Remote management system for home care of chronic patient

Abstract

The invention relates to the technical field of medical information management, in particular to a chronic disease patient home nursing remote management system which comprises a volatility analysis module, a symbol recognition module, a correlation analysis module, a task decoupling module and a nursing task generation module. According to the invention, by analyzing the fluctuation amplitude of the continuously collected physical sign data in the specific time window, the short-term sudden change trend which cannot be revealed by single-time manually reported data can be explored, and the data change direction is further symbolized to identify a continuous change mode with clinical significance; and the time overlapping degree of the continuous sudden change time periods of different physical sign data can be calculated, so that the synchronous association reaction among various physical sign indexes is accurately identified, the potential composite physiological risk is revealed, and the dependency decoupling and execution sequence reconstruction of the high-risk task are realized based on the quantitative evaluation of the task emergency degree and the delay degree.

Description

Technical Field

[0001] This invention relates to the field of medical information management technology, and in particular to a remote management system for home care of patients with chronic diseases. Background Technology

[0002] The field of medical information management technology is a comprehensive technical field that intersects information science and medical engineering. Its core aspects include medical data collection, patient health record management, remote information transmission, doctor-patient interaction records, and health monitoring information processing.

[0003] Among them, the remote management system for home care of patients with chronic diseases refers to the system that records physiological data through information terminals in the home environment for patients with hypertension, diabetes, coronary heart disease and other conditions that require long-term monitoring and care, and periodically reports the data to medical institutions manually to achieve information management.

[0004] Current technology relies on patients recording and manually reporting physiological data periodically at home via information terminals. This discontinuous operation mode prevents medical institutions from capturing rapid or brief fluctuations in vital signs between reporting periods. For example, a diabetic patient may experience a brief but dangerous hypoglycemic event between two blood glucose reports, but the system can only receive seemingly stable report data, thus missing critical opportunities for early warning and intervention. Furthermore, this technology manages different vital signs in isolation, lacking the ability to analyze the potential correlations between multiple physiological indicators. Even if a patient reports abnormalities in both blood glucose and blood pressure, the system cannot automatically identify their temporal synchronicity and the specific complication risks they may indicate. This results in nursing recommendations being based on only a single indicator, lacking comprehensiveness and foresight. Ultimately, all nursing instructions are based on fixed schedules and manual judgment, unable to dynamically adjust task priorities according to real-time monitored complex risks, leading to significant delays in responding to emergencies. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a remote management system for home care of patients with chronic diseases.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A remote management system for home care of patients with chronic diseases includes:

[0007] The volatility analysis module collects continuous vital sign data from patients, analyzes the volatility of each vital sign data by setting time windows, and filters out periods of sudden changes.

[0008] The symbol recognition module acquires the vital sign data corresponding to the mutation time period, performs symbol conversion on the change direction of each vital sign data, compares it with the set continuous fluctuation range, and filters the continuous mutation time period.

[0009] The correlation analysis module counts the duration of continuous mutations for each vital sign data, calculates the degree of temporal overlap of the duration of continuous mutations for each vital sign data, identifies the mutation correlation time between vital sign data, and outputs the mutation correlation analysis results.

[0010] The task decoupling module determines the execution time of each nursing task, and based on the mutation association analysis results, determines whether there are any nursing tasks that need to be executed immediately or have their execution order adjusted when a mutation event occurs, and outputs the decoupling task dependencies.

[0011] The nursing task generation module re-evaluates the execution time of nursing tasks based on the decoupled task dependencies, arranges the execution order of nursing tasks, and obtains nursing task management results.

[0012] As a further aspect of the present invention, the mutation time period specifically refers to the start time, end time, and fluctuation range of the vital signs data; the continuous mutation time period specifically refers to the symbol consistency, fluctuation range, and duration; the mutation association analysis results include the overlap between vital signs, the association direction, and the time synchronization; the decoupling task dependency relationship includes the dependency difference between tasks, the task execution order, and the task execution time period; and the nursing task management results include the task execution order, the task adjustment time period, and the task execution priority.

[0013] As a further aspect of the present invention, the volatility analysis module includes:

[0014] The continuous data acquisition submodule collects continuous vital sign data of patients, including blood glucose, blood pressure, and body temperature. Within a preset time window, it records and integrates each vital sign data and its corresponding acquisition time, and arranges them in chronological order to generate a vital sign data sequence.

[0015] The data fluctuation calculation submodule calculates the average value of each of the aforementioned vital sign data sequences within a time window, obtains the absolute value of the difference between each vital sign data and its corresponding average value within the time window, and performs cumulative summation to obtain the magnitude of the vital sign data change.

[0016] The mutation period filtering submodule compares the magnitude of the change in the vital signs data in each time window with a preset fluctuation threshold. When the magnitude of the change is greater than the fluctuation threshold, it is determined that there is a mutation trend in the data within the corresponding time window. All time windows with mutation trends are filtered and aggregated to establish a mutation period.

[0017] As a further aspect of the present invention, the symbol recognition module includes:

[0018] The change direction symbolization submodule calls the vital sign data within the abrupt change time period, compares the values ​​of adjacent vital sign data, and represents the comparison results with symbols to generate a change direction symbol sequence.

[0019] The continuous trend determination submodule checks the length of consecutive identical symbols in the symbol sequence of the changing direction and compares it with the set continuous fluctuation range. When the length of consecutive identical symbols meets the continuous fluctuation range, it is determined to be a continuous fluctuation trend and the symbol consistency data segment is obtained.

[0020] The mutation period filtering submodule identifies and extracts the start and end time points of each symbolic consistency data segment in the original vital sign data, pairs all identified start and end time points, and establishes a continuous mutation period.

[0021] As a further aspect of the present invention, the process of representing the comparison result with a symbol is as follows: when comparing the values ​​of adjacent vital signs, if the value of the later data point is greater than the value of the earlier data point, the comparison result is represented as symbol +1; if the value of the later data point is less than the value of the earlier data point, the comparison result is represented as symbol -1.

[0022] As a further aspect of the present invention, the process of comparing with the set continuous fluctuation range specifically involves: determining whether the length of consecutive occurrences of the same symbol in the change direction symbol sequence is greater than or equal to a preset minimum continuous count, and whether the length of consecutive occurrences of the same symbol is less than or equal to a preset maximum continuous count.

[0023] As a further aspect of the present invention, the correlation analysis module includes:

[0024] The mutation period statistics submodule classifies and summarizes the continuous mutation period of each vital sign data, records the vital sign data category to which each period belongs, and establishes a set of vital sign mutation period.

[0025] The time overlap calculation submodule, based on the set of abrupt change periods of vital signs, pairs the abrupt change periods corresponding to each vital sign data item, calculates the intersection duration of the paired time periods on the time axis, and obtains the ratio of the intersection duration to the total duration of the union of the two time periods to obtain the degree of time overlap.

[0026] The correlation identification submodule compares the degree of time overlap with a preset overlap threshold, filters out vital sign data pairs whose overlap exceeds the threshold, and determines whether the change direction signs of the vital sign data pairs are consistent within the overlap time based on the change direction sign sequence. The overlap time corresponding to the data pairs with consistent signs is identified as the mutation correlation time, and mutation correlation analysis results are established.

[0027] As a further aspect of the present invention, the task decoupling module includes:

[0028] The task time period matching submodule obtains the preset execution time of each nursing task, matches the execution time range of the nursing task with the mutation association time in the mutation association analysis results, filters out nursing tasks whose execution time and mutation association time have an intersection, and generates an overlapping nursing task set.

[0029] The dependency difference assessment submodule obtains the preset execution order dependency relationship of each nursing task in the overlapping nursing task set, and analyzes the degree of conflict between the original execution order and the immediate execution needs caused by the mutation association time based on the overlap between the nursing task execution time and the mutation association time, and obtains the task dependency difference degree.

[0030] The task order reconstruction submodule compares the task dependency difference of each nursing task with a preset dependency difference threshold. When the task dependency difference of a nursing task exceeds the dependency difference threshold, it determines that the nursing task needs to be executed immediately or the execution order needs to be adjusted, and removes the original order dependency between the nursing task and other tasks, establishing a decoupled task dependency relationship.

[0031] As a further aspect of the present invention, the nursing task generation module includes:

[0032] The execution time period reassessment submodule, for nursing tasks whose sequential dependencies have been removed in the decoupled task dependency relationship, and in combination with the mutation association analysis results, reassesses the execution time period of each nursing task, matches the nursing tasks with unoccupied time windows, and filters the set of optional execution time periods;

[0033] The task order reordering submodule calls the optional execution time set and the decoupled task dependency relationship, inserts the nursing tasks that need to be adjusted into the unaffected nursing task time schedule, and sorts them according to the remaining inter-task dependencies to establish an optimized task execution sequence;

[0034] The nursing plan generation submodule specifies the execution time and order for each nursing task in the optimized task execution sequence, and establishes nursing task management results.

[0035] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0036] In this invention, by analyzing the fluctuation range of continuously collected vital sign data within a specific time window, short-term mutation trends that cannot be revealed by a single manually reported data can be discovered. Furthermore, the direction of data change is symbolized to identify clinically significant persistent change patterns, and the degree of temporal overlap of persistent mutation periods of different vital sign data can be calculated. This allows for the accurate identification of synchronous correlation responses between multiple vital sign indicators, revealing potential complex physiological risks. Based on a quantitative assessment of the urgency and delay of tasks, the invention dynamically judges the execution conflict between preset nursing tasks and immediate physiological events, achieving decoupling of dependencies and reconstruction of execution order for high-risk tasks. Ultimately, it generates an optimized nursing plan that is automatically adjusted according to real-time risks, significantly improving the early warning and rapid response capabilities for acute risk events during home isolation for patients with chronic diseases. Attached Figure Description

[0037] Figure 1 This is a system flowchart of the present invention;

[0038] Figure 2 This is a flowchart of the volatility analysis module of the present invention;

[0039] Figure 3 This is a flowchart of the symbol recognition module of the present invention;

[0040] Figure 4 This is a flowchart of the correlation analysis module of the present invention;

[0041] Figure 5 This is a flowchart of the task decoupling module of the present invention;

[0042] Figure 6 This is a flowchart of the nursing task generation module of the present invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0044] Please see Figure 1 A remote management system for home care of patients with chronic diseases includes:

[0045] The volatility analysis module collects continuous vital sign data from patients, analyzes the volatility of each vital sign data by setting time windows, and filters out periods of sudden changes.

[0046] The symbol recognition module acquires vital sign data corresponding to the mutation time period, performs symbol conversion on the change direction of each vital sign data, compares it with the set continuous fluctuation range, and filters the continuous mutation time period.

[0047] The correlation analysis module counts the duration of continuous mutations for each vital sign data, calculates the degree of temporal overlap of the duration of continuous mutations for each vital sign data, identifies the mutation correlation time between vital sign data, and outputs the mutation correlation analysis results.

[0048] The task decoupling module determines the execution time of each nursing task, and based on the mutation association analysis results, determines whether there are any nursing tasks that need to be executed immediately or have their execution order adjusted when a mutation event occurs, and outputs the decoupling task dependencies.

[0049] The nursing task generation module re-evaluates the execution time of nursing tasks based on the decoupled task dependencies, arranges the execution order of nursing tasks, and obtains nursing task management results.

[0050] The mutation time period specifically refers to the start time, end time, and fluctuation range of the vital signs data. The continuous mutation time period specifically refers to the sign consistency, fluctuation range, and duration. The mutation association analysis results include the overlap between vital signs, the direction of association, and the time synchronization. The decoupling of task dependencies includes the degree of dependency difference between tasks, the order of task execution, and the time period of task execution. The nursing task management results include the order of task execution, the time period of task adjustment, and the priority of task execution.

[0051] Please see Figure 2 The volatility analysis module includes:

[0052] The continuous data acquisition submodule collects continuous vital sign data of patients, including blood glucose, blood pressure, and body temperature. Within a preset time window, it records and integrates each vital sign data and its corresponding acquisition time, and arranges them in chronological order to generate a vital sign data sequence.

[0053] A preset time window is set to 30 minutes. Continuous vital sign data monitoring is performed on a hospitalized patient, including blood glucose, blood pressure, and body temperature. Within a specific 30-minute time window, blood glucose data is collected every 5 minutes using a blood glucose monitoring device, blood pressure data every 10 minutes using an electronic blood pressure monitor, and body temperature data every 15 minutes using an electronic thermometer. Within the first 30-minute time window, the blood glucose monitoring device collects blood glucose values ​​at minutes 0, 5, 10, 15, 20, 25, and 30, forming a set of 7 blood glucose data points. The blood pressure monitoring device collects systolic and diastolic blood pressure values ​​at minutes 0, 10, 20, and 30, forming a set of 4 blood pressure data points. The body temperature monitoring device collects body temperature values ​​at minutes 0, 15, and 30, forming a set of 3 body temperature data points. Each collected vital sign data point is recorded and integrated with its corresponding collection time. For example, blood glucose data points are recorded as (collection time, blood glucose value). Subsequently, the data points for each vital sign item are arranged according to the chronological order of collection time to generate blood glucose vital sign data sequences, blood pressure vital sign data sequences, and body temperature vital sign data sequences.

[0054] The data fluctuation calculation submodule calculates the average value of each vital sign data sequence within the time window, obtains the absolute value of the difference between each vital sign data and its corresponding average value within the time window, and performs cumulative summation to obtain the magnitude of the vital sign data change.

[0055] First, the blood glucose monitoring data sequence is processed. The average blood glucose level within a 30-minute time window is calculated. This is done by summing the values ​​of the seven blood glucose data points in the sequence and then dividing the sum by 7. After obtaining the average blood glucose level, the difference between the value of each blood glucose data point within the time window and the calculated average blood glucose level is calculated, and the absolute value of the difference is taken. For example, the average blood glucose level is subtracted from the value of the first blood glucose data point; if the difference is negative, the opposite number is taken. This process of calculating the absolute value of the difference is repeated for all seven blood glucose data points, resulting in seven absolute difference values. Finally, the seven calculated absolute difference values ​​are summed, and the sum represents the amplitude of the blood glucose monitoring data. Similarly, the same calculation process is performed on the blood pressure and body temperature data sequences, that is, the average value of each data point in the sequence is calculated, the absolute value of the difference between each data point and the corresponding average value is calculated, and finally the sum is accumulated to obtain the amplitude values ​​of blood pressure and body temperature data changes, respectively.

[0056] The mutation period filtering submodule compares the magnitude of the change in vital signs data in each time window with a preset fluctuation threshold. When the magnitude of the change in data is greater than the fluctuation threshold, it is determined that there is a mutation trend in the data within the corresponding time window. All time windows with mutation trends are filtered and aggregated to establish a mutation period.

[0057] The fluctuation threshold was set through statistical analysis of vital sign data from at least 1000 patients under different health conditions. The analysis included collecting continuous vital sign data during stable and acute exacerbation phases, calculating the amplitude of data changes within each time window, and using a 95% confidence interval as the boundary. The upper limit of the amplitude of data changes during the stable phase was used as a reference for the fluctuation threshold. For example, analysis of historical blood glucose data revealed that the amplitude of blood glucose data changes within 30 minutes was typically below a certain value in a stable state, while the amplitude of changes significantly increased during hypoglycemia or hyperglycemia events. Based on this historical blood glucose data analysis, a specific value was set for the blood glucose fluctuation threshold. The amplitude of blood glucose data changes calculated in the previous steps was directly compared with the preset blood glucose fluctuation threshold. When the amplitude of blood glucose data changes exceeded the blood glucose fluctuation threshold, it was determined that there was a sudden change in blood glucose data within the current 30-minute time window. The same comparison and judgment were performed on the amplitude of changes in blood pressure and body temperature. Finally, all time windows identified as having a tendency to change were selected and aggregated to establish a time period of change.

[0058] Please see Figure 3 The symbol recognition module includes:

[0059] The symbolization of change direction submodule calls the vital sign data within the abrupt change time period, compares the values ​​of adjacent vital sign data, and represents the comparison results with symbols to generate a symbol sequence of change direction;

[0060] The process of representing the comparison result with symbols is as follows: when comparing the values ​​of adjacent vital signs, if the value of the later data point is greater than the value of the earlier data point, the comparison result is represented by the symbol +1; if the value of the later data point is less than the value of the earlier data point, the comparison result is represented by the symbol -1.

[0061] First, the vital sign data contained within the mutation period are retrieved. Taking a selected blood glucose mutation period as an example, this period contains a sequence of five consecutively collected blood glucose data points. The numerical values ​​of adjacent vital sign data points in the sequence are compared sequentially. First, the blood glucose value of the second data point is compared with that of the first. If the value of the second data point is greater than that of the first, the comparison result is represented by the symbol "+1". Next, the blood glucose value of the third data point is compared with that of the second. If the value of the third data point is less than that of the second, the comparison result is represented by the symbol "-1". If the values ​​of two adjacent data points are equal, no symbol is generated. This same comparison operation is performed on all subsequent adjacent data points until the end of the sequence. After all comparisons are completed, the generated symbols are arranged in the order of comparison, forming a sequence of change direction symbols. For example, if the values ​​of a blood glucose data sequence are successively increasing, then decreasing, and then decreasing again, the final generated sequence of change direction symbols would be "+1, +1, -1, -1".

[0062] The continuous trend determination submodule checks the length of consecutive identical symbols in the symbol sequence of changing direction and compares it with the set continuous fluctuation range. When the length of consecutive identical symbols meets the continuous fluctuation range, it is determined to be a continuous fluctuation trend and the symbol consistency data segment is obtained.

[0063] The process of comparing with the set continuous fluctuation range is as follows: determine whether the length of the consecutive occurrence of the same symbol in the symbol sequence of the change direction is greater than or equal to the preset minimum continuous count, and whether the length of the consecutive occurrence of the same symbol is less than or equal to the preset maximum continuous count.

[0064] The range of sustained fluctuations is defined by two parameters: the minimum sustained count and the maximum sustained count. The minimum sustained count is set based on the clinical definition of the trend of vital signs. For example, three or more consecutive data points changing in the same direction are considered to constitute a valid trend; therefore, the minimum sustained count is set to 3. The maximum sustained count is set to exclude long-term, slow physiological changes. The maximum sustained count value is based on the analysis of data change patterns before acute events in historical data. For example, most acute blood glucose fluctuations reach their peak within eight consecutive changes; therefore, the maximum sustained count is set to 8. The specific determination process involves calculating the trend significance score for each detected continuous segment using the following formula:

[0065] ;

[0066] in, The trend significance score is a dimensionless composite score used to quantify whether a trend in a physical sign has clinical significance. This represents the length of consecutive identical symbols detected, i.e., the number of times consecutive data points change in the same direction. In the "-1, -1, -1" sequence generated by the "Change Direction Symbolization Submodule", The value is 3. This is the preset maximum duration count, the value of which has been explained above in this paragraph. It is set based on historical data analysis and is 8 in this scenario. This parameter is used to normalize the actual duration of the trend. Representing length weight, it is a dimensionless value used to adjust the importance of the "persistence" of a trend in the total score. This represents the average absolute value of the change between adjacent data points during the trend. It is calculated by summing the magnitude of each change within the trend (e.g., blood sugar dropping from 6.0 to 5.5, with an absolute change of 0.5) and then taking the average. It reflects the drasticness or "steepness" of the trend change. It is a normalization constant with the same dimensions as the original vital sign data (e.g., mmol / L). It represents a unit of change that is clinically considered typical and significant for that vital sign. This value is set based on clinical guidelines and expert experience; for example, for blood glucose, a rapid change of 1 mmol / L may be considered a meaningful benchmark. The magnitude weight is a dimensionless value, and is related to... These factors complement each other, and are used to adjust the "severity" of a trend in the overall score. and The setting is based on the monitoring objective: if the objective is to identify and issue early warnings for acute events that change rapidly but are not lasting, then... The weight of certain trends will be set higher; conversely, if more attention is paid to trends that are long-lasting and relatively mild, then... The weight will be higher. Both weights are dimensionless values, and are set... This is to ensure that the composition of the total score is standardized.

[0067] Taking blood glucose data as an example, suppose that during a period of sudden change, the consecutive blood glucose values ​​are 6.0, 5.5, 5.1 and 4.8 (unit: mmol / L), and the corresponding direction of change is indicated by "-1, -1, -1".

[0068] 1. Parameter setting: Trend length (Because there are three consecutive "-1"s). Maximum duration count (Based on the previous settings). To prioritize capturing rapidly changing trends, magnitude weights are set. Therefore, length weight Calculate the average change. The absolute values ​​of the three consecutive decreases were respectively , , .therefore, mmol / L. Define the typical clinical range of change. mmol / L (assuming a clinically significant change in blood glucose is defined as 1.0 mmol / L).

[0069] 2. Substitute into the formula to calculate: The calculated trend significance score T is 0.39. Subsequently, this score is compared with a preset trend score threshold (e.g., 0.3). If it is greater than the threshold, the data segment is determined to be a continuous fluctuation trend, and the original vital sign data corresponding to this continuous fluctuation trend is extracted as the sign consistency data segment.

[0070] The mutation period filtering submodule identifies and extracts the start and end time points of each symbolic consistency data segment in the original vital sign data, pairs all identified start and end time points, and establishes continuous mutation period segments.

[0071] For each identified sign-consistency data segment, a search is performed within the original, timestamped sequence of vital signs data. The search identifies and extracts the acquisition time corresponding to the first data point constituting the sign-consistency segment as the start time point, and the acquisition time corresponding to the last data point as the end time point. For example, in a blood glucose data segment identified as exhibiting a continuous fluctuation trend, the acquisition time of the first data point in the original sequence is 10:15 AM, and the acquisition time of the last data point is 10:25 AM. Therefore, 10:15 AM and 10:25 AM are extracted. Subsequently, the identified start and end time points are paired to form a defined time interval, such as (10:15 AM, 10:25 AM). The same start and end time point identification, extraction, and pairing operation is performed on all sign-consistency data segments filtered through the continuous trend determination. Finally, all paired time intervals are aggregated to form a set; this aggregated set of time intervals represents the continuous fluctuation period.

[0072] Please see Figure 4 The correlation analysis module includes:

[0073] The mutation period statistics submodule classifies and summarizes the continuous mutation periods of each vital sign data, records the category of vital sign data to which each time period belongs, and establishes a set of vital sign mutation periods.

[0074] The set of persistent mutation time periods is processed. This set contains multiple persistent mutation time periods from different vital signs (blood glucose, blood pressure, body temperature). The processing first involves iterating through each persistent mutation time period in the set and identifying and recording the vital sign data category to which each persistent mutation time period belongs. For example, if a time period (10:15 AM, 10:25 AM) is obtained by analyzing blood glucose data, then this time period is labeled as "blood glucose". Another time period (10:20 AM, 10:30 AM) is obtained by analyzing blood pressure data, then this time period is labeled as "blood pressure". All persistent mutation time periods are then categorized and summarized according to their respective vital sign data categories. The final result is a structured set of vital sign mutation time periods, where each entry clearly indicates a specific time interval and the corresponding vital sign item, such as {blood glucose: [(time period one), (time period two)], blood pressure: [(time period three)]}.

[0075] The time overlap calculation submodule, based on the set of abrupt changes in vital signs, pairs the abrupt change time periods corresponding to each vital sign data point, calculates the intersection duration of the paired time periods on the time axis, and obtains the ratio of the intersection duration to the total duration of the union of the two time periods to get the degree of time overlap.

[0076] First, pair the time periods corresponding to the abrupt changes in different vital signs. For example, select a time period A of sustained blood glucose abrupt change and a time period B of sustained blood pressure abrupt change from the set and pair them. Time period A is from 8:00 AM to 8:30 AM, and time period B is from 8:20 AM to 8:50 AM. Calculate the intersection duration of the two paired time periods on the time axis. In this example, the intersection of time periods A and B is from 8:20 AM to 8:30 AM, with an intersection duration of 10 minutes. Next, calculate the union duration of these two time periods. The union is from 8:00 AM to 8:50 AM, with a total duration of 50 minutes. Then, divide the calculated intersection duration by the union duration, i.e., divide 10 minutes by 50 minutes, resulting in a value of 0.2. The calculated ratio represents the degree of time overlap. For all combinations of time periods with persistent mutations from different sources in the set, such as blood glucose and body temperature, blood pressure and body temperature, this process of pairing, calculating the intersection duration, calculating the union total duration, and calculating the ratio is repeated.

[0077] The correlation identification submodule compares the degree of time overlap with a preset overlap threshold, filters out the symptom data pairs whose overlap exceeds the threshold, and determines whether the change direction signs of the symptom data pairs are consistent within the overlap time based on the change direction sign sequence. The overlap time corresponding to the data pairs with consistent signs is identified as the mutation correlation time, and the mutation correlation analysis results are established.

[0078] The overlap threshold is set based on a retrospective analysis of a large number of historical complication case data, statistically analyzing the temporal overlap patterns of different abnormal signs occurring simultaneously. A threshold is selected that can effectively distinguish between accidental overlap and strongly correlated events. For example, the analysis shows that when the temporal overlap of the time periods of two signs' mutations exceeds 0.5, the probability of a physiological correlation between the two signs exceeds 90%, therefore the overlap threshold is set to 0.5. The specific identification process involves comparing the temporal overlap of each pair of sign data calculated in the previous steps with the preset overlap threshold of 0.5. All sign data pairs with a temporal overlap greater than 0.5 are selected. For each selected data pair, such as blood glucose and blood pressure, the direction of change of the two signs is determined based on the symbolic sequence of the change direction within the overlapping time period. The determination method is to check whether the symbolic sequence of the change direction of blood glucose and the symbolic sequence of the change direction of blood pressure show consistency within the overlapping time interval, for example, both showing synchronous increase or synchronous decrease. When the direction of the symbols is determined to be consistent, the overlapping time corresponding to the vital sign data pair is identified as the mutation association time, and the association is recorded to establish the final mutation association analysis result.

[0079] Please see Figure 5 The task decoupling module includes:

[0080] The task time period matching submodule obtains the preset execution time of each nursing task, matches the execution time range of the nursing task with the mutation association time in the mutation association analysis results, filters out nursing tasks whose execution time and mutation association time have an intersection, and generates an overlapping nursing task set.

[0081] First, all pre-set nursing tasks and their planned execution time ranges are retrieved from the nursing information database. For example, one nursing task is "measuring pre-lunch blood glucose," with a pre-set execution time of 11:00 AM to 11:30 AM; another task is "injecting routine insulin," with a pre-set execution time of 11:45 AM to 12:00 PM. The execution time range of each nursing task is then matched against all mutation association times in the mutation association analysis results generated in the previous steps. The matching process involves checking whether there is an intersection between the two time ranges on the time axis. For example, the mutation association analysis results show a mutation association time of 10:50 AM to 11:20 AM, indicating a rapid and synchronized rise in blood glucose and blood pressure. Comparing the execution time of the "measuring pre-lunch blood glucose" task (11:00 AM to 11:30 AM) with the mutation association time of 10:50 AM to 11:20 AM, an intersection is found between 11:00 AM and 11:20 AM. Therefore, the nursing task of "measuring pre-lunch blood glucose" is selected and added to a set called the Overlapping Nursing Task Set. This matching process is performed on all preset nursing tasks.

[0082] The dependency difference assessment submodule obtains the preset execution order dependency relationship of each nursing task in the overlapping nursing task set, and analyzes the degree of conflict between the original execution order and the immediate execution needs caused by the mutation association time based on the overlap between the nursing task execution time and the mutation association time, and obtains the task dependency difference degree.

[0083] First, the pre-defined execution order dependencies of each nursing task in the overlapping nursing task set are obtained. For example, the nursing process stipulates that the task of "measuring pre-lunch blood glucose" must be performed before the task of "injecting routine insulin". Next, based on the overlap between the execution time of the nursing tasks and the time of mutation-related events, the degree of conflict between the original execution order and the immediate execution needs caused by the mutation-related events is analyzed. The degree of conflict between the original execution order and the immediate execution needs is quantified by the task dependency difference, the specific quantification process of which is defined by the following formula:

[0084] ;

[0085] in, The task-dependency score is a dimensionless composite score. The higher the score, the greater the conflict between the original plan for the care task and the immediate needs arising from the current physiological condition, and the more it needs to be reassessed and adjusted. Representing task urgency, this is a dimensionless score. Its value originates from the analysis results of preceding modules, such as the severity of abrupt events discovered by the "correlation identification submodule" (e.g., "time overlap" or "trend significance score"), which is then transformed into a standardized scoring range (e.g., 0 to 10) using a mapping function. In the current scenario, blood sugar and blood pressure are rising rapidly and simultaneously, which is a high-risk signal, therefore... The score will be very high. This represents the degree of task delay and is a dimensionless score. Its value is calculated based on the difference between the planned execution time of the nursing task and the current time, and is standardized to the same scoring range (e.g., 0 to 10 points). The further the planned execution time is from the current time, the longer the delay in intervention will be if the original plan is followed. The higher the score, the better. and These represent the task urgency weight and task delay weight, respectively, used to adjust... and Its importance in the overall score, and The weighting is determined based on the nature of the task and its clinical risk: for intervention tasks directly related to patient safety (such as insulin injection to manage a hyperglycemic crisis), the urgency is far greater than following a routine schedule, therefore... The value will be set very high, close to 1. The interaction coefficient is a dimensionless improved parameter. Its function is to capture and amplify the coupling effect when both task urgency and delay are high. In clinical practice, the risk increase of a highly urgent task delayed for a long time is not linear, but exponential. This is precisely to simulate such a disproportionate increase in risk. The value is derived from regression analysis of a large number of historical high-risk medical error events, and its value is usually between 0.1 and 0.5.

[0086] Taking the routine insulin injection task as an example, let's say the current time is 11:00 AM, while the task was originally scheduled to be performed at 11:45 AM.

[0087] 1. Parameter settings: Task urgency (Within the 0-10 score range, a score of 9.0 indicates a very urgent situation due to the detection of a severe event involving a simultaneous rapid rise in blood sugar and blood pressure.) Task Delay (Within the 0-10 range, there are still 45 minutes until the scheduled execution, which is too long for this emergency, resulting in a score of 7.0). Because this is a rescue mission for acute hyperglycemia, an urgency weight is assigned. Then the task latency weight Set the interaction impact coefficient. This is to reflect the additional risks that arise when high urgency and high delay coexist.

[0088] 2. Substitute into the formula to calculate: The calculated task dependency variance D is 21.2. This score is then compared to a preset dependency variance threshold (e.g., 7). Since 21.2 is significantly greater than 7, this indicates a substantial execution variance in the task, necessitating immediate breaking of existing dependencies and task refactoring.

[0089] The task order restructuring submodule compares the task dependency difference of each nursing task with the preset dependency difference threshold. When the task dependency difference of a nursing task exceeds the dependency difference threshold, it determines that the nursing task needs to be executed immediately or the execution order needs to be adjusted, and removes the original order dependency between the nursing task and other tasks, establishing a decoupled task dependency relationship.

[0090] The dependency difference threshold was set with reference to emergency plans for different levels of medical risk events, using the lower limit of the task dependency difference corresponding to events requiring immediate intervention as the threshold. For example, according to hospital risk management regulations, any delayed operation that may lead to serious consequences must break the routine process if the task dependency difference score exceeds 7; therefore, the dependency difference threshold is set to 7. The specific restructuring process involves comparing the task dependency difference of each nursing task calculated in the previous steps with the dependency difference threshold of 7. When a nursing task, such as "injecting routine insulin," has a calculated task dependency difference of 21.2, exceeding the threshold of 7, it is determined that the "injecting routine insulin" nursing task needs to be executed immediately or significantly earlier. To this end, the original sequential dependency between the "injecting routine insulin" task and other tasks is removed. For example, it is no longer required that the "injecting routine insulin" task must be executed after a non-urgent task. By removing the original sequential constraints that are inapplicable in the current emergency situation, a new, temporary, decoupled task dependency relationship is established, in which the "injecting routine insulin" task becomes a high-priority independent task.

[0091] Please see Figure 6 The nursing task generation module includes:

[0092] The execution time reassessment submodule targets nursing tasks whose sequential dependencies have been removed in decoupled task dependencies, and reassesses the execution time of each nursing task based on the results of mutation association analysis. It matches nursing tasks with unoccupied time windows and filters out a set of available execution time periods.

[0093] For nursing tasks whose sequential dependencies have been removed from decoupled task dependencies, such as the aforementioned "injecting routine insulin" task, the optimal execution time for the "injecting routine insulin" task is reassessed by combining the specific timing information of the synchronous rise in blood glucose and blood pressure from the mutation association analysis results. The assessment process involves matching the nature of the task (injecting insulin to lower blood glucose) with the physiological event (rapid rise in blood glucose) to determine that the task needs to be executed as soon as possible. Subsequently, the real-time task schedules of all nursing staff are reviewed to find unoccupied time windows. The required execution time of the task is matched with the found free time windows to filter out all available time periods with sufficient length and the earliest start time. For example, if a nurse is found to have no task scheduled in the next 15 minutes, a 15-minute free time window is selected. All selected time windows are aggregated to form a set of optional execution time periods for selection in subsequent steps.

[0094] The task order reordering submodule calls the optional execution time set and decouples task dependencies, inserts the nursing tasks that need to be adjusted into the unaffected nursing task schedule, and sorts them according to the remaining inter-task dependencies to establish an optimized task execution sequence.

[0095] The system invokes the set of available execution time slots and decouples task dependencies. For the nursing task requiring adjustment, namely the "inject routine insulin" task, it selects the most suitable time slot from the set of available execution time slots (usually the first available slot) and inserts the "inject routine insulin" task into the unaffected nursing task schedule. For example, it schedules the "inject routine insulin" task to be performed 5 minutes after the current time. Inserting the urgent "inject routine insulin" task disrupts the original task sequence. Other tasks are then reordered based on remaining inter-task dependencies. For example, the "observe patient's postprandial response" task, originally planned to be performed after the "inject routine insulin" task, remains after the "inject routine insulin" task, but its specific execution time will be adjusted according to the new time point after the insulin injection. Other patients' routine tasks that are not directly dependent on the "inject routine insulin" task maintain their original order, but their execution times may be adjusted due to the overall schedule shift. By inserting the urgent task and reordering, an optimized task execution sequence is established.

[0096] The nursing plan generation submodule clarifies the execution time and order of each nursing task in the optimized task execution sequence, and establishes nursing task management results.

[0097] To optimize each nursing task in the task execution sequence, the execution time and order of each task are clearly defined. For example, the first item in the sequence is to "immediately administer insulin injection" to a patient, with the execution time being 5 to 10 minutes from the current time; the second item is to "turn the patient over" for another patient, with the execution time remaining unchanged; the third item is to "retest blood glucose in 30 minutes" for the patient who has received insulin, with the "retest blood glucose in 30 minutes" task generated based on a new situation. This process assigns specific time and sequence attributes to all tasks in the entire sequence, forming a logically clear and time-precise list of action instructions. Finally, the list containing all tasks and their clearly defined execution times and orders is integrated to create the final nursing task management result. The nursing task management result is then distributed to the relevant nursing staff electronically.

[0098] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications 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 protection scope of the present invention.

Claims

1. A remote management system for home care of patients with chronic diseases, characterized in that, The system includes: The volatility analysis module collects continuous vital sign data from patients, analyzes the volatility of each vital sign data by setting time windows, and filters out periods of sudden changes. The symbol recognition module acquires the vital sign data corresponding to the mutation time period, performs symbol conversion on the change direction of each vital sign data, compares it with the set continuous fluctuation range, and filters the continuous mutation time period. The correlation analysis module counts the duration of continuous mutations for each vital sign data, calculates the degree of temporal overlap of the duration of continuous mutations for each vital sign data, identifies the mutation correlation time between vital sign data, and outputs the mutation correlation analysis results. The task decoupling module determines the execution time of each nursing task, and based on the mutation association analysis results, determines whether there are any nursing tasks that need to be executed immediately or have their execution order adjusted when a mutation event occurs, and outputs the decoupling task dependencies. The nursing task generation module re-evaluates the execution time of nursing tasks based on the decoupled task dependencies, arranges the execution order of nursing tasks, and obtains nursing task management results.

2. The remote management system for home care of patients with chronic diseases according to claim 1, characterized in that, The mutation time period specifically refers to the start time, end time, and fluctuation range of the vital signs data. The continuous mutation time period specifically refers to the symbol consistency, fluctuation range, and duration. The mutation association analysis results include the overlap between vital signs, the association direction, and the time synchronization. The decoupling task dependency relationship includes the dependency difference between tasks, the task execution order, and the task execution time period. The nursing task management results include the task execution order, the task adjustment time period, and the task execution priority.

3. The remote management system for home care of patients with chronic diseases according to claim 1, characterized in that, The volatility analysis module includes: The continuous data acquisition submodule collects continuous vital sign data of patients, including blood glucose, blood pressure, and body temperature. Within a preset time window, it records and integrates each vital sign data and its corresponding acquisition time, and arranges them in chronological order to generate a vital sign data sequence. The data fluctuation calculation submodule calculates the average value of each of the aforementioned vital sign data sequences within a time window, obtains the absolute value of the difference between each vital sign data and its corresponding average value within the time window, and performs cumulative summation to obtain the magnitude of the vital sign data change. The mutation period filtering submodule compares the magnitude of the change in the vital signs data in each time window with a preset fluctuation threshold. When the magnitude of the change is greater than the fluctuation threshold, it is determined that there is a mutation trend in the data within the corresponding time window. All time windows with mutation trends are filtered and aggregated to establish a mutation period.

4. The remote management system for home care of patients with chronic diseases according to claim 3, characterized in that, The symbol recognition module includes: The change direction symbolization submodule calls the vital sign data within the abrupt change time period, compares the values ​​of adjacent vital sign data, and represents the comparison results with symbols to generate a change direction symbol sequence. The continuous trend determination submodule checks the length of consecutive identical symbols in the symbol sequence of the changing direction and compares it with the set continuous fluctuation range. When the length of consecutive identical symbols meets the continuous fluctuation range, it is determined to be a continuous fluctuation trend and the symbol consistency data segment is obtained. The mutation period filtering submodule identifies and extracts the start and end time points of each symbolic consistency data segment in the original vital sign data, pairs all identified start and end time points, and establishes a continuous mutation period.

5. The remote management system for home care of patients with chronic diseases according to claim 4, characterized in that, The process of representing the comparison result with symbols is as follows: when comparing the values ​​of adjacent vital signs, if the value of the later data point is greater than the value of the earlier data point, the comparison result is represented by the symbol +1; if the value of the later data point is less than the value of the earlier data point, the comparison result is represented by the symbol -1.

6. The remote management system for home care of patients with chronic diseases according to claim 4, characterized in that, The process of comparing with the set continuous fluctuation range is as follows: determine whether the length of the consecutive occurrence of the same symbol in the change direction symbol sequence is greater than or equal to the preset minimum continuous count, and the length of the consecutive occurrence of the same symbol is less than or equal to the preset maximum continuous count.

7. The remote management system for home care of patients with chronic diseases according to claim 4, characterized in that, The correlation analysis module includes: The mutation period statistics submodule classifies and summarizes the continuous mutation period of each vital sign data, records the vital sign data category to which each period belongs, and establishes a set of vital sign mutation period. The time overlap calculation submodule, based on the set of abrupt change periods of vital signs, pairs the abrupt change periods corresponding to each vital sign data item, calculates the intersection duration of the paired time periods on the time axis, and obtains the ratio of the intersection duration to the total duration of the union of the two time periods to obtain the degree of time overlap. The correlation identification submodule compares the degree of time overlap with a preset overlap threshold, filters out vital sign data pairs whose overlap exceeds the threshold, and determines whether the change direction signs of the vital sign data pairs are consistent within the overlap time based on the change direction sign sequence. The overlap time corresponding to the data pairs with consistent signs is identified as the mutation correlation time, and mutation correlation analysis results are established.

8. The remote management system for home care of patients with chronic diseases according to claim 7, characterized in that, The task decoupling module includes: The task time period matching submodule obtains the preset execution time of each nursing task, matches the execution time range of the nursing task with the mutation association time in the mutation association analysis results, filters out nursing tasks whose execution time and mutation association time have an intersection, and generates an overlapping nursing task set. The dependency difference assessment submodule obtains the preset execution order dependency relationship of each nursing task in the overlapping nursing task set, and analyzes the degree of conflict between the original execution order and the immediate execution needs caused by the mutation association time based on the overlap between the nursing task execution time and the mutation association time, and obtains the task dependency difference degree. The task order reconstruction submodule compares the task dependency difference of each nursing task with a preset dependency difference threshold. When the task dependency difference of a nursing task exceeds the dependency difference threshold, it determines that the nursing task needs to be executed immediately or the execution order needs to be adjusted, and removes the original order dependency between the nursing task and other tasks, establishing a decoupled task dependency relationship.

9. The remote management system for home care of patients with chronic diseases according to claim 8, characterized in that, To analyze the degree of conflict between the original execution order and the immediate execution demand caused by the mutation-related time, the following formula is used: ； Calculate task dependency variance Based on task dependency differences Quantify the degree of conflict between the original execution order and the immediate execution requirements caused by the time associated with the mutation; in, Represents the urgency of the task. Represents task latency. and These represent the task urgency weight and the task delay weight, respectively. This represents the interaction coefficient.

10. The remote management system for home care of patients with chronic diseases according to claim 8, characterized in that, The nursing task generation module includes: The execution time period reassessment submodule, for nursing tasks whose sequential dependencies have been removed in the decoupled task dependency relationship, and in combination with the mutation association analysis results, reassesses the execution time period of each nursing task, matches the nursing tasks with unoccupied time windows, and filters the set of optional execution time periods; The task order reordering submodule calls the optional execution time set and the decoupled task dependency relationship, inserts the nursing tasks that need to be adjusted into the unaffected nursing task time schedule, and sorts them according to the remaining inter-task dependencies to establish an optimized task execution sequence; The nursing plan generation submodule specifies the execution time and order for each nursing task in the optimized task execution sequence, and establishes nursing task management results.

Images