Call distribution method and system based on nursing needs

Abstract

The present application relates to the technical field of nursing information management, in particular to a call distribution method and system based on nursing needs, comprising the following steps: collecting bedside heart rate, blood oxygen and blood pressure measurement values and constructing an abnormal trigger identifier based on a threshold value, calling a Boolean logic function to generate an emergency code for a patient bed call, calculating the remaining time of a single item based on the task start time to obtain a task load value, generating a response candidate path based on the spatial mapping node and the topological connection relationship, and performing multi-dimensional weighted comparison to generate a nursing call task distribution sequence. In the present application, by real-time monitoring of abnormal signs and establishing a rigorous emergency classification, automatic classification of call requests is realized, dynamic assessment of resource load and passing cost weight is performed, and multi-dimensional weighted logic is used to optimize the task distribution sequence, solving the problems of information perception lag, unreasonable response sequence and resource scheduling imbalance in the traditional mode, and achieving the improvement of nursing response speed, utilization rate and accuracy in the ward.

Description

Technical Field

[0001] This invention relates to the field of nursing information management technology, and in particular to a call distribution method and system based on nursing needs. Background Technology

[0002] The field of nursing information management technology involves the reception, identification, transmission, and processing of patient nursing service requests within hospital wards. It encompasses core components such as ward call devices, bedside terminals, nurse station receiving devices, communication lines, and nursing task recording systems. The overall technical system typically revolves around patients issuing nursing request signals in their wards via bedside buttons, pull cord devices, or bedside touch terminals. The call terminals deployed in the wards transmit the electrical signals to the nurse station host device via the ward communication lines. The nurse station host displays the called bed number, call time, and basic patient information on a screen. On-duty nursing staff then manually judge and respond based on the displayed information, while simultaneously recording the call processing status in the nursing record terminal, thus forming a mechanism for the transmission and response of patient nursing request information in the ward.

[0003] The traditional call distribution method refers to the processing method of receiving and distributing nursing request information issued by patients in the hospital ward nursing call system. It is mainly for the task of receiving and distributing nursing calls from multiple patients in the ward at the same time. The traditional method usually installs a mechanical push-button call or pull-cord trigger switch at each bed. When the patient triggers the device, the bedside call terminal generates an electrical signal and transmits it to the host device at the nurse station through the ward wiring. The host at the nurse station arranges the corresponding bed number in the display terminal list according to the time sequence of the received signal and is accompanied by an audible prompt. The on-duty nursing staff manually checks the patient's basic nursing level record, bed distribution table and the on-duty nursing staff's duty area table according to the bed information on the display screen. Then, the nursing station staff assigns a specific nursing staff to the corresponding bed to handle the call through the intercom or manual notification. At the same time, the time of receipt, the person handling the call and the handling status are recorded in the paper nursing record sheet or nursing information terminal, thereby completing the process of receiving and distributing nursing call information in the ward.

[0004] Traditional dispatch methods rely on mechanical pressure to trigger a single call signal, with nurses manually grading the list displayed on the main unit. This results in a lag in the perception of patients' vital signs and an inability to capture potential clinical risk indicators in real time. Furthermore, in scenarios with multiple beds calling concurrently, prioritizing tasks based solely on time makes it difficult to prioritize urgent and critical care requests. In addition, the reliance on manual communication and coordination for nursing task allocation can easily lead to information transmission errors and uneven resource allocation, resulting in a low overall response efficiency in the ward. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a call distribution method and system based on nursing needs.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a call distribution method based on nursing needs, comprising the following steps: S1: Obtain heart rate, blood oxygen, and blood pressure measurements, compare them with preset vital sign thresholds, construct heart rate abnormality trigger flags, blood oxygen abnormality trigger flags, and blood pressure abnormality trigger flags, and obtain abnormal state identification information. S2: Based on the abnormal state identification information, call the Boolean logic function to perform logical analysis and calculation on the heart rate abnormality trigger flag, blood oxygen abnormality trigger flag and blood pressure abnormality trigger flag, determine the emergency level trigger score, and generate the bed call emergency code; S3: Obtain the task type classification field and task start time parameter, match the standard task processing time, use the current system time to subtract the task start time to construct the task execution time, subtract the standard task processing time from the task execution time to calculate the remaining time of a single task, accumulate the remaining time of a single task, and generate the task load value. S4: Obtain the corridor node coordinate array and node topology connection relationship, use the passage status of the node coordinates to correct the basic path length parameter to construct the corrected topology path length, use the corridor node coordinate array, node topology connection relationship and corrected topology path length to perform pathfinding analysis calculation, and generate candidate response paths for the nurse station; S5: Obtain the estimated processing time of the nursing call, combine it with the task load value, calculate the load capacity assessment score, map and compare the emergency code of the bed call, the load capacity assessment score and the path length value associated with the nurse station response candidate path, and generate a nursing call task distribution sequence.

[0007] The present invention improves upon the following: the emergency call code for hospital beds includes a red emergency response code, an orange priority intervention code, and a yellow routine standby code; the task load value includes a high-intensity saturation index, a medium load margin, and a slight idle weight; the nurse station response candidate path includes a main road direct corridor, a ward area backup passage, and a fire-prevention bypass corridor; and the nursing call task distribution sequence includes an initial emergency dispatch work order, a regular queuing queue, and a delayed observation and allocation list.

[0008] The present invention is improved in that the step of obtaining the abnormal state identification information is specifically as follows: S111: Monitor the communication messages transmitted by the bedside interface node, perform delimitation and interception processing on the communication messages, extract the heart rate measurement value, blood oxygen measurement value and blood pressure measurement value embedded in the communication messages, perform timestamp alignment and arrangement on the heart rate measurement value, blood oxygen measurement value and blood pressure measurement value, and establish a vital sign feature matrix. S112: Based on the preset vital sign thresholds, call the upper limit benchmark parameter and the lower limit benchmark parameter, and use the upper limit benchmark parameter and the lower limit benchmark parameter to perform difference operation on the heart rate measurement value, blood oxygen measurement value and blood pressure measurement value embedded in the vital sign feature matrix. Extract the out-of-bounds value sites associated with crossing the upper limit benchmark parameter or the lower limit benchmark parameter, perform logical labeling processing on the out-of-bounds value sites, and construct the heart rate abnormality triggering flag, blood oxygen abnormality triggering flag and blood pressure abnormality triggering flag. S113: Obtain the heart rate abnormality trigger identifier, blood oxygen abnormality trigger identifier and blood pressure abnormality trigger identifier, and perform timestamp alignment and sorting on the heart rate abnormality trigger identifier, blood oxygen abnormality trigger identifier and blood pressure abnormality trigger identifier to obtain abnormal state identification information.

[0009] The present invention is improved in that the step of obtaining the emergency call code for the hospital bed is specifically as follows: S211: Based on the abnormal state identification information, and based on the logical algebra operation rules, perform full-space logical intersection and combination operations on the heart rate abnormality triggering identifier, blood oxygen abnormality triggering identifier and blood pressure abnormality triggering identifier, monitor the linkage triggering coupling relationship of multiple categories of vital sign abnormality features, extract signal component vectors with discrete Boolean logic state features, and establish a multi-dimensional vital sign linkage logic matrix. S212: Obtain the preset level trigger judgment rule set, call the Boolean truth value of the logic site embedded in the multi-dimensional vital sign linkage logic matrix, count the total number of heart rate abnormality trigger markers, blood oxygen abnormality trigger markers and blood pressure abnormality trigger markers in the active state, perform a ladder mapping comparison judgment on the total number and the preset level trigger judgment rule set, extract the corresponding emergency level quantitative assessment level, and generate an emergency level trigger score. S213: Obtain the priority coding communication protocol, retrieve the preset level identifier character index library based on the emergency level trigger score, perform character sequence encapsulation processing on the emergency level trigger score, construct the priority code sequence, write the priority code sequence into the emergency priority control field position in the head of the bed call data packet, associate the corresponding bed equipment hardware identifier with the call logic weight level, establish the mapping relationship between call record data nodes and emergency codes, and generate the bed call emergency code.

[0010] The present invention is improved in that the step of obtaining the task load value is specifically as follows: S311: Monitor task management node messages, extract task type classification field and task start time parameter, perform pre-configured time mapping table retrieval for task type classification field, extract the standard task processing time associated with matching task type classification field, obtain system timestamp, perform time axis difference operation for system timestamp and task start time parameter, extract time difference parameter, and establish task execution time. S312: Subtract the standard task processing time from the task execution time to extract the time difference parameter, traverse the unfinished task sequence to perform a loop subtraction, extract the time difference parameter of each node to be processed, aggregate the extracted parameters to construct a countdown array, and obtain the remaining time of each task item. S313: Based on the nursing staff identification code, extract the remaining time of each task item in the pending work queue, perform a linear summation operation on the remaining time of all the tasks, extract the total time surplus feature parameter, perform a weighted mapping comparison between the total time surplus feature parameter and the preset carrying capacity benchmark value, establish a staff work leeway feature index, and generate task load values.

[0011] The present invention is improved in that the step of obtaining the candidate path response at the nurse station is specifically as follows: S411: Obtain the corridor node coordinate array and node topology connection relationship transmitted by the spatial mapping node, monitor the traffic status indicator value associated with the corridor node coordinate array, perform numerical quantization conversion processing on the traffic status indicator value based on the mapping result of the preset congestion level mapping table, extract the traffic resistance influencing factor, perform discrete interval normalization mapping on the traffic resistance influencing factor and extract the mapping result value, and establish a traffic weighted correction coefficient. S412: Obtain the local storage basic path length parameter, perform multiplication weighting operation on the toll weighting correction coefficient and the basic path length parameter, extract the path loss gain component, perform linear compensation summation adjustment on the basic path length parameter and the path loss gain component, determine the actual toll cost weight of each road segment, and construct the corrected topology path length. S413: Call the corridor node coordinate array, node topology connection relationship and corrected topology path length, perform weight assignment on the node topology connection relationship based on graph theory traversal rules, construct a weighted topology adjacency matrix, retrieve the starting node position of the nurse station and the target node position of the call source, perform multi-round node relaxation iterative optimization calculation on the weighted topology adjacency matrix, extract the connected node sequence with the minimum total weight score, perform spatial coordinate mapping processing on the connected node sequence, and generate candidate response paths for the nurse station.

[0012] The present invention is improved in that the step of obtaining the nursing call task distribution sequence is specifically as follows: S511: Obtain the task load value and the estimated processing time of nursing calls from the task management node; extract the local storage load weight coefficient, duration adjustment factor, and time decay correction constant; perform a product summation and weighted operation on the task load value, the estimated processing time of nursing calls, and the load weight coefficient; extract the real-time load characteristics of discrete nodes; perform nonlinear smoothing filtering on the load characteristics based on the time decay correction constant; map the processed data to a linear evaluation interval; and establish a load capacity evaluation score. S512: Based on the emergency code of the bed call, the load capacity assessment score and the path length value associated with the candidate response path of the nurse station, and combined with the simulated business dataset to obtain the equipment integrity benchmark value, the nurse qualification level score and the ward occupancy density coefficient, calculate the scheduling response priority assessment value of the task to be assigned, extract the coupling feature of task urgency and resource matching degree, perform discrete interval outlier removal for the scheduling response priority assessment value, and establish a task assessment feature vector; S513: Call the task evaluation feature vector, perform multi-dimensional weight normalization mapping on the embedded data components, extract standardized evaluation indicators, search the preset priority sorting index library, perform descending sorting comparison analysis on the standardized evaluation indicators, determine the execution order position of each nursing task in the global pending task pool, associate the execution order position with the corresponding ward node hardware address code, establish a call task and nursing human resource allocation relationship matrix, and obtain the nursing call task distribution sequence.

[0013] The present invention is improved in that the formula for calculating the scheduling response priority evaluation value of the task to be dispatched is specifically as follows: ; in, This represents the numerical value used to evaluate the priority of the scheduling response. The normalized value representing the emergency call code for hospital beds. The normalized value representing the path length. This represents the baseline value for equipment availability. Represents the ward occupancy density coefficient. The normalized value representing the nurse's qualification level score. The normalized value representing the qualification benchmark reference value, This represents the load capacity assessment score.

[0014] A call distribution system based on nursing needs, wherein the call distribution system based on nursing needs is used to implement the above-mentioned call distribution method based on nursing needs, the system comprising: The abnormal state analysis module acquires heart rate, blood oxygen, and blood pressure measurements, compares them with preset vital sign thresholds, constructs heart rate abnormality trigger flags, blood oxygen abnormality trigger flags, and blood pressure abnormality trigger flags, and obtains abnormal state identification information. The emergency level classification module, based on the abnormal state identification information, calls a Boolean logic function to perform logical analysis and calculation on the heart rate abnormality trigger flag, blood oxygen abnormality trigger flag and blood pressure abnormality trigger flag, determines the emergency level trigger score, and generates a bed call emergency code; The task load analysis module obtains the task type classification field and task start time parameter, matches the standard task processing time, uses the current system time to subtract the task start time to construct the task execution time, subtracts the standard task processing time from the task execution time to calculate the remaining time of each task, and accumulates the remaining time of each task to generate the task load value. The response filtering module obtains the corridor node coordinate array and node topology connection relationship, uses the passage status of the node coordinates to correct the basic path length parameter to construct the corrected topology path length, and uses the corridor node coordinate array, node topology connection relationship and corrected topology path length to perform pathfinding analysis calculation to generate nurse station response candidate paths; The task distribution and processing module obtains the estimated processing time of nursing calls, calculates the load capacity assessment score based on the task load value, and maps and compares the path length value associated with the emergency code of the bed call, the load capacity assessment score, and the candidate response path of the nurse station to generate a nursing call task distribution sequence.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by comparing bedside monitoring vital signs in real time and constructing multidimensional abnormal trigger markers, the automatic identification of clinical risk characteristics is achieved. By combining Boolean logic analysis to establish emergency codes for bedside calls, the problems of strong subjectivity in manual assessment and ambiguity in response priority are solved. At the same time, a dynamic assessment mechanism for task execution time and load capacity is introduced, and the path response path generated by spatial mapping nodes is used. By normalizing and mapping urgency, load, and passage cost and performing multidimensional weighted comparison, the optimal distribution of nursing resources from a global perspective is achieved, which effectively shortens the waiting time for critical care calls, reduces the workload of staff, and improves the accuracy and real-time response of automated collaboration in ward nursing service processes. Attached Figure Description

[0016] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a flowchart illustrating the process of obtaining abnormal state identification information according to the present invention; Figure 3 This is a flowchart illustrating the process of obtaining the emergency code for a hospital bed call according to the present invention; Figure 4 This is a flowchart illustrating how the present invention obtains task load values; Figure 5 This is a flowchart illustrating the process of obtaining candidate response paths for the nurse station according to the present invention; Figure 6 This is a flowchart illustrating the process of obtaining the nursing call task distribution sequence according to the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0019] Please see Figure 1 This invention provides a technical solution for a call distribution method based on nursing needs, comprising the following steps: S1: Obtain heart rate, blood oxygen, and blood pressure measurements through bedside monitoring equipment. Perform comparative calculations on the heart rate, blood oxygen, and blood pressure measurements based on preset vital sign thresholds to construct heart rate abnormality trigger flags, blood oxygen abnormality trigger flags, and blood pressure abnormality trigger flags, thereby obtaining abnormal state identification information. S2: Based on the abnormal state identification information, call the Boolean logic function to perform logical analysis and calculation on the heart rate abnormality trigger flag, blood oxygen abnormality trigger flag and blood pressure abnormality trigger flag, determine the emergency level trigger score based on the logical analysis and calculation results, and generate the bed call emergency code; S3: Obtain the task type classification field and task start time parameter through the task management node, call the standard task processing time matching the task type classification field, perform subtraction comparison between the current system time and the task start time parameter to calculate the task's executed time, subtract the standard task processing time from the task's executed time to calculate the remaining time of each task, and perform cumulative calculation on the remaining time of each task in the queue to generate the task load value. S4: Obtain the corridor node coordinate array and node topology connection relationship through spatial mapping nodes, monitor the passage status indicator value associated with the corridor node coordinate array, perform multiplication correction calculation on the basic path length parameter based on the passage status indicator value to construct the corrected topology path length, and input the corridor node coordinate array, node topology connection relationship and corrected topology path length into the Dijkstra algorithm to perform pathfinding analysis calculation to generate nurse station response candidate paths; S5: Obtain the estimated processing time of nursing calls, perform a weighted summation calculation on the task load value and the estimated processing time of nursing calls to construct a load capacity assessment score, perform a normalization mapping calculation on the emergency code of the bed call, the load capacity assessment score and the path length value associated with the candidate response path of the nurse station, perform a multi-dimensional weighted comparison calculation on the normalization mapping calculation result, and generate a nursing call task distribution sequence based on the multi-dimensional weighted comparison calculation result.

[0020] Heart rate abnormality trigger indicators include sinus tachycardia warning symbol, ventricular fibrillation warning code, and heart rate variability abnormality symbol. Blood oxygen abnormality trigger indicators specifically include hypoxemia status code, tissue hypoxia characteristic code, and peripheral circulation disorder indicator. Blood pressure abnormality trigger indicators specifically include systolic blood pressure trough warning code, diastolic blood pressure abnormality identifier, and pulse pressure systolic status code. Bedside emergency call codes include red emergency response code, orange priority intervention code, and yellow routine standby code. Task load values ​​include high intensity saturation index, medium load margin, and mild idle weight. Nurse station response candidate paths include main road direct corridor, ward area backup passage, and fire zone bypass corridor. Nursing call task distribution sequence includes initial rescue dispatch work order, regular queuing queue, and delayed observation and allocation list.

[0021] Please see Figure 2 The specific steps for obtaining abnormal state identification information are as follows: S111: Monitor the communication messages transmitted by the bedside interface node, perform delimitation and interception processing on the communication messages, extract the heart rate measurement value, blood oxygen measurement value and blood pressure measurement value embedded in the communication messages, perform timestamp alignment and arrangement on the heart rate measurement value, blood oxygen measurement value and blood pressure measurement value, and establish a vital sign feature matrix. The system connects to the bedside monitoring device's transmission port in real time, establishing a long-lived connection using the transmission control protocol, and receives device communication messages at 1000-millisecond intervals. It performs delimitation and truncation processing on the received binary data stream, reading the payload between the header identifier 0xAA and the tail identifier 0x55, and extracting the embedded physical feature fields. It reads heart rate measurement values ​​at offsets of 12 to 15 bytes, blood oxygen measurement values ​​at offsets of 16 to 19 bytes, and blood pressure measurement values ​​at offsets of 20 to 27 bytes, with the blood pressure data split into 4 bytes of systolic pressure and 4 bytes of diastolic pressure. The extracted hexadecimal data stream is converted to decimal floating-point numbers to obtain the raw vital signs parameters. A mean filtering mechanism with a sliding time window size of 5 is introduced to smooth and denoise the heart rate, blood oxygen, and blood pressure measurement values ​​over five consecutive sampling periods, removing abnormal high-frequency fluctuations that deviate from the current mean by more than 20%. The system invokes the high-precision clock service of the local system kernel to extract the current absolute system timestamp. It then performs timestamp alignment for the smoothed heart rate, blood oxygen, and blood pressure measurements. A time alignment tolerance interval of 50 milliseconds is set. If the difference between the original timestamp and the absolute system timestamp of the vital signs data extracted from different sensor channels falls within this tolerance interval, the data is determined to belong to the same synchronized time segment. A dynamic contiguous address space is allocated in system memory. Using the synchronized time segment as the column index and the heart rate, blood oxygen, systolic blood pressure, and diastolic blood pressure values ​​as row vectors, a 4-row, 120-column vital sign feature matrix is ​​constructed, cyclically storing continuous vital sign data from the past 120 seconds. Taking the No. 3 intensive care unit bed as an example, its heart rate measurement value at a specific synchronous time section is 85, blood oxygen measurement value is 98, systolic blood pressure value is 120, and diastolic blood pressure value is 80, corresponding to an absolute system timestamp of 1678886400000. This set of measured data is accurately written into the latest column position of the vital signs feature matrix.

[0022] S112: Based on the preset vital sign thresholds, call the upper limit benchmark parameter and the lower limit benchmark parameter, use the upper limit benchmark parameter and the lower limit benchmark parameter to perform difference operation on the heart rate measurement value, blood oxygen measurement value and blood pressure measurement value embedded in the vital sign feature matrix, extract the out-of-bounds value sites associated with the upper limit benchmark parameter or the lower limit benchmark parameter, perform logical labeling processing on the out-of-bounds value sites, and construct heart rate abnormality triggering flag, blood oxygen abnormality triggering flag and blood pressure abnormality triggering flag; The system retrieves preset vital sign threshold parameters configured in the storage medium, loading the upper limit benchmark parameter of heart rate (100) and the lower limit benchmark parameter of heart rate (60), the lower limit benchmark parameter of blood oxygen (94), the upper limit benchmark parameter of systolic blood pressure (140) and the lower limit benchmark parameter of systolic blood pressure (90), and the upper limit benchmark parameter of diastolic blood pressure (90) and the lower limit benchmark parameter of diastolic blood pressure (60). It then iterates through and reads the embedded heart rate, blood oxygen, and blood pressure measurements in the latest column vector of the vital sign feature matrix, performing independent and parallel difference calculations on these measurements using the upper and lower limit benchmark parameters. The difference calculation process is as follows: Heart rate difference = 85 - 100 = -15, heart rate difference = 85 - 60 = 25; Based on the difference calculation results, out-of-bounds numerical points associated with either the upper or lower limit benchmark parameters are extracted. Since 85 is less than 100 and greater than 60, no out-of-bounds numerical points are generated; logical labeling is performed on these points, assigning them a Boolean false value of 0. If the heart rate measurement value at another time point is 105, the difference calculation is performed: Heart rate difference = 105 - 100 = 5; Since the difference result is greater than 0, the out-of-bounds value is truncated, and a logical labeling process is performed on the out-of-bounds value, assigning it a Boolean truth value of 1. This logical operation is performed in parallel for all vital signs, constructing independent trigger flags for abnormal heart rate, abnormal blood oxygenation, and abnormal blood pressure. In the actual calculation scenario, when the measured blood oxygenation value drops to 92, the out-of-bounds value is triggered, generating a blood oxygenation abnormality trigger flag with a state value of 1; when the systolic blood pressure rises to 145, a blood pressure abnormality trigger flag with a state value of 1 is generated.

[0023] S113: Obtain the heart rate abnormality trigger flag, blood oxygen abnormality trigger flag, and blood pressure abnormality trigger flag; perform timestamp alignment and organization on the heart rate abnormality trigger flag, blood oxygen abnormality trigger flag, and blood pressure abnormality trigger flag to obtain abnormal state identification information; Read the heart rate abnormality trigger flags, blood oxygen abnormality trigger flags, and blood pressure abnormality trigger flags from the memory cache queue, and extract the hardware crystal oscillator timestamps bound to these flags when they were generated. Using milliseconds as the minimum precision unit, perform timestamp alignment and sorting on the heart rate abnormality trigger flags, blood oxygen abnormality trigger flags, and blood pressure abnormality trigger flags, comparing the timestamp differences among the three types of flags. Set the associated time window to 200 milliseconds. If the timestamp of the heart rate abnormality trigger flag is 1678886400100, the timestamp of the blood oxygen abnormality trigger flag is 1678886400150, and the timestamp of the blood pressure abnormality trigger flag is 1678886400200, since the maximum time difference of 100 milliseconds is less than the associated time window of 200 milliseconds, then perform spatial dimensionality reduction encapsulation on the above three flags, aggregating them into a state vector containing 3 Boolean data bits to obtain the abnormal state identification information. Recorded in vector form as [0, 1, 1], representing that no heart rate abnormality occurred in the current time window, but concurrent blood oxygen and blood pressure out-of-bounds abnormalities occurred.

[0024] Please see Figure 3 The specific steps for obtaining the emergency call code for a hospital bed are as follows: S211: Based on abnormal state identification information, and based on the rules of logical algebra, perform full-space logical intersection and combination operations on the heart rate abnormality triggering flag, blood oxygen abnormality triggering flag and blood pressure abnormality triggering flag, monitor the linkage triggering coupling relationship of multiple categories of vital sign abnormality features, extract signal component vectors with discrete Boolean logic state features, and establish a multi-dimensional vital sign linkage logic matrix. The Boolean vectors contained in the abnormal state identification information are analyzed. Based on the rules of Boolean algebra, a full-space logical intersection and combination operation is performed on the heart rate abnormality trigger flags, blood oxygen abnormality trigger flags, and blood pressure abnormality trigger flags. The heart rate abnormality trigger flag corresponding to vector point 0, the blood oxygen abnormality trigger flag corresponding to point 1, and the blood pressure abnormality trigger flag corresponding to point 2 are extracted. A composite logic gate circuit mapping model composed of bitwise AND and bitwise OR operations is used to monitor the linkage and triggering coupling relationship of multiple categories of vital sign abnormality features. The operation logic is defined as follows: Linkage status = (heart rate abnormality trigger flag AND blood oxygen abnormality trigger flag) OR (heart rate abnormality trigger flag AND blood pressure abnormality trigger flag) OR (blood oxygen abnormality trigger flag AND blood pressure abnormality trigger flag); Substitute the abnormal state identification information [0, 1, 1] into the above logical expression for calculation: Linkage state = (0 AND 1) OR (0 AND 1) OR (1 AND 1) = 0 OR 0 OR 1 = 1; A result of 1 indicates the presence of two or more abnormal physiological symptom linkages. Signal component vectors with discrete Boolean logic state characteristics are extracted. All isolated trigger identifiers and linkage state results are then extended and concatenated in two dimensions to establish a multi-dimensional physiological symptom linkage logic matrix with dimensions of 3 rows and 3 columns. The diagonal elements of the matrix store the trigger state of a single physiological symptom, while the off-diagonal elements store the linkage trigger states resulting from pairwise intersection operations.

[0025] S212: Obtain the preset level trigger judgment rule set, call the Boolean truth value of the logic site embedded in the multi-dimensional vital sign linkage logic matrix, count the total number of active state heart rate abnormality trigger markers, blood oxygen abnormality trigger markers and blood pressure abnormality trigger markers, perform a ladder mapping comparison judgment on the total number and the preset level trigger judgment rule set, extract the corresponding emergency level quantitative assessment level, and generate an emergency level trigger score. Connect to the local rule database to obtain the preset level trigger judgment rule set, and call the Boolean truth values ​​of the logic points embedded in the multi-dimensional vital sign linkage logic matrix. By looping through all valid points in the matrix, perform a Boolean state bit accumulation operation to count the total number of active state trigger indicators for abnormal heart rate, abnormal blood oxygen, and abnormal blood pressure. The judgment rule set is set as follows: a total quantity of 1 maps to a level 3 emergency state, a total quantity of 2 maps to a level 2 emergency state, and a total quantity greater than or equal to 3 maps to a level 1 emergency state. Extract the total quantity parameter; here, the total number of active states is 2. Perform a step-by-step mapping comparison judgment on the total quantity 2 and the preset level trigger judgment rule set. Hard-code the parameter 2 with the conditional branches in the judgment rule set, accurately matching it to the level 2 emergency state branch, and extract the corresponding emergency level quantitative assessment level as level 2. Generate an emergency level trigger score using a weighted quantification formula: Emergency level trigger score = (total number * 30) + (linkage status * 20) = (2 * 30) + (1 * 20) = 80; An emergency level trigger score of 80 is generated. This score clearly quantifies the urgency of the current physiological alarm in the hospital bed.

[0026] S213: Obtain the priority coding communication protocol, retrieve the preset level identifier character index library based on the emergency level trigger score, perform character sequence encapsulation processing for the emergency level trigger score, construct the priority code sequence, write the priority code sequence into the emergency priority control field position in the head of the bed call data packet, associate the corresponding bed equipment hardware identifier with the call logic weight level, establish the mapping relationship between call record data nodes and emergency codes, and generate the bed call emergency code; The system loads the priority-encoded communication protocol stack running in kernel mode and retrieves a pre-defined priority identifier character index based on the emergency level trigger score of 80. The index uses a hash table structure, with score ranges as keys and hexadecimal encoded characters as values. A score of 80 falls within the range of 70 to 90, and is mapped to the hexadecimal character 0x02. Character sequence encapsulation is performed on the emergency level trigger score, concatenating the hexadecimal character 0x02 with the frame header 0xEE and the frame tail 0xFF to construct the priority code sequence 0xEE02xFF. The underlying network communication socket is opened, and the priority code sequence 0xEE02xFF is written to the emergency priority control field in the header of the bed call data packet, occupying bytes 3 to 5 of the message. The MAC address of the local device is extracted as the bed device hardware identifier, and the corresponding bed device hardware identifier is associated with the call logical weight level to establish a mapping relationship between call record data nodes and emergency codes. Table 1 shows an example of the data structure of the emergency code mapping table.

[0027] Table 1. Emergency Code Mapping Relationship

[0028] As shown in Table 1, the generated emergency priority control field is bound and stored with the bed equipment hardware identifier, and finally encapsulated to generate a complete emergency call code message for the hospital bed and pushed to the message bus.

[0029] Please see Figure 4 The specific steps for obtaining the task load value are as follows: S311: Monitor task management node messages, extract task type classification field and task start time parameter, perform pre-configured time mapping table retrieval for task type classification field, extract the standard task processing time associated with matching task type classification field, obtain system timestamp, perform time axis difference operation for system timestamp and task start time parameter, extract time difference parameter, and establish task execution time. A background daemon process monitors the task flow messages distributed by the task management node in real time and parses the JSON-formatted message payload structure. It extracts the task type classification field and task start time parameter from the message key-value pairs. The extracted task type classification field is "intravenous infusion," and the extracted task start time parameter is 1678886400000 (corresponding to absolute time). For the task type classification field "intravenous infusion," a pre-configured time mapping table is retrieved to locate the operation standard for intravenous infusion in the mapping table. The standard task processing time associated with the matching task type classification field is extracted, and the table shows that the standard task processing time for intravenous infusion is 1800000 milliseconds (i.e., 30 minutes). The operating system's underlying real-time clock interface is called to obtain the current system timestamp accurate to milliseconds, and the current system timestamp is set to 1678886800000. A time axis difference operation is performed on the system timestamp and the task start time parameter. The calculation process is as follows: Time difference parameter = 1678886800000 - 1678886400000 = 400000; Extract the time difference parameter of 400,000 milliseconds. This value physically represents the absolute time span from task allocation to the current moment. The execution time of the task in memory is 400,000 milliseconds.

[0030] S312: Subtract the standard task processing time from the task execution time to extract the time difference parameter, traverse the unfinished task sequence to perform a loop subtraction, extract the time difference parameter of each node to be processed, aggregate the extracted parameters to build a countdown array, and obtain the remaining time of each task item. The system retrieves the standard task processing time (1,800,000 milliseconds) and the task execution time (400,000 milliseconds) locked in memory, and then subtracts these values. The calculation process is as follows: Timing differential parameter = 1,800,000 - 400,000 = 1,400,000; Extract the time difference parameter of 1400000 milliseconds. Obtain the sequence of incomplete tasks bound to the current nurse node, which is stored using a doubly linked list data structure. Traverse the sequence of incomplete tasks and perform a loop subtraction operation, performing the standard difference operation logic described above on each task node in the linked list, and extract the time difference parameter mapped to each node to be processed. Assume that there are 3 nodes to be processed in the current linked list of the nurse, and the time difference parameters obtained by extracting them in sequence are 1400000 milliseconds, 600000 milliseconds, and 900000 milliseconds. Dynamically allocate array space in a contiguous area of ​​memory, aggregate the extracted parameters to construct a countdown array [1400000, 600000, 900000], and obtain the remaining time of each task. Taking a real ward scenario as an example, after the nurse completes the infusion operation for the current bed, the above loop calculation is completed, and it is determined that there are three countdown tasks remaining under the nurse's name.

[0031] S313: Based on the nursing staff identification code, extract the remaining time of each task in the pending work queue, perform linear accumulation on the remaining time of all tasks, extract the total time surplus feature parameter, perform weight mapping comparison on the total time surplus feature parameter and the preset carrying capacity benchmark value, establish the staff work leeway feature index, and generate task load value. Read the nursing staff identification code synchronized from the access control and attendance device, and retrieve the in-memory database using this code as the unique primary key. Extract the remaining time for each task in the to-do queue, i.e., obtain the countdown array [1400000, 600000, 900000]. Perform a linear accumulation operation on the remaining time for all tasks. The calculation process is as follows: Total remaining time = 1,400,000 + 600,000 + 900,000 = 2,900,000; The total remaining time feature parameter is extracted as 2,900,000 milliseconds. A pre-configured baseline load capacity value in the kernel is loaded, which is set to 14,400,000 milliseconds (i.e., the upper limit of 4-hour effective workload) based on hospital nursing scheduling specifications. A weighted mapping comparison is performed between the total remaining time feature parameter and the baseline load capacity value. The calculation process is as follows: Personnel workload leeway characteristic index = (14,400,000 - 2,900,000) / 14,400,000 = 0.798; The established staff workload leeway characteristic index is 0.798. This index reflects the current level of idle time among nursing staff. Task load values ​​are generated using inverse normalization logic. Task load value = 1 - 0.798 = 0.202; The task load value is 0.202. The closer this value is to 1, the closer the nursing staff are to being under full workload.

[0032] Please see Figure 5 The specific steps for obtaining candidate paths at the nurse station response are as follows: S411: Obtain the corridor node coordinate array and node topology connection relationship transmitted by the spatial mapping node, monitor the traffic status indicator value associated with the corridor node coordinate array, perform numerical quantization conversion processing on the traffic status indicator value based on the mapping result of the preset congestion level mapping table, extract the traffic resistance influencing factor, perform discrete interval normalization mapping on the traffic resistance influencing factor and extract the mapping result value, and establish a traffic weighted correction coefficient. A Hypertext Transfer Protocol (HTTP) request is initiated to access the hospital infrastructure spatial mapping nodes, obtaining the returned JSON-formatted corridor node coordinate array and node topology connections. The corridor node coordinate array is defined as a set of points in a three-dimensional Cartesian coordinate system, in the form of [[0, 0, 0], [10, 0, 0], [10, 20, 0]]. The node topology connections are presented as an edge list, recorded as [[0, 1], [1, 2]], representing the physical connectivity between nodes. Real-time monitoring of infrared thermal imaging and RFID dual-mode passenger flow sensor data associated with the corridor node coordinate array is conducted via an IoT gateway to obtain passage status indicator values. This value is an integer variable from 0 to 5, representing the personnel density per square meter within the spatial area. Taking the path from corridor node 1 to node 2 as an example, its passage status indicator value is detected as 4. Based on the passage status indicator value, a preset congestion level mapping table is retrieved. Table 2 lists an example of the data structure of the preset congestion level mapping table.

[0033] Table 2 Preset Congestion Level Mapping Table

[0034] As shown in Table 2, the system accurately locates the record corresponding to congestion level 3 based on the traffic status indicator value of 4. The traffic status indicator value undergoes numerical quantization conversion, and the traffic resistance influence factor is extracted from the table as 2.5. Discrete interval normalization mapping is performed on the traffic resistance influence factor, setting the global maximum resistance factor limit to 5.0, and the extracted mapping result value is 0.5 (calculation process: 2.5 / 5.0=0.5). Combined with the baseline offset constant of 0.8, a traffic weighted correction coefficient is established: The weighted adjustment factor for passage is 0.5 + 0.8 = 1.3; The final pass weighting correction factor is set at 1.3.

[0035] S412: Obtain the basic path length parameter stored locally, perform a multiplicative weighted operation on the toll weight correction coefficient and the basic path length parameter, extract the path loss gain component, perform linear compensation summation adjustment on the basic path length parameter and the path loss gain component, determine the actual toll cost weight of each road segment, and construct the corrected topology path length. Access the pre-installed Building Information Modeling (BIM) database on the local solid-state drive and execute an SQL query to obtain the locally stored basic path length parameter. For the road segment from node 1 to node 2 in the aforementioned corridor, the basic path length parameter is extracted as 20,000 mm. Retrieve the traffic weighting correction coefficient of 1.3 residing in memory, and perform a multiplicative weighted operation on the traffic weighting correction coefficient and the basic path length parameter. The calculation process is as follows: Path loss gain component = 20000 * 1.3 = 26000; A path loss gain component of 26,000 mm is extracted, which converts the time delay caused by congestion into a virtual physical path extension. Furthermore, considering the physical occupancy compensation of fixed equipment such as medical carts, a compensation base constant of 1,500 mm is set. Linear compensation and summation adjustments are performed on the basic path length parameter and the path loss gain component. The calculation process is as follows: Corrected topology path length = 26000 + 1500 = 27500; The actual passage cost weight for each road segment is determined to be 27,500 mm. This value is used to cover the original physical length, and a corrected topology path length is constructed. In actual dispatching scenarios, a corridor that is originally very short but is full of family members and obstacles will have a significantly increased corrected topology path length, and will therefore be automatically downgraded in subsequent algorithms.

[0036] S413: Call the corridor node coordinate array, node topology connection relationship and corrected topology path length, perform weight assignment on the node topology connection relationship based on graph theory traversal rules, construct a weighted topology adjacency matrix, retrieve the starting node position of the nurse station and the target node position of the call source, perform multi-round node relaxation iterative optimization calculation on the weighted topology adjacency matrix, extract the connected node sequence with the minimum total weight score, perform spatial coordinate mapping processing on the connected node sequence, and generate nurse station response candidate paths; Extract the corridor node coordinate array, node topological connections, and the fully updated corrected topological path lengths stored in memory. Initialize a two-dimensional floating-point array structure of size N multiplied by N, where N represents the total number of spatially mapped nodes. Based on graph theory traversal rules, assign weights to the node topological connections, accurately filling all corrected topological path length values ​​into the corresponding x and y coordinate intersection points of the two-dimensional array. If two nodes are not directly connected, fill in an infinite constant, thus constructing a weighted topological adjacency matrix. Connect to the scheduling database to retrieve the starting node location identifier code of the currently idle nurse station, and simultaneously parse alarm messages to retrieve the target node location identifier code of the call source. Use the starting node and target node as the algorithm input source, loading the Dijkstra single-source shortest path algorithm kernel. Perform multi-round node relaxation iterative optimization calculations on the weighted topological adjacency matrix. In each iteration, extract the unvisited node with the smallest current distance value, traverse all its adjacent edges, and if the cumulative weight of reaching neighboring nodes through this node is less than the weight already recorded by the neighboring node, update the weight value and record the predecessor node. The process is repeated until the target node is marked as visited. The predecessor node array is traced back in reverse to extract the connected node sequence with the lowest total weight score, for example, [node 1, node 5, node 8, node 12]. Spatial coordinate mapping is then performed on the connected node sequence, reversing the node identifier code into a three-dimensional Cartesian coordinate set to generate candidate paths for the nurse station response.

[0037] Please see Figure 6 The specific steps for obtaining the nursing call task distribution sequence are as follows: S511: Obtain the task load value and the estimated processing time of nursing calls from the task management node, extract the local storage load weight coefficient, duration adjustment factor and time decay correction constant, perform product summation and weighted operation on the task load value, the estimated processing time of nursing calls and the load weight coefficient, extract the real-time load feature of discrete nodes, perform nonlinear smoothing filtering on the load feature based on the time decay correction constant, map the processed data to the linear evaluation interval, and establish a load capacity evaluation score; An asynchronous message subscription mechanism is established to obtain the task load value and estimated processing time of nursing calls from the task management node in real time. The currently captured task load value is 0.202, and the estimated processing time of the nursing call associated with the parsed call request category packet is 600,000 milliseconds (converted to a quantized standard value of 0.6). The local configuration file directory is mounted, and the local storage load weight coefficient of 0.4, duration adjustment factor of 0.6, and time decay correction constant of 0.1 are extracted. The weighted calculation engine is started to perform a product summation weighted operation on the task load value, the estimated processing time of the nursing call, and the load weight coefficient. The calculation process is as follows: Initial feature quantity = (0.202 * 0.4) + (0.6 * 0.6) = 0.0808 + 0.36 = 0.4408; The real-time load characteristic value of discrete nodes is extracted as 0.4408. To suppress the drastic fluctuations in load assessment caused by sudden high-concurrency calls, an exponential smoothing filter algorithm is introduced. The smoothing characteristic value of historical periods is extracted and cached as 0.4100. Based on the time decay correction constant of 0.1, nonlinear smoothing filtering is performed on the load characteristic value. The filtering calculation process is as follows: Smoothed load characteristic = (0.1 * 0.4408) + ((1 - 0.1) * 0.4100) =0.04408 + 0.369 = 0.41308; Call the linear mapping function, set the lower limit benchmark value of the evaluation interval to 0 and the upper limit benchmark value to 1, and map the processed data to the linear evaluation interval. Since 0.41308 is directly within this interval, perform rounding to three decimal places and establish a load capacity evaluation score of 0.413.

[0038] S512: Based on the emergency code for bedside calls, load capacity assessment scores, and path length values ​​associated with candidate response paths at nurse stations, combined with simulated business datasets, the baseline value of equipment availability, nurse qualification level scores, and ward occupancy density coefficients are obtained, using the following formula: ; The scheduling response priority evaluation value of the task to be distributed is obtained by calculation, the coupling feature of task urgency and resource matching degree is extracted, the discrete interval outlier removal is performed on the scheduling response priority evaluation value, and the task evaluation feature vector is established. in, This represents the numerical value used to evaluate the priority of the scheduling response. The normalized value representing the emergency code for a hospital bed call is obtained by retrieving the normalized mapping from historical steps. The normalized value representing the path length is obtained by acquiring the execution normalization mapping from historical steps. The baseline value representing the equipment availability rate is extracted from the simulated business dataset. The ward occupancy density coefficient was extracted from the simulated business dataset. The normalized value representing the nurse's qualification level score is obtained from the execution quantification mapping obtained from the simulated business dataset. The normalized value representing the qualification benchmark reference value is extracted from the simulated business dataset. This represents the load capacity assessment score; The system utilizes global shared memory to read normalized values ​​obtained from parsing the emergency code for bed calls generated in the preceding workflow, the load capacity assessment score of 0.413, and the normalized values ​​obtained from parsing the associated path length of the nurse station response candidate path. It then connects to the simulated business dataset server and executes a data scraping script to obtain the equipment availability baseline, nurse qualification level scores, and ward occupancy density coefficient.

[0039] Table 3 shows the data collection and normalization results of various parameters calculated by the formula;

[0040] As shown in Table 3, all parameters have been standardized and quantized. The core formula is invoked using the underlying hardware-accelerated computing unit: ; The scheduling response priority evaluation values ​​of the tasks to be distributed are obtained through calculation, and the coupling features between task urgency and resource matching degree are extracted. The specific values ​​in Table 3 are strictly substituted into the formula to perform the calculation.

[0041] Prefix denominator operations: ; Pre-calculation of molecules: ; Division within the square root: ; Square root operation: ; Absolute value numerator operations: ; Post-division operation: ; Addition operation: ; The dispatch response priority assessment value P was obtained as 1.979. This result indicates that the current dispatch response priority assessment value is in the extremely high priority sequence (much greater than the threshold of 1.0), suggesting that the system must intervene immediately under conditions of high urgency, short path, and nurses with qualifications far exceeding the baseline. The advantage of the formula lies in balancing the extreme value effects of urgency (A, C) and resistance terms (B, D) by introducing a square root nonlinear decay mechanism, and by using the absolute value term... The deviation of nurses' qualifications from the benchmark was forcibly compensated. Discrete interval outlier removal was performed on the scheduling response priority evaluation value, and the removal upper limit was set to 5.0. Since 1.979 did not exceed the limit, it was incorporated into the array structure to establish the task evaluation feature vector [1.979, 0.80, 0.413].

[0042] S513: Call the task evaluation feature vector, perform multi-dimensional weight normalization mapping on the embedded data components, extract standardized evaluation indicators, search the preset priority sorting index library, perform descending sorting comparison analysis on the standardized evaluation indicators, determine the execution order position of each nursing task in the global pending task pool, associate the execution order position with the corresponding ward node hardware address code, establish the call task and nursing human resource allocation relationship matrix, and obtain the nursing call task distribution sequence. The task evaluation feature vector [1.979, 0.80, 0.413] is retrieved from shared memory, initiating the data standardization pipeline. Multidimensional weighted normalization mapping is performed on the embedded data components. The weights for the evaluation values ​​are set to 0.6, the urgency component to 0.3, and the workload component to 0.1. Vector dot product operations are performed to extract standardized evaluation metrics. Standardized evaluation index = (1.979*0.6) + (0.80*0.3) + (0.413*0.1) =1.1874+0.240+0.0413=1.4687; The standardized evaluation index 1.4687 is extracted. A preset priority sorting index library resident in main memory is connected. A QuickSort algorithm is executed on the currently generated standardized evaluation index 1.4687 and the evaluation indices of other tasks in the pool (e.g., 1.210, 0.985, 1.650). A descending order comparison analysis is performed on the standardized evaluation index. The sorting result is [1.650, 1.4687, 1.210, 0.985], thus determining that each nursing task with the current nurse evaluation index 1.4687 has the second execution order position in the global task pool. The binding attribute of this execution order position is parsed, and the corresponding ward node hardware address code is associated with the execution order position. For example, the infusion pump terminal with MAC address 00-1A-2B-3C-4D-60 is extracted. A call task and nursing human resource allocation relationship matrix is ​​established in the background, rigidly binding the task ID, nurse employee number, device MAC address, and sorting position. Finally, the formatted output yields the nursing call task distribution sequence.

[0043] A call distribution system based on nursing needs is used to implement the above-mentioned call distribution method based on nursing needs. The system includes: The abnormal state analysis module acquires heart rate, blood oxygen, and blood pressure measurements, compares them with preset vital sign thresholds, constructs heart rate abnormality trigger flags, blood oxygen abnormality trigger flags, and blood pressure abnormality trigger flags, and obtains abnormal state identification information. The emergency level classification module, based on abnormal state identification information, calls Boolean logic functions to perform logical analysis and calculation on the heart rate abnormality trigger flag, blood oxygen abnormality trigger flag, and blood pressure abnormality trigger flag, determines the emergency level trigger score, and generates a bed call emergency code; The task load analysis module obtains the task type classification field and task start time parameter, matches the standard task processing time, uses the current system time to subtract the task start time to construct the task execution time, subtracts the standard task processing time from the task execution time to calculate the remaining time of each task, and accumulates the remaining time of each task to generate the task load value. The response filtering module obtains the corridor node coordinate array and node topology connection relationship, uses the passage status of the node coordinates to correct the basic path length parameter to construct the corrected topology path length, and uses the corridor node coordinate array, node topology connection relationship and corrected topology path length to perform pathfinding analysis calculation to generate nurse station response candidate paths; The task distribution and processing module obtains the estimated processing time of nursing calls, calculates the load capacity assessment score by combining the task load value, maps and compares the emergency code of the bed call, the load capacity assessment score and the path length value associated with the candidate response path of the nurse station, and generates a nursing call task distribution sequence.

[0044] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A call distribution method based on nursing needs, characterized in that, Includes the following steps: S1: Obtain heart rate, blood oxygen, and blood pressure measurements, compare them with preset vital sign thresholds, construct heart rate abnormality trigger flags, blood oxygen abnormality trigger flags, and blood pressure abnormality trigger flags, and obtain abnormal state identification information. S2: Based on the abnormal state identification information, call the Boolean logic function to perform logical analysis and calculation on the heart rate abnormality trigger flag, blood oxygen abnormality trigger flag and blood pressure abnormality trigger flag, determine the emergency level trigger score, and generate the bed call emergency code; S3: Obtain the task type classification field and task start time parameter, match the standard task processing time, use the current system time to subtract the task start time to construct the task execution time, subtract the standard task processing time from the task execution time to calculate the remaining time of a single task, accumulate the remaining time of a single task, and generate the task load value. S4: Obtain the corridor node coordinate array and node topology connection relationship, use the passage status of the node coordinates to correct the basic path length parameter to construct the corrected topology path length, use the corridor node coordinate array, node topology connection relationship and corrected topology path length to perform pathfinding analysis calculation, and generate candidate response paths for the nurse station; S5: Obtain the estimated processing time of the nursing call, combine it with the task load value, calculate the load capacity assessment score, map and compare the emergency code of the bed call, the load capacity assessment score and the path length value associated with the nurse station response candidate path, and generate a nursing call task distribution sequence.

2. The call distribution method based on nursing needs according to claim 1, characterized in that, The emergency call codes for hospital beds include red emergency response codes, orange priority intervention codes, and yellow routine standby codes. The task load values ​​include high intensity saturation index, medium load margin, and light idle weight. The nurse station response candidate paths include main road direct corridors, ward area backup passages, and fire zone bypass corridors. The nursing call task distribution sequence includes initial emergency dispatch work orders, regular queuing queues, and delayed observation and allocation lists.

3. The call distribution method based on nursing needs according to claim 1, characterized in that, The specific steps for obtaining the abnormal state identification information are as follows: S111: Monitor the communication messages transmitted by the bedside interface node, perform delimitation and interception processing on the communication messages, extract the heart rate measurement value, blood oxygen measurement value and blood pressure measurement value embedded in the communication messages, perform timestamp alignment and arrangement on the heart rate measurement value, blood oxygen measurement value and blood pressure measurement value, and establish a vital sign feature matrix. S112: Based on the preset vital sign thresholds, call the upper limit benchmark parameter and the lower limit benchmark parameter, and use the upper limit benchmark parameter and the lower limit benchmark parameter to perform difference operation on the heart rate measurement value, blood oxygen measurement value and blood pressure measurement value embedded in the vital sign feature matrix. Extract the out-of-bounds value sites associated with crossing the upper limit benchmark parameter or the lower limit benchmark parameter, perform logical labeling processing on the out-of-bounds value sites, and construct the heart rate abnormality triggering flag, blood oxygen abnormality triggering flag and blood pressure abnormality triggering flag. S113: Obtain the heart rate abnormality trigger identifier, blood oxygen abnormality trigger identifier and blood pressure abnormality trigger identifier, and perform timestamp alignment and sorting on the heart rate abnormality trigger identifier, blood oxygen abnormality trigger identifier and blood pressure abnormality trigger identifier to obtain abnormal state identification information.

4. The call distribution method based on nursing needs according to claim 3, characterized in that, The specific steps for obtaining the emergency call code for the hospital bed are as follows: S211: Based on the abnormal state identification information, and based on the logical algebra operation rules, perform full-space logical intersection and combination operations on the heart rate abnormality triggering identifier, blood oxygen abnormality triggering identifier and blood pressure abnormality triggering identifier, monitor the linkage triggering coupling relationship of multiple categories of vital sign abnormality features, extract signal component vectors with discrete Boolean logic state features, and establish a multi-dimensional vital sign linkage logic matrix. S212: Obtain the preset level trigger judgment rule set, call the Boolean truth value of the logic site embedded in the multi-dimensional vital sign linkage logic matrix, count the total number of heart rate abnormality trigger markers, blood oxygen abnormality trigger markers and blood pressure abnormality trigger markers in the active state, perform a ladder mapping comparison judgment on the total number and the preset level trigger judgment rule set, extract the corresponding emergency level quantitative assessment level, and generate an emergency level trigger score. S213: Obtain the priority coding communication protocol, retrieve the preset level identifier character index library based on the emergency level trigger score, perform character sequence encapsulation processing on the emergency level trigger score, construct the priority code sequence, write the priority code sequence into the emergency priority control field position in the head of the bed call data packet, associate the corresponding bed equipment hardware identifier with the call logic weight level, establish the mapping relationship between call record data nodes and emergency codes, and generate the bed call emergency code.

5. The call distribution method based on nursing needs according to claim 4, characterized in that, The specific steps for obtaining the task load value are as follows: S311: Monitor task management node messages, extract task type classification field and task start time parameter, perform pre-configured time mapping table retrieval for task type classification field, extract the standard task processing time associated with matching task type classification field, obtain system timestamp, perform time axis difference operation for system timestamp and task start time parameter, extract time difference parameter, and establish task execution time. S312: Subtract the standard task processing time from the task execution time to extract the time difference parameter, traverse the unfinished task sequence to perform a loop subtraction, extract the time difference parameter of each node to be processed, aggregate the extracted parameters to construct a countdown array, and obtain the remaining time of each task item. S313: Based on the nursing staff identification code, extract the remaining time of each task item in the pending work queue, perform a linear summation operation on the remaining time of all the tasks, extract the total time surplus feature parameter, perform a weighted mapping comparison between the total time surplus feature parameter and the preset carrying capacity benchmark value, establish a staff work leeway feature index, and generate task load values.

6. The call distribution method based on nursing needs according to claim 5, characterized in that, The specific steps for obtaining the candidate response path at the nursing station are as follows: S411: Obtain the corridor node coordinate array and node topology connection relationship transmitted by the spatial mapping node, monitor the traffic status indicator value associated with the corridor node coordinate array, perform numerical quantization conversion processing on the traffic status indicator value based on the mapping result of the preset congestion level mapping table, extract the traffic resistance influencing factor, perform discrete interval normalization mapping on the traffic resistance influencing factor and extract the mapping result value, and establish a traffic weighted correction coefficient. S412: Obtain the local storage basic path length parameter, perform multiplication weighting operation on the toll weighting correction coefficient and the basic path length parameter, extract the path loss gain component, perform linear compensation summation adjustment on the basic path length parameter and the path loss gain component, determine the actual toll cost weight of each road segment, and construct the corrected topology path length. S413: Call the corridor node coordinate array, node topology connection relationship and corrected topology path length, perform weight assignment on the node topology connection relationship based on graph theory traversal rules, construct a weighted topology adjacency matrix, retrieve the starting node position of the nurse station and the target node position of the call source, perform multi-round node relaxation iterative optimization calculation on the weighted topology adjacency matrix, extract the connected node sequence with the minimum total weight score, perform spatial coordinate mapping processing on the connected node sequence, and generate candidate response paths for the nurse station.

7. The call distribution method based on nursing needs according to claim 6, characterized in that, The specific steps for obtaining the nursing call task distribution sequence are as follows: S511: Obtain the task load value and the estimated processing time of nursing calls from the task management node; extract the local storage load weight coefficient, duration adjustment factor, and time decay correction constant; perform a product summation and weighted operation on the task load value, the estimated processing time of nursing calls, and the load weight coefficient; extract the real-time load characteristics of discrete nodes; perform nonlinear smoothing filtering on the load characteristics based on the time decay correction constant; map the processed data to a linear evaluation interval; and establish a load capacity evaluation score. S512: Based on the emergency code of the bed call, the load capacity assessment score and the path length value associated with the candidate response path of the nurse station, and combined with the simulated business dataset to obtain the equipment integrity benchmark value, the nurse qualification level score and the ward occupancy density coefficient, calculate the scheduling response priority assessment value of the task to be assigned, extract the coupling feature of task urgency and resource matching degree, perform discrete interval outlier removal for the scheduling response priority assessment value, and establish a task assessment feature vector; S513: Call the task evaluation feature vector, perform multi-dimensional weight normalization mapping on the embedded data components, extract standardized evaluation indicators, search the preset priority sorting index library, perform descending sorting comparison analysis on the standardized evaluation indicators, determine the execution order position of each nursing task in the global pending task pool, associate the execution order position with the corresponding ward node hardware address code, establish a call task and nursing human resource allocation relationship matrix, and obtain the nursing call task distribution sequence.

8. The call distribution method based on nursing needs according to claim 7, characterized in that, The specific formula for obtaining the scheduling response priority evaluation value of the task to be distributed is as follows: ； in, This represents the numerical value used to evaluate the priority of the scheduling response. The normalized value representing the emergency call code for hospital beds. The normalized value representing the path length. This represents the baseline value for equipment availability. Represents the ward occupancy density coefficient. The normalized value representing the nurse's qualification level score. The normalized value representing the qualification benchmark reference value, This represents the load capacity assessment score.

9. A call distribution system based on nursing needs, characterized in that, The system is used to implement the call distribution method based on nursing needs as described in any one of claims 1-8, the system comprising: The abnormal state analysis module acquires heart rate, blood oxygen, and blood pressure measurements, compares them with preset vital sign thresholds, constructs heart rate abnormality trigger flags, blood oxygen abnormality trigger flags, and blood pressure abnormality trigger flags, and obtains abnormal state identification information. The emergency level classification module, based on the abnormal state identification information, calls a Boolean logic function to perform logical analysis and calculation on the heart rate abnormality trigger flag, blood oxygen abnormality trigger flag and blood pressure abnormality trigger flag, determines the emergency level trigger score, and generates a bed call emergency code; The task load analysis module obtains the task type classification field and task start time parameter, matches the standard task processing time, uses the current system time to subtract the task start time to construct the task execution time, subtracts the standard task processing time from the task execution time to calculate the remaining time of each task, and accumulates the remaining time of each task to generate the task load value. The response filtering module obtains the corridor node coordinate array and node topology connection relationship, uses the passage status of the node coordinates to correct the basic path length parameter to construct the corrected topology path length, and uses the corridor node coordinate array, node topology connection relationship and corrected topology path length to perform pathfinding analysis calculation to generate nurse station response candidate paths; The task distribution and processing module obtains the estimated processing time of nursing calls, calculates the load capacity assessment score based on the task load value, and maps and compares the path length value associated with the emergency code of the bed call, the load capacity assessment score, and the candidate response path of the nurse station to generate a nursing call task distribution sequence.

Images