participate in the formulation or modification of modulation treatment plans through network mutual assistance and collaboration; (3) Review and distribution of the plan: After the regional edge layer experts and / or physician management physicians review or adjust the individualized sound stimulation neuromodulation treatment plan, it is distributed to the patient terminal layer; (4) Protocol execution feedback: During the protocol execution process, patient feedback information data is collected and uploaded in real time; (5) Closed-loop optimization: The AI ​​algorithm and big data analysis module updates or optimizes the individualized acoustic stimulation neuromodulation treatment plan based on feedback information data to achieve dynamic iterative optimization; (6) Synchronized supervision: All process data is synchronized to the medical data center through the national medical data supervision interface module for supervision, realizing safe, intelligent and closed-loop remote neuromodulation therapy.

Images