A health index abnormal trend early warning method and system for homebound elderly people
Patent Information
- Application Number
- CN202611126270.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-28
- Publication Date
- 2026-10-27
AI Technical Summary
现有健康监测系统虽然能够检测到生理指标的异常变化，但无法区分异常是由疾病进展引起还是药物干预所致，更无法通过服药后指标恢复速度这一关键信息评估药物有效性，现有系统在触发预警时，未将老人的用药行为纳入异常分析框架，居家老人常需服用降压药、降糖药等慢性病治疗药物，服药后生理指标逐渐恢复正常是预期的正常响应过程，而服药后指标持续不恢复则可能是药物失效、病情加重或用药依从性问题的信号
该一种面向居家老人的健康指标异常趋势预警方法及系统，通过获取异常事件发生前后预设时间窗口内的用药行为数据，以服药时间戳为锚点，精确计算从服药时刻至生理指标恢复正常范围的时间间隔作为药后异常持续时间，将异常事件与用药行为进行时间维度的精准关联分析，通过构建历史恢复时间分布参数，将药后异常持续时间与该参数进行统计比对，计算偏离度得分，根据偏离度得分的偏离方向和偏离程度生成差异化的预警，通过偏离度得分计算当药后恢复时间短于历史基线时提示药物敏感性增强风险，长于个体历史基线时提示药物有效性下降风险，实现异常归因的精准判别，通过个体基线将药物有效的判断基准从群体标准转变为老人自身的纵向历史表现，实现药物响应基线的动态追踪与自适应更新，将长时间窗口内多次异常的持续时间加权累加作为异常累积负荷 ，将多次服药后恢复时间与个体基线的比值作为药物响应效率，当异常累积负荷持续上升与药物响应效率持续下降同时发生时，生成药效衰减健康衰退预警，联合预警机制能够捕获整体生理功能下降的早期信号，实现异常累积负荷与药物响应效率的联合预警。
Smart Images
Figure CN122991448A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of health monitoring technology, and in particular to a method and system for early warning of abnormal trends in health indicators for elderly people living at home. Background Technology
[0002] Currently, there is considerable technological accumulation in the field of home-based elderly health monitoring. Wireless network-based health monitoring systems in smart spaces can collect real-time data such as heart rate, sleep patterns, and activity levels through wearable wristbands, mattress pressure sensors, and other devices. This data is pre-processed by edge computing nodes and uploaded to a cloud platform, where AI health assessment algorithms automatically identify abnormal conditions and trigger multi-level warnings. Non-contact monitoring devices such as millimeter-wave radar have also been deployed, enabling 24-hour unobtrusive collection of vital signs data such as respiratory rate, heart rate, and nighttime bed-leaning frequency. AI systems then provide real-time analysis and tiered warnings.
[0003] However, existing home health early warning technologies have the following shortcomings: While existing health monitoring systems can detect abnormal changes in physiological indicators, they cannot distinguish whether the abnormalities are caused by disease progression or medication intervention. Furthermore, they cannot assess drug effectiveness using the crucial information of the speed of indicator recovery after medication. When triggering alerts, existing systems do not include elderly individuals' medication behavior in their anomaly analysis framework. Elderly people living at home often need to take medications for chronic diseases such as antihypertensive and hypoglycemic agents. A gradual return of physiological indicators to normal after medication is an expected and normal response process. However, a persistent failure to recover after medication may signal drug ineffectiveness, worsening of the condition, or medication adherence problems. Current technology conflates the gradual return of indicators to normal after medication with persistent abnormalities when medication is not taken, failing to distinguish the root cause of the abnormality and leading to many abnormal events that could be attributed to medication behavior being misjudged as health deterioration, resulting in ineffective alerts.
[0004] To address the aforementioned technical shortcomings, a solution is proposed. Summary of the Invention
[0005] This invention adopts the following technical solution: a method for early warning of abnormal trends in health indicators for elderly people living at home, comprising the following steps: Step 1: Continuously collect multi-dimensional health monitoring data of elderly people living at home. The multi-dimensional health monitoring data should include at least vital signs data, daily activity data, and medication record data. Step 2: Construct drug effect curves and establish typical response morphology baselines for each vital sign index after medication for each common drug. The typical response morphology baselines include the historical recovery time distribution parameters of the elderly person's physiological indicators from abnormal state to normal range after taking a specific drug. Step 3: Perform real-time anomaly detection on the collected physiological monitoring data. When the physiological indicators deviate from the normal fluctuation range in each dynamic baseline database, mark the start time of the abnormality of the physiological indicators, continuously track the abnormal state of the physiological indicators until they return to the normal range, and record the natural abnormal duration of this abnormal event. Step 4: Obtain medication behavior data within a preset time window before and after the occurrence of the abnormal event. When there is a corresponding medication record after the occurrence of the abnormal event, use the medication timestamp of the medication record as the anchor point to calculate the time interval from the time of medication administration to the recovery of physiological indicators to the normal range, which is used as the post-medication abnormality duration of the abnormal event. Step 5: Statistically compare the abnormal duration of medication with the historical recovery time distribution parameters of the corresponding drugs and physiological indicators of the elderly, calculate the deviation score of the abnormal duration of medication, and generate the first type of warning signal and the second type of warning signal.
[0006] Furthermore, continuous collection of multi-dimensional health monitoring data from elderly people living at home, including at least vital signs data, daily activity data, and medication records, is conducted as follows: Vital signs data are automatically collected through health terminal devices. After each device completes the measurement, it automatically sends the data via Bluetooth broadcast, and receives and transmits it back to the data management center. Daily activity data is collected by environmental sensors and wearable devices deployed in the residence. The daily activity data includes at least the time spent in bed, the frequency of getting out of bed at night, and the speed of movement. The environmental sensors continuously record the activity trajectory by sensing the presence of the human body, and the wearable devices collect posture changes through built-in accelerometers and gyroscopes. The terminal application collects medication record data, including drug name, medication time, medication dosage and administration route. The terminal application supports automatic entry of medication information by taking pictures of drug packaging and prescriptions, and generates medication timeline records. The collected vital signs data, daily activity data, and medication records are synchronized and aligned according to timestamps, and then merged to generate multi-dimensional health monitoring time series data based on the individual elderly person.
[0007] Furthermore, drug effect curves were constructed, and baselines of typical response patterns for various vital signs were established for each common drug after administration. These baselines included the historical recovery time distribution parameters of the elderly patient's physiological indicators from abnormal states to normal ranges after taking the specific drug. The specific process is as follows: Acquire the historical medication records of elderly people living at home and their corresponding vital signs time-series monitoring data. Vital signs include at least heart rate, blood oxygen saturation and blood pressure. Using a single medication event as the anchor point, extract the time-series change trajectory of each vital sign indicator within a preset time window after medication to form the effect curve sample corresponding to this medication event. The effect curve sample contains the numerical sequence of each indicator changing over time from the time of medication. Using the time of medication administration as the zero point, each sample is aligned and normalized on the time axis. Using the steady-state mean of the index during the preset period before medication administration as the benchmark, the measured values at each time after medication administration are converted into deviations from the baseline to generate a normalized effect curve. For multiple normalized response curves of the same drug and the same indicator, fit a typical response curve of the population. The typical response curve includes parameters such as onset time, time to peak, recovery to baseline time, and duration of effect. The mean and standard deviation of the historical recovery time of the physiological indicator from an abnormal state to the normal range after taking the drug are calculated, and stored in association with the typical response curve of the population to form a drug effect curve library.
[0008] Furthermore, real-time anomaly detection is performed on the collected physiological monitoring data. When a physiological indicator deviates from the normal fluctuation range in each dynamic baseline database, the onset time of the abnormality is marked, and the abnormal state of the physiological indicator is continuously tracked until it returns to the normal range. The natural duration of this abnormal event is recorded. The specific process is as follows: The system receives physiological monitoring data in real time from multimodal sensors deployed in the home environment at a preset sampling frequency. The physiological monitoring data includes at least heart rate, blood oxygen saturation, systolic blood pressure, diastolic blood pressure, and nighttime bed-leaning frequency. Based on the current time, the dynamic baseline interval of the corresponding physiological indicator for the elderly is retrieved from the individual dynamic baseline database. The dynamic baseline interval includes the upper limit and lower limit of the normal fluctuation range of the physiological indicator within the current time period. The dynamic baseline database stores the historical statistical distribution parameters of each physiological indicator at different time periods. When the current real-time physiological indicator measurement value is greater than the lower limit of the normal fluctuation range, it is determined that the current physiological indicator has deviated, the current timestamp is marked as the abnormal start time, and the physiological indicator is marked as an abnormal state; In an abnormal state, subsequent measurements of the physiological indicator are continuously received, and each subsequent measurement is compared with the dynamic baseline interval. If the measurement is still outside the dynamic baseline interval, the abnormal state is maintained and the current duration of the abnormality is recorded. When the measured values are detected to have returned to the dynamic baseline range for a preset number of consecutive times, it is determined that the physiological indicators have returned to the normal range, the current timestamp is marked as the abnormal recovery time, the natural abnormal duration of this abnormal event is recorded, and the abnormal event and the natural abnormal duration are stored in the elderly person's historical abnormal event database.
[0009] Furthermore, using the medication timestamp in the medication record as the anchor point, the time interval from the time of medication administration to the return of physiological indicators to the normal range is calculated as the post-medication abnormality duration of this abnormal event. The specific process is as follows: Obtain recorded abnormal event information, which includes the abnormal start time, abnormal recovery time, natural abnormal duration, and the type of physiological indicator that caused the abnormality; Using the moment of the anomaly as the search center point, a preset time window is extended forward and backward to generate a time interval for searching medication behavior. Within the medication behavior retrieval time interval, retrieve the elderly person's medication behavior records. The medication behavior records include at least a medication timestamp, drug name, drug dosage, and route of administration. Determine whether there are any valid medication records whose medication timestamp falls within the retrieval time interval. When there are one or more valid medication records, the medication record whose medication timestamp is closest to and greater than or equal to the abnormal start time is selected as the associated medication record, and the medication timestamp of the associated medication record is extracted. Using the medication timestamp as the anchor point, the time interval from the time of medication administration to the time of abnormal recovery is calculated as the post-medication abnormality duration of this abnormal event; When there are no valid medication records within the search time interval, the abnormal event is marked as a medication-free associated abnormal event, and the medication-free associated abnormal event and its natural abnormal duration are stored in the historical abnormal event database.
[0010] Furthermore, the duration of abnormal drug use after medication was statistically compared with the historical recovery time distribution parameters of the corresponding drugs and physiological indicators in the elderly, and the deviation score of the duration of abnormal drug use after medication was calculated. The specific process is as follows: From the drug response baseline database, retrieve historical recovery time distribution parameters that match the drug name and the type of current abnormal physiological indicator in the currently associated medication record; The historical recovery time distribution parameters include the historical mean and standard deviation of the historical recovery time for the physiological indicator. The historical mean recovery time is the statistical average of the time required for the physiological indicator to return to the normal range after the elderly person took the drug in the past. The historical standard deviation of the recovery time is the statistical standard deviation of the corresponding time. The deviation score of the abnormal duration after medication is calculated based on the mean of historical recovery time and the standard deviation of historical recovery time.
[0011] Furthermore, the generation of the first and second type of early warning signals proceeds as follows: When the deviation score of abnormal duration after medication is between the first threshold and the second threshold, a level 3 warning is output. The level 3 warning is used to indicate a slight deviation in drug response and suggests that close observation is recommended. When the deviation score is greater than the second threshold, a level 2 warning is output. The level 2 warning is used to indicate abnormal drug response and suggests contacting medical staff to evaluate the medication plan. When the increasing trend of consecutive abnormal events of the same physiological indicator reaches a preset number of times within a preset observation window, a Level 1 warning is output. The Level 1 warning is used to indicate that there is a risk of systemic deterioration of health status and recommends immediate medical evaluation.
[0012] A health indicator abnormality trend early warning system for elderly people living at home includes a data acquisition module for continuously collecting multi-dimensional health monitoring data of elderly people living at home. The multi-dimensional health monitoring data includes at least vital signs data, daily activity data and medication record data. The drug baseline construction module is used to establish a typical response pattern baseline for each vital sign index after taking each common drug. The typical response pattern baseline includes the historical recovery time distribution parameters of the physiological indicators of the elderly person after taking a specific drug, from an abnormal state to the normal range. The anomaly monitoring module is used to detect anomalies in the collected physiological monitoring data in real time. When a physiological indicator deviates from the normal fluctuation range in the individual dynamic baseline database, the module marks the start time of the anomaly and continuously tracks the abnormal state of the physiological indicator until it returns to the normal range. The module also records the natural duration of the anomaly event. The anomaly analysis module is used to obtain medication behavior data within a preset time window before and after an abnormal event. When there is a corresponding medication record after an abnormal event, the time interval from the time of medication to the recovery of physiological indicators to the normal range is calculated using the medication timestamp of the medication record as the anchor point. This time interval is used as the post-medication abnormality duration of the abnormal event. The warning signal generation module is used to statistically compare the abnormal duration after medication with the historical recovery time distribution parameters of the corresponding drugs and physiological indicators of the elderly, calculate the deviation score of the abnormal duration after medication, and generate the first type of warning signal and the second type of warning signal based on the deviation score.
[0013] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: This method and system for early warning of abnormal health indicators in elderly people living at home acquires medication behavior data within a preset time window before and after an abnormal event. Using the medication timestamp as an anchor, it accurately calculates the time interval from the medication time to the return of physiological indicators to the normal range as the duration of post-medication abnormality. It performs precise temporal correlation analysis between abnormal events and medication behavior, constructs historical recovery time distribution parameters, statistically compares the duration of post-medication abnormality with these parameters, calculates a deviation score, and generates differentiated warnings based on the direction and degree of deviation. The deviation score indicates a risk of increased drug sensitivity when the post-medication recovery time is shorter than the historical baseline, and a risk when it is longer than the historical baseline. Individual historical baseline data indicates the risk of declining drug efficacy, enabling accurate identification of abnormal attribution. By using individual baseline data, the criterion for judging drug efficacy is transformed from a group standard to the elderly person's own longitudinal historical performance, enabling dynamic tracking and adaptive updating of the drug response baseline. The weighted sum of the durations of multiple abnormalities within a long time window is used as the abnormal cumulative load, and the ratio of recovery time after multiple medications to the individual baseline is used as the drug response efficiency. When the abnormal cumulative load continues to rise and the drug response efficiency continues to decline simultaneously, an early warning of drug efficacy decline and health deterioration is generated. The joint early warning mechanism can capture early signals of overall physiological function decline, achieving joint early warning of abnormal cumulative load and drug response efficiency. Attached Figure Description
[0014] Figure 1 A schematic diagram of the overall structure of the method steps of the present invention is shown; Figure 2 A schematic diagram of the overall system flow structure of the present invention is shown. Detailed Implementation
[0015] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.
[0016] Example 1: like Figure 1 As shown, a method for early warning of abnormal trends in health indicators for elderly people living at home includes the following steps: Step 1: Continuously collect multi-dimensional health monitoring data of elderly people living at home. The multi-dimensional health monitoring data should include at least vital signs data, daily activity data, and medication record data. Step 2: Construct drug effect curves and establish typical response morphology baselines for each vital sign index after medication for each common drug. The typical response morphology baselines include the historical recovery time distribution parameters of the elderly person's physiological indicators from abnormal state to normal range after taking a specific drug. Step 3: Perform real-time anomaly detection on the collected physiological monitoring data. When the physiological indicators deviate from the normal fluctuation range in each dynamic baseline database, mark the start time of the abnormality of the physiological indicators, continuously track the abnormal state of the physiological indicators until they return to the normal range, and record the natural abnormal duration of this abnormal event. Step 4: Obtain medication behavior data within a preset time window before and after the occurrence of the abnormal event. When there is a corresponding medication record after the occurrence of the abnormal event, use the medication timestamp of the medication record as the anchor point to calculate the time interval from the time of medication administration to the recovery of physiological indicators to the normal range, which is used as the post-medication abnormality duration of the abnormal event. Step 5: Statistically compare the abnormal duration of medication with the historical recovery time distribution parameters of the corresponding drugs and physiological indicators of the elderly, calculate the deviation score of the abnormal duration of medication, and generate the first type of warning signal and the second type of warning signal.
[0017] Continuous collection of multi-dimensional health monitoring data from elderly people living at home, including at least vital sign data, daily activity data, and medication records, is carried out as follows: Vital signs data are automatically collected through health terminal devices. After each device completes the measurement, it automatically sends the data via Bluetooth broadcast, and receives and transmits it back to the data management center. Daily activity data is collected by environmental sensors and wearable devices deployed in the residence. The daily activity data includes at least the time spent in bed, the frequency of getting out of bed at night, and the speed of movement. The environmental sensors continuously record the activity trajectory by sensing the presence of the human body, and the wearable devices collect posture changes through built-in accelerometers and gyroscopes. The terminal application collects medication record data, including drug name, medication time, medication dosage and administration route. The terminal application supports automatic entry of medication information by taking pictures of drug packaging and prescriptions, and generates medication timeline records. The collected vital signs data, daily activity data, and medication records are synchronized and aligned according to timestamps, and then merged to generate multi-dimensional health monitoring time series data based on the individual elderly person.
[0018] Drug effect curves were constructed, and baselines of typical response patterns for various vital signs were established for each common drug after administration. These baselines included the historical recovery time distribution parameters of the elderly patient's physiological indicators from abnormal states to normal ranges after taking the specific drug. The specific process is as follows: Obtain the historical medication records of elderly people living at home and their corresponding vital signs time-series monitoring data. Vital signs include at least heart rate, blood oxygen saturation and blood pressure. Using a single medication event as the anchor point, extract the time-series change trajectory of each vital sign indicator within a preset time window after medication to form the effect curve sample corresponding to this medication event. The effect curve sample contains the numerical sequence of each indicator changing over time from the time of medication. Using the time of medication administration as the zero point, each sample is aligned and normalized on the time axis. Using the steady-state mean of the index during the preset period before medication administration as the benchmark, the measured values at each time after medication administration are converted into deviations from the baseline to generate a normalized effect curve. For multiple normalized response curves of the same drug and the same indicator, fit a typical response curve of the population. The typical response curve includes parameters such as onset time, time to peak, recovery to baseline time, and duration of effect. The mean and standard deviation of the historical recovery time of the physiological indicator from an abnormal state to the normal range after taking the drug are calculated, and stored in association with the typical response curve of the population to form a drug effect curve library.
[0019] Real-time anomaly detection is performed on the collected physiological monitoring data. When a physiological indicator deviates from the normal fluctuation range in each dynamic baseline database, the onset time of the anomaly is marked. The abnormal state of the physiological indicator is continuously tracked until it returns to the normal range, and the natural duration of this abnormal event is recorded. The specific process is as follows: The system receives physiological monitoring data in real time from multimodal sensors deployed in the home environment at a preset sampling frequency. The physiological monitoring data includes at least heart rate, blood oxygen saturation, systolic blood pressure, diastolic blood pressure, and nighttime bed-leaning frequency. Based on the current time, the dynamic baseline interval of the corresponding physiological indicator for the elderly is retrieved from the individual dynamic baseline database. The dynamic baseline interval includes the upper limit and lower limit of the normal fluctuation range of the physiological indicator within the current time period. The dynamic baseline database stores the historical statistical distribution parameters of each physiological indicator at different time periods. When the current real-time physiological indicator measurement value is greater than the lower limit of the normal fluctuation range, it is determined that the current physiological indicator has deviated, the current timestamp is marked as the abnormal start time, and the physiological indicator is marked as an abnormal state; In an abnormal state, subsequent measurements of the physiological indicator are continuously received, and each subsequent measurement is compared with the dynamic baseline interval. If the measurement is still outside the dynamic baseline interval, the abnormal state is maintained and the current duration of the abnormality is recorded. When the measured values are detected to have returned to the dynamic baseline range for a preset number of consecutive times, it is determined that the physiological indicators have returned to the normal range, the current timestamp is marked as the abnormal recovery time, the natural abnormal duration of this abnormal event is recorded, and the abnormal event and the natural abnormal duration are stored in the elderly person's historical abnormal event database.
[0020] Using the medication timestamp in the medication record as the anchor point, the time interval from the time of medication administration to the return of physiological indicators to the normal range is calculated as the post-medication abnormality duration of this abnormal event. The specific process is as follows: Obtain recorded abnormal event information, which includes the abnormal start time, abnormal recovery time, natural abnormal duration, and the type of physiological indicator that caused the abnormality; Using the moment of the anomaly as the search center point, a preset time window is extended forward and backward to generate a time interval for searching medication behavior. Within the medication behavior retrieval time interval, retrieve the elderly person's medication behavior records. The medication behavior records include at least a medication timestamp, drug name, drug dosage, and route of administration. Determine whether there are any valid medication records whose medication timestamp falls within the retrieval time interval. When there are one or more valid medication records, the medication record whose medication timestamp is closest to and greater than or equal to the abnormal start time is selected as the associated medication record, and the medication timestamp of the associated medication record is extracted. Using the medication timestamp as the anchor point, the time interval from the time of medication administration to the time of abnormal recovery is calculated as the post-medication abnormality duration of this abnormal event; When there are no valid medication records within the search time interval, the abnormal event is marked as a medication-free associated abnormal event, and the medication-free associated abnormal event and its natural abnormal duration are stored in the historical abnormal event database.
[0021] The deviation score of the abnormal duration after medication was calculated by statistically comparing it with the historical recovery time distribution parameters of the corresponding drugs and physiological indicators in the elderly. The specific process is as follows: From the drug response baseline database, retrieve historical recovery time distribution parameters that match the drug name and the type of current abnormal physiological indicator in the currently associated medication record; The historical recovery time distribution parameters include the historical mean and standard deviation of the historical recovery time for the physiological indicator. The historical mean recovery time is the statistical average of the time required for the physiological indicator to return to the normal range after the elderly person took the drug in the past. The historical standard deviation of the recovery time is the statistical standard deviation of the corresponding time. The deviation score of the abnormal duration after medication is calculated based on the mean of historical recovery time and the standard deviation of historical recovery time.
[0022] The specific process for generating the first and second type of early warning signals is as follows: When the deviation score of abnormal duration after medication is between the first threshold and the second threshold, a level 3 warning is output. The level 3 warning is used to indicate a slight deviation in drug response and suggests that close observation is recommended. When the deviation score is greater than the second threshold, a level 2 warning is output. The level 2 warning is used to indicate abnormal drug response and suggests contacting medical staff to evaluate the medication plan. When the increasing trend of consecutive abnormal events of the same physiological indicator reaches a preset number of times within a preset observation window, a Level 1 warning is output. The Level 1 warning is used to indicate that there is a risk of systemic deterioration of health status and recommends immediate medical evaluation.
[0023] Bluetooth health terminal devices include at least a Bluetooth blood pressure monitor, a Bluetooth electrocardiogram (ECG) monitor, a Bluetooth pulse oximeter, and a Bluetooth thermometer. Through microcirculation detection devices, they analyze the user's microcirculation status to help assess cardiovascular health. The pulse oximeter integrates a pulse oximeter sensor to detect blood oxygen saturation and ensure timely prevention of hypoxia risks. It also integrates a heart rate sensor to detect the user's heart rate.
[0024] Example 2: like Figure 2 As shown, a health indicator abnormality trend early warning system for elderly people living at home includes a data acquisition module for continuously collecting multi-dimensional health monitoring data of elderly people living at home. The multi-dimensional health monitoring data includes at least vital sign data, daily activity data and medication record data. The drug baseline construction module is used to establish a typical response pattern baseline for each vital sign index after taking each common drug. The typical response pattern baseline includes the historical recovery time distribution parameters of the physiological indicators of the elderly person after taking a specific drug, from an abnormal state to the normal range. The anomaly monitoring module is used to detect anomalies in the collected physiological monitoring data in real time. When a physiological indicator deviates from the normal fluctuation range in the individual dynamic baseline database, the module marks the start time of the anomaly and continuously tracks the abnormal state of the physiological indicator until it returns to the normal range. The module also records the natural duration of the anomaly event. The anomaly analysis module is used to obtain medication behavior data within a preset time window before and after an abnormal event. When there is a corresponding medication record after an abnormal event, the time interval from the time of medication to the recovery of physiological indicators to the normal range is calculated using the medication timestamp of the medication record as the anchor point. This time interval is used as the post-medication abnormality duration of the abnormal event. The warning signal generation module is used to statistically compare the abnormal duration after medication with the historical recovery time distribution parameters of the corresponding drugs and physiological indicators of the elderly, calculate the deviation score of the abnormal duration after medication, and generate the first type of warning signal and the second type of warning signal based on the deviation score.
[0025] The size of the interval and threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value.
[0026] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation. In the two embodiments provided in this application, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the method embodiments described above are merely illustrative. For example, the division of modules is merely a logical functional division, and there may be other division methods in actual implementation. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection between the shown or discussed components may be indirect coupling or communication connection through some interfaces, devices or modules, and may be electrical, mechanical or other forms. The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.
Claims
1. A method for early warning of abnormal trends in health indicators for elderly people living at home, characterized in that, Includes the following steps: Step 1: Continuously collect multi-dimensional health monitoring data of elderly people living at home. The multi-dimensional health monitoring data should include at least vital signs data, daily activity data, and medication record data. Step 2: Construct drug effect curves and establish typical response morphology baselines for each vital sign index after medication for each common drug. The typical response morphology baselines include the historical recovery time distribution parameters of the elderly person's physiological indicators from abnormal state to normal range after taking a specific drug. Step 3: Perform real-time anomaly detection on the collected physiological monitoring data. When the physiological indicators deviate from the normal fluctuation range in each dynamic baseline database, mark the start time of the abnormality of the physiological indicators, continuously track the abnormal state of the physiological indicators until they return to the normal range, and record the natural abnormal duration of this abnormal event. Step 4: Obtain medication behavior data within a preset time window before and after the occurrence of the abnormal event. When there is a corresponding medication record after the occurrence of the abnormal event, use the medication timestamp of the medication record as the anchor point to calculate the time interval from the time of medication administration to the recovery of physiological indicators to the normal range, which is used as the post-medication abnormality duration of the abnormal event. Step 5: Statistically compare the abnormal duration of medication with the historical recovery time distribution parameters of the corresponding drugs and physiological indicators of the elderly, calculate the deviation score of the abnormal duration of medication, and generate the first type of warning signal and the second type of warning signal.
2. The method for early warning of abnormal trends in health indicators for elderly people living at home, as described in claim 1, is characterized in that... Continuous collection of multi-dimensional health monitoring data from elderly people living at home, including at least vital sign data, daily activity data, and medication records, is carried out as follows: Vital signs data are automatically collected through health terminal devices. After each device completes the measurement, it automatically sends the data via Bluetooth broadcast, and receives and transmits it back to the data management center. Daily activity data is collected by environmental sensors and wearable devices deployed in the residence. The daily activity data includes at least the time spent in bed, the frequency of getting out of bed at night, and the speed of movement. The environmental sensors continuously record the activity trajectory by sensing the presence of the human body, and the wearable devices collect posture changes through built-in accelerometers and gyroscopes. The terminal application collects medication record data, including drug name, medication time, medication dosage and administration route. The terminal application supports automatic entry of medication information by taking pictures of drug packaging and prescriptions, and generates medication timeline records. The collected vital signs data, daily activity data, and medication records are synchronized and aligned according to timestamps, and then merged to generate multi-dimensional health monitoring time series data based on the individual elderly person.
3. The method for early warning of abnormal trends in health indicators for elderly people living at home, as described in claim 1, is characterized in that... Drug effect curves were constructed, and baselines of typical response patterns for various vital signs were established for each common drug after administration. These baselines included the historical recovery time distribution parameters of the elderly patient's physiological indicators from abnormal states to normal ranges after taking the specific drug. The specific process is as follows: Acquire the historical medication records of elderly people living at home and their corresponding vital signs time-series monitoring data. Vital signs include at least heart rate, blood oxygen saturation and blood pressure. Using a single medication event as the anchor point, extract the time-series change trajectory of each vital sign indicator within a preset time window after medication to form the effect curve sample corresponding to this medication event. The effect curve sample contains the numerical sequence of each indicator changing over time from the time of medication. Using the time of medication administration as the zero point, each sample is aligned and normalized on the time axis. Using the steady-state mean of the index during the preset period before medication administration as the benchmark, the measured values at each time after medication administration are converted into deviations from the baseline to generate a normalized effect curve. For multiple normalized response curves of the same drug and the same indicator, fit a typical response curve of the population. The typical response curve includes parameters such as onset time, time to peak, recovery to baseline time, and duration of effect. The mean and standard deviation of the historical recovery time of the physiological indicator from an abnormal state to the normal range after taking the drug are calculated, and stored in association with the typical response curve of the population to form a drug effect curve library.
4. The method and system for early warning of abnormal trends in health indicators for elderly people living at home, as described in claim 1, is characterized in that... Real-time anomaly detection is performed on the collected physiological monitoring data. When a physiological indicator deviates from the normal fluctuation range in each dynamic baseline database, the onset time of the anomaly is marked. The abnormal state of the physiological indicator is continuously tracked until it returns to the normal range, and the natural duration of this abnormal event is recorded. The specific process is as follows: The system receives physiological monitoring data in real time from multimodal sensors deployed in the home environment at a preset sampling frequency. The physiological monitoring data includes at least heart rate, blood oxygen saturation, systolic blood pressure, diastolic blood pressure, and nighttime bed-leaning frequency. Based on the current time, the dynamic baseline interval of the corresponding physiological indicator for the elderly is retrieved from the individual dynamic baseline database. The dynamic baseline interval includes the upper limit and lower limit of the normal fluctuation range of the physiological indicator within the current time period. The dynamic baseline database stores the historical statistical distribution parameters of each physiological indicator at different time periods. When the current real-time physiological indicator measurement value is greater than the lower limit of the normal fluctuation range, it is determined that the current physiological indicator has deviated, the current timestamp is marked as the abnormal start time, and the physiological indicator is marked as an abnormal state; In an abnormal state, subsequent measurements of the physiological indicator are continuously received, and each subsequent measurement is compared with the dynamic baseline interval. If the measurement is still outside the dynamic baseline interval, the abnormal state is maintained and the current duration of the abnormality is recorded. When the measured values are detected to have returned to the dynamic baseline range for a preset number of consecutive times, it is determined that the physiological indicators have returned to the normal range, the current timestamp is marked as the abnormal recovery time, the natural abnormal duration of this abnormal event is recorded, and the abnormal event and the natural abnormal duration are stored in the elderly person's historical abnormal event database.
5. The method and system for early warning of abnormal trends in health indicators for elderly people living at home, as described in claim 1, is characterized in that... Using the medication timestamp in the medication record as the anchor point, the time interval from the time of medication administration to the return of physiological indicators to the normal range is calculated as the post-medication abnormality duration of this abnormal event. The specific process is as follows: Obtain recorded abnormal event information, which includes the abnormal start time, abnormal recovery time, natural abnormal duration, and the type of physiological indicator that caused the abnormality; Using the moment of the anomaly as the search center point, a preset time window is extended forward and backward to generate a time interval for searching medication behavior. Within the medication behavior retrieval time interval, retrieve the elderly person's medication behavior records. The medication behavior records include at least a medication timestamp, drug name, drug dosage, and route of administration. Determine whether there are any valid medication records whose medication timestamp falls within the retrieval time interval. When there are one or more valid medication records, the medication record whose medication timestamp is closest to and greater than or equal to the abnormal start time is selected as the associated medication record, and the medication timestamp of the associated medication record is extracted. Using the medication timestamp as the anchor point, the time interval from the time of medication administration to the time of abnormal recovery is calculated as the post-medication abnormality duration of this abnormal event; When no valid medication records are found within the search time interval, the abnormal event is marked as a medication-free associated abnormal event, and the medication-free associated abnormal event and its natural abnormal duration are stored in the historical abnormal event database.
6. The method for early warning of abnormal trends in health indicators for elderly people living at home, as described in claim 1, is characterized in that... The deviation score of the abnormal duration after medication was calculated by statistically comparing it with the historical recovery time distribution parameters of the corresponding drugs and physiological indicators in the elderly. The specific process is as follows: From the drug response baseline database, retrieve historical recovery time distribution parameters that match the drug name and the type of current abnormal physiological indicator in the currently associated medication record; The historical recovery time distribution parameters include the historical mean and standard deviation of the historical recovery time for the physiological indicator. The historical mean recovery time is the statistical average of the time required for the physiological indicator to return to the normal range after the elderly person took the drug in the past. The historical standard deviation of the recovery time is the statistical standard deviation of the corresponding time. The deviation score of the abnormal duration after medication is calculated based on the mean of historical recovery time and the standard deviation of historical recovery time.
7. The method for early warning of abnormal trends in health indicators for elderly people living at home according to claim 1, characterized in that, The specific process for generating the first and second type of early warning signals is as follows: When the deviation score of abnormal duration after medication is between the first threshold and the second threshold, a level 3 warning is output. The level 3 warning is used to indicate a slight deviation in drug response and suggests that close observation is recommended. When the deviation score is greater than the second threshold, a level 2 warning is output. The level 2 warning is used to indicate abnormal drug response and suggests contacting medical staff to evaluate the medication plan. When the increasing trend of consecutive abnormal events of the same physiological indicator reaches a preset number of times within a preset observation window, a Level 1 warning is output. The Level 1 warning is used to indicate that there is a risk of systemic deterioration of health status and recommends immediate medical evaluation.
8. A health indicator abnormality trend early warning system for elderly people living at home, used in accordance with the health indicator abnormality trend early warning method for elderly people living at home as described in any one of claims 1-7, characterized in that, It includes a data acquisition module for continuously collecting multi-dimensional health monitoring data of elderly people living at home. The multi-dimensional health monitoring data includes at least vital signs data, daily activity data, and medication record data. The drug baseline construction module is used to establish a typical response pattern baseline for each vital sign index after taking each common drug. The typical response pattern baseline includes the historical recovery time distribution parameters of the elderly person's physiological indicators from abnormal state to normal range after taking a specific drug. The anomaly monitoring module is used to detect anomalies in the collected physiological monitoring data in real time. When a physiological indicator deviates from the normal fluctuation range in the individual dynamic baseline database, the module marks the start time of the anomaly and continuously tracks the abnormal state of the physiological indicator until it returns to the normal range. The module also records the natural duration of the anomaly event. The anomaly analysis module is used to obtain medication behavior data within a preset time window before and after an abnormal event. When there is a corresponding medication record after an abnormal event, the time interval from the time of medication to the recovery of physiological indicators to the normal range is calculated using the medication timestamp of the medication record as the anchor point. This time interval is used as the post-medication abnormality duration of the abnormal event. The warning signal generation module is used to statistically compare the abnormal duration after medication with the historical recovery time distribution parameters of the corresponding drugs and physiological indicators of the elderly, calculate the deviation score of the abnormal duration after medication, and generate the first type of warning signal and the second type of warning signal based on the deviation score.