Postoperative walking safety management method and system based on multi-source data fusion
By fusion of multi-source data to assess the fall risk level of postoperative patients, dynamically setting fall monitoring tags, and optimizing monitoring management, the problems of monitoring reliability and nursing staff burden in postoperative fall monitoring have been solved, achieving efficient and reliable fall risk management.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- THE FIRST AFFILIATED HOSPITAL ZHEJIANG UNIV COLLEGE OF MEDICINE
- Filing Date
- 2026-01-28
- Publication Date
- 2026-05-19
AI Technical Summary
In postoperative patient fall risk monitoring, existing technologies often result in false alarms as the number of patients increases, leading to excessive workload for nursing staff and making it difficult to effectively reduce their workload. At the same time, the monitoring reliability is insufficient.
By using a multi-source data fusion approach, the fall risk level of postoperative patients is assessed, the ward distribution of high-risk patients is dynamically determined, fall monitoring labels are rationally set, monitoring management is implemented for different ward areas, the setting method of fall monitoring labels is optimized, and unnecessary label setting is reduced.
While ensuring the reliability of monitoring high-risk patients, the nursing difficulty for nursing staff is reduced, the reliability of fall monitoring is improved, the problem of excessive labor intensity caused by too many labels is avoided, and the effectiveness of monitoring and the efficiency of resource utilization are improved.
Smart Images
Figure CN122067765A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of data processing technology, and in particular relates to a postoperative walking safety management method and system based on multi-source data fusion. Background Technology
[0002] Postoperative patients frequently experience falls, which can lead to secondary risks. Therefore, to identify patients at high fall risk and provide targeted care, a method is needed. For example, the invention patent application CN202111192225.8, "A Comprehensive Assessment and Health Intervention System for Fall Risk in the Elderly," describes a method for classifying health status and diagnosing risk based on physiological conditions. This method categorizes falls into low, medium, and high risk levels, enabling early prediction, prevention, intervention, and timely treatment. However, it has the following technical problems: To monitor and manage falls in patients at risk of falling, it is often necessary to use fall monitoring tags and monitoring devices to identify and process falls. However, with the increase in the number of patients, setting up too many fall monitoring tags can inevitably lead to excessive workload for nursing staff if false alarms occur. Therefore, how to accurately identify patients for fall monitoring tag setting and further integrate with monitoring devices to monitor abnormal falls, thereby improving monitoring reliability and reducing the workload of nursing staff, has become an urgent technical problem to be solved.
[0003] Therefore, there is an urgent need for a postoperative walking safety management method and system based on multi-source data fusion. Summary of the Invention
[0004] To achieve the objectives of this invention, the following technical solution is adopted: Specifically, this application provides a postoperative walking safety management method based on multi-source data fusion, which includes: S1 uses the postoperative fall risk level assessment results to identify high-risk patients. Based on the ward distribution data of the high-risk patients, it determines when to set fall monitoring tags for other patients with fall risk levels. Based on the high-risk patient data, it determines the target for setting fall monitoring tags for other patients with fall risk levels. S2 determines the monitoring and management methods for patients' walking safety after surgery based on the setting goals of fall monitoring tags, and determines the ward areas for monitoring and analyzing walking safety based on the monitoring and processing methods of different ward monitoring areas. S3 determines the setting method for fall monitoring tags for newly added patients with different fall risk levels based on patient data from the ward monitoring area that monitors and analyzes walking safety.
[0005] The beneficial effects of this invention are as follows: Based on data from high-risk patients, the target for setting fall monitoring tags in other fall risk levels is determined. This ensures the reliability of monitoring for high-risk patients while also taking into full account the overlap between other fall risk levels and the wards of high-risk patients. As a result, the reliability of fall monitoring is improved while reducing the difficulty of nursing care for nursing staff.
[0006] Based on patient data from ward monitoring areas used for walking safety monitoring and analysis, the method for setting fall monitoring tags for patients with different fall risk levels was determined. This fully considers the differences in the number of patients in the ward monitoring areas, which leads to differences in the reliability of fall monitoring for patients with different fall risk levels. Therefore, from the perspective of the reliability of fall monitoring for patients with different fall risk levels, the method for setting fall monitoring tags for patients in different ward monitoring areas was determined. This improves the reliability of fall monitoring and analysis while avoiding the technical problem of excessive workload caused by setting too many fall monitoring tags.
[0007] Furthermore, the fall risk levels include high risk, low risk, and medium risk.
[0008] Furthermore, the assessment result of the fall risk level is determined based on the monitoring data of the patient's postoperative physical indicators.
[0009] Furthermore, it was determined that fall monitoring labels could be set up for patients with other fall risk levels, specifically including: Based on the distribution data of the wards of the high-risk patients, the wards containing the high-risk patients are identified; Using data on high-risk patients in wards with existing high-risk patients, we can determine whether fall monitoring tags can be set up for other patients at different fall risk levels.
[0010] Furthermore, the method for determining the setting method of the fall monitoring tags for patients with different newly added fall risk levels is as follows: The ward monitoring area, which is used to monitor and analyze walking safety, is designated as the monitoring area. Based on the patient data in the monitoring area, the number of patients in the monitoring area is determined. Based on the monitoring and management methods for the monitoring areas, determine the number of patients in the monitoring areas under different monitoring and management methods; Based on the number of patients in monitoring areas with different monitoring and management methods, determine how to set fall monitoring tags for newly added patients with different fall risk levels.
[0011] It should be noted that when the patient is classified as a high-risk patient, then it is determined that the patient needs to have a fall monitoring tag set.
[0012] In a second aspect, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the above-described method for postoperative walking safety management based on multi-source data fusion when running the computer program.
[0013] Other features and advantages will be set forth in the following description, and the objects and other advantages of the invention are realized and obtained through the structures particularly pointed out in the description and the drawings.
[0014] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description
[0015] The above and other features and advantages of the present invention will become more apparent from a detailed description of exemplary embodiments thereof with reference to the accompanying drawings.
[0016] Figure 1 This is a flowchart of a postoperative walking safety management method based on multi-source data fusion; Figure 2 This is a flowchart illustrating a method for determining how to set up fall monitoring tags for patients at other fall risk levels; Figure 3 This is a flowchart illustrating the method for determining the target setting of fall monitoring labels in patients with other fall risk levels; Figure 4 A flowchart outlining the methods for determining the monitoring and management of patient walking safety after surgery. Detailed Implementation
[0017] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.
[0018] Example 1 like Figure 1 As shown, this application provides a postoperative walking safety management method based on multi-source data fusion, specifically including: S1 uses the postoperative fall risk level assessment results to identify high-risk patients. Based on the ward distribution data of the high-risk patients, it determines when to set fall monitoring tags for other patients with fall risk levels. Based on the high-risk patient data, it determines the target for setting fall monitoring tags for other patients with fall risk levels. S2 determines the monitoring and management methods for patients' walking safety after surgery based on the setting goals of fall monitoring tags, and determines the ward areas for monitoring and analyzing walking safety based on the monitoring and processing methods of different ward monitoring areas. S3 determines the setting method for fall monitoring tags for newly added patients with different fall risk levels based on patient data from the ward monitoring area that monitors and analyzes walking safety.
[0019] Furthermore, the fall risk levels include high risk, low risk, and medium risk.
[0020] Furthermore, the assessment result of the fall risk level is determined based on the monitoring data of the patient's postoperative physical indicators.
[0021] Specifically, the approach shifts from a single subjective questionnaire to the fusion and analysis of multi-dimensional objective data. The system will continuously collect key physiological and activity data from patients after surgery, dynamically calculate their fall risk value using a weighted scoring model, and automatically classify the risk level.
[0022] High-risk level: This indicates that the patient is extremely unstable, has a high probability of falling, and requires the highest level of monitoring and intervention.
[0023] Medium risk level: This indicates that the patient has significant risk factors and requires targeted preventive measures.
[0024] Low risk level: This indicates that the patient is relatively stable, but basic health education and monitoring are still required.
[0025] II. Data Collection and Indicator Selection; The system will collect data from the following multiple dimensions to ensure the comprehensiveness of the evaluation: 1. Physiological and consciousness indicators; Blood pressure stability: In particular, monitor orthostatic hypotension (a decrease in systolic blood pressure of ≥20 mmHg or a decrease in diastolic blood pressure of ≥10 mmHg when moving from a lying to a sitting / standing position).
[0026] Heart rate and rhythm: Is there tachycardia, bradycardia or arrhythmia (such as atrial fibrillation)?
[0027] Blood oxygen saturation: Whether there is persistent hypoxemia after surgery.
[0028] Pain level: Using a numerical rating scale, moderate to severe pain (≥4 points) significantly affects mobility.
[0029] Consciousness and cognitive status: Use simple methods for assessing confusion, such as 3D-CAM. Postoperative delirium is a significant risk factor for falls.
[0030] 2. Indicators of muscle strength and activity level; Lower limb muscle strength: The clinically common test is the "supine hip raise" test (points are deducted for those who cannot complete the test or cannot maintain it steadily).
[0031] Balance function: "Stand-walk" timed test: the time required to stand up from a chair, walk 3 meters, and return to a seated position.
[0032] Eyes-closed standing test: Can you stand with your eyes closed for 30 seconds?
[0033] Gait analysis: Using wearable sensors or nurses to observe, assess walking speed, stride length, and whether the gait is unsteady or disjointed.
[0034] 3. Drug and treatment factors; High-risk medication use: Whether sedatives, opioid analgesics, antipsychotics, antihypertensives, diuretics, etc. are used.
[0035] Tubing and Obstacles: Whether the person is carrying multiple tubes (such as drainage tubes, infusion pumps) that may affect their mobility.
[0036] III. Risk Assessment Model and Quantification Rules We designed a weighted scoring card that assigns different scores based on the degree of abnormality of the indicators.
[0037] Table 1 Fall Risk Scoring Table
[0038] Risk level classification rules: Low risk: Total score 0-2 points; Medium risk: Total score 3-5 points; High risk: Total score ≥ 6 points. Special veto item: Any of the following situations will be automatically deemed high-risk: The patient had fallen within 24 hours, exhibited a clear postoperative delirium, and suffered from severe orthostatic hypotension that prevented him from standing independently.
[0039] Specifically, such as Figure 2As shown, it was determined that fall monitoring labels could be set up for patients with other fall risk levels, specifically including: Based on the distribution data of the wards of the high-risk patients, the wards containing the high-risk patients are identified; Using data on high-risk patients in wards with existing high-risk patients, we can determine whether fall monitoring tags can be set up for other patients at different fall risk levels.
[0040] It is understandable that when the number of patients at the high-risk level is within a preset range, in order to reduce the monitoring and management difficulty for nursing staff, it is determined that patients at other fall risk levels do not need to have fall monitoring tags set.
[0041] Additionally, it should be noted that when the number of patients with high risk levels is not within the preset range, the average number of patients with high risk levels in the wards containing high-risk patients is determined based on the data of high-risk patients in those wards. If the average number of high-risk patients in the wards containing high-risk patients is greater than the preset patient number threshold, then the distribution of high-risk patients is relatively clustered. Therefore, it is necessary to set fall monitoring tags for patients with other fall risk levels.
[0042] Monitoring patient fall risk is a crucial aspect of routine management in hospital wards. To optimize the allocation of nursing resources, we have adopted a data-driven decision-making process. The core of this process is to dynamically determine whether to assign fall monitoring tags to patients of other risk levels based on the distribution of high-risk patients.
[0043] Scenario 1: The number of high-risk patients is too large, so it is decided not to set labels. Data Acquisition: There are currently 75 high-risk fall patients in the hospital, distributed across 35 different wards.
[0044] Decision Analysis: Step 1: Determine the total number of high-risk patients. The total number is 75, far exceeding the upper limit of the preset range (30 people). This means that the nursing team is facing enormous hospital-wide fall prevention and control pressure.
[0045] Step Two: Draw a direct conclusion. Nursing management resources must be fully allocated to these 75 high-risk patients. To ensure core risks are controlled and to avoid diverting nursing staff's attention, it was decided not to assign fall monitoring tags to other medium- and low-risk patients.
[0046] Conclusion: Do not assign fall monitoring labels to patients at other fall risk levels. The core strategy is to "prioritize key patients."
[0047] Scenario 2: The number of high-risk patients is controllable, but their distribution is highly clustered, so it is decided to set up labels. Data Acquisition: There are currently 45 high-risk fall patients in the hospital. (Note: Although 45 seems like a lot, it is considered "manageable" relative to the total number of beds of 240 and the threshold of 30, because we need to further analyze its distribution.)
[0048] The distribution of these 45 patients is extremely uneven: Wards 201, 202, 305, 306, and 410: each ward has 3 high-risk patients (all beds are full); Wards 108, 209, 311, and 407: each ward has 2 high-risk patients; Wards 101, 115, 208, 312, and 409: each ward has 1 high-risk patient; the total number of wards with high-risk patients is: 5 (full) + 4 (2 people) + 5 (1 person) = 14 wards.
[0049] Decision Analysis: Step 1: Determine the total number of high-risk patients. The total number is 45, exceeding 30. However, please note that in more complex models, this "preset range" can be a soft reference, or we can define it as "when the total number exceeds X, further analysis of the distribution is required." Here we move on to the second step for in-depth analysis.
[0050] Step 2: Calculate the average number of patients in high-risk wards.
[0051] Average number = Total number of high-risk patients / Number of wards with high-risk patients, Avg = 45 / 14 ≈ 3.21 patients / ward Step 3: Determine if the average number exceeds the threshold.
[0052] The calculated Avg ≈ 3.21 is much higher than the preset threshold of 1.8.
[0053] Step 4: Drawing Conclusions. The data revealed a serious problem: a high concentration of high-risk patients. Five wards were completely occupied by high-risk patients, becoming "extremely high-risk units." Nurses working in these wards faced immense workload and psychological stress, and the probability of falls was significantly higher than in other wards. Therefore, intensive measures were necessary.
[0054] Decision: Fall monitoring tags need to be set up for other patients with fall risk levels in these "high-risk cluster wards". Reasons: Reducing nurse workload: Tagging and monitoring low- and medium-risk patients in the same ward can create a systematic safety net, preventing nurses from neglecting other patients due to being busy caring for high-risk patients. Sharing environmental risks: In a ward with 2-3 high-risk patients, potential risks from the environment (such as bathrooms, floors) (water accumulation, obstacles) exist for everyone.
[0055] Conclusion: Labels need to be set for other patients of different risk levels within high-risk cluster wards. This is a strategy of "focusing on key areas while comprehensively improving overall coverage."
[0056] Scenario 3: The number of high-risk patients is controllable and their distribution is scattered, so it is decided not to set labels; Data Acquisition: There are currently 25 high-risk fall patients in the hospital, and these 25 patients are distributed in 23 different wards (21 wards have 1 patient and 2 wards have 2 patients).
[0057] Decision Analysis: Step 1: Determine the total number of high-risk patients. The total number is 25, which is within the preset range [0,30]. Step 2: Calculate the average number of patients in the high-risk ward, Avg = 25 / 23 ≈ 1.09 patients / ward; Step 3: Determine if the average number exceeds the threshold, Avg ≈ 1.09 < preset threshold 1.8.
[0058] Step 4: Conclusion. The total number of high-risk patients is small, and most wards have only one high-risk patient, with a very dispersed distribution. No obvious nursing stress points were observed. Each responsible nurse managed the fall risk in their unit well. Therefore, there is no need to initiate an expanded monitoring tagging system.
[0059] Conclusion: Do not assign fall monitoring labels to patients with other fall risk levels. The strategy is "routine monitoring, with resources on standby".
[0060] The core value of this model lies in its data-driven approach, which scientifically allocates limited nursing resources to the highest-risk and most needed areas.
[0061] Specifically, such as Figure 3 As shown, the method for determining the target for setting fall monitoring labels in patients with other fall risk levels is as follows: Based on the data of patients at the high-risk level, determine the number of patients at the high-risk level; The high-risk patient ratio is determined based on the ratio of the number of high-risk patients to the number of nursing staff. Based on the high-risk patient ratio, the target for setting fall monitoring tags in patients with other fall risk levels is determined.
[0062] It is understandable that when the ratio of high-risk patients is greater than the preset patient ratio threshold, only medium-risk patients in the same ward as the high-risk patients will be used as targets for setting fall monitoring tags.
[0063] It should also be noted that when the ratio of high-risk patients is not greater than the preset patient ratio threshold, if the ratio of high-risk patients is within the preset patient ratio range, the ratio of high-risk patients is relatively large. Therefore, only other patients with fall risk levels in the same ward as the high-risk patients are considered as targets for setting fall risk labels.
[0064] Furthermore, if the ratio of high-risk patients is not within the preset range, then the ratio of high-risk patients is relatively large, and therefore all medium-risk patients are used as the target for setting fall risk labels.
[0065] 1. Setting decision-making rules; Key Indicator: High-risk patient ratio (R) = Total number of high-risk patients / Total number of available caregivers; Preset patient ratio threshold (R1): 1.5 (each nurse is responsible for 1.5 high-risk patients as the stress limit); Preset patient ratio range: [0.8, 1.2] (moderate stress range); 2. Application Scenarios and Decision-Making; Scenario A: R > 1.5 (pressure limit); Data: 30 high-risk patients, 18 nursing staff. R = 30 / 18 ≈ 1.67; Decision: The ratio exceeds 1.5, indicating extreme strain on nursing resources. To ensure core safety, the monitoring target must be the most convergent.
[0066] Implementation: Only target intermediate-risk patients sharing a ward with high-risk patients. Low-risk patients and intermediate-risk patients not in high-risk wards are not targeted.
[0067] Scenario B: 0.8 ≤ R ≤ 1.2 (moderate pressure); Data: 20 high-risk patients and 18 nursing staff. R = 20 / 18 ≈ 1.11. Decision: The ratio is within a manageable range, but there is still pressure. The risk mainly exists in the local environment where the high-risk patients are located.
[0068] Implementation: Only all other patients (intermediate and low risk) in the same ward as the high-risk patient are targeted. This measure aims to alleviate nurses' stress in "high-risk cluster wards" and create a safety net within the ward.
[0069] Scenario C: R < 0.8 (relatively low pressure); Data: 10 high-risk patients, 18 nursing staff. R = 10 / 18 ≈ 0.56. Decision: Nursing staff have sufficient capacity for broader monitoring. Therefore, the focus of prevention should be placed on all medium-risk patients throughout the hospital.
[0070] Implementation: Target all medium-risk patients in the hospital, regardless of whether they shared a room with a high-risk patient.
[0071] Implementation Method 2: Dynamic Weight Adjustment Strategy; This approach introduces more variables into the basic classification, making decisions more precise and suitable for hospitals with a high level of information technology.
[0072] 1. Upgraded decision-making rules; Key indicator: Weighted nursing stress index (P); Calculation formula: P = (Number of high-risk patients * 3 + Number of medium-risk patients * 1) / Total number of available nursing staff. Weighting explanation: High-risk patients are assigned a higher weight (e.g., 3), medium-risk patients are assigned a lower weight (e.g., 1), and low-risk patients are not counted. This more accurately reflects the workload of nursing staff.
[0073] 2. Application Scenarios and Decision-Making; Scene: High-risk patients: 15, medium-risk patients: 25, nursing staff: 20; P = (15 * 3 + 25 * 1) / 20 = (45 + 25) / 20 = 70 / 20 = 3.5; Decision-making rules: If P > 4.0, the target is limited to high-risk patients in the same ward.
[0074] If 2.5 ≤ P ≤ 4.0, then the target is set as: all other patients in the same ward.
[0075] If P < 2.5, the target is set as: all intermediate-risk patients in the hospital.
[0076] Execution: The calculated P = 3.5, falling within the second interval. Therefore, the system decided to target only medium- and low-risk patients sharing a ward with the 15 high-risk patients as targets for fall monitoring tags.
[0077] Specifically, such as Figure 4 As shown, the method for determining the postoperative walking safety monitoring and management method for the patient is as follows: Based on the target setting of fall monitoring tags, determine the proportion of all patients in the ward monitoring area that have the target setting of fall monitoring tags, and use this proportion as the monitoring ratio; Based on the fall monitoring tag setting data of medium-risk patients in the ward monitoring area, medium-risk patients who have not set fall monitoring tags in the ward monitoring area are identified and treated as unmonitored patients; Based on the number of unmonitored patients in the ward monitoring area, a preset threshold corresponding to the ward monitoring area is determined, and a monitoring and processing method for the ward monitoring area is determined based on the preset threshold and the monitoring ratio.
[0078] It is understood that if the monitoring ratio of the ward monitoring area is less than the preset threshold, then the monitoring processing method of the ward monitoring area is determined to be to monitor all patients, thereby using the monitoring images to determine whether the patients have experienced accidents such as falls.
[0079] Additionally, it is understood that if the monitoring ratio of the ward monitoring area is not less than the preset threshold, then the monitoring processing method for the ward monitoring area is determined to be that no monitoring processing is required, and only fall detection tags are needed to detect and process patient falls. This avoids the technical problem that the timeliness of identification processing may be affected by monitoring processing in all ward monitoring areas.
[0080] The following are several corresponding embodiments: Implementation Method 1: In-depth application and analysis of the hierarchical dynamic threshold method; (a) Strategic considerations for background and parameter setting; The core of using the "hierarchical dynamic threshold method" to achieve intelligent monitoring resource scheduling lies in our recognition that risks are not evenly distributed and that we must respond in a hierarchical manner based on the key indicator of "unmonitored patients".
[0081] Scientific classification of risk levels: Based on retrospective analysis of historical fall event data, we determined the correlation between risk levels and the number of unmonitored patients (U): Low-risk zone (U ≤ 2): Historical data shows that the probability of falls caused by unmonitored patients is less than 0.5% within this range. Our goal in this zone is "efficient operation and maintenance," setting the monitoring ratio threshold (T) to 0.5. This means that as long as half of the patients are covered by the tagging system, we consider routine monitoring safe.
[0082] Medium-risk area (3 ≤ U ≤ 4): Within this range, the probability of falling risk rises to 1.5% - 3%. Our strategy changes to "active defense", and the threshold T is significantly increased to 0.7. This means that the system's requirements for label coverage become strict to prevent possible risk aggregation.
[0083] High-risk area (U ≥ 5): This is our warning area, where the risk probability exceeds 5%. The strategy is "full coverage", and a very high threshold of 0.9 is set. This almost means that as long as there are any slight loopholes in the label coverage, the system will automatically start the highest level of monitoring.
[0084] (2) In-depth deduction of specific application scenarios; Let's focus on the "Third Nursing Unit" in the orthopedic ward area. This unit has a total of 8 wards and 40 beds.
[0085] Data situation: There are 40 patients in this unit. After preliminary evaluation, there are 10 high-risk patients and 15 medium-risk patients.
[0086] According to the label setting strategy in the previous stage (for example, only set labels for medium-risk patients in the same ward as high-risk patients), finally only 10 high-risk patients and 7 medium-risk patients are set with fall monitoring labels. Therefore, the total number of patients with labels set is 17, and the number of medium-risk patients not monitored (U) = 15 - 7 = 8.
[0087] Operation and logical deduction of the decision tree: First step: Risk level determination. The system detects that U = 8, immediately designates this unit as a "high-risk area", and calls the corresponding monitoring ratio threshold T = 0.9. Second step: Monitoring ratio calculation. The monitoring ratio (M) = 17 / 40 = 0.425. Third step: Core decision-making judgment. The system makes a comparison: M (0.425) < T (0.9). The condition is met. Fourth step: Instruction output and execution. The system automatically generates an instruction: "Start full image monitoring and processing for the Third Nursing Unit."
[0088] Subsequent impacts and value analysis; Resource consideration: Although full monitoring increases the system computing power burden, compared with the potential medical disputes, additional treatment costs, and sharp consumption of human resources after a patient falls, this preventive resource investment has proven to be highly cost-effective.
[0089] Implementation method 2: Fine operation of the continuous function model method; (1) Strategic considerations for background and parameter setting; For our neurology department, patients' conditions are complex and fluctuate greatly, and a simple three-level classification may not be accurate enough. We introduced a dynamic threshold model based on continuous functions, aiming to achieve precise management in a "stepless" manner.
[0090] Model building: Baseline threshold (T0): Set to 0.6. This represents the baseline we consider 60% label coverage to be ideal. Adjustment factor (k): Set to 0.02 after data fitting. This means that for each additional unmonitored patient, the baseline threshold decreases by 0.02, reflecting the system's sensitivity to the expansion of risk blind spots. Dynamic threshold formula: T = T0 - (U × k) = 0.6 - (U × 0.02).
[0091] (II) In-depth analysis of specific application scenarios; Taking the "cerebrovascular disease ward" in the Department of Neurology as an example, the ward has 5 rooms and a total of 25 patients.
[0092] Data status: There are currently 25 patients. 6 high-risk patients and 8 medium-risk patients have been successfully tagged, so the total number of tagged patients is 14. The total number of medium-risk patients is 12. Therefore, the number of unmonitored medium-risk patients (U) is 12 - 8 = 4.
[0093] Precise calculations of the decision model: Step 1: Calculate the dynamic threshold. T = 0.6 - (4 × 0.02) = 0.6 - 0.08 = 0.52. Step 2: Calculate the monitoring ratio. M = 14 / 25 = 0.56.
[0094] Step 3: Core Decision Judgment. System Comparison: M (0.56) > T (0.52). Condition not met. Step 4: Instruction Output and Execution. System generates instruction: Only fall detection tags are needed, greatly reducing the analysis pressure on the backend artificial intelligence and improving the accuracy and response speed of alarms.
[0095] Subsequent impact and value analysis: Efficiency Improvement: This model avoids a "one-size-fits-all" approach to comprehensive monitoring, saving approximately 44% of computing resources ((25-14) / 25), allowing the information center to allocate computing power to other wards that need it more. Management Refinement: Head nurses reported that this "targeted" monitoring makes the alarms at the nurses' station more "credible," reducing false alarms caused by the normal activities of low-risk patients, and increasing the nursing team's trust in and willingness to use the intelligent system.
[0096] Optionally, the method for determining the postoperative walking safety monitoring and management method for the patient is as follows: Based on the target setting of fall monitoring tags, determine the proportion of all patients in the ward monitoring area that have the target setting of fall monitoring tags, and use this proportion as the monitoring ratio; Based on the fall monitoring tag setting data of medium-risk patients in the ward monitoring area, medium-risk patients who have not set fall monitoring tags in the ward monitoring area are identified and treated as unmonitored patients; Based on the monitoring ratio of the ward monitoring area and the number of unmonitored patients, the monitoring and processing method for the ward monitoring area is determined.
[0097] It is understandable that when the monitoring ratio is less than the preset monitoring ratio threshold, the number of fall monitoring targets set in the ward monitoring area is small. Therefore, the monitoring and processing method for the ward monitoring area is determined to be to monitor all patients, thereby using the monitoring images to determine whether the patient has experienced a fall or other accident.
[0098] Furthermore, if the monitoring ratio is not less than a preset monitoring ratio threshold, the number of unmonitored patients in the ward monitoring area is further determined. If there are no unmonitored patients in the ward monitoring area, the monitoring processing method for the ward monitoring area is determined to be that no monitoring processing is required, and only fall detection processing of patients needs to be achieved using fall detection tags. This avoids the technical problem that the timeliness of identification processing may be affected by monitoring processing in all ward monitoring areas.
[0099] Additionally, it should be noted that if there are unmonitored patients in the ward monitoring area, the number of unmonitored patients is determined. If the number of unmonitored patients is greater than a preset threshold for the number of monitored patients, the monitoring process for the ward monitoring area is determined to be monitoring all patients, thereby using the monitoring images to determine whether the patients have experienced accidents such as falls.
[0100] Furthermore, if the number of unmonitored patients is not greater than a preset threshold for the number of monitored patients, a preset threshold corresponding to the ward monitoring area is determined based on the number of unmonitored patients. If the monitoring ratio of the ward monitoring area is less than the preset threshold, the monitoring processing method for the ward monitoring area is determined to be to monitor all patients, thereby using the monitoring images to determine whether the patients have experienced accidents such as falls.
[0101] It should be noted that the preset threshold is determined based on the number of unmonitored patients, and the larger the number of unmonitored patients, the larger the preset threshold.
[0102] Additionally, it can be understood that if the monitoring ratio of the ward monitoring area is not less than the preset threshold, then the monitoring processing method of the ward monitoring area is determined to be to monitor all patients, thereby using the monitoring images to determine whether the patients have experienced accidents such as falls.
[0103] Furthermore, the method for determining the setting method of the fall monitoring tags for patients with different newly added fall risk levels is as follows: The ward monitoring area, which is used to monitor and analyze walking safety, is designated as the monitoring area. Based on the patient data in the monitoring area, the number of patients in the monitoring area is determined. Based on the monitoring and management methods for the monitoring areas, determine the number of patients in the monitoring areas under different monitoring and management methods; Based on the number of patients in monitoring areas with different monitoring and management methods, determine how to set fall monitoring tags for newly added patients with different fall risk levels.
[0104] It should be noted that when the patient is classified as a high-risk patient, then it is determined that the patient needs to have a fall monitoring tag set.
[0105] It should be noted that the method for setting fall monitoring labels for newly added patients with different fall risk levels is determined based on the number of patients in the monitoring areas of different monitoring and management methods. Specifically, this includes: If the patient is not a high-risk patient, and if the patient is in the monitoring area, then it is determined that the patient does not need to set a fall monitoring tag. However, if the patient is not in the monitoring area, and if the number of medium-risk patients in the monitoring area accounts for a greater than a preset percentage threshold among all medium-risk patients, then it is determined that the patient needs to set a fall monitoring tag.
[0106] Furthermore, if the number of patients at medium risk in the monitoring area does not exceed a preset percentage threshold as a percentage of all patients at medium risk, then the patients who need to be assigned fall monitoring tags are determined based on the percentage of patients in the monitoring areas of different monitoring and management methods as a percentage of all patients at medium risk.
[0107] Understandably, the number of patients in monitoring areas using different monitoring and management methods is used as a basis to determine the proportion of patients in all medium-risk areas, thus identifying the patients who require fall monitoring labeling. Specifically, this includes: Based on the proportion of the number of patients in the monitoring areas of different monitoring and management methods to the number of patients in all medium-risk areas, a monitoring deviation coefficient for medium-risk patients is determined. If the monitoring deviation coefficient is greater than a preset deviation coefficient threshold, it is determined that the patient needs to be set with a fall monitoring tag. If the monitoring deviation coefficient is not greater than the preset deviation coefficient threshold, and if the monitoring ratio in the monitoring area is within the preset monitoring ratio range, that is, the number of patients with fall monitoring tags in the monitoring area is small, then it is determined that the patient needs to have a fall monitoring tag set. If the monitoring ratio in the monitoring area is not within the preset monitoring ratio range, then it is determined that the patient does not need to have a fall monitoring tag set.
[0108] Implementation plan for setting up labels for medium-risk patients based on the global monitoring resource distribution; I. Initial configuration of the hospital's monitoring environment; Our hospital's inpatient department currently has a total of 500 beds. According to the real-time assessment of the intelligent operation and maintenance system, the current risk level distribution of all patients in the hospital is as follows: 50 high-risk patients, 100 medium-risk patients, and 350 low-risk patients. Following preliminary assessment and decision-making, three key monitoring areas have been designated, and comprehensive image monitoring and analysis have been initiated in all of these areas. The orthopedic ward monitoring area admitted 80 patients, including 25 medium-risk patients; the neurology monitoring area admitted 70 patients, including 10 medium-risk patients; and the critical care rehabilitation monitoring area admitted 50 patients, including 5 medium-risk patients. The monitored area admitted a total of 200 patients, including 40 medium-risk patients. The parameters of the hospital-wide monitoring system were set as follows: preset percentage threshold of 60%, preset deviation coefficient threshold of 1.5, and preset monitoring ratio range of [0.3, 0.7].
[0109] II. Decision-making process for setting labels for newly admitted patients; Scenario 1: Handling of new patients in the monitoring area; When a newly admitted medium-risk patient is assigned to the monitoring area of the orthopedic ward, the system immediately initiates a decision analysis process. First, the patient's risk level is confirmed as medium-risk, and then their location is identified as the monitoring area. Based on the rule that "no tag is needed if the patient is within the monitoring area," the system automatically generates a decision: do not assign a fall monitoring tag to Ms. Li. The decision is based on the fact that comprehensive image surveillance is already in place in this area, enabling real-time monitoring of her behavior through intelligent video analytics, eliminating the need for additional electronic tag monitoring.
[0110] Scenario 2: Management of patients in non-monitoring areas; Another newly admitted medium-risk patient was assigned to a general internal medicine ward (non-monitoring area), and the system initiated a more complex multi-dimensional decision analysis: Global distribution analysis: There are currently 40 medium-risk patients in the monitoring area, accounting for 40% of the total number of medium-risk patients in the hospital (100). This proportion is lower than the preset threshold of 60%, indicating that the distribution of medium-risk patients is relatively balanced throughout the hospital, and there is no excessive concentration in the monitoring area.
[0111] Resource distribution balance assessment: The system further calculates the monitoring deviation coefficient. The proportion of patients in the monitored area to all patients in the hospital is 200 / 500 = 0.4; the proportion of high-risk patients in the monitored area to all medium-risk patients in the hospital is 40 / 100 = 0.4. The monitoring deviation coefficient is 0.4 / 0.4 = 1.0, which is lower than the preset threshold of 1.5. This data indicates that the distribution of monitoring resources is basically matched with the risk distribution, and there is no significant imbalance in resource allocation.
[0112] Monitoring area effectiveness assessment: A thorough analysis of the label coverage in the monitoring area revealed that 50 patients had been tagged within the area, out of a total of 200 patients, resulting in a monitoring ratio of 50 / 200 = 0.25. This figure is lower than the lower limit of the preset monitoring ratio range [0.3, 0.7], highlighting the insufficient basic label coverage within the monitoring area.
[0113] Based on the above analysis, the system made the final decision: to set up a fall detection tag for Mr. Zhang. The underlying logic of this decision is that when the basic tag coverage of the monitoring area itself is insufficient, it indicates that the risk management foundation of the entire hospital is weak, and it is necessary to strengthen the hospital's safety network by expanding the tag coverage of ordinary areas.
[0114] Through the implementation of this solution, the hospital has achieved precise allocation of monitoring resources: the monitoring area makes full use of image monitoring technology to achieve full coverage monitoring of key areas, avoiding the waste of resources caused by repeated labeling; non-monitoring areas are dynamically evaluated through intelligent algorithms, and labels are only added when needed for global risk management, ensuring that the effectiveness of safety investment is maximized; the system can promptly identify structural problems in the hospital's risk management by continuously monitoring various indicators and automatically adjust strategies.
[0115] Example 2 In a second aspect, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the above-described method for postoperative walking safety management based on multi-source data fusion when running the computer program.
[0116] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, devices, and non-volatile computer storage media are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.
[0117] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.
[0118] The above description is merely one or more embodiments of this specification and is not intended to limit this specification. Various modifications and variations can be made to the one or more embodiments of this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of one or more embodiments of this specification should be included within the scope of the claims of this specification.
Claims
1. A postoperative walking safety management method based on multi-source data fusion, characterized in that, Specifically, it includes: Based on the postoperative fall risk level assessment results, high-risk patients are identified. Based on the ward distribution data of the high-risk patients, it is determined whether fall monitoring tags can be set for other patients with fall risk levels. Based on the data of high-risk patients, the target for setting fall monitoring tags among the other patients with fall risk levels is determined. Based on the goals of fall monitoring tag setting, determine the monitoring and management methods for patients' walking safety after surgery, and based on the monitoring and processing methods of different ward monitoring areas, determine the ward areas for monitoring and analyzing walking safety. Based on patient data from ward monitoring areas that monitor and analyze walking safety, the method for setting fall monitoring tags for newly added patients with different fall risk levels was determined.
2. The postoperative walking safety management method based on multi-source data fusion as described in claim 1, characterized in that, The fall risk levels are categorized into high risk, low risk, and medium risk.
3. The postoperative walking safety management method based on multi-source data fusion as described in claim 1, characterized in that, The assessment of the fall risk level is determined based on the monitoring data of the patient's postoperative physical indicators.
4. The postoperative walking safety management method based on multi-source data fusion as described in claim 1, characterized in that, It was determined that fall monitoring labels could be set up for patients at other fall risk levels, specifically including: Based on the distribution data of the wards of the high-risk patients, the wards containing the high-risk patients are identified; Using data on high-risk patients in wards with existing high-risk patients, we can determine whether fall monitoring tags can be set up for other patients at different fall risk levels.
5. The postoperative walking safety management method based on multi-source data fusion as described in claim 1, characterized in that, When the number of patients at the high-risk level is within a preset range, it is determined that patients at other fall risk levels do not need to have fall monitoring tags set.
6. The postoperative walking safety management method based on multi-source data fusion as described in claim 1, characterized in that, The method for determining the target for setting fall monitoring tags in patients with other fall risk levels is as follows: Based on the data of patients at the high-risk level, determine the number of patients at the high-risk level; The high-risk patient ratio is determined based on the ratio of the number of high-risk patients to the number of nursing staff. Based on the high-risk patient ratio, the target for setting fall monitoring tags in patients with other fall risk levels is determined.
7. The postoperative walking safety management method based on multi-source data fusion as described in claim 6, characterized in that, When the ratio of high-risk patients is greater than the preset patient ratio threshold, only medium-risk patients in the same ward as the high-risk patients will be used as targets for setting fall monitoring tags.
8. The postoperative walking safety management method based on multi-source data fusion as described in claim 1, characterized in that, The method for determining the setting method of fall monitoring tags for patients with different newly added fall risk levels is as follows: The ward monitoring area, which is used to monitor and analyze walking safety, is designated as the monitoring area. Based on the patient data in the monitoring area, the number of patients in the monitoring area is determined. Based on the monitoring and management methods for the monitoring areas, determine the number of patients in the monitoring areas under different monitoring and management methods; Based on the number of patients in monitoring areas with different monitoring and management methods, determine how to set fall monitoring tags for newly added patients with different fall risk levels.
9. The postoperative walking safety management method based on multi-source data fusion as described in claim 8, characterized in that, When the patient is classified as a high-risk patient, it is determined that the patient needs to have a fall monitoring tag set.
10. A computer system, comprising: A memory and processor connected in communication, and a computer program stored in the memory and capable of running on the processor, characterized in that, when the processor runs the computer program, it executes a postoperative walking safety management method based on multi-source data fusion as described in any one of claims 1-9.