A cognitive impairment chronic disease health management system and method based on the Internet of Things

Abstract

The application belongs to the technical field of the Internet of Things and old-age health care, and particularly relates to a cognitive impairment chronic disease health management system and method based on the Internet of Things. The method is applied to a smart old-age care scene and comprises the following steps: a data acquisition step, in which an active acquisition and passive acquisition combination is adopted; a data transmission step, in which a 5G, WiFi and LoRa multi-network fusion transmission mode is adopted; an intelligent analysis step, in which a weighted fusion algorithm is adopted to fuse multi-source data and identify cognitive function abnormality features; a grading evaluation step, in which a four-dimensional weighted scoring model is used to calculate grading evaluation; and a grading intervention step, in which corresponding intervention measures are executed according to the grading evaluation result. The application realizes full-process automation of cognitive impairment chronic disease management, improves early screening coverage, delays disease progression, reduces care burden and improves the quality of life of the elderly.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of Internet of Things and elderly health care, specifically relating to an Internet of Things-based health management system and method for chronic diseases of cognitive impairment. Background Technology

[0002] With the accelerating aging of the population, the incidence of chronic cognitive impairment diseases (such as Alzheimer's disease) is rising year by year, becoming a significant type of chronic disease affecting the quality of life of the elderly and increasing the care burden on families and society. It is also one of the core health issues that urgently need to be addressed in the smart elderly care service system. Currently, the management of chronic cognitive impairment diseases still has many shortcomings: On the one hand, traditional detection methods rely on professional medical personnel and large-scale testing equipment, and are mostly concentrated in hospital settings, making it difficult to achieve routine and convenient early screening in home, community, and other elderly care settings. As a result, most patients are diagnosed at an advanced stage of the disease, missing the best intervention opportunity. On the other hand, existing health management models lack systematicity and stratification, failing to form a full-process management system adapted to the characteristics of the progression of chronic cognitive impairment diseases in the elderly. Furthermore, it is difficult to combine intelligent technology to achieve real-time collection and dynamic monitoring of patients' physiological and behavioral data, making it impossible to implement precise and personalized graded interventions based on the severity of the patient's condition, and failing to meet the refined care needs of the elderly with cognitive impairment in smart elderly care scenarios. Summary of the Invention

[0003] The purpose of this invention is to provide a health management system and method for chronic diseases of cognitive impairment based on the Internet of Things (IoT). This invention integrates IoT sensing and monitoring, intelligent data analysis, and tiered health intervention technologies into an age-friendly elderly care service system. It enables early screening, dynamic monitoring, and full-process tiered management of cognitive impairment in the elderly population, and is applicable to various elderly care application scenarios such as home-based care, community-based care, and institutional care.

[0004] The implementation process of this invention is as follows: A method for managing chronic diseases of cognitive impairment based on the Internet of Things includes the following steps: (1) Physiological, behavioral and environmental data are collected regularly by the IoT sensing device in the IoT sensing and acquisition module. Simple cognitive test tasks are pushed regularly by the cognitive assessment terminal. The elderly complete the test to realize the collection of cognitive data. The collected physiological, behavioral, environmental and cognitive data are denoised, deduplicated and standardized to remove abnormal data. (2) The processed physiological, behavioral, environmental and cognitive data are transmitted to the data transmission and storage module. The data transmission adopts the integration of 5G, WiFi and LoRa networks and dynamically switches the communication link according to the RSSI signal strength. WiFi is used for home scenarios, 5G or LoRa is used for community and institutional scenarios, and data caching function is set for remote areas or areas with weak signals. The data is automatically synchronized after the signal is restored. The storage includes local storage and cloud storage. Local storage is used to store recent real-time data for quick access and emergency retrieval. Cloud storage uses AES encrypted cloud servers to store long-term data. The AES encryption algorithm ensures data security and prevents the leakage of personal health information. It also supports data backup and recovery functions to ensure data integrity. (3) The data transmission and storage module transmits physiological, behavioral, environmental, and cognitive data to the intelligent analysis and hierarchical evaluation module, and uses a weighted fusion algorithm to assign weights to different types of data. The merged data is obtained, a decision tree is constructed, and random forest classification is performed on the merged data. The final classification result is output as anomaly probability using a voting method. According to the probability of anomalies The warning is divided into three levels, with a minimum warning level of 0.7. <0.8, moderate warning 0.8≤ <0.9, severe warning ≥0.9, when An alert is triggered when the value is ≥0.7; Based on the progression characteristics of chronic cognitive impairment and combined with early warning levels, a three-tiered grading assessment method called the Maim assessment was established, and a total score for the grading assessment was constructed. Based on cognitive function scores Daily Living Skills Score Physiological index scores Behavior performance The weighted calculation yields weights of 0.4, 0.3, 0.2, and 0.1, respectively. (Formula follows) The total score is 100 points, and the condition is graded into three levels: Level 1 mild cognitive impairment, Level 2 moderate cognitive impairment, and Level 3 severe cognitive impairment. (4) Based on the disease severity assessment results evaluated by the intelligent analysis and grading assessment module, the three-level intervention units are matched to implement the corresponding intervention measures to achieve a closed-loop intervention process.

[0005] Furthermore, in step (3), the condition is classified into three levels. Level 1 mild cognitive impairment, characterized by a slight decline in cognitive function, with generally normal daily living abilities, as determined by a score on the Mini-Cognitive Assessment Scale, and no significant abnormal behavior or fluctuations in physiological indicators; the quantitative standard is... Score: No record of abnormal behavior; Level II moderate cognitive impairment, characterized by a significant decline in cognitive function, partial impairment of daily living abilities, and marked behavioral or physiological abnormalities; the quantitative standard is... One or two abnormal behaviors or physiological indicators are present. Level 3 severe cognitive impairment, with severely impaired cognitive function, loss of basic self-care ability, requiring 24-hour care, and significant and fluctuating physiological indicators; the quantitative standard is... The score indicates the presence of three or more abnormal behaviors or severely abnormal physiological indicators.

[0006] Furthermore, in step (3), the intelligent analysis and grading assessment module periodically re-analyzes and assesses the collected data, dynamically adjusts the grading results based on changes in the patient's cognitive function, physiological state, and behavioral performance, ensures the timeliness and accuracy of the grading assessment, and calculates the score difference between the two grading assessments. , For the current score, For the last score, when Time-sharing triggers tiered adjustments; when The original grading system is maintained throughout the time period, while the trend of score changes is recorded. The formula is: T represents the number of days between two assessments, used to predict the direction of disease progression.

[0007] Furthermore, in step (4), the three-level intervention units are the first-level mild cognitive impairment intervention units: (I) Cognitive training intervention, specifically based on a personalized training task matching algorithm, which regularly pushes personalized cognitive training tasks through a cognitive assessment terminal, encouraging the elderly to complete them daily to strengthen cognitive function; according to the cognitive assessment score The formula for the algorithm to match training difficulty and personalized training tasks is as follows: Difficulty range , The higher the level, the lower the difficulty, and the longer the daily training time. (i) The unit is minutes; (ii) Lifestyle intervention, which is based on the collected physiological and behavioral data to push personalized life suggestions to the elderly, and family members assist in supervising the implementation; The Level 2 moderate cognitive impairment intervention unit provides professional cognitive intervention, rehabilitation training, daily living assistance intervention, and regular assessments through community healthcare workers or nursing staff in elderly care institutions. Level 3 Severe Cognitive Impairment Intervention Unit: (S1) 24-hour professional care from nursing home staff or home caregivers to assist in daily living activities and prevent complications; (S2) Establishing a green channel for diagnosis and treatment in conjunction with hospitals, with regular home visits by professional doctors to monitor changes in the condition, adjust medication plans, and provide targeted medical support to achieve medical intervention; (S3) Nursing staff and family members cooperate to provide emotional companionship and psychological counseling to alleviate emotions, improve quality of life, and achieve emotional intervention; (S4) Real-time monitoring: IoT sensing devices collect physiological and behavioral data. When abnormal data occurs, an emergency warning is immediately triggered to notify medical staff and family members for emergency handling.

[0008] Furthermore, in the Level 2 moderate cognitive impairment intervention unit, professional cognitive intervention and rehabilitation training specifically involve community healthcare workers or nursing home caregivers visiting 1-2 times per week to conduct professional cognitive intervention and rehabilitation training, regularly monitoring physiological indicators, and adjusting the intervention plan according to the graded scores. Determine the frequency of home visits ,formula: When 50≤ <65 hours, Times / week; when 65≤ <80 hours, Once per week; Assisted living intervention specifically uses IoT devices to provide smart reminders, allowing family members to assist with daily care while receiving professional care guidance and medication reminders. Based on the doctor's orders, the trigger conditions for the reminder are: the difference between the current time and the preset medication time is ≤10 minutes, and no medication behavior is detected by the vital signs monitoring device; the trigger conditions for the missing person reminder are: the activity trajectory is located outside the preset safe area via GPS, and the radius of the safe area is [missing information]. ,formula: The unit is meters; Regular assessments involve monthly cognitive evaluations and physiological indicator testing to dynamically adjust the grading and intervention plan, and, when necessary, coordinate with the hospital for further diagnosis and treatment. Assessments are triggered on a fixed date each month, and temporary assessments are triggered when abnormal physiological indicators persist for ≥3 days. The abnormality criterion is a physiological indicator exceeding the normal range by 20%, as shown in the formula: These are normal reference values ​​for physiological indicators.

[0009] Furthermore, in (S1), the interval of care. Based on the degree of limb movement impairment Sure, The score ranges from 0 to 10, with higher scores indicating more severe obstacles. The formula is: The unit is minutes. hour, minute; (S2) Consultation cycle Based on graded scores Confirmed, formula: The unit is days. hour, sky; In (S3), emotional feature values ​​are extracted from behavioral data such as facial expressions, activity levels, and tone of voice. The value of E ranges from 0 to 100 points, with lower scores indicating worse moods and longer emotional companionship durations. formula: The unit is minutes / day. hour, minutes / day; In (S4), the physiological data abnormality trigger condition is: heart rate. satisfy times / minute or Blood oxygen saturation (Hz / min) <90%, duration ≥30 seconds; Abnormal behavior triggering conditions: activity level 0 and no human body sensor response, duration ≥10 minutes; Warning response time Within seconds, medical staff and family members were notified.

[0010] A cognitive impairment chronic disease health management system based on the Internet of Things (IoT) includes an IoT sensing and acquisition module, a data transmission and storage module, an intelligent analysis and hierarchical assessment module, and a Mainm three-level intervention execution module. The IoT sensing and acquisition module is connected to the data transmission and storage module via a data connection. The data transmission and storage module is connected to the intelligent analysis and hierarchical assessment module via a data connection. The intelligent analysis and hierarchical assessment module is connected to the Mainm three-level intervention execution module via a signal connection. The four modules are interconnected through standardized interfaces to achieve data interoperability and command linkage. The IoT sensing and acquisition module is used to periodically collect four types of multi-source data—physiological, behavioral, cognitive, and environmental—through IoT sensing devices and cognitive assessment terminals, and to perform noise reduction, deduplication, and standardization on the collected multi-source data, and to remove abnormal data. The data transmission and storage module adopts a multi-network convergence of 5G, WiFi and LoRa, dynamically switches the optimal communication link based on RSSI signal strength, and combines a local cache synchronization mechanism to ensure data integrity in weak network environments. At the same time, it ensures data security and recoverability through a dual storage architecture of local and cloud storage combined with AES-128 encryption and incremental backup strategy. The intelligent analysis and hierarchical evaluation module uses a weighted fusion algorithm to integrate multi-source data acquired by the IoT sensing and acquisition module, and outputs anomaly probabilities based on random forest classification. According to the probability of anomalies The warning is divided into three levels: mild warning ( 0.7≤ <0.8 Moderate warning （0.8≤ <0.9） Severe warning （ ≥0.9） ,when The system triggers an early warning, combined with the four-dimensional weighted scoring formula of the Maim three-level grading assessment. The total score is 100 points. A quantitative total score is generated, categorized into first-level... Level 2 Level 3 A graded assessment of the condition was implemented, classifying the degree of cognitive impairment into three levels: mild, moderate, and severe, and introducing... A dynamic adjustment mechanism for scores avoids rigid evaluation; The Maim three-level intervention execution module matches corresponding differentiated intervention paths and executes intervention measures based on the grading results of the intelligent analysis and grading assessment module.

[0011] Furthermore, in the data transmission and storage module, the handover timing is determined based on the RSSI signal strength, assuming the current network RSSI is... The preset threshold is ,when If the condition persists for more than 3 seconds, a network switch will be triggered. WiFi threshold ≥ -60dBm 5G threshold ≥ -70dBm, LoRa threshold ≥ -80dBm .

[0012] Furthermore, in the intelligent analysis and grading assessment module, the specific disease grading is divided into three levels. Level 1 mild cognitive impairment, characterized by a slight decline in cognitive function, with generally normal daily living abilities, as determined by a score on the Mini-Cognitive Assessment Scale, and no significant abnormal behavior or fluctuations in physiological indicators; the quantitative standard is... Score: No record of abnormal behavior; Level II moderate cognitive impairment, characterized by a significant decline in cognitive function, partial impairment of daily living abilities, and marked behavioral or physiological abnormalities; the quantitative standard is... One or two abnormal behaviors or physiological indicators are present. Level 3 severe cognitive impairment, with severely impaired cognitive function, loss of basic self-care ability, requiring 24-hour care, and significant and fluctuating physiological indicators; the quantitative standard is... A score indicating three or more abnormal behaviors or severely abnormal physiological indicators; Dynamic grading updates involve regularly re-analyzing and evaluating the collected data. The grading results are dynamically adjusted based on changes in the patient's cognitive function, physiological state, and behavioral performance to ensure the timeliness and accuracy of the grading assessment and provide a basis for subsequent grading interventions. Calculate the score difference between the two graded assessments. ,in For the current score, For the last score, when Time-sharing triggers tiered adjustments; when The original grading system is maintained throughout the time period, while the trend of score changes is recorded. The formula is: Where T is the number of days between two assessments, used to predict the direction of disease progression.

[0013] Furthermore, the Maimu three-level intervention execution module includes three-level intervention units corresponding to the disease grading in the intelligent analysis and grading assessment module; The intervention units for mild cognitive impairment include cognitive training intervention and lifestyle intervention; the intervention units for moderate cognitive impairment include professional cognitive intervention, rehabilitation training, life assistance intervention and regular assessment by community medical staff or nursing staff in elderly care institutions; and the intervention units for severe cognitive impairment include 24-hour professional care, medical intervention, emotional intervention and real-time monitoring by nursing staff in elderly care institutions or home caregivers.

[0014] The positive effects of this invention: (1) The system described in this invention realizes the full-process automated closed loop of chronic disease management of cognitive impairment; from data collection, intelligent analysis, graded judgment to intervention execution, no excessive human intervention is required, which greatly reduces the workload of caregivers and medical staff and reduces the investment of human resources. At the same time, through the personalized intervention plan matching algorithm, it ensures that the intervention measures are accurately matched with the patient's condition, improves the intervention effect, and slows down the progression of cognitive impairment.

[0015] (2) The system described in this invention is suitable for various elderly care scenarios such as home, community, and elderly care institutions; the device adopts a lightweight and age-friendly design, supports touch and voice dual interaction modes, does not require professional personnel to operate, adapts to the usage habits of the elderly, realizes the normalized and convenient collection of cognitive impairment-related data, solves the problems of limited monitoring scenarios and high difficulty of operation for the elderly in the prior art, and improves the universality and ease of use of the system.

[0016] (3) The dual early warning mechanism established by the system described in this invention, namely, graded early warning + emergency early warning, can identify abnormal cognitive function and abnormal physiological indicators in advance, and quickly push early warning information to caregivers and family members, so as to buy time for emergency response, effectively reduce the safety risks of elderly people getting lost and sudden illness, and protect the health and safety of the elderly population.

[0017] (4) The system described in this invention improves and enhances the quality of life for the elderly; it promotes the transformation of chronic disease management of cognitive impairment from "passive treatment" to "active prevention and precise intervention," effectively increasing the coverage rate of early screening for cognitive impairment, delaying disease progression, reducing the incidence of severe cognitive impairment, alleviating the care burden on families and society, and saving medical resources. Integrating intelligent Internet of Things technology with the Maiem hierarchical management concept, it improves the smart elderly care service system, providing elderly people with cognitive impairment with full-process, precise, and intelligent health management services, enhancing the quality of life and well-being of the elderly population, and helping to address the trend of population aging. The standardized hierarchical assessment system and intervention plan can realize the standardized and large-scale promotion of chronic disease management of cognitive impairment, providing standardized management tools for elderly care institutions and community service centers, promoting the standardized development of the smart elderly care industry, and has broad application prospects and promotional value. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the overall architecture of the IoT-based chronic disease health management system for cognitive impairment described in Example 1; Figure 2 This is a schematic diagram of the Maim three-level grading assessment system in Example 1; Figure 3 This is a flowchart of the Maim three-level intervention execution module in Example 1; Figure 4 This is a flowchart of the IoT-based chronic disease health management method for cognitive impairment described in Example 5. Detailed Implementation

[0019] The present invention will be further described below with reference to the embodiments.

[0020] The IoT-based chronic disease health management system and method for cognitive impairment described in this invention achieves a fully automated closed-loop management process for chronic cognitive impairment, from data collection, intelligent analysis, and tiered assessment to intervention execution. It requires minimal human intervention, significantly reducing the workload of caregivers and medical staff, and lowering labor costs. Simultaneously, through personalized intervention matching algorithms, it ensures that intervention measures are accurately matched to the patient's condition, improving intervention effectiveness and slowing the progression of cognitive impairment. The device features a lightweight, age-friendly design, supporting both touch and voice interaction modes. It requires no professional personnel to operate and is adapted to the usage habits of the elderly, enabling routine and convenient collection of cognitive impairment-related data. This solves the problems of limited monitoring scenarios and difficult operation for the elderly in existing technologies, improving the system's universality and ease of use. The established dual early warning mechanism includes tiered early warning and emergency early warning, which can identify cognitive function abnormalities and physiological indicator abnormalities in advance, quickly pushing early warning information to caregivers and family members, buying time for emergency response, effectively reducing the safety risks of the elderly getting lost or experiencing sudden illness, and ensuring the health and safety of the elderly population. This initiative promotes a shift in chronic disease management for cognitive impairment from "passive treatment" to "proactive prevention and precise intervention," effectively increasing the coverage of early screening for cognitive impairment, slowing disease progression, reducing the incidence of severe cognitive impairment, alleviating the care burden on families and society, and conserving medical resources. By integrating smart IoT technology with the Mainm hierarchical management concept, it has improved the smart elderly care service system, providing comprehensive, precise, and intelligent health management services for elderly people with cognitive impairment, thereby enhancing their quality of life and well-being and contributing to addressing the aging population trend. Example 1 A health management system for chronic diseases of cognitive impairment based on the Internet of Things (IoT), see [link / reference]. Figure 1 The system includes an IoT sensing and acquisition module, a data transmission and storage module, an intelligent analysis and hierarchical assessment module, and a Mainm three-level intervention execution module. The IoT sensing and acquisition module is connected to the data transmission and storage module via a data connection. The data transmission and storage module is connected to the intelligent analysis and hierarchical assessment module via a data connection. The intelligent analysis and hierarchical assessment module is connected to the Mainm three-level intervention execution module via a signal connection. The four modules are connected through standardized interfaces to achieve data interoperability and command linkage, ensuring the real-time and continuous nature of detection, assessment, and intervention, and adapting to the deployment needs of different elderly care scenarios. The IoT sensing and acquisition module is used to periodically collect four types of multi-source data—physiological, behavioral, cognitive, and environmental—through IoT sensing devices and cognitive assessment terminals, and to perform noise reduction, deduplication, and standardization on the collected multi-source data, and to remove abnormal data. Deploy lightweight, age-friendly IoT sensing devices adapted for the elderly population. No professional personnel are required for operation. They can be flexibly deployed in scenarios such as home-based elderly care (key areas such as bedrooms, living rooms, and kitchens), community elderly care service centers, and elderly care institutions. These include, but are not limited to: sleep belts (collecting physiological data such as heart rate, sleep, activity level, and sleep quality), environmental sensing devices (collecting environmental data such as temperature, humidity, human body sensing, and gas leaks), cognitive assessment terminals (tablets and wall-mounted terminals with built-in simple cognitive assessment scales and interactive testing tools), and vital sign monitoring devices (which can monitor users' physiological characteristics for a long time, such as blood sugar, blood pressure, uric acid, heart rate, and blood oxygen). All devices adopt the LoRa low-power protocol, ensuring a battery life of ≥30 days. Multi-device collaborative synchronization is based on a timestamp synchronization algorithm, achieving a data collection time deviation of ≤1 second across multiple devices.

[0021] In this embodiment, the IoT sensing device includes a sleep belt deployed in the bedroom to collect physiological data such as heart rate, sleep quality, and activity level; human body sensors and temperature and humidity sensors deployed in key areas such as the living room and kitchen to capture behavioral and environmental data in real time; and vital sign monitoring devices, including a blood pressure monitor and a heart rate meter, for long-term monitoring of physiological characteristics such as blood pressure, heart rate, and blood oxygen, with a passive sampling frequency of 10Hz to ensure data continuity.

[0022] The cognitive assessment terminal is a device designed for elderly users to assess their cognitive abilities. It features a built-in simplified cognitive assessment scale and interactive testing tools with a recognition accuracy of ≥95%. It supports both touch and voice interaction modes, lowering the barrier to entry for elderly users and adapting to their operating habits. The click response time is ≤0.5s.

[0023] All devices are connected to the home WiFi network, and initialization and debugging are completed to ensure that the device battery life is ≥30 days. Based on the timestamp synchronization algorithm, the data collection time deviation of multiple devices is ≤1 second.

[0024] During the data acquisition phase, a combination of active and passive acquisition methods was adopted.

[0025] Passive data collection uses a sleep belt, human body sensor, temperature and humidity sensor, and vital sign monitoring device 1015 to collect physiological data every hour, as well as real-time behavioral and environmental data, without requiring active operation by the elderly.

[0026] Active data collection is achieved through a cognitive assessment terminal, which pushes a simple cognitive test task once a week, including memory test, attention test, and language ability test. The elderly can complete the test by touch or voice, realizing regular screening and dynamic monitoring of cognitive function.

[0027] In the data preprocessing stage, the system performs noise reduction, deduplication, and standardization on the collected multi-source data. Specifically: The collected multi-source data undergoes noise reduction, deduplication, and standardization processes to remove outlier data (such as invalid data caused by equipment malfunctions). Data of different types and formats is uniformly converted into a standardized data format to provide high-quality data support for subsequent intelligent analysis. Calculation process and formulas: (1) Abnormal data removal: The 3σ criterion is adopted. Let a certain type of collected data sequence be... Calculate the mean and standard deviation When the data meets When this happens, the data is identified as abnormal and removed. formula: , ; (2) Data standardization: The min-max standardization method is used to map the data to the [0,1] interval to eliminate the influence of units; formula: ,in For standardized data, This is the original data.

[0028] The data transmission and storage module employs a convergence of 5G, WiFi, and LoRa networks to transmit pre-processed data to local and cloud servers. It dynamically switches to the optimal communication link based on RSSI signal strength and utilizes a local cache synchronization mechanism to ensure data integrity in weak network environments. Simultaneously, a dual local and cloud storage architecture, coupled with AES-128 encryption and incremental backup strategies, ensures data security and recoverability. The switching timing is determined based on RSSI signal strength; assuming the current network RSSI is [value missing], [further details missing]. The preset threshold is , WiFi threshold ≥ -60dBm, 5G threshold ≥ -70dBm, LoRa threshold ≥ -80dBm ,when If the condition persists for more than 3 seconds, a network switch will be triggered.

[0029] Transmission delay calculation: Transmission delay Transmission time Processing time Cache time Composition, formula: ;in ( For data volume, (for network transmission rate), requirements (Real-time data) (Non-real-time data).

[0030] For home use, WiFi transmission is preferred, while 5G or LoRa transmission is used in communities and institutions. For remote or weak signal areas, a data caching function is set up to automatically synchronize data after the signal is restored, thus avoiding data loss.

[0031] Local storage (such as community servers and home gateways) is used to store recent (e.g., 7 days) real-time data for quick retrieval and emergency queries, ensuring data transmission latency ≤ 5 seconds and a success rate ≥ 99.8%. Cloud storage uses encrypted cloud servers to store long-term data (e.g., over 1 year), including collected data, evaluation results, and intervention records. Encryption algorithms (such as AES encryption) ensure data security, preventing the leakage of personal health information. Data backup and recovery functions are also supported to ensure data integrity. Calculation process and formula: (1) AES encryption calculation: The AES-128 encryption algorithm is used, with a key length of 128 bits. The encryption process is as follows: initial round key addition → byte substitution → row shift → column mixing → round key addition, repeating 10 rounds of iteration; Encryption formula (simplified): C = E(K, P) Deciphering the formula: ,in For encryption key, Plain text data This is the encrypted ciphertext; (2) Data backup frequency: An incremental backup strategy is adopted, and newly added data is backed up at 2:00 AM every day. The backup integrity verification formula is as follows: Compare the verification values ​​before and after the backup to ensure that the backup data is consistent with the original data.

[0032] The intelligent analysis and hierarchical evaluation module can be found here. Figure 2 A weighted fusion algorithm is used to fuse multi-source data acquired by IoT sensing and acquisition modules, and anomaly probabilities are output based on random forest classification. According to the probability of anomalies The warning is divided into three levels: mild warning ( 0.7≤ <0.8 Moderate warning （0.8≤ <0.9） Severe warning （ ≥0.9） ,when The system triggers an early warning, combined with the four-dimensional weighted scoring formula of the Maim three-level grading assessment. The total score is 100 points. A quantitative total score is generated, categorized into first-level... Level 2 Level 3 A graded assessment of the condition was implemented, classifying the degree of cognitive impairment into three levels: mild, moderate, and severe, and introducing... A dynamic adjustment mechanism for scores avoids rigid evaluation; The specific calculation process of the weighted fusion algorithm is as follows: assign weights to different types of data. (physiological data) Behavioral data Cognitive test data Environmental data ), integrated data after fusion formula: ( (For each type of standardized data). The specific process of random forest classification is as follows: 100 decision trees are constructed, each tree classifies the samples, and the final classification result is determined by a voting method, with anomaly probability... formula: ， when When this happens, an alert is triggered; Risk warning level: based on The warning is divided into three levels: mild warning ( 0.7≤ <0.8 Moderate warning （0.8 ≤ <0.9） Severe warning （ ≥0.9） .

[0033] The Maiem Three-Tier Classification Assessment: Based on the progression characteristics of chronic cognitive impairment and the above analysis results, a Maiem three-tier classification assessment system is established, clearly defining the assessment criteria for each level to achieve accurate classification of the severity of the patient's condition. Calculation process and formula: Construct a tiered assessment total score Based on cognitive function scores Daily Living Skills Score Physiological index scores Behavior performance The weighted calculation yields weights of 0.4, 0.3, 0.2, and 0.1, using the following formula: (Total score: 100 points) The specific classifications are as follows: Level 1 Mild Cognitive Impairment: Mild decline in cognitive function, with generally normal daily living abilities, only mild memory loss and poor concentration, assessed by a score on the Mini-Cognitive Assessment Scale, and no significant abnormal behavior or fluctuations in physiological indicators; Quantitative Standards: The patient scored [score missing], with no record of abnormal behavior. The diagnosis is Level 1 mild cognitive impairment, triggering a mild warning.

[0034] Level 2 Moderate Cognitive Impairment: Significant decline in cognitive function, including marked memory loss, language impairment, and orientation difficulties; partial impairment of daily living abilities (e.g., inability to eat or dress independently); significant behavioral abnormalities (e.g., tendency to wander off, mood swings); or abnormal physiological indicators; Quantitative Standards: There are 1-2 abnormal behaviors or abnormal physiological indicators.

[0035] Level 3 Severe Cognitive Impairment: Cognitive function is severely impaired, resulting in loss of basic self-care abilities. Symptoms include severe memory loss, confusion, and limb movement disorders, requiring 24-hour care. Physiological indicators show significant abnormalities and large fluctuations. Quantitative Standards: The score indicates the presence of three or more abnormal behaviors or severely abnormal physiological indicators.

[0036] Dynamic grading updates: The intelligent analysis model regularly (e.g., once a month) re-analyzes and evaluates the latest collected data, dynamically adjusting the grading results based on changes in the patient's cognitive function, physiological state, and behavioral performance, ensuring the timeliness and accuracy of the grading assessment, and providing a basis for subsequent grading interventions.

[0037] Calculate the score difference between the two graded assessments. ( For the current score, (for the previous score), when Time-sharing triggers tiered adjustments; when The original grading system is maintained throughout the time period, while the trend of score changes is recorded. The formula is: (T is the number of days between two assessments), used to predict the direction of disease progression.

[0038] The Mainm three-tiered intervention execution module matches corresponding differentiated intervention paths and implements intervention measures based on the grading results from the intelligent analysis and grading assessment module. Specifically, the Level 1 mild cognitive impairment intervention unit includes cognitive training intervention and lifestyle intervention; the Level 2 moderate cognitive impairment intervention unit includes professional cognitive intervention, rehabilitation training, daily living assistance intervention, and regular assessment by community healthcare workers or nursing home caregivers; and the Level 3 severe cognitive impairment intervention unit includes 24-hour professional care, medical intervention, emotional intervention, and real-time monitoring by nursing home caregivers or home caregivers.

[0039] Specifically: Level 1 Mild Cognitive Impairment Intervention Unit: Focusing on "preventing progression and strengthening cognition," this unit implements basic intervention measures without requiring full-time professional medical staff involvement. Specific measures include: (1) Cognitive training intervention: Personalized cognitive training tasks (such as memory games, language training, and logical reasoning training) are regularly pushed through a cognitive assessment terminal to encourage the elderly to complete them daily, thereby strengthening cognitive function. The core technology is a personalized training task matching algorithm; based on cognitive assessment scores... Matching training difficulty, formula: (Difficulty range) , (The higher the level, the lower the difficulty) Daily training time (Unit: minutes)

[0040] (2) Lifestyle intervention: Based on collected physiological and behavioral data, personalized lifestyle suggestions (such as dietary recommendations, exercise plans, and sleep adjustments) are provided to the elderly, with family members assisting in monitoring their implementation; Taking sleep intervention as an example, based on sleep quality data... (0-100 points) Recommended sleep duration ,formula: (Unit: hour); Exercise intervention is based on activity level. (Steps / day), Recommended daily exercise duration ,formula: (Unit: minutes), when hour, minute.

[0041] Level 2 Moderate Cognitive Impairment Intervention Unit: Focusing on "delaying progression and assisting daily living," this unit implements a combination of professional and family-based intervention measures, specifically including: (1) Professional intervention: Community medical staff or nursing staff of elderly care institutions visit 1-2 times per week to conduct professional cognitive intervention and rehabilitation training (such as language rehabilitation and physical rehabilitation), monitor physiological indicators regularly, and adjust the intervention plan according to the graded score. Determine the frequency of home visits ,formula: When 50≤ <65 hours, Times / week; when 65≤ <80 hours, Once per week.

[0042] (2) Daily living assistance intervention: Smart reminders (such as medication reminders, meal reminders, and outing reminders) are implemented through IoT devices to prevent people from getting lost or missing medications; family members assist in daily living care and receive professional care guidance; medication reminder intervals Based on the doctor's orders, the trigger conditions for the reminder are: the difference between the current time and the preset medication time is ≤10 minutes, and no medication behavior is detected (confirmed by vital sign monitoring equipment); the trigger conditions for the missing person reminder are: the activity trajectory exceeds the preset safe area (based on GPS positioning, safe area radius). ,formula: (Unit: meters)

[0043] (3) Regular assessment: A cognitive assessment and physiological indicator test will be conducted monthly to dynamically adjust the grading and intervention plan. Further diagnosis and treatment will be conducted in conjunction with the hospital when necessary. Assessments will be triggered on a fixed date each month. Additionally, a temporary assessment will be triggered if abnormal physiological indicators persist for ≥3 days. The abnormality criterion is: physiological indicators exceeding the normal range by 20%. Formula: ( (These are normal reference values ​​for physiological indicators).

[0044] Level 3 Severe Cognitive Impairment Intervention Unit: With "maintaining quality of life and providing comprehensive care" as its core, this unit implements professional and all-round intervention measures, specifically including: (1) Professional care: Nursing staff in elderly care institutions or home caregivers provide 24-hour care, assisting with daily living activities such as eating, dressing, and washing, and preventing complications such as pressure sores and pneumonia; care intervals Based on the degree of limb movement impairment (0-10 points, higher scores indicate more severe obstacles) Confirmed, formula: (Unit: minutes), when hour, minute.

[0045] (2) Medical intervention: Establish a green channel for diagnosis and treatment in conjunction with hospitals, and have professional doctors conduct regular home visits to monitor changes in the patient's condition, adjust medication plans, and provide targeted medical support; consultation cycle Based on graded scores Confirmed, formula: (Unit: day), when hour, sky.

[0046] (3) Emotional intervention: Caregivers work with family members to provide emotional support and psychological counseling to the elderly, alleviate anxiety, depression and other emotions, and improve their quality of life; emotional feature values ​​are extracted through behavioral data (facial expressions, activity levels, tone of voice). (0-100 points, the lower the score, the worse the mood), duration of emotional companionship formula: (Unit: minutes / day), when hour, minutes / day.

[0047] (4) Real-time monitoring: IoT devices collect physiological and behavioral data in real time. Once an abnormality occurs (such as sudden changes in heart rate or abnormal consciousness), an emergency warning is immediately triggered to notify medical staff and family members for emergency treatment. The triggering condition for abnormal physiological indicators is: heart rate satisfy times / minute or Blood oxygen saturation (Hz / min) <90%, duration ≥30 seconds; Triggering conditions for abnormal consciousness: activity level is 0 and there is no response from human senses, duration ≥10 minutes; Warning response time Seconds (to notify medical staff and family members, etc.).

[0048] In the system described in this embodiment, the IoT sensing and acquisition module addresses the technical bottlenecks of "difficult early screening and limited monitoring scenarios" in existing technologies, enabling the routine and convenient collection of cognitive impairment-related data in daily elderly care scenarios such as home and community settings. The data transmission and storage module solves the problems of unstable multi-source data transmission and insecure storage, achieving real-time data transmission and long-term secure storage. The intelligent analysis and hierarchical assessment module addresses the technical shortcomings of "insufficient data fusion and intelligent decision-making" and "lack of a systematic hierarchical management system" in existing technologies, enabling accurate screening, risk warning, and Mainm three-level hierarchical assessment of cognitive impairment. The Mainm three-level intervention execution module implements the hierarchical management concept, providing precise and personalized hierarchical interventions for patients at different levels, addressing the problem of "low matching degree between intervention methods and disease conditions" in existing technologies. Specific intervention plans correspond to the Mainm three-level hierarchical classification, achieving a closed-loop intervention process.

[0049] Example 2 This embodiment is applied to the home-based elderly care scenario, targeting people with mild cognitive impairment, to achieve data collection, early screening and basic intervention.

[0050] During the equipment deployment phase, sleep belts are deployed in bedrooms at home to collect physiological data such as heart rate, sleep quality, and activity levels.

[0051] Human body sensors and temperature and humidity sensors are deployed in key areas such as the living room and kitchen to capture behavioral and environmental data in real time.

[0052] Place cognitive assessment terminals suitable for the elderly. These terminals have built-in simple cognitive assessment scales and interactive testing tools, and support both touch and voice interaction modes, reducing the barrier to use for the elderly.

[0053] Equipped with vital sign monitoring equipment, including a blood pressure monitor and a heart rate meter, for long-term monitoring of physiological characteristics such as blood pressure, heart rate, and blood oxygen.

[0054] All devices are connected to the home WiFi network, and initialization and debugging are completed to ensure that the device battery life is ≥30 days. Based on the timestamp synchronization algorithm, the data collection time deviation of multiple devices is ≤1 second.

[0055] During the data acquisition phase, a combination of active and passive acquisition methods was adopted.

[0056] The passive data acquisition module collects physiological data every hour through a sleep belt, human body sensor, temperature and humidity sensor, and vital sign monitoring equipment, and collects behavioral and environmental data in real time without requiring active operation by the elderly.

[0057] The active data collection module pushes a simple cognitive test task once a week through the cognitive assessment terminal, including memory test, attention test, and language ability test. The elderly can complete the test by touch or voice.

[0058] During the data preprocessing stage, the system performs noise reduction, deduplication, and standardization on the collected multi-source data.

[0059] During the data transmission and storage phase, WiFi transmission mode is used to transmit the pre-processed data to the local gateway and cloud server in real time.

[0060] The cloud uses the AES-128 encryption algorithm for data encryption and storage. The key length is 128 bits. The encryption process is as follows: initial round key addition → byte substitution → row shifting → column mixing → round key addition, which is repeated for 10 rounds to ensure data security and prevent the leakage of personal health information.

[0061] The local gateway stores data from the past 7 days for quick access and emergency queries, ensuring data transmission latency of ≤5 seconds and a transmission success rate of ≥99.8%.

[0062] In the intelligent analysis phase, the system uses a weighted fusion algorithm to fuse multi-source data.

[0063] The comprehensive data F and anomaly probability are calculated using a multi-source data fusion algorithm. Combined with the four-dimensional weighted scoring formula ,when Anomaly probability When the value is less than 0.8, it is determined to be a level 1 mild cognitive impairment, triggering a mild warning.

[0064] In the tiered assessment phase, such as Figure 2 As shown, the system calculates the total score S for the graded evaluation based on a four-dimensional weighted scoring model.

[0065] The model is based on cognitive function scores. Daily Living Skills Score Physiological index scores Behavior performance The weighted calculation yields weights of 0.4, 0.3, 0.2, and 0.1, respectively. (Formula follows) Total score: 100 points.

[0066] when point, 0.7≤ <0.8， If there is no record of abnormal behavior, it is determined to be a level one mild cognitive impairment, triggering a mild warning.

[0067] Tiered intervention stage, such as Figure 3 As shown, the system automatically pushes personalized intervention plans based on the grading results.

[0068] Level 1 mild cognitive impairment intervention focuses on "preventing progression and strengthening cognition," implementing basic intervention measures that do not require full-time professional medical staff intervention. These measures include: (1) Cognitive training intervention: Personalized cognitive training tasks (such as memory games, language training, and logical reasoning training) are regularly pushed through a cognitive assessment terminal to encourage the elderly to complete them daily, thereby strengthening cognitive function. The core technology is a personalized training task matching algorithm; based on cognitive assessment scores... Matching training difficulty, formula: (Difficulty range) , (The higher the level, the lower the difficulty) Daily training time (Unit: minutes)

[0069] (2) Lifestyle intervention: Based on collected physiological and behavioral data, personalized lifestyle suggestions (such as dietary recommendations, exercise plans, and sleep adjustments) are provided to the elderly, with family members assisting in monitoring their implementation; Taking sleep intervention as an example, based on sleep quality data... (0-100 points) Recommended sleep duration ,formula: (Unit: hour); Exercise intervention is based on activity level. (Steps / day), Recommended daily exercise duration ,formula: (Unit: minutes), when hour, minute.

[0070] A comprehensive cognitive assessment is triggered every 3 months, forming a closed management loop of "collection → analysis → assessment → intervention → update".

[0071] Through the above implementation methods, routine monitoring and basic intervention for mild cognitive impairment in home-based elderly care scenarios have been achieved. By adopting a data collection method that combines active and passive approaches, the problem of traditional detection methods relying on professional personnel and large-scale equipment has been solved, thereby improving the coverage rate of early screening.

[0072] Example 3 This embodiment is applied to community-based elderly care scenarios, targeting individuals with moderate cognitive impairment, to achieve collaborative intervention combining professional and family-based approaches.

[0073] During the equipment deployment phase, wall-mounted cognitive assessment terminals and vital sign monitoring equipment are deployed in community elderly care service centers, and portable vital sign monitoring equipment is provided for elderly people with mild cognitive impairment living at home.

[0074] The device uses 5G or LoRa transmission mode to connect with the community server and cloud.

[0075] Data transmission adopts a multi-network convergence transmission method of "5G+WiFi+LoRa", and automatically switches the transmission mode according to the signal strength of the scene.

[0076] Specifically, the handover timing is determined based on the RSSI signal strength, assuming the current network RSSI is... The preset threshold is ,when If the condition persists for more than 3 seconds, a network switch will be triggered. WiFi threshold ≥ -60dBm, 5G threshold ≥ -70dBm, LoRa threshold ≥-80dBm .

[0077] For areas with weak signal, a data caching function is set up to automatically synchronize data after the signal is restored, ensuring that the real-time data transmission delay T≤5 seconds and the non-real-time data transmission delay T≤30 seconds.

[0078] During the data collection and analysis phase, community medical staff assist the elderly in completing a cognitive assessment once a month, and home devices continuously collect physiological and behavioral data and transmit them in real time.

[0079] The system completes data fusion analysis through random forest classification, and calculates the hierarchical score S. , 0.8≤ <0.9， If one or two abnormal behaviors or physiological indicators are present, the condition is classified as Level II moderate cognitive impairment.

[0080] In the tiered intervention phase, according to the formula N represents the number of home visits per week. Community medical staff will visit homes 1-2 times per week to conduct professional rehabilitation training. They will also use IoT devices to send reminders for medication and meals, and guide family members in completing daily care. The tiered update and intervention plan will be adjusted monthly.

[0081] Intervention for Level 2 Moderate Cognitive Impairment: Focusing on "delaying progression and assisting daily living," intervention measures combining professional and family-based approaches are implemented, specifically including: Professional intervention: Community healthcare workers or nursing home staff visit 1-2 times per week to provide professional cognitive intervention and rehabilitation training (such as speech and physical rehabilitation), regularly monitor physiological indicators, and adjust the intervention plan accordingly; based on the graded score. Determine the frequency of home visits ,formula: When 50≤ <65 hours, Times / week; when 65≤ <80 hours, Once per week.

[0082] Assisted living intervention: Smart reminders (such as medication reminders, meal reminders, and departure reminders) are implemented through IoT devices to prevent getting lost or missing medications; family members assist with daily care while receiving professional care guidance; medication reminder intervals... Based on the doctor's orders, the trigger conditions for the reminder are: the difference between the current time and the preset medication time is ≤10 minutes, and no medication behavior is detected (confirmed by vital sign monitoring equipment); the trigger conditions for the missing person reminder are: the activity trajectory exceeds the preset safe area (based on GPS positioning, safe area radius). ,formula: (Unit: meters)

[0083] Regular assessments: Cognitive assessments and physiological indicator tests are conducted monthly to dynamically adjust the grading and intervention plan, and to coordinate with hospitals for further diagnosis and treatment when necessary; assessments are triggered on fixed dates each month, and temporary assessments are triggered when abnormal physiological indicators persist for ≥3 days. The abnormality criterion is: physiological indicators exceeding the normal range by 20%, formula: ( (These are normal reference values ​​for physiological indicators).

[0084] Guide family members to complete daily care, and complete the graded update and intervention plan adjustment once a month.

[0085] Through the above implementation methods, professional intervention for moderate cognitive impairment in community-based elderly care scenarios has been achieved. By adopting multi-network fusion transmission technology, the problem of unstable network signals in community scenarios has been solved, ensuring the real-time transmission of data.

[0086] Example 4 This embodiment is applied to the setting of elderly care institutions, and provides 24-hour professional care and medical intervention for people with severe cognitive impairment.

[0087] During the equipment deployment phase, full-scene sensing equipment, 24-hour vital sign monitoring equipment, and emergency early warning terminals are deployed in the wards and activity areas of elderly care institutions.

[0088] Employing a multi-network converged transmission mode ensures no signal blind spots, enabling the device to collect physiological and behavioral data in real time.

[0089] Intelligent analysis and tiered assessment: The system analyzes and collects data in real time, and assigns a tiered score. point, ≥0.9， When three or more abnormal behaviors or severely abnormal physiological indicators are present, the individual is diagnosed with level three severe cognitive impairment.

[0090] An emergency warning is triggered immediately when the heart rate (HR) is <50 beats / min or >120 beats / min and the blood oxygen level is <90% for ≥30 seconds.

[0091] When the activity level is 0 and there is no response from the human body sensor, the duration is ≥10 minutes; Early warning response time Within seconds, the system automatically pushes early warning information to medical staff and their families.

[0092] In the tiered intervention phase, nursing staff use the formula Determine care intervals and provide 24-hour continuous care; medical staff follow the formula Determine the consultation cycle and adjust the medication plan; arrange daily emotional support based on the emotional characteristic value E to ensure the quality of life of the elderly.

[0093] Intervention for Level 3 Severe Cognitive Impairment: Centered on "maintaining quality of life and providing comprehensive care," this intervention involves professional and all-encompassing measures, specifically including: Professional care: 24-hour care from nursing home staff or home caregivers, assisting with daily living activities such as eating, dressing, and washing, and preventing complications such as pressure sores and pneumonia; care intervals... Based on the degree of limb movement impairment (0-10 points, higher scores indicate more severe obstacles) Confirmed, formula: (Unit: minutes), when hour, minute.

[0094] Medical intervention: Establish a green channel for diagnosis and treatment in conjunction with hospitals, with regular home visits by professional doctors to monitor changes in the patient's condition, adjust medication regimens, and provide targeted medical support; consultation cycle. Based on graded scores Confirmed, formula: (Unit: day), when hour, sky.

[0095] Emotional intervention: Caregivers work with family members to provide emotional support and psychological counseling to the elderly, alleviating anxiety, depression, and other emotions, and improving their quality of life; emotional characteristic values ​​are extracted through behavioral data (facial expressions, activity levels, tone of voice). (0-100 points, the lower the score, the worse the mood), duration of emotional companionship formula: (Unit: minutes / day), when hour, minutes / day.

[0096] Real-time monitoring: IoT devices collect physiological and behavioral data in real time. In the event of any abnormality (such as sudden changes in heart rate or altered consciousness), an emergency alert is immediately triggered, notifying medical staff and family members for emergency treatment. Triggering conditions for abnormal physiological indicators: heart rate... satisfy times / minute or Blood oxygen saturation (Hz / min) <90%, duration ≥30 seconds; Triggering conditions for abnormal consciousness: activity level is 0 and there is no response from human senses, duration ≥10 minutes; Warning response time Seconds (notifying medical staff and family members).

[0097] Through the above implementation methods, comprehensive care for severe cognitive impairment in elderly care institutions has been achieved. By automatically calculating care intervals, consultation cycles, and duration of emotional companionship based on the grading score, precise and comprehensive care for severe cognitive impairment has been realized, improving the quality of life for the elderly.

[0098] Example 5 A method for managing chronic diseases of cognitive impairment based on the Internet of Things includes the following steps: (2) Physiological, behavioral and environmental data are collected regularly through the IoT sensing device in the IoT sensing and acquisition module. Simple cognitive test tasks are pushed regularly through the cognitive assessment terminal. The elderly complete the test to realize the collection of cognitive data. The collected physiological, behavioral, environmental and cognitive data are denoised, deduplicated and standardized to remove abnormal data. (2) The processed physiological, behavioral, environmental and cognitive data are transmitted to the data transmission and storage module. The data transmission adopts the integration of 5G, WiFi and LoRa networks and dynamically switches the communication link according to the RSSI signal strength. WiFi is used for home scenarios, 5G or LoRa is used for community and institutional scenarios, and data caching function is set for remote areas or areas with weak signals. The data is automatically synchronized after the signal is restored. The storage includes local storage and cloud storage. Local storage is used to store recent real-time data for quick access and emergency retrieval. Cloud storage uses AES encrypted cloud servers to store long-term data. The AES encryption algorithm ensures data security and prevents the leakage of personal health information. It also supports data backup and recovery functions to ensure data integrity. (3) The data transmission and storage module transmits physiological, behavioral, environmental, and cognitive data to the intelligent analysis and hierarchical evaluation module, and uses a weighted fusion algorithm to assign weights to different types of data. The merged data is obtained, a decision tree is constructed, and random forest classification is performed on the merged data. The final classification result is output using a voting method to determine the anomaly probability P. abn According to P abn The warning is divided into three levels, with a mild warning level of 0.7 ≤ P. abn <0.8, moderate warning 0.8≤P abn <0.9, Severe Warning P abn ≥0.9, when P abn An alert is triggered when the value is ≥0.7; Based on the progression characteristics of chronic cognitive impairment and combined with early warning levels, a three-tiered grading assessment method called the Maim assessment was established, and a total score for the grading assessment was constructed. Based on cognitive function scores Daily Living Skills Score Physiological index scores Behavior performance The weighted calculation yields weights of 0.4, 0.3, 0.2, and 0.1, respectively. (Formula follows) The total score is 100 points, and the condition is graded into three levels: Level 1 mild cognitive impairment, Level 2 moderate cognitive impairment, and Level 3 severe cognitive impairment. The condition is classified into three levels. Level 1 mild cognitive impairment, with a slight decline in cognitive function and basically normal daily living abilities, determined by the score of the Simplified Cognitive Assessment Scale, and no obvious abnormal behavior or fluctuations in physiological indicators; the quantitative standard is 80≤S≤100 points, S1≥75 points, S3≥75 points, and no record of abnormal behavior. Level 2 moderate cognitive impairment, with significant decline in cognitive function, partial impairment of daily living abilities, and obvious behavioral or physiological abnormalities; the quantitative standard is 50≤S<80 points, 50≤S1<75 points, 60≤S3<80 points, with abnormal behavior or physiological abnormalities. Level 3 severe cognitive impairment, with severely impaired cognitive function, loss of basic self-care ability, requiring 24-hour care, and significant abnormalities and large fluctuations in physiological indicators; the quantitative standard is S < 50 points, S1 < 50 points, S3 < 60 points, with abnormal behavior or severely abnormal physiological indicators.

[0099] The intelligent analysis and grading assessment module periodically re-analyzes and evaluates the collected data, dynamically adjusting the grading results based on changes in the patient's cognitive function, physiological state, and behavioral performance. This ensures the timeliness and accuracy of the grading assessment and calculates the score difference between two grading assessments. , For the current score, For the last score, when Time-sharing triggers tiered adjustments; when The original grading system is maintained throughout the time period, while the trend of score changes is recorded. The formula is: T represents the number of days between two assessments, used to predict the direction of disease progression.

[0100] (4) Based on the disease severity assessment results evaluated by the intelligent analysis and grading assessment module, the three-level intervention units are matched to implement the corresponding intervention measures to achieve a closed-loop intervention process.

[0101] The three-level intervention units are the first-level mild cognitive impairment intervention units: (I) Cognitive training intervention, specifically based on a personalized training task matching algorithm, which regularly pushes personalized cognitive training tasks through a cognitive assessment terminal, encouraging the elderly to complete them daily to strengthen cognitive function; based on cognitive assessment scores... The formula for the algorithm to match training difficulty and personalized training tasks is as follows: Difficulty range , The higher the level, the lower the difficulty, and the longer the daily training time. (i) The unit is minutes; (ii) Lifestyle intervention, which is based on the collected physiological and behavioral data to push personalized life suggestions to the elderly, and family members assist in supervising the implementation; The Level 2 moderate cognitive impairment intervention unit provides professional cognitive intervention, rehabilitation training, daily living assistance intervention, and regular assessments through community healthcare workers or nursing staff in elderly care institutions. Specifically, professional cognitive intervention and rehabilitation training involve community healthcare workers or nursing home staff visiting homes 1-2 times per week to conduct professional cognitive intervention and rehabilitation training, regularly monitoring physiological indicators, and adjusting the intervention plan accordingly; based on the graded score... Determine the frequency of home visits ,formula: When 50≤ <65 hours, Times / week; when 65≤ <80 hours, Once per week; Assisted living intervention specifically uses IoT devices to provide smart reminders, allowing family members to assist with daily care while receiving professional care guidance and medication reminders. Based on the doctor's orders, the trigger conditions for the reminder are: the difference between the current time and the preset medication time is ≤10 minutes, and no medication behavior is detected by the vital signs monitoring device; the trigger conditions for the missing person reminder are: the activity trajectory is located outside the preset safe area via GPS, and the radius of the safe area is [missing information]. ,formula: The unit is meters; Regular assessments involve monthly cognitive evaluations and physiological indicator testing to dynamically adjust the grading and intervention plan, and, when necessary, coordinate with the hospital for further diagnosis and treatment. Assessments are triggered on a fixed date each month, and temporary assessments are triggered when abnormal physiological indicators persist for ≥3 days. The abnormality criterion is a physiological indicator exceeding the normal range by 20%, as shown in the formula: These are normal reference values ​​for physiological indicators.

[0102] Level 3 Severe Cognitive Impairment Intervention Unit: (S1) Nursing staff in elderly care institutions or home caregivers provide 24-hour professional care, assisting with daily living activities and preventing complications; the interval between care sessions... Based on the degree of limb movement impairment Sure, The score ranges from 0 to 10, with higher scores indicating more severe obstacles. The formula is: The unit is minutes. hour, minute; (S2) Establish a green channel for diagnosis and treatment in conjunction with hospitals, with professional doctors making regular home visits to monitor changes in the patient's condition, adjust medication plans, and provide targeted medical support to achieve medical intervention; the consultation cycle Based on graded scores Confirmed, formula: The unit is days. hour, sky; (S3) Caregivers work with family members to provide emotional support and psychological counseling to the elderly, alleviate their emotions, improve their quality of life, and achieve emotional intervention; emotional feature values ​​are extracted from behavioral data such as facial expressions, activity levels, and tone of voice. The value of E ranges from 0 to 100 points, with lower scores indicating worse moods and longer emotional companionship durations. formula: The unit is minutes / day. hour, minutes / day; (S4) Real-time monitoring: IoT sensing devices collect physiological and behavioral data. When abnormal data is detected, an emergency alert is immediately triggered, notifying medical staff and family members for emergency handling. Physiological data abnormality trigger condition: Heart rate. satisfy times / minute or Blood oxygen saturation (Hz / min) <90%, duration ≥30 seconds; Abnormal behavior triggering conditions: activity level 0 and no human body sensor response, duration ≥10 minutes; Warning response time Within seconds, medical staff and family members were notified.

[0103] The cognitive impairment chronic disease health management system based on the Internet of Things described in this invention takes into account both technical logic and practicality. It achieves early screening, accurate classification and personalized intervention of cognitive impairment through a closed-loop process of "Internet of Things sensing - data transmission and storage - intelligent analysis and classification - classification intervention" and combines multi-source data fusion, machine learning and the Myem classification concept.

[0104] Perception principle: Utilizing low-power IoT sensors, cognitive assessment terminals and other devices, the system captures multi-dimensional data on physiology, behavior, cognition and environment through active and passive acquisition modes. After noise reduction and standardization preprocessing, invalid data is removed to provide high-quality data support for subsequent analysis. Transmission and storage principle: It adopts the "5G+WiFi+LoRa" multi-network integrated transmission, automatically switches modes according to the signal strength of the scene, and is equipped with local + cloud dual storage. It uses AES encryption algorithm to ensure data security and realizes real-time data transmission and long-term retention. Then, based on algorithms such as random forest, multi-source standardized data is fused to extract cognitive abnormality features. A weighted calculation is then performed to obtain a total score for graded assessment. Combined with the Maim three-level grading standard, precise grading of the condition is achieved. A dynamic update algorithm adjusts the grading results according to data changes. The principle of graded intervention: Based on the Maim three-level grading results, corresponding intervention strategies are matched. Through personalized algorithms (such as training difficulty matching and care frequency optimization), precise adaptation of "mild prevention, moderate delay, and severe care" is achieved, forming a closed-loop management of "assessment-intervention-update".

[0105] This invention requires three stages of use: "initial deployment - daily operation - post-maintenance," and is suitable for three major scenarios: home, community, and elderly care institutions. The specific steps are as follows: Initial deployment (completed in one go) Device Deployment: Based on scenario requirements, install sleep belts, environmental sensing devices, cognitive assessment terminals, and vital sign monitoring devices in homes (key areas such as bedrooms and living rooms), community service centers, and elderly care institutions. Complete device debugging, ensure device network connectivity (preferably WiFi / 5G / LoRa compatibility), and normal battery life (≥30 days). System initialization: Set up a local server and an encrypted cloud server, complete the data transmission interface connection, set network switching thresholds, abnormal data judgment criteria, graded assessment parameters and intervention plan benchmark values, and enter basic user information (age, basic medical history, etc.). Personnel training: Provide basic operational training to caregivers and their families, including the use of the cognitive assessment terminal, the reception and processing of early warning information, and the daily inspection of basic equipment.

[0106] Routine operation (standardized procedures) Data collection: The device automatically initiates "active + passive" data collection. Passive data collection requires no manual operation and captures physiological, behavioral, and environmental data in real time. Active data collection involves the cognitive assessment terminal periodically pushing test tasks, which are then completed by caregivers with the assistance of the elderly (touch / voice operation). Data processing and analysis: The system automatically preprocesses the collected data (noise reduction, deduplication, and standardization), transmits it to the server for storage and encryption, and the intelligent analysis module automatically runs algorithms to complete data fusion, anomaly identification and hierarchical evaluation, and generates an evaluation report; Early warning and intervention: When abnormal data is detected (such as abnormal physiological indicators or decreased cognitive scores), the system automatically triggers the corresponding level of early warning and pushes it to caregivers / family members; caregivers implement corresponding intervention measures (such as mild cognitive training, moderate home intervention, and severe 24-hour care) according to the classification results. Dynamic updates: The system automatically re-analyzes the latest data monthly and adjusts the grading results and intervention plans; caregivers complete regular assessments and records as required (every 3 months for mild cases, monthly for moderate cases, and real-time for severe cases).

[0107] Post-maintenance (performed periodically) Equipment maintenance: Check the operating status of the sensing equipment weekly, replace the battery in time (if the battery life is less than 7 days), clean the dust from the equipment, and ensure the accuracy of the collected data; Data maintenance: Check the synchronization status of local and cloud data monthly, complete data backup verification, and ensure data integrity; update encryption keys every six months to ensure data security; System maintenance: The system is upgraded quarterly to optimize algorithm parameters (such as tiered evaluation weights and intervention scheme matching logic), fix operational faults, and ensure smooth system operation.

[0108] The innovation of this invention: (1) Improvement of the “Full Quantification” Grading Evaluation System Based on the Maim Concept Pain points of traditional technologies: Traditional health and wellness assessments are mostly textual descriptions or vague classifications, lacking objective quantitative indicators, making it difficult to accurately guide intervention plans.

[0109] The innovation of this invention is the quantitative scoring model: a four-dimensional weighted scoring model (formula S=0.4S1+0.3S2+0.2S3+0.1S4) is constructed with cognitive function, living ability, physiological indicators and behavioral performance as the core, which transforms the abstract degree of cognitive impairment into a specific score of 0-100.

[0110] Dynamic adaptive grading: A dynamic grading update algorithm is introduced. By calculating the difference ΔS between two assessment scores, the automatic identification of disease progression and real-time dynamic adjustment of grading results are achieved, which solves the static defect of traditional management that "one assessment determines life".

[0111] (2) Integrated innovation of the “assessment-intervention-feedback” closed-loop system Pain points of traditional technologies: Existing technologies often "emphasize monitoring but neglect intervention", resulting in a disconnect between data and actual care actions, and a lack of executable implementation procedures.

[0112] This invention features innovative precision intervention matching: breaking away from generic elderly care solutions, it designs differentiated and quantifiable intervention execution modules tailored to the Mainm three-level classification (mild / moderate / severe). For example, it automatically calculates the frequency of home visits N based on the score S and recommends sleep duration Ts based on sleep quality Qs. Full-link closed-loop control: It achieves a fully automated closed loop from data collection, intelligent analysis, and classification determination to automatically pushing specific intervention tasks (such as medication reminders and cognitive training), ensuring that assessment results can be directly translated into actual care actions.

[0113] (3) Improvement of aging-friendly multi-terminal and heterogeneous network integration technology Traditional technical pain points: The operation threshold is high for elderly users, and the network signal is unstable in remote communities or home environments, resulting in serious data packet loss.

[0114] This invention features innovative low-power, lightweight deployment: the device adopts an age-friendly, minimalist interactive design, supporting both touch and voice modes to reduce the difficulty of use for the elderly; the hardware employs low-power sensing technology to ensure long battery life.

[0115] Multi-network fault-tolerant transmission: innovative adoption "5G + WiFi + LoRa" Tri-network Convergence Automatic Switching Technology The system intelligently switches networks based on RSSI signal strength and configures a local cache synchronization mechanism, completely solving the common industry problems of unstable data transmission and data loss in complex elderly care scenarios, and ensuring data integrity and real-time performance.

[0116] In summary, through the above embodiments, this invention achieves accurate detection, classification, and intervention of cognitive impairment in multiple scenarios, solves many shortcomings of existing technologies, and can be widely applied in the field of smart elderly care to realize closed-loop management of chronic diseases such as cognitive impairment.

[0117] It should be noted that the aforementioned cognitive assessment terminal is not limited to tablets or wall-mounted terminals, but can also be other forms of smart terminals such as smartwatches and smart speakers, as long as they can push cognitive test tasks and support touch / voice interaction.

[0118] Obviously, the above algorithm is not limited to the random forest algorithm. Other machine learning algorithms such as neural networks and support vector machines can also be used, as long as they can fuse and analyze multi-source data and identify cognitive dysfunction.

[0119] It is understandable that the above-mentioned multi-network convergence transmission method is not limited to the "5G+WiFi+LoRa" three-network combination, but can also use "4G+WiFi+NB-IoT" or other network combinations, as long as the transmission mode can be automatically switched according to the signal strength of the scene.

[0120] Obviously, the specific parameters in the quantitative matching algorithm for the above-mentioned graded intervention are not limited to the specific values ​​given in this application. The weight coefficients and thresholds can be adjusted according to the actual application scenario, as long as the intervention parameters can be automatically calculated based on the graded results and the intervention plan can be accurately matched with the severity of the disease.

[0121] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

Claims

1. A method for health management of chronic diseases with cognitive impairment based on the Internet of Things, characterized in that, Includes the following steps: (1) Physiological, behavioral and environmental data are collected regularly by the IoT sensing device in the IoT sensing and acquisition module. Simple cognitive test tasks are pushed regularly by the cognitive assessment terminal. The elderly complete the test to realize the collection of cognitive data. The collected physiological, behavioral, environmental and cognitive data are denoised, deduplicated and standardized to remove abnormal data. (2) The processed physiological, behavioral, environmental and cognitive data are transmitted to the data transmission and storage module. The data transmission adopts the integration of 5G, WiFi and LoRa networks and dynamically switches the communication link according to the RSSI signal strength. WiFi is used for home scenarios, 5G or LoRa is used for community and institutional scenarios, and data caching function is set for remote areas or areas with weak signals. The data is automatically synchronized after the signal is restored. The storage includes local storage and cloud storage. Local storage is used to store recent real-time data for quick access and emergency retrieval. Cloud storage uses AES encrypted cloud servers to store long-term data. The AES encryption algorithm ensures data security and prevents the leakage of personal health information. It also supports data backup and recovery functions to ensure data integrity. (3) The data transmission and storage module transmits physiological, behavioral, environmental, and cognitive data to the intelligent analysis and hierarchical evaluation module, and uses a weighted fusion algorithm to assign weights to different types of data. The merged data is obtained, a decision tree is constructed, and random forest classification is performed on the merged data. The final classification result is output as anomaly probability using a voting method. According to the probability of anomalies The warning is divided into three levels, with a minimum warning level of 0.

7. <0.8, moderate warning 0.8≤ <0.9, severe warning ≥0.9, when An alert is triggered when the value is ≥0.7; Based on the progression characteristics of chronic cognitive impairment and combined with early warning levels, a three-tiered grading assessment method called the Maim assessment was established, and a total score for the grading assessment was constructed. Based on cognitive function scores Daily Living Skills Score Physiological index scores Behavior performance The weighted calculation yields weights of 0.4, 0.3, 0.2, and 0.1, respectively. (Formula follows) The total score is 100 points, and the condition is graded into three levels: Level 1 mild cognitive impairment, Level 2 moderate cognitive impairment, and Level 3 severe cognitive impairment. (4) Based on the disease severity assessment results evaluated by the intelligent analysis and grading assessment module, the three-level intervention units are matched to implement the corresponding intervention measures to achieve a closed-loop intervention process.

2. The method according to claim 1, characterized in that: In step (3), the condition is classified into three levels. Level 1 mild cognitive impairment, characterized by a slight decline in cognitive function, with generally normal daily living abilities, as determined by a score on the Mini-Cognitive Assessment Scale, and no significant abnormal behavior or fluctuations in physiological indicators; the quantitative standard is... Score: No record of abnormal behavior; Level II moderate cognitive impairment, characterized by a significant decline in cognitive function, partial impairment of daily living abilities, and marked behavioral or physiological abnormalities; the quantitative standard is... One or two abnormal behaviors or physiological indicators are present. Level 3 severe cognitive impairment, with severely impaired cognitive function, loss of basic self-care ability, requiring 24-hour care, and significant and fluctuating physiological indicators; the quantitative standard is... The score indicates the presence of three or more abnormal behaviors or severely abnormal physiological indicators.

3. The method according to claim 1, characterized in that: In step (3), the intelligent analysis and grading assessment module periodically re-analyzes and assesses the collected data, dynamically adjusts the grading results based on changes in the patient's cognitive function, physiological state, and behavioral performance, ensures the timeliness and accuracy of the grading assessment, and calculates the score difference between the two grading assessments. , For the current score, For the last score, when Time-sharing triggers tiered adjustments; when The original grading system is maintained throughout the time period, while the trend of score changes is recorded. The formula is: T represents the number of days between two assessments, used to predict the direction of disease progression.

4. The method according to claim 1, characterized in that: In step (4), the three levels of intervention units are the first-level mild cognitive impairment intervention units: (I) Cognitive training intervention, specifically based on a personalized training task matching algorithm, which regularly pushes personalized cognitive training tasks through a cognitive assessment terminal, encouraging the elderly to complete them daily to strengthen cognitive function; based on cognitive assessment scores The formula for the algorithm to match training difficulty and personalized training tasks is as follows: Difficulty range , The higher the level, the lower the difficulty, and the longer the daily training time. (i) The unit is minutes; (ii) Lifestyle intervention, which is based on the collected physiological and behavioral data to push personalized life suggestions to the elderly, and family members assist in supervising the implementation; The Level 2 moderate cognitive impairment intervention unit provides professional cognitive intervention, rehabilitation training, daily living assistance intervention, and regular assessments through community healthcare workers or nursing staff in elderly care institutions. Level 3 Severe Cognitive Impairment Intervention Unit: (S1) 24-hour professional care from nursing home staff or home caregivers to assist in daily living activities and prevent complications; (S2) Establishing a green channel for diagnosis and treatment in conjunction with hospitals, with regular home visits by professional doctors to monitor changes in the condition, adjust medication plans, and provide targeted medical support to achieve medical intervention; (S3) Nursing staff and family members cooperate to provide emotional companionship and psychological counseling to alleviate emotions, improve quality of life, and achieve emotional intervention; (S4) Real-time monitoring: IoT sensing devices collect physiological and behavioral data. When abnormal data occurs, an emergency warning is immediately triggered to notify medical staff and family members for emergency handling.

5. The method according to claim 4, characterized in that: In the Level 2 moderate cognitive impairment intervention unit, professional cognitive intervention and rehabilitation training specifically involve community healthcare workers or nursing home caregivers visiting the patient 1-2 times per week to conduct professional cognitive intervention and rehabilitation training, regularly monitoring physiological indicators, and adjusting the intervention plan according to the graded score. Determine the frequency of home visits ,formula: When 50≤ <65 hours, Times / week; when 65≤ <80 hours, Once per week; Assisted living intervention specifically uses IoT devices to provide smart reminders, allowing family members to assist with daily care while receiving professional care guidance and medication reminders. Based on the doctor's orders, the trigger conditions for the reminder are: the difference between the current time and the preset medication time is ≤10 minutes, and no medication behavior is detected by the vital signs monitoring device; the trigger conditions for the missing person reminder are: the activity trajectory is located outside the preset safe area via GPS, and the radius of the safe area is [missing information]. ,formula: The unit is meters; Regular assessments involve monthly cognitive evaluations and physiological indicator testing to dynamically adjust the grading and intervention plan, and, when necessary, coordinate with the hospital for further diagnosis and treatment. Assessments are triggered on a fixed date each month, and temporary assessments are triggered when abnormal physiological indicators persist for ≥3 days. The abnormality criterion is a physiological indicator exceeding the normal range by 20%, as shown in the formula: These are normal reference values ​​for physiological indicators.

6. The method according to claim 4, characterized in that: In (S1), the interval of care. Based on the degree of limb movement impairment Sure, The score ranges from 0 to 10, with higher scores indicating more severe obstacles. The formula is: The unit is minutes. hour, minute; (S2) Consultation cycle Based on graded scores Confirmed, formula: The unit is days. hour, sky; In (S3), emotional feature values ​​are extracted from behavioral data such as facial expressions, activity levels, and tone of voice. The value of E ranges from 0 to 100 points, with lower scores indicating worse moods and longer emotional companionship durations. formula: The unit is minutes / day. hour, minutes / day; In (S4), the physiological data abnormality trigger condition is: heart rate. satisfy times / minute or Blood oxygen saturation (Hz / min) <90%, duration ≥30 seconds; Abnormal behavior triggering conditions: activity level 0 and no human body sensor response, duration ≥10 minutes; Warning response time Within seconds, medical staff and family members were notified.

7. A health management system for chronic diseases of cognitive impairment based on the Internet of Things, characterized in that: It includes an IoT sensing and acquisition module, a data transmission and storage module, an intelligent analysis and hierarchical evaluation module, and a Mainm three-level intervention execution module. The IoT sensing and acquisition module is connected to the data transmission and storage module via a data connection. The data transmission and storage module is connected to the intelligent analysis and hierarchical evaluation module via a data connection. The intelligent analysis and hierarchical evaluation module is connected to the Mainm three-level intervention execution module via a signal connection. The four modules are connected through standardized interfaces to achieve data interoperability and command linkage. The IoT sensing and acquisition module is used to periodically collect four types of multi-source data—physiological, behavioral, cognitive, and environmental—through IoT sensing devices and cognitive assessment terminals, and to perform noise reduction, deduplication, and standardization on the collected multi-source data, and to remove abnormal data. The data transmission and storage module adopts a multi-network convergence of 5G, WiFi and LoRa, dynamically switches the optimal communication link based on RSSI signal strength, and combines a local cache synchronization mechanism to ensure data integrity in weak network environments. At the same time, it ensures data security and recoverability through a dual storage architecture of local and cloud storage combined with AES-128 encryption and incremental backup strategy. The intelligent analysis and hierarchical evaluation module uses a weighted fusion algorithm to integrate multi-source data acquired by the IoT sensing and acquisition module, and outputs anomaly probabilities based on random forest classification. According to the probability of anomalies The warning is divided into three levels: mild warning ( 0.7≤ <0.8 Moderate warning （0.8≤ <0.9） Severe warning （ ≥0.9） ,when The system triggers an early warning, combined with the four-dimensional weighted scoring formula of the Maim three-level grading assessment. The total score is 100 points. A quantitative total score is generated, categorized into first-level... Level 2 Level 3 A graded assessment of the condition was implemented, classifying the degree of cognitive impairment into three levels: mild, moderate, and severe, and introducing... A dynamic adjustment mechanism for scores avoids rigid evaluation; The Maim three-level intervention execution module matches corresponding differentiated intervention paths and executes intervention measures based on the grading results of the intelligent analysis and grading assessment module.

8. The IoT-based chronic disease health management system for cognitive impairment according to claim 7, characterized in that: In the data transmission and storage module, the handover timing is determined based on the RSSI signal strength. Let the current network RSSI be... The preset threshold is ,when If the condition persists for more than 3 seconds, a network switch will be triggered. WiFi threshold ≥ -60dBm, 5G threshold ≥ -70dBm, LoRa threshold ≥-80dBm .

9. The IoT-based chronic disease health management system for cognitive impairment according to claim 7, characterized in that: In the intelligent analysis and grading assessment module, the specific condition is graded into three levels. Level 1 mild cognitive impairment, characterized by a slight decline in cognitive function, with generally normal daily living abilities, as determined by a score on the Mini-Cognitive Assessment Scale, and no significant abnormal behavior or fluctuations in physiological indicators; the quantitative standard is... Score: No record of abnormal behavior; Level II moderate cognitive impairment, characterized by a significant decline in cognitive function, partial impairment of daily living abilities, and marked behavioral or physiological abnormalities; the quantitative standard is... One or two abnormal behaviors or physiological indicators are present. Level 3 severe cognitive impairment, with severely impaired cognitive function, loss of basic self-care ability, requiring 24-hour care, and significant and fluctuating physiological indicators; the quantitative standard is... A score indicating three or more abnormal behaviors or severely abnormal physiological indicators; Dynamic grading updates involve regularly re-analyzing and evaluating the collected data. The grading results are dynamically adjusted based on changes in the patient's cognitive function, physiological state, and behavioral performance to ensure the timeliness and accuracy of the grading assessment and provide a basis for subsequent grading interventions. Calculate the score difference between the two graded assessments. ,in For the current score, For the last score, when Time-sharing triggers tiered adjustments; when The original grading system is maintained throughout the time period, while the trend of score changes is recorded. The formula is: Where T is the number of days between two assessments, used to predict the direction of disease progression.

10. The IoT-based chronic disease health management system for cognitive impairment according to claim 7, characterized in that: The Mainm three-level intervention execution module includes three-level intervention units corresponding to the disease grading in the intelligent analysis and grading assessment module; The intervention units for mild cognitive impairment include cognitive training intervention and lifestyle intervention; the intervention units for moderate cognitive impairment include professional cognitive intervention, rehabilitation training, life assistance intervention and regular assessment by community medical staff or nursing staff in elderly care institutions; and the intervention units for severe cognitive impairment include 24-hour professional care, medical intervention, emotional intervention and real-time monitoring by nursing staff in elderly care institutions or home caregivers.

Images