Orthopedic patient care plan assistance selection system

Abstract

The present application relates to the technical field of nursing decision, specifically to a nursing scheme auxiliary selection system for orthopedic patients, in the present application, in the rehabilitation training process, the moving track of the patient is collected in real time and deviation analysis is conducted with the ideal path, the deviation risk fragment set is established by means of the spatial coordinate sequence and the Hurst index calculation mode, the identification and positioning of the deviation trend of the rehabilitation action are strengthened, and the accuracy of the walking control stability evaluation is enhanced. The heart rate signal is mapped to a multi-dimensional density space, the local abnormal interval in the heart rate change is accurately calibrated through the local abnormal factor operation mode under the sliding window, and the identification of the slight physiological change is facilitated. The heart rate abnormal interval is classified and processed according to the duration, a heart rate fluctuation typing system is constructed, the transient physiological fluctuation and the continuous stress state can be effectively distinguished, and the identification degree of the pain inducement is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of nursing decision-making, in particular to a nursing scheme assisted selection system for orthopedic patients. BACKGROUND

[0002] The nursing scheme assisted selection system for orthopedic patients refers to a system technical scheme for providing standardized nursing path recommendation and nursing measure combination support based on information such as disease characteristics, treatment methods, rehabilitation cycles and individual health conditions of orthopedic patients, and specifically taking patient diagnosis and treatment records, assessment scale data and vital sign parameters as the basis, assisting nursing personnel to select schemes by setting standard nursing action lists and matters needing attention corresponding to the nursing path.

[0003] The prior art relies on standardized scheme matching in the nursing path recommendation for orthopedic patients, mainly sets static paths around diagnosis and treatment records and assessment scales, which leads to insufficient recognition of subtle changes in the recovery process. The recognition of mobile path control deviation relies on subjective observation or single-point monitoring methods, which is difficult to depict the trend of continuous behavior change, so as to accurately identify the potential deviation risk period. The assessment of heart rate and blood pressure stays at the level of basic parameters such as mean value and extreme value, ignoring the dynamic characteristics in the local time domain, which leads to lag in the recognition of risk states such as pain, stress or low perfusion. During night nursing, there is a lack of continuous assessment means for venous return, and passive response is often triggered only after the patient shows obvious swelling or blood circulation disorder symptoms, missing the early intervention opportunity. In addition, the nursing path suggestion often adopts a unified process setting, which fails to accurately adjust according to the physiological fluctuation state and real-time performance of the patient. For example, when the heart rate fluctuates abnormally but there is no clear behavior abnormality, it is often ignored, delaying pain intervention; when the blood pressure continues to drop at night but there is no nursing record intervention, it is also not recognized by the system, increasing the nursing blind area and risk exposure point, and affecting the safety, comfort and rehabilitation quality of the patient in the recovery process. SUMMARY

[0004] The purpose of the present application is to solve the problems existing in the prior art, and to provide a nursing scheme assisted selection system for orthopedic patients.

[0005] In order to achieve the above-mentioned purpose, the technical scheme adopted by the present application is as follows: the nursing scheme assisted selection system for orthopedic patients comprises:

[0006] The gait deviation monitoring module collects the moving path of the patient during the daytime rehabilitation training process, compares it with the ideal moving path, extracts the deviation distance between the paths according to the comparison result, forms a deviation time sequence, screens the risk time period by calculating the Hurst index of the amplitude change of the deviation distance in the deviation time sequence, and forms a deviation risk segment set;

[0007] The heart rate abnormality recognition module collects the heart rate of the patient during daytime rehabilitation training and maps it to a multi-dimensional density space, and marks the continuous local abnormal time points of the heart rate in the multi-dimensional density space as a heart rate local abnormal interval;

[0008] The heart rate fluctuation classification module classifies the interval type according to the duration of the heart rate local abnormal interval in each time period, and establishes a heart rate fluctuation classification set;

[0009] The blood pressure trend analysis module collects the continuous systolic blood pressure pulse waveform signal of the patient at night after the operation, analyzes the rhythm change of the systolic blood pressure pulse in the specified period, and generates a blood pressure trend analysis result;

[0010] The body position intervention discrimination module marks the target period segment without nursing intervention according to the blood pressure trend analysis result and the operation data in the same time period in the nursing record, and establishes a venous reflux abnormality non-response record;

[0011] The rehabilitation guidance module recommends a corresponding nursing scheme to the patient according to the offset risk segment set, the heart rate fluctuation classification set, and the venous reflux abnormality non-response record, and obtains a multi-maintenance nursing path recommendation result.

[0012] As a further scheme of the present application, the offset risk segment set acquisition step is specifically:

[0013] The movement path of the patient during daytime rehabilitation training is collected through a wearable device on the patient, the latitude and longitude coordinate points recorded every second in the movement path are converted into an equal-interval continuous spatial coordinate sequence, and the deviation distance of the coordinate points of the same time continuous spatial coordinate sequence relative to the corresponding position coordinate points of the ideal movement path is calculated and recorded according to the current postoperative recovery period of the patient and the ideal movement path of the rehabilitation training plan;

[0014] The deviation distance of each time is counted to form an offset time sequence, and the HURST index of the deviation distance amplitude change in the offset time sequence is calculated by analyzing the power law relationship between the change amplitude of the deviation distance at different times and the length of the entire offset time sequence using the HURST index algorithm;

[0015] The HURST index is compared with a preset reference index threshold to evaluate the stability of the direction control of the patient during walking, and the corresponding risk time period in the offset time sequence that is lower than the reference index threshold is marked to form an offset risk segment set.

[0016] As a further scheme of the present application, the heart rate local abnormal interval acquisition step is specifically:

[0017] Collecting heart rate per minute of a patient in a daytime rehabilitation training process through a wearable device on the patient, and dividing according to a fixed sliding window, extracting continuous heart rate in each sliding window to form a heart rate information set;

[0018] Selecting each time point and heart rate information of adjacent regions before and after a target time point in the heart rate information set, mapping all the heart rate information to a multidimensional density space, and calculating the local density of each time point in the corresponding adjacent region in the multidimensional density space by using a local anomaly factor algorithm;

[0019] Comparing the local density with a preset density threshold value, screening local anomaly time points with a local density lower than the density threshold value, and marking continuous local anomaly time points as a heart rate local anomaly interval.

[0020] As a further scheme of the present application, the obtaining step of the heart rate fluctuation classification set is specifically:

[0021] Statistically analyzing the duration of the heart rate local anomaly interval of each time period, comparing the statistical duration with a set time threshold value respectively, marking a heart rate local anomaly interval higher than the time threshold value as a sustained fluctuation section, and marking a heart rate local anomaly interval lower than the time threshold value as an instantaneous fluctuation section;

[0022] Classifying and marking according to the section types corresponding to the sustained fluctuation section and the instantaneous fluctuation section to establish a heart rate fluctuation classification set.

[0023] As a further scheme of the present application, the obtaining step of the blood pressure trend analysis result is specifically:

[0024] Collecting continuous systolic pressure pulse waveform signals of a patient at night after an operation through a blood pressure monitoring device, dividing into multiple equal-length time signal windows, constructing a continuous signal curve with the corresponding time as the horizontal axis and the systolic pressure pulse as the vertical axis for each systolic pressure pulse waveform in the time signal window;

[0025] Performing Hilbert transform on the continuous signal curve to analyze the instantaneous frequency and phase trajectory in each systolic pressure pulse corresponding period, analyzing the rhythm change in the systolic pressure pulse corresponding period according to the instantaneous frequency and phase trajectory, identifying and calibrating a period section with a rhythm decline and no rebound trend, and generating a blood pressure trend analysis result.

[0026] As a further scheme of the present application, the obtaining step of the venous reflux abnormality non-response record is specifically:

[0027] According to the blood pressure trend analysis result, whether the patient has lower limb venous return slowing down in the target period is inferred, wherein the period corresponding to the rhythm decrease and no rebound trend corresponds to a stable slow process of continuously weakened blood flow dynamics and interrupted by body position change, indicating that there is a long time of venous return efficiency reduction in static state, and a lower limb venous return analysis result is obtained;

[0028] Based on the lower limb venous return analysis result, whether there is nursing intervention of lifting lower limbs and body position guidance in the operation data of the same time period in the linkage nursing record is identified, if no nursing intervention is performed, the target period is marked again, and a venous return abnormality non-response record is established.

[0029] As a further scheme of the application, the acquisition step of the multi-dimensional nursing path recommendation result is specifically:

[0030] According to the offset risk segment set, the recommendation of daytime moving path setting and guiding mode is performed, including wall leaning, setting boundary auxiliary objects, and arranging accompanying personnel to guide movement at a specified distance;

[0031] According to the heart rate fluctuation classification set, the transient fluctuation segment is regarded as a non-continuous reaction, and no nursing scheme adjustment is performed, the continuous fluctuation segment is regarded as existing a pain state, and the increase of cold intervention prompt and nursing patrol video frequency can be recommended;

[0032] According to the venous return abnormality non-response record, the time point setting of increasing the lifting of lower limbs or guiding the turning over in the marked nighttime target period is recommended, a multi-dimensional nursing path recommendation result is obtained, and decision support for the auxiliary selection of the nursing scheme of orthopedic patients is provided.

[0033] Compared with the prior art, the application has the advantages and positive effects that:

[0034] In the present application, during the rehabilitation training process, the patient's moving track is collected in real time and analyzed for deviation from the ideal path, and the deviation risk segment set is established by means of spatial coordinate sequence and Hurst index calculation method, so as to strengthen the identification and positioning of the deviation trend of rehabilitation movement, and enhance the accuracy of walking control stability evaluation. The heart rate signal is mapped to a multi-dimensional density space, and the local abnormal interval in the heart rate change is accurately marked by means of local anomaly factor operation under the sliding window, which is beneficial to identifying the slight physiological change. The duration classification processing is carried out on the heart rate abnormal interval, and the heart rate fluctuation typing system is constructed, which can effectively distinguish the transient physiological fluctuation and the continuous stress state, and improve the identification degree of the pain inducement. In the night blood pressure monitoring, based on the continuous curve construction of the systolic pressure pulse waveform, the instantaneous frequency and phase trajectory are obtained by Hilbert transform, so as to realize the dynamic trend capture of the blood pressure rhythm and venous return lag process. Combined with the nursing record period data, the rhythm attenuation period which is not interrupted by operation is determined, so as to realize the automatic identification and marking of the missing link of nursing intervention. The risk deviation segment, heart rate fluctuation type and venous return response state are fused, and the nursing suggestion content is matched according to the recovery stage, such as guidance mode, cold intervention frequency and body position adjustment setting, so as to strengthen the individualization adaptability and decision rationality of the path recommendation. In summary, through the multi-signal linkage analysis and dynamic identification mechanism, the response speed and execution accuracy of the rehabilitation scheme selection can be effectively improved. BRIEF DESCRIPTION OF DRAWINGS

[0035] Figure 1 The system flowchart of the present application. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application, and are not used to limit the present application.

[0037] Please refer to Figure 1 The nursing scheme auxiliary selection system for orthopedic patients comprises:

[0038] The gait deviation monitoring module collects the moving path of the patient during the daytime rehabilitation training process, compares it with the ideal moving path, extracts the deviation distance between the paths according to the comparison result, forms a deviation time sequence, and screens the risk time period by calculating the Hurst index of the amplitude change of the deviation distance in the deviation time sequence, thereby forming a deviation risk segment set;

[0039] The heart rate abnormality identification module collects the heart rate of the patient during the daytime rehabilitation training process and maps it to a multi-dimensional density space, and marks the continuous local abnormal time points of the multi-dimensional density space heart rate as the heart rate local abnormal interval;

[0040] The heart rate fluctuation classification module divides the interval type according to the duration of the local abnormal interval of the heart rate in each time period, and establishes a heart rate fluctuation classification set;

[0041] The blood pressure trend analysis module collects the continuous systolic blood pressure waveform signal of the patient at night after the operation, analyzes the rhythm change of the systolic blood pressure pulse in the specified period, and generates a blood pressure trend analysis result;

[0042] The body position intervention discrimination module marks the target period segment without nursing intervention according to the blood pressure trend analysis result and the operation data in the same time period in the nursing record, and establishes a venous reflux abnormality non-response record;

[0043] The rehabilitation guidance module recommends a corresponding nursing scheme to the patient according to the offset risk segment set, the heart rate fluctuation classification set and the venous reflux abnormality non-response record, and obtains a multi-maintenance nursing path recommendation result.

[0044] The offset risk segment set is obtained by the following steps:

[0045] The moving path of the patient during the daytime rehabilitation training is collected by a wearable device on the patient's body, the latitude and longitude coordinate points recorded every second in the moving path are converted into an equal-interval continuous spatial coordinate sequence, the ideal moving path of the rehabilitation training plan is preset according to the current postoperative recovery period of the patient, and the deviation distance of the coordinate points of the same time continuous spatial coordinate sequence relative to the corresponding position coordinate points of the ideal moving path is calculated and recorded.

[0046] The moving path of the patient during the daytime rehabilitation training is collected by a wearable device on the patient's body, the device includes a smart wearable terminal with high-precision GNSS positioning capability, which can sample and record the latitude and longitude coordinate points of the patient every second in the moving process in real time at a frequency of 1 Hz, the collected coordinate data is expressed in WGS-84 coordinate system, and ArcGISPro software is used to process the whole process, including batch conversion of original latitude and longitude coordinates into XY plane coordinate points in the projection coordinate system, conversion unit is meter, and the spatial scale of distance calculation is unified; then, the rehabilitation training plan of different stages is set by the rehabilitation expert according to the postoperative recovery period of the patient, wherein the ideal moving path can be determined based on the trajectory data of the patients with the same type of surgery and similar signs in the recovery period after statistical analysis, the continuous coordinate point sequence of the ideal moving path is output according to the time point marked in each second by importing the standard path layer in ArcGISPro; after the actual trajectory and the ideal path trajectory of the two groups of time points are aligned and exported, the deviation distance of the coordinate points of the same time continuous spatial coordinate sequence relative to the corresponding position coordinate points of the ideal moving path is calculated, and the formula is as follows:

[0047]

[0048] wherein, represents the patient's first The deviation distance between the moving path in seconds and the ideal moving path, in meters, represents the patient's first The coordinate of the moving path in seconds on the X axis, in meters, is derived from the projection coordinates extracted in ArcGIS Pro. represents the patient's first The coordinate of the second moving path on the Y axis, in meters, is obtained in the same way; Indicates that the patient The coordinate of the ideal moving path on the X axis in meters, which is extracted after conversion from the training trajectory defined by rehabilitation. Indicates that the patient The coordinate of the ideal moving path on the Y axis in meters. It is the time point index, which increases from 1 and represents the time series point corresponding to the sampling frequency per second. represents the patient's movement path in a continuous sequence of spatial coordinates, Indicates the ideal movement path.

[0049] Assume that a test training is conducted in the postoperative rehabilitation department of a hospital. The patient's wearable device is used to record movement data and match it with the ideal path set by the doctor in ArcGIS Pro. Assume that in the early postoperative rehabilitation stage, the patient's walking speed is controlled between 0.3-0.7 m / s and the path fluctuation is small. If the actual coordinates exported in ArcGIS Pro at the 10th second are rice, meters, the coordinates of the ideal path are rice, Meters, the calculation is as follows:

[0050]

[0051] The results show that at the 10th second, the patient's actual walking position deviated by approximately 1.945 meters relative to the preset ideal path.

[0052] The formula calculates the distance between the coordinate position of the patient's actual movement trajectory at each time point and the coordinate position of the ideal movement path at that time point. This offset reflects the degree of deviation of the patient's actual position from the ideal path at that moment. This calculation process is repeated at all time points, forming a numerical sequence that describes the patient's continuous offset during training.

[0053] The deviation distance of each time is counted to form a shift time sequence, and the Heston index algorithm is adopted to analyze the power law relationship between the change amplitude of the deviation distance of different times in the shift time sequence and the length of the entire shift time sequence, so as to calculate the Heston index of the amplitude change of the deviation distance in the shift time sequence;

[0054] According to the collection condition of the sampling frequency 1Hz, all the deviation distances are arranged in time sequence to form a continuous time sequence, and a shift time sequence is formed ; the sequence is used to reflect the spatial shift change of the patient in the training process.

[0055] Suppose the current shift time sequence is (5 shift distance data): , the average shift distance in the shift time sequence is calculated : meters.

[0056] In the calculation of the Heston index, the dispersion degree of each point in the shift time sequence only reflects the static volatility. In order to reveal the fluctuation trend change of the patient's gait shift in the time dimension, the shift distance sequence needs to be centralized and the difference value needs to be accumulated according to time, so the centralized cumulative deviation sequence needs to be constructed. The construction of the centralized cumulative deviation sequence is to reveal whether there is a persistent or trend fluctuation in the shift time sequence based on the overall shift mean value. First, the average shift distance of the entire shift time sequence is counted, and the difference between the shift value of each time point and the average value is calculated to obtain the centralized shift value. Then, the centralized shift value of each time point is sequentially accumulated to form a cumulative time sequence. The construction logic of the sequence is based on: if the shift trend has persistence, the cumulative value will continuously rise or fall, otherwise it will fluctuate around zero.

[0057] According to the average shift distance in the shift time sequence, the centralized cumulative deviation sequence is constructed :

[0058] ; ; ; ; , the centralized cumulative deviation sequence is .

[0059] The range of the centralized cumulative deviation sequence is calculated : The range is the difference between the maximum and minimum values ​​in a centered cumulative deviation sequence and is typically used to measure the maximum fluctuation range of a sequence over the observation period. In Hurst exponent calculations, the range represents the maximum cumulative change in the patient's deviation trend over the entire training period. This value can be used to determine whether the deviation exhibits significant unidirectional fluctuations over time, thereby reflecting the patient's ability to control trend deviations. Combined with the standard deviation, the range as a numerator can express the degree of trend diffusion and is an important measure of long-term dependent characteristics.

[0060] Calculate the standard deviation of a time series with an offset : Meters. The standard deviation is a statistical indicator used to measure the degree of dispersion of a set of data relative to its mean. In an offset time series, the standard deviation reflects the fluctuation of the offset distance per second relative to the average offset distance for the entire training session. Larger values ​​indicate a wider range of offset distance variations and more dramatic fluctuations; smaller values ​​indicate a more stable offset distance and less fluctuation. In the calculation of the Hurst exponent, the standard deviation is used as the denominator to normalize the range of the centralized cumulative deviation, making the result unaffected by the absolute size of the offset value, thereby focusing on the trend fluctuation characteristics themselves.

[0061] Recovery period weight factor It is a dimensionless parameter used to adjust the sensitivity of the Hurst exponent to offset fluctuations, and its setting needs to be based on the patient's current postoperative recovery stage. The setting process is based on the training data of patients with the same surgical procedure collected from the hospital rehabilitation database, and the standard deviation range of the offset time series and the cumulative deviation range in different postoperative days are counted. Combined with the objective score of gait control ability by rehabilitation experts (MRC muscle strength grade assessment), the patient's rehabilitation stage can be divided into the following three categories: Early postoperative stage (1st to 7th day after surgery): Assume that the standard deviation of the patient's offset time series in this stage is concentrated between 0.18 meters and 0.35 meters, and the centralized cumulative deviation range is concentrated between 0.38 meters and 0.70 meters. Therefore, the weight factor is set. , used to enhance the ability to respond to fluctuations and make the Hurst index have a higher sensitivity for identifying excursions. In the mid-term postoperative period (8-14 days after surgery): it is assumed that the standard deviation of the excursion of patients in this period is generally between 0.10m and 0.18m, with a range of 0.25m to 0.40m, which is in the median 50% interval of the full data statistics. The fluctuation range is between the initial stage and the stable stage, so it is set , and moderately adjust the fluctuation characteristics. Postoperative stabilization stage (postoperative day 15 and later): Assume that the offset standard deviation is reduced to the range of 0.05m to 0.10m, and the range of values ​​is concentrated between 0.12m and 0.25m, which is the lowest 25% fluctuation range in the database. To avoid the Hurst exponent from over-responding to small disturbances, set , to reduce the sensitivity to micro-fluctuations and ensure that the index reflects the actual deviation trend rather than data noise. Assuming that a patient is on the 5th day after surgery, the muscle strength assessment level is M2, and the activity ability is limited. The doctor assesses the gait deviation on site and finds that there is a visible continuous deviation trend. The measured standard deviation of the deviation is 0.24 meters, and the cumulative deviation range is 0.52 meters. The standard deviation is within the range of [0.18, 0.35] meters, and the range is within the range of [0.38, 0.70] meters, both of which belong to the preset early postoperative fluctuation interval. It is judged that it is in the early postoperative stage, which meets the set conditions. According to the predefined standard, the recovery period weight factor is reasonably set as , to enhance the identification ability of the Hurst index to the deviation trend in this stage.

[0062] The Hurst index of the change in deviation distance amplitude in the deviation time series is calculated using the Hurst index algorithm The formula is as follows:

[0063]

[0064] wherein is a dimensionless quantity, used to measure the fluctuation persistence or repeatability of the deviation time series, is a logarithmic function with base 10, used to express the power-law relationship, is the range of the centralized cumulative deviation sequence, with a unit of meters (m), is the standard deviation of the deviation time series, with a unit of meters (m), is the recovery period weight factor (dimensionless) used to adjust the Hurst index setting, is the length of the entire deviation time series, with a unit of seconds (s), representing the number of data points in the sequence.

[0065] The above data is brought into the Hurst index formula to calculate:

[0066]

[0067] The results show that the Hurst index of the change in deviation distance amplitude in the deviation time series is .

[0068] The Hurst index is compared with the preset reference index threshold to evaluate the stability of the patient's direction control during walking. The corresponding risk period in the deviation time series that is lower than the reference index threshold is marked to form a set of deviation risk segments;

[0069] The benchmark index threshold of HURST index is used as a reference standard for determining whether the gait deviation stability meets the standard. The setting basis is the historical HURST index data distribution in the training records of patients after orthopedic surgery in the clinical rehabilitation database, and the grading evaluation results of the rehabilitation experts on the gait control state are double calibrated.

[0070] Assuming that a plurality of patients after orthopedic surgery collected from the hospital rehabilitation department, the training trajectory data of each patient at the 15th day, the 30th day and the 45th day after surgery is extracted, the HURST index of the corresponding deviation time sequence is calculated, and more than 1000 groups of effective sample data are formed. Each group of training data is independently judged by two experienced rehabilitation doctors through video evaluation and functional scoring tool (such as Tinetti score table) on the same day whether the gait is stable, and stable is defined as the patient can independently control the gait direction, the number of obvious direction drift is not more than 2 times, and the average trajectory error is less than 2 meters. The HURST index of the training records in the above samples is extracted, and the statistical distribution range is mainly concentrated in [0.60, 0.78], the mean is 0.68, the median is 0.66, and the standard deviation is 0.045. The lower quartile (25th percentile) is 0.60. In order to ensure the identification sensitivity and avoid excessive misjudgment, the lower quartile 0.60 of the HURST index distribution in the stable state is set as the minimum stability limit, and is set as the benchmark index threshold. The HURST index below this value represents that the gait deviation trend fluctuation exceeds the normal range, and there may be a risk of unstable direction control or aggravation over time, therefore, the current benchmark index threshold is reasonably set as .

[0071] The core of the comparison process is to compare the HURST index calculated in the current patient deviation time sequence with the preset benchmark index threshold to identify whether there is a period of insufficient direction control stability. The whole process is divided into the following steps: assuming that the whole deviation time sequence is divided into continuous and non-overlapping analysis windows (for example, each segment is 60 seconds long), ensure that each segment of data has enough sample number (such as n=60) for calculating HURST index, and guarantee the statistical stability. The setting of the window length is based on the previous experience analysis, which has been verified that the trajectory change in 60 seconds can better reflect the small-scale trend fluctuation characteristics. The HURST index of the deviation distance sequence in each window is calculated independently, and the corresponding time period number and its start and end time stamp are recorded. The numerical comparison of each segment HURST index and the benchmark index threshold : if : it is judged that the gait direction control has trend stability, and is not marked as risk; if : it is judged that there is a risk of unstable gait direction control, and needs to be marked. All The time period number and the timestamp are integrated to form a set of offset risk segments. The segment set is composed of time period start and end time, Hurst value, recovery period label, etc., and can be used for retrospective analysis and subsequent training program adjustment. For example, the current training process is 300 seconds, divided into 5 windows, each 60 seconds: suppose the calculated , , , , ; then the Hurst index in windows 3 and 4 is lower than the benchmark value 0.60, and the time range of 120-180 seconds and 180-240 seconds is identified as a risk period; the final offset risk segment set is recorded as: ; this comparison process not only ensures continuous analysis on the time scale, but also retains the quantitative judgment ability of gait offset trend, making the identification of risk time period more targeted and accurate, providing key reference for the rehabilitation training monitoring system.

[0072] By continuously counting the spatial deviation of the patient's movement coordinates per second from the ideal path, and using the deviation amplitude change trend for index analysis, the time period of unstable direction control during walking can be dynamically identified, potential gait abnormalities or fall risks can be detected in advance, and the perception accuracy of behavior deviation and the timeliness of nursing intervention during rehabilitation training can be improved.

[0073] The steps for obtaining the local abnormal interval of heart rate are as follows:

[0074] The heart rate of the patient during daytime rehabilitation training is collected by the wearable device on the patient's body, and the continuous heart rate in each sliding window is extracted to form a heart rate information set.

[0075] The heart rate data of the patient during daytime rehabilitation training is collected by the intelligent wearable terminal on the patient's body in real time. The photoplethysmogram sensor built-in the terminal can continuously obtain heart rate change information. All collected data are transmitted to the edge processing unit synchronously. The heart rate sequence is processed using Matlab software. The original heart rate sequence is segmented according to the set sliding window parameters, including window length and sliding step. The program automatically traverses the original data and constructs multiple continuous subsequences to form a sliding window sequence set. Then, the heart rate data in each window is classified, numbered, and arranged in parallel using the Pandas library in Python. Finally, the heart rate information set is output for subsequent density analysis and abnormality detection operations.

[0076] Select each time point and the heart rate information of the adjacent region before and after the target time point in the heart rate information set in turn, map all the heart rate information to a multidimensional density space, and calculate the local density of each time point in the corresponding adjacent region in the multidimensional density space by using a local anomaly factor algorithm;

[0077] First, the heart rate data is reconstructed by a sliding window in chronological order. Each time point corresponds to a fixed-dimension heart rate vector, which represents the continuous heart rate change characteristics within several minutes before and after the time point. Then, taking each time point as the center, the heart rate vector thereof is combined with the heart rate vectors of the surrounding time points to form a data set. The spatial embedding of the multidimensional vector is completed by MATLAB, i.e., all the heart rate vectors are represented as data points in a unified multidimensional Euclidean space, so as to realize the spatial expression of the heart rate information.

[0078] The local density of each time point in the multidimensional density space is calculated The formula of the local density of each time point in the corresponding adjacent region is as follows:

[0079]

[0080] Among them, represents the target time point in the multidimensional density space The local density of the corresponding adjacent region, is the target time point currently analyzed in the multidimensional density space, and the heart rate vector thereof is derived from the sliding window in the heart rate information set, is the target time point in the multidimensional density space The corresponding adjacent region set, represents the set composed of the closest time points in the time sequence before and after the target time point, and the size of the neighborhood is Through experimental setting (for example, selecting to represent the nearest 3 time points), represents the neighborhood point compared with the target time point in the adjacent region set, , represents the traversal summation of all neighborhood points in the adjacent region set, and the traversal variable is , and the summation object is each , is the reachable distance from the target time point to the neighborhood point , which is defined as: , represents the Euclidean distance between the target time point and the neighborhood point , which is the distance between two heart rate sliding window vectors in the multidimensional space, represents the neighborhood point the first nearest neighbor Euclidean distance, which means calculating the distance with other points in the dataset and sorting them, and then selecting the distance.

[0081] Assume that the original 5-dimensional sliding window vector of each heart rate time point has been standardized by Z-score standardization, which is calculated based on the mean and standard deviation of the daily heart rate full sample data. After standardization, all data are dimensionless values, with a center of 0 and a standard deviation of 1.

[0082] Define the heart rate vector of the analysis target time point and the neighborhood point: assume that the target time point is the standardized heart rate vector , assume that the neighborhood point is the standardized heart rate vector , assume that the neighborhood point is the standardized heart rate vector , assume that the neighborhood point is the standardized heart rate vector , where the neighborhood size is .

[0083] Calculate the Euclidean distance between the target time point and each neighborhood point in the multidimensional density space:

[0084]

[0085] where is the standardized heart rate vector of the target time point , is the standardized heart rate vector of the neighborhood point , is the standardized heart rate vector index, the value range is , represents the heart rate value, is the dimension of the standardized heart rate vector, which is equal to the width of the sliding window, and the acquisition method is to set the sliding window time range (such as 5 minutes), is the difference between the standardized heart rate vectors of the target time point and the neighborhood point. Calculate : the Euclidean distance

[0086] between the target time point and the neighborhood point ; ​​​

[0087] Compute : Euclidean distance between target time point and neighborhood point .

[0088] Compute : Euclidean distance between target time point and neighborhood point .

[0089] Results show that, , , .

[0090] Determine the first neighbor distance of each neighborhood point :

[0091] The first neighbor distance formula is: ;

[0092] Compute the Euclidean distance between neighborhood point and each time point except neighborhood point :

[0093]

[0094] where, is the first normalized heart rate vector in each time point except neighborhood point .

[0095] Set in neighborhood size , take the second smallest Euclidean distance between each point and other points as its first neighbor distance.

[0096] Compute : Euclidean distance between current neighborhood point and target time point ;

[0097] Compute : Euclidean distance between current neighborhood point and neighborhood point ;

[0098] Compute : Euclidean distance between neighborhood point and neighborhood point ; ​​​​​​

[0099] Sort and take the 2nd smallest value: .

[0100] Compute : Euclidean distance between neighborhood point and target time point ; ;

[0101] Compute : Euclidean distance between neighborhood point and neighborhood point ; ;

[0102] Compute : Euclidean distance between neighborhood point and neighborhood point ; ;

[0103] Sort and take the 2nd smallest value: .

[0104] Compute : Euclidean distance between neighborhood point and target time point ; ;

[0105] Compute : Euclidean distance between neighborhood point and neighborhood point ; ;

[0106] Compute : Euclidean distance between neighborhood point and neighborhood point ; ;

[0107] Sort and take the 2nd smallest value: .

[0108] The results show that , , .

[0109] It is known that , , ; , , .

[0110] According to the formula:

[0111] Compute : target time point With neighboring points The reachable distance ;

[0112] calculate : Target time point With neighboring points The reachable distance ;

[0113] calculate : Target time point With neighboring points The reachable distance .

[0114] Calculate target time point The local density in the corresponding adjacent area ; The results show that in the standardized heart rate data, the target time point The local density of the neighborhood is .

[0115] Calculate the first Neighborhood points To the neighboring point except the target The local density of the adjacent area :

[0116]

[0117] in, is the neighborhood size, is the gth neighborhood point in the multidimensional density space The corresponding set of adjacent regions, Indicates the adjacent region set with Neighborhood points Other time points for comparison The heart rate vector in , It is Neighborhood points To the corresponding time point The reachable distance is defined as: , It is Neighborhood points With other time points The Euclidean distance between Indicates other time points No. The Euclidean distance of the nearest neighbor is After calculating the distances with other points in the dataset and sorting them, select the one ranked highest. distance, is the neighborhood point index.

[0118] According to the formula ,for :where z is 、 、 ; Get : Neighborhood points Target time point The Euclidean distance between Target point The set of Euclidean distances to other points is: ; After sorting, take the second smallest value: ; According to the formula: .

[0119] Get : Neighborhood points With neighboring points Euclidean distance between ; :point The distance to other points is: ; After sorting, take the second smallest value: ; According to the formula: .

[0120] Get : Neighborhood points With neighboring points Euclidean distance between ; :point The distance to other points is: ; After sorting, take the second smallest value: ; According to the formula: .

[0121] calculate : Target neighborhood point The local density of: .

[0122] for :where z is 、 、 ; Get : Neighborhood points With the target point The Euclidean distance between: ; Get : Target point The Euclidean distance to other points is: ; After sorting, take the second smallest value: ; Substitute into the formula: .

[0123] Get : neighborhood points Euclidean distance between neighborhood points : neighborhood points ; get : distance between point and other points ; take the second smallest value after sorting ; substitute into the formula .

[0124] get : neighborhood points Euclidean distance between neighborhood points : neighborhood points ; get : distance between point and other points ; take the second smallest value after sorting ; substitute into the formula .

[0125] calculate : local density of neighborhood points .

[0126] for : where z is , , ; get : Euclidean distance between neighborhood points and target point : neighborhood points ; get : ; substitute into the formula .

[0127] get : Euclidean distance between neighborhood points and neighborhood points : neighborhood points ; get : ; substitute into the formula .

[0128] get : Euclidean distance between neighborhood points and neighborhood points : neighborhood points ; get : ; substitute into the formula .

[0129] calculate : local density of neighborhood points .​ .

[0130] Local outlier factor for measuring the relative ratio of the density of the target point to its neighborhood points , defined as follows:

[0131]

[0132] Where: , , , , is the neighborhood size, is the neighborhood point index.

[0133] Each ratio is calculated term by term: , , .

[0134] Result: ;

[0135] The results show that the target time point has a local density of in the corresponding adjacent area.

[0136] LOF (Local Outlier Factor) is an anomaly detection method based on density ratio. Its core logic is not directly dependent on an absolute threshold or distance size, but rather on comparing the local density difference between the target point and its neighborhood to determine whether it is abnormal. First, the Euclidean distance is used to obtain the reachable distance between the target point and the neighborhood points, and then the density of each point in its local space (i.e., local reachable density LRD) is reflected. Next, LOF compares the density of the target point with the average density of the neighborhood points: if the target point density is significantly lower than the average neighborhood density (i.e., LOF value greater than 1), it is identified as a local outlier point; otherwise, if LOF is close to or less than 1, it indicates that the point is in a consistent or crowded area, and is a normal point. The effect is to identify time points that are inconsistent with the surrounding environment density, thereby accurately identifying local abnormal segments in the heart rate sliding sequence, rather than relying solely on global statistical features or single-point fluctuations.

[0137] Compare the local density with the preset density threshold to filter out local abnormal time points with a local density lower than the density threshold, and mark consecutive local abnormal time points as heart rate local abnormal intervals;

[0138] Specifically, the LOF value is equal to the ratio of the average local density of the neighborhood points to the local density of the current point itself. If the density of a point is much lower than the density of its surrounding points, then this ratio will be greater than 1, otherwise it will be less than 1. Therefore: LOF=1 means that the density of the point is comparable to that of its surrounding neighbors, and it is an ordinary point in the density structure; LOF>1 means that the density of the point is less than the average density of the neighborhood, and it may fall in a relatively sparse or isolated location, and there is a risk of local anomaly; LOF<1 means that the point is in a dense area, its density is higher than that of the neighborhood, and it is usually not considered an anomaly. Based on this feature, when the LOF value of a certain time point is greater than 1, it means that its relative density is low, and it can be regarded as a potential local anomaly time point. By scanning the entire time series, all time points that meet LOF>1 are selected and preliminarily used as candidate points for local anomalies. These points may correspond to mutations, disorders or other atypical fluctuations in the heart rate signal. Furthermore, in order to enhance the stability of detection, these discrete anomaly points can be connected and analyzed in the time dimension: if multiple anomaly points are adjacent in time or the intervals are very short, they can be merged into a continuous local heart rate anomaly interval and uniformly marked as a local heart rate anomaly interval.

[0139] By performing sliding window processing on continuous heart rate data during rehabilitation training and combining it with multidimensional density analysis and local abnormality factor algorithm, it is possible to accurately identify continuous abnormal fluctuations in heart rate within a local time period, effectively distinguish short-term abnormalities from physiological fluctuations in a stable state, improve the ability to identify patients' stress reactions or potential abnormal physiological states, and enhance the sensitivity and discrimination accuracy of heart rate monitoring.

[0140] The specific steps for obtaining the heart rate fluctuation classification set are:

[0141] The duration of the local abnormal heart rate interval in each time period is counted and compared with the set time threshold. The local abnormal heart rate interval above the time threshold is marked as a continuous fluctuation segment, and the local abnormal heart rate interval below the time threshold is marked as an instantaneous fluctuation segment.

[0142] According to the acquired local abnormal time point and its specific timestamp position in the heart rate time series, a time interval array is established ,in, Indicates the Sampling time points that are judged as local anomalies are traversed in sequence The interval difference between the time points in the middle, if there is an interval between adjacent time points is a sampling period , then merge them into the same local abnormal heart rate interval and record the starting timestamp of each local abnormal heart rate interval With end timestamp For each local abnormal heart rate interval, its duration is calculated according to the following formula: ;in, Indicates the The duration of a local abnormal heart rate interval (in seconds), For a single sampling period (as read through the device configuration ), and They are the start and end timestamps (unit: seconds) of the local abnormal heart rate interval. The duration of all segments constitutes a duration set in turn, and then the time threshold set in the system parameters is extracted. (For example, obtained from clinical experience or statistical analysis of training samples, the unit is also seconds). Perform the comparison operation when the following conditions are met: ; then the The segment is marked as continuous fluctuation, otherwise it is marked as instantaneous fluctuation. The comparison process is completed in the segment index loop. The segment marking result and time index are saved in the array mapping through the key-value pair structure for subsequent heart rate fluctuation type classification processing.

[0143] Classify and mark the segment types corresponding to the continuous fluctuation segment and the instantaneous fluctuation segment to establish a heart rate fluctuation classification set;

[0144] Extract the time range and corresponding fluctuation type label of each local abnormal segment, build a segment classification comparison table, traverse each item in the comparison table, read the start timestamp and end timestamp of the current segment, and classify all heart rate data segments in the time interval according to the fluctuation label type corresponding to the segment. If the segment label is continuous fluctuation, the corresponding data segment is added to the continuous fluctuation classification set. If the segment label is instantaneous fluctuation, its data segment is classified into the instantaneous fluctuation classification set. The data segment needs to be numbered each time it is classified to mark its corresponding segment number and source. After completing all segment traversal and data classification, the two fluctuation sets are assembled into a unified data set structure according to the structure definition method, named heart rate fluctuation classification set.

[0145] By grading the duration of local abnormal heart rate intervals and dividing them into two categories: continuous fluctuations and instantaneous fluctuations, we can achieve classified management of different types of heart rate abnormalities, help distinguish short-term emotional fluctuations from potential pathological states, provide a stratified basis for subsequent nursing responses, and improve the accuracy of abnormal heart rate treatment and the targeted intervention.

[0146] The specific steps for obtaining blood pressure trend analysis results are as follows:

[0147] The continuous systolic pulse waveform signal of the patient after the operation at night is collected by the blood pressure monitoring device, and is divided into multiple equal-length time signal windows. For each systolic pulse waveform in the time signal window, a continuous signal curve is constructed with the corresponding time as the horizontal axis and the systolic pulse as the vertical axis;

[0148] The continuous systolic pulse waveform signal of the patient after the operation at night is collected by the blood pressure monitoring device, and is divided into multiple equal-length time signal windows. For each systolic pulse waveform in the time signal window, a continuous signal curve is constructed with the corresponding time as the horizontal axis and the systolic pulse as the vertical axis;

[0149] The continuous signal curve is subjected to Hilbert transform to analyze the instantaneous frequency and phase trajectory in each systolic pulse corresponding period, analyze the rhythm change in the systolic pulse corresponding period according to the instantaneous frequency and phase trajectory, identify and calibrate the period segment with rhythm decline and no rebound trend, and generate blood pressure trend analysis result;

[0150] After preprocessing the continuous signal curve, the Hilbert transform operation is performed using the SciPy library in Python. The transformation process converts the original real-valued systolic pulse waveform signal into a complex analytical signal, where the real part is the original signal itself and the imaginary part is the Hilbert transform component corresponding to the original signal. The analytical signal generated after the transformation can further extract the instantaneous frequency and phase trajectory. The instantaneous frequency reflects the speed change of the pulse signal oscillation per unit time, and the unit is Hertz. The instantaneous phase indicates the phase position of the waveform on the time axis, and the unit is radian. SciPy obtains the phase sequence and the corresponding instantaneous frequency sequence by performing angle function and derivative processing on the analytical signal. Statistics on the instantaneous frequency changes of each signal segment can identify the segments where the frequency decreases over time. After the Hilbert transform is completed, the instantaneous frequency obtained represents: the speed of oscillation of the systolic pulse signal at each moment in unit time, and its numerical unit is Hz, which corresponds to the number of pulse waveform cycles contained in each second; and the phase trajectory represents: the instantaneous phase corresponding to the signal at each moment, in radians, which is an indicator of the position of the waveform evolution on the time axis, and is often used to describe rhythmic progress and waveform alignment. For example, assuming that in the analysis of blood pressure waveform data, after applying the Hilbert transform to each signal window segment, its corresponding instantaneous frequency and phase sequence are extracted. For rhythm change analysis, the system sets the physiologically reasonable range of instantaneous frequency to be , that is, the normal heart rhythm frequency should be between 0.5 and 2.0 times per second. Rhythm maintenance segment: If the time interval The instantaneous frequency sequence is kept within range, and the fluctuation amplitude does not exceed ±0.1Hz. At the same time, the phase trajectory increases linearly with time, and its first-order derivative remains in the positive range, indicating that this segment is a rhythm stable segment; rhythm fluctuation segment: if in the time period In the period, the instantaneous frequency has multiple ups and downs, and there is an interval rebound, such as: from 1.4Hz to 1.0Hz, and then rises back to 1.3Hz. Combined with the phase trajectory, there are multiple inflection points, which is determined to be a rhythm fluctuation segment; rhythm decline segment (calibration object): in the time period The instantaneous frequency continues to decrease, and the initial value is , the final value is , and there is no more than The rising point (according to the preset rebound judgment threshold) is considered to have no rebound trend, and the phase trajectory keeps increasing continuously in the whole interval without phase reversal or interruption, and its derivative is always greater than zero but the absolute value gradually decreases. Then, this section is marked as a periodic segment with a declining rhythm and no rebound trend according to the standard.

[0151] By time segmentation and signal modeling of the continuous systolic pulse waveform signal at night, and by introducing instantaneous frequency and phase trajectory analysis, abnormal periods with rhythm decline and no recovery trend can be identified, effectively capturing early signs of reduced blood flow dynamics in a static state, and improving the identification ability and early warning value of night blood pressure monitoring for chronic hidden risks.

[0152] The acquisition step of the venous return abnormality non-response record is specifically:

[0153] According to the blood pressure trend analysis result, it is inferred whether there is a decrease in lower limb venous return in the target period, wherein the period with rhythm decline and no rebound trend corresponds to a stable and slow process of continuous blood flow dynamics reduction without being interrupted by body position change, indicating a decrease in venous return efficiency in a long static state, and a lower limb venous return analysis result is obtained.

[0154] The period labeled as rhythm decline and no rebound trend is selected as the target paragraph, and according to the continuous decline of instantaneous frequency and the linear evolution state of phase trajectory in this paragraph, it is inferred that this section represents a blood flow dynamics slow process that is not interrupted by body position change. In specific analysis, if the instantaneous frequency value in a certain period decreases from 1.5 Hz to 0.7 Hz, the system detects the derivative sequence of the continuous fluctuation, sets the standard for judging the downward trend as the frequency value in not less than 5 consecutive windows satisfies the negative derivative, and the maximum frequency rise in this section does not exceed 0.05 Hz. The threshold is determined according to the normal fluctuation range of postoperative night blood pressure rhythm, i.e. the normal frequency fluctuation range should not exceed ±0.1 Hz. If the actual fluctuation amplitude is less than the lower limit of this range, it is considered to have no significant rebound. At the same time, the phase trajectory derivative always remains positive and has no local slope mutation, and the change curvature fluctuation does not exceed 0.05 rad / s², which is the standard deviation statistical result of the fluctuation stable section in the same batch of resting samples. If the target period meets the above two standards of instantaneous frequency decline and phase trajectory stability, and the duration of the slow state is greater than 120 seconds, which is the minimum length required for the clinical standard to maintain a non-active body position that may affect venous return efficiency, it is inferred that there is a risk of reduced venous return efficiency in this paragraph. The system accordingly labels this period as a suspected state section of reduced venous return.

[0155] Based on the lower limb venous return analysis result, through the operation data in the same time period in the linked nursing record, it is identified whether there is a nursing intervention of elevating the lower limbs or guiding the body position. If no nursing intervention is performed, the target period is marked again, and a venous return abnormality non-response record is established.

[0156] The start and end times of the period segment determined to have slow venous return are read, and the corresponding patient's daily nursing operation record data is called, the operation entries with overlapping record times and target period segments are screened out, the record time field in the nursing record is compared with the target period segment one by one, it is judged whether the record time is between the start and end times of the period segment, the matched record content field is extracted, and it is identified whether there is an operation description containing "elevate lower limbs", "body position adjustment", "leg elevation pad", "pad high double lower limbs" and the like in the identification process, the identification process is matched by a preset keyword dictionary, if the matching is successful, it is considered that there is nursing intervention behavior in the period segment, and no further processing is done, if no nursing intervention keyword record is identified in the whole period segment, the period segment is determined to be "venous return abnormality not responded", the system marks the period segment again as abnormal, adds it to the abnormality non-response record set, and records the period segment number, corresponding timestamp and non-matched nursing intervention field content for subsequent analysis and responsibility archiving.

[0157] By analyzing the period with decreased blood pressure rhythm and no rebound, the possible problem of slow venous return of lower limbs of the patient in the static state is identified, and whether the intervention measures such as leg lifting or body position guidance are implemented is judged by combining the nursing record in the same period, so as to mark the risk period not responded, improve the identification accuracy of night venous return abnormality and the perception ability of nursing deficiency, and help to prevent complications such as thrombosis.

[0158] The acquisition steps of the multi-maintenance nursing path recommendation result are specifically:

[0159] According to the offset risk segment set, the daytime moving path setting and guiding mode recommendation are performed, including wall leaning, setting boundary auxiliary objects, and arranging accompanying personnel to guide movement within a specified distance;

[0160] The start and end time information of all time periods marked as having risk offset are extracted, and the monitoring room number and patient number information corresponding to the corresponding segment are obtained, the corresponding room structure drawing in the field environment layout data is called and loaded with the current bed number and traffic route information of the patient, the initial moving trajectory line segment is generated according to the actual positioning point of each offset risk segment, and the potential offset direction angle is marked, according to the relationship between the direction angle and the room boundary, it is judged whether it is close to the door, the corridor, the corner and other boundary areas, if the direction points to the open area, the auxiliary recommendation is set to "arrange accompanying personnel to follow within one meter from the target position", if the direction is close to the wall area, the auxiliary mode is set to "move along the wall", if the direction points to the bed boundary position, it is set to "place auxiliary objects at the boundary", at the same time, the time tag, target point coordinate, recommendation guiding mode and auxiliary explanation of each offset risk path are recorded in the system, and finally the guiding mode suggestion summary table of all offset segments is output as the auxiliary recommendation content of daytime path setting.

[0161] According to the heart rate fluctuation classification set, the instantaneous fluctuation segment is regarded as a non-continuous reaction, no nursing plan adjustment is made, and the continuous fluctuation segment is regarded as the presence of pain state, and the cold intervention prompt and nursing patrol video are recommended to be increased;

[0162] Extract all time period numbers and corresponding start and end times marked as instantaneous fluctuations in the classification set, and directly label this type of segment as a reference area that does not require intervention, skip the subsequent nursing configuration processing flow, and then extract all segments marked as continuous fluctuations in the classification set, count the duration of each segment and the corresponding patient number, call the cold intervention suggestion rule, and determine whether the set prompt condition is met. The condition is that the continuous fluctuation duration is greater than or equal to 180 seconds. This threshold is obtained by analyzing the correspondence between the postoperative patient self-reported pain and continuous heart rate changes. If the condition is met, the current time period is marked as a cold prompt time period, and the patient's past daily patrol frequency record in the nursing log is searched. Compare whether the daily patrol interval is more than 30 minutes. If the patrol interval is greater than the time length, add a suggestion item to the record table, prompt to increase the nursing patrol operation once before and after the start time of the current fluctuation segment, and write the cold prompt and nursing patrol adjustment suggestions into the nursing plan suggestion list for manual review and confirmation.

[0163] According to the abnormal venous return non-response record, recommend increasing the time point setting of lifting the lower limbs or guiding the turning over in the marked night target period segment, obtain the multi-maintenance nursing path recommendation result, and provide decision support for the auxiliary selection of the nursing plan of orthopedic patients;

[0164] Read the start and end times of all marked night target period segments, and process them by patient number grouping. Arrange the abnormal period segments of each patient in chronological order, establish a period segment list and filter out duplicate labels, call the preset nursing intervention time interval standard, which stipulates that between the night period 23:00 and 05:00, if there is no body position adjustment record for more than 120 minutes and there is an abnormal non-response period segment, an operation of lifting the lower limbs or guiding the turning over needs to be added in this segment. According to this standard, the system calculates the duration of each period segment and compares it with the interval before and after. If the interval between the current period segment and the previous segment is greater than 60 minutes, a suggestion record of lifting the lower limbs is inserted at the start time of the current period segment. If the duration of the current period segment itself is greater than 180 seconds, a suggestion record of guiding the turning over is inserted at the midpoint of the period. Encapsulate all generated time points, operation types, and corresponding patient numbers into a recommended action set and write it into the nursing path suggestion file to form the multi-maintenance nursing path recommendation result.

[0165] Based on the results of multi-source data such as gait deviation, heart rate fluctuation type and venous return abnormalities, specific nursing recommendations are formulated, such as adjusting walking guidance methods, increasing the frequency of cold compresses and inspections, arranging nighttime turning interventions, etc., to achieve personalized, time-based and targeted recommendations for nursing plans, enhance the matching degree between nursing measures and patient status, and improve the effectiveness of nursing interventions and the initiative of risk response.

[0166] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A nursing plan auxiliary selection system for orthopedic patients, characterized by: The system includes: The gait deviation monitoring module collects the patient's movement path during daytime rehabilitation training, compares it with the ideal movement path, and extracts the deviation distance between the paths based on the comparison results to form a deviation time series. The risk time period is screened by calculating the Hurst exponent of the deviation distance amplitude change in the deviation time series to form a deviation risk segment set. The heart rate anomaly recognition module collects the patient's heart rate during daytime rehabilitation training and maps it to a multidimensional density space. Using a local anomaly factor algorithm, it marks the consecutive local abnormal time points of the heart rate in the multidimensional density space as local abnormal heart rate intervals. The heart rate fluctuation classification module divides the interval types according to the duration of the local abnormal heart rate interval in each time period and establishes a heart rate fluctuation classification set; The blood pressure trend analysis module collects the patient's continuous systolic blood pressure pulse waveform signal at night after surgery, analyzes the rhythm changes of the systolic blood pressure pulse within a specified period, and generates blood pressure trend analysis results; The postural intervention discrimination module links the operation data of the same time period in the nursing records based on the blood pressure trend analysis results, marks the target period segments where no nursing intervention has been performed, and establishes a record of abnormal venous return and no response; The rehabilitation guidance module recommends corresponding nursing plans to patients based on the deviation risk segment set, heart rate fluctuation classification set, and venous return abnormality unresponsiveness record, and obtains multi-dimensional nursing path recommendation results; The steps for obtaining the blood pressure trend analysis results are specifically as follows: The patient's continuous systolic blood pressure pulse waveform signal is collected through a blood pressure monitoring device during the night after surgery and divided into multiple equal-time signal windows. For each systolic blood pressure pulse waveform in the time signal window, a continuous signal curve is constructed with the corresponding time as the horizontal axis and the systolic blood pressure pulse as the vertical axis. Performing a Hilbert transform on the continuous signal curve to analyze the instantaneous frequency and phase trajectory within the corresponding cycle of each systolic blood pressure pulse, analyzing the rhythm changes within the corresponding cycle of the systolic blood pressure pulse based on the instantaneous frequency and phase trajectory, identifying and calibrating the periodic segments with a rhythm decline and no rebound trend, and generating a blood pressure trend analysis result; The specific steps for obtaining the record of abnormal venous return without response are as follows: Based on the blood pressure trend analysis results, it is inferred whether the patient has a slowing of lower limb venous return in the target cycle segment. The cycle segment with a declining rhythm and no rebound trend corresponds to a stable stagnation process in which the hemodynamics continues to weaken and is not interrupted by changes in body position, indicating that the venous return efficiency is reduced in a long-term static state, and the lower limb venous return analysis results are obtained; Based on the lower limb venous return analysis results, by linking the operation data of the same time period in the nursing records, it is identified whether there is nursing intervention such as raising the lower limbs and posture guidance. If no nursing intervention is performed, the target cycle segment is marked again and a record of abnormal venous return and no response is established.

2. The nursing plan auxiliary selection system for orthopedic patients according to claim 1 is characterized in that: The steps for obtaining the offset risk segment set are specifically as follows: The patient's movement path during daytime rehabilitation training is collected through wearable devices on the patient. The latitude and longitude coordinate points recorded every second in the movement path are converted into a continuous spatial coordinate sequence with equal intervals. The ideal movement path of the rehabilitation training plan is preset based on the patient's current postoperative recovery period. The deviation distance of the coordinate points in the continuous spatial coordinate sequence at the same time relative to the corresponding position coordinate points on the ideal movement path is calculated and recorded; The deviation distance at each time is counted to form a migration time series, and the Hurst exponent algorithm is used to calculate the Hurst exponent of the deviation distance amplitude change in the migration time series by analyzing the power law relationship between the variation amplitude of the deviation distance at different times in the migration time series and the length of the entire migration time series; The Hurst index is compared with a preset benchmark index threshold to evaluate the stability of the patient's directional control during walking, and the corresponding risk time periods in the deviation time series that are lower than the benchmark index threshold are marked to form a deviation risk segment set.

3. The nursing plan auxiliary selection system for orthopedic patients according to claim 1 is characterized in that: The steps for obtaining the local abnormal heart rate interval are specifically as follows: The wearable device on the patient collects the patient's heart rate every minute during the daytime rehabilitation training, divides it into fixed sliding windows, and extracts the continuous heart rate within each sliding window to form a heart rate information set; Selecting heart rate information at each time point and the heart rate information of adjacent areas before and after the target time point in the heart rate information set, mapping all heart rate information into a multidimensional density space, and using a local anomaly factor algorithm to calculate the local density of each time point in the corresponding adjacent area in the multidimensional density space; The local density is compared with a preset density threshold, local abnormal time points where the local density is lower than the density threshold are screened, and consecutive local abnormal time points are marked as local abnormal heart rate intervals.

4. The nursing plan auxiliary selection system for orthopedic patients according to claim 1, characterized in that: The steps for obtaining the heart rate fluctuation classification set are specifically as follows: The duration of the local abnormal heart rate interval in each time period is counted, and the counted duration is compared with the set time threshold. The local abnormal heart rate interval above the time threshold is marked as a continuous fluctuation segment, and the local abnormal heart rate interval below the time threshold is marked as an instantaneous fluctuation segment; Classification and labeling are performed according to the segment types corresponding to the continuous fluctuation segment and the instantaneous fluctuation segment to establish a heart rate fluctuation classification set.

5. The nursing plan auxiliary selection system for orthopedic patients according to claim 1 is characterized in that: The steps for obtaining the multi-dimensional management path recommendation result are specifically as follows: Based on the set of excursion risk segments, recommendations are made on daytime movement path settings and guidance methods, including leaning against walls, setting up boundary aids, and arranging accompanying personnel to guide movement at designated distances; According to the heart rate fluctuation classification set, transient fluctuation segments are considered non-sustained reactions and no nursing plan adjustment is made. Sustained fluctuation segments are considered to be painful and additional cold compress intervention reminders and nursing rounds are recommended. Based on the record of abnormal venous return without response, it is recommended to add the time point setting of raising the lower limbs or guiding turning over within the calibrated night target cycle period, and obtain the multi-faceted care path recommendation results to provide decision support for auxiliary selection of care plans for orthopedic patients.

Images