Patient sign real-time monitoring and intervention response system for smart tumor ward

Abstract

The present application relates to the technical field of vital sign monitoring, in particular to a patient vital sign real-time monitoring and intervention response system for smart tumor ward, which comprises a skin thermal state recognition module, a communication abnormal period screening module, a vital sign event screening module, an intervention task generation module and an intervention response execution state construction module.In the present application, through joint analysis of temperature trend and communication abnormal period, the timing correspondence relationship between vital sign fluctuation and channel interruption is established, the data section of closed loop interruption risk is screened, the continuous change trend of body temperature, heart rate and respiratory rate is combined, the abnormal section and task state are labeled, the nursing terminal activation condition and control equipment number are matched, the hierarchical generation, state identification and response matching of the task are completed, the vital sign recognition is changed from static index to trend linkage, the task scheduling is extended from artificial driving to closed loop control, and the continuity of vital sign perception, the identification ability of communication interference and the connection efficiency of intervention execution in the ward scene are enhanced.

Description

Technical Field

[0001] This invention relates to the field of vital sign monitoring technology, and in particular to a real-time monitoring and intervention response system for patient vital signs in smart oncology wards. Background Technology

[0002] The field of vital sign monitoring technology involves the continuous acquisition and data processing of physiological parameters such as heart rate, blood pressure, body temperature, respiratory rate, and blood oxygen saturation. Core aspects include the sensing methods of physiological signals, data acquisition cycles, transmission paths, abnormality identification criteria, and processing response mechanisms. Vital sign data is typically acquired through attached sensors, wearable devices, or bedside devices, and transmitted to the processing end via wired or wireless means. Indicator judgments are made based on signal amplitude, rate of change, thresholds, or time window statistical results, triggering early warning responses. Traditional real-time patient vital sign monitoring and intervention response systems for smart oncology wards refer to the use of sensors such as attached ECG leads, temperature sensors, and pulse oximeters in the oncology ward environment to collect patients' electrophysiological signals, temperature, and blood oxygen data. This data is transmitted to a mobile terminal via Bluetooth, where nursing staff log in to view the data and manually trigger interventions such as telephone notifications, ward rounds, and medical record recording based on whether the indicators exceed static preset thresholds. This type of system is mainly used to address the problem of rapid changes in vital signs and insufficient frequency of manual monitoring in hospitalized cancer patients. Traditional methods generally rely on timed data collection and passive response, lacking logical judgment and dynamic adjustment mechanisms between data.

[0003] Existing technologies rely on attached sensors to periodically acquire single vital signs. If sudden interruptions or data discontinuities occur during signal transmission, it is impossible to effectively identify the potential correlation between abnormal acquisition links and sudden changes in physiological parameters. In scenarios with high-frequency pathological changes, the lack of trend judgment logic and data cross-analysis mechanisms can easily lead to passive response and delayed intervention in clinical events. For example, when body temperature fluctuations are accompanied by communication abnormalities, it is impossible to accurately reconstruct the evolution of the event or trigger task scheduling in a timely manner, resulting in missed intervention nodes or failure of closed-loop status, which affects the stable operation of the ward monitoring system and the efficiency of task implementation. Summary of the Invention

[0004] To address the technical problems existing in the prior art, embodiments of the present invention provide a real-time monitoring and intervention response system for patient vital signs in smart oncology wards.

[0005] On the one hand, a real-time monitoring and intervention response system for patient vital signs in smart oncology wards is provided, which includes: The skin thermal state recognition module acquires temperature sensor data of the subclavian and ventral regions of patients in the smart oncology ward, analyzes the direction of continuous periodic temperature change of the thermal nodes, and obtains the set of monitoring nodes. The communication anomaly screening module, based on the set of monitoring nodes, identifies the continuity of the periodic signal of the frequency hopping channel, extracts the time periods corresponding to sudden interruptions, data packet loss, and signal drops, and pairs them with channel numbers to obtain the identification structure of the trend of vital signs. Based on the vital sign change trend recognition structure, the vital sign event screening module extracts continuous sequences of body temperature, heart rate, and respiratory rate, determines the time segments of continuous rise and reverse mutation, pairs the vital sign name with the time segment, and obtains a set of vital sign trend mutation data. The intervention task generation module, based on the data pairs of the vital signs time periods in the set of data on sudden changes in vital signs trends, assigns an unclosed state to be intervened when any vital sign becomes abnormal, thus obtaining the data structure of the task to be closed. The intervention response execution status construction module, based on the data structure of the task to be closed, matches the activation status of the nursing execution terminal, divides the response tasks according to the response status, and obtains a set of task execution response statuses.

[0006] As a further embodiment of the present invention, the monitoring node set includes node number and temperature rise trend status structure; the vital sign change trend identification structure includes time period, channel number pairing information, and signal anomaly type; the vital sign trend mutation data set includes vital sign name, time segment pairing, and change trend type; the task data structure to be closed-loop includes vital sign channel number, time period, and status field; and the task execution response status set includes channel number, device number, and device response status.

[0007] As a further aspect of the present invention, the continuous periodic temperature of the thermal node refers to the trend of continuous temperature rise of the target monitoring point on the body surface in a continuous time period. The continuity of the frequency hopping channel periodic signal refers to the measure of whether the signal of the frequency hopping channel is uninterrupted within a continuous time.

[0008] As a further aspect of the present invention, the cross-coverage interval refers to the intersection segment where abnormal vital signs and communication interference overlap in time; The term "unclosed-loop state to be intervened" refers to the status of vital signs and events for which the intervention task has not yet been completed.

[0009] As a further aspect of the present invention, the skin thermal sensitivity state recognition module includes: The data stream receiving submodule acquires the placement point numbers in the subclavian and ventral regions of the skin thermal state recognition module, detects the temperature sensor output data corresponding to the placement point number in the current cycle, distinguishes the region to which the temperature data belongs according to the number, and binds the temperature value corresponding to the number to obtain the placement point temperature corresponding list. The temperature change submodule, based on the list of temperature correspondences of the distribution points, calls the temperature values ​​of the same distribution point number in two consecutive time periods, compares the direction of temperature change in the two periods according to the number order, and extracts the distribution point numbers that show an upward change in both time periods to obtain a set of continuously rising distribution point numbers. The temperature rise trend extraction submodule, based on the continuously rising distribution point number set, calls the temperature change direction data within two time periods, associates the direction value corresponding to the number, extracts the number and direction information with the temperature rise trend, and obtains the monitoring node set.

[0010] As a further aspect of the present invention, the communication anomaly screening module includes: The device mapping and positioning submodule, based on the node number in the monitoring node set, detects the identification of the connected vital sign acquisition device according to the structure corresponding to the number, extracts the frequency hopping communication device number associated with each node number, verifies the correspondence between the device and the node, and obtains the node communication device index table. The signal continuity detection submodule extracts the signal data stream sequence within a continuous time period under the frequency hopping channel according to the device number based on the node communication device index table, detects the interval changes between data according to the time sequence, identifies the periodic points where the signal transmission is interrupted, the value is missing, and the intensity drops sharply, and obtains a set of abnormal signal time segments. The abnormal time period identification submodule matches the corresponding channel number based on the abnormal signal time segment set, and combines the data positions according to the association order between each channel number and the corresponding time segment to obtain the vital sign change trend identification structure.

[0011] As a further aspect of the present invention, during the process of the interval change between the detection data: when there is a time interval between two adjacent data points that is greater than a preset threshold, and at least one of the adjacent data points has a signal strength value that is attenuated, it is determined to be a periodic point where signal transmission is abnormal. In the process of matching the corresponding channel number: each channel number and the corresponding time segment are matched one by one in time order. When the same time segment appears in multiple consecutive association processes, it is incorporated into the vital sign change trend recognition structure.

[0012] As a further aspect of the present invention, the vital sign event screening module includes: The channel index association submodule searches for the corresponding patient's vital sign acquisition channel information according to the channel number in the vital sign change trend identification structure, identifies the channel number associated with body temperature, heart rate, and respiratory rate, and matches the vital sign name with the channel number to obtain a set of vital sign channel numbers. The vital signs sequence extraction submodule calls the set of vital signs channel numbers, reads the sampled values ​​of body temperature, heart rate and respiratory rate in a continuous time segment according to the channel number, connects the order of the sampled values, and obtains a set of vital signs time series. The trend change identification submodule, based on the set of vital signs time series, determines the direction of change of each vital sign over time, identifies the numerical change segment that reverses direction after a continuous upward change segment, and pairs the vital sign name with the corresponding time segment to obtain the set of vital sign trend change data.

[0013] As a further aspect of the present invention, the intervention task generation module includes: The time segment matching submodule, based on the time segment corresponding to the vital signs in the set of data on sudden changes in vital signs trends, calls the interference time period recorded in the vital sign change trend identification structure, compares the start and end boundaries of the time segment, and obtains an abnormal vital sign number table when any vital sign becomes abnormal. The task status assignment submodule, based on each vital sign record in the abnormal vital sign number table, assigns a status field as pending intervention and unclosed loop for records with overlapping entries, and synchronously binds the status value with the vital sign number and the time interval to obtain a set of task status tags. The task field aggregation submodule extracts the vital sign channel number, corresponding time period, and status value from each data entry based on the task status label set, and incorporates the content into the same structure to obtain the task data structure to be closed.

[0014] As a further aspect of the present invention, the intervention response execution state construction module includes: The channel status matching submodule, based on the vital sign channel number recorded in the data structure of the task to be closed, queries the list of currently active channels of the nursing execution terminal, extracts the data item that matches the task channel number, and determines the activation status of the task channel based on the matching relationship of the numbers, thereby obtaining the task channel activation status list. The control device identification submodule calls the channel number in the task channel activation status list, searches for the associated device number in the control path configuration, extracts the pairing information between the device and the channel, and adds it to the task structure field to obtain the task control device mapping table. The task status determination submodule collects the current response status field of the device based on the device number in the task control device mapping table, compares the device number bound to each task with the response value, assigns the corresponding status label to the task data, and obtains a set of task execution response statuses.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by jointly analyzing temperature trends and communication anomalies over a period of time, a temporal correspondence between vital sign fluctuations and channel interruptions is established. Data segments with closed-loop interruption risks are screened, and abnormal segments are marked with task status by combining the continuous change trends of body temperature, heart rate, and respiratory rate. The activation status of nursing terminals is paired with the control device number to complete the hierarchical generation of tasks, status identification, and response matching. This promotes the transformation of vital sign recognition from static indicators to trend linkage, and extends task scheduling from manual drive to closed-loop control. It enhances the continuity of vital sign perception, the ability to identify communication interference, and the efficiency of intervention execution in ward scenarios. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a system flowchart of the present invention; Figure 2 This is a system block diagram of the present invention; Figure 3 This is a flowchart of the skin thermal sensitivity recognition module in this invention; Figure 4 This is a flowchart of the communication anomaly screening module in this invention; Figure 5 This is a flowchart of the vital sign event screening module in this invention; Figure 6 This is a flowchart of the intervention task generation module in this invention; Figure 7 This is a flowchart of the intervention response execution state construction module in this invention. Detailed Implementation

[0018] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0019] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0020] This invention provides a real-time monitoring and intervention response system for patient vital signs in smart oncology wards, such as... Figure 1-2 The diagram shown illustrates a real-time monitoring and intervention response system for patient vital signs in a smart oncology ward. The system includes: The skin thermal state recognition module acquires temperature sensor data of the subclavian and ventral regions of patients in the smart oncology ward, analyzes the temperature change direction of the thermal nodes in two consecutive time periods, filters the node numbers that continuously rise in temperature in the two periods, and combines the node numbers with the temperature rise trend state structure to obtain the monitoring node set. The communication anomaly screening module locates the connected vital sign acquisition device based on the node number in the monitoring node set, identifies the signal continuity in each time period in the frequency hopping channel of the device, determines the time period corresponding to sudden interruption, data packet loss and signal drop, and pairs the time period with the corresponding channel number to obtain the vital sign change trend identification structure. The vital sign event screening module associates the corresponding patient’s vital sign data channels with the channel number recorded in the vital sign change trend identification structure. It extracts continuous data sequences of body temperature, heart rate and respiratory rate from the associated channels, identifies the change trend of each vital sign parameter, determines whether there are continuous rises and reverse mutations in segments, and pairs the vital sign name with the corresponding time segment to obtain a set of vital sign trend mutation data. The intervention task generation module compares the time intervals of the data pairs in the data set of sudden changes in vital signs with the time intervals recorded in the structure for identifying trends in vital signs changes to determine whether there are overlapping coverage intervals. For the data pairs of vital signs that have overlap, the module assigns an unclosed state to be intervened. The module pairs the vital sign channel number, time interval, and state field to obtain the data structure of the task to be closed. The intervention response execution status construction module is based on the vital sign channel number recorded in the data structure of the task to be closed, matches the current activation status of the nursing execution terminal in the smart oncology ward, obtains the device number that has a control relationship with the vital sign channel, matches the channel number and device number of each task, divides the response tasks according to the device response status, and obtains the task execution response status set.

[0021] The monitoring node set includes node number and temperature rise trend status structure; the vital sign change trend identification structure includes time period, channel number pairing information, and signal anomaly type; the vital sign trend change data set includes vital sign name, time segment pairing, and change trend type; the task data structure to be closed-loop includes vital sign channel number, time period, and status field; and the task execution response status set includes channel number, device number, and device response status.

[0022] Specifically, such as Figure 2 , 3 As shown, the skin thermal sensitivity recognition module includes: The data stream receiving submodule acquires the placement point numbers in the subclavian and ventral regions of the skin thermal state recognition module, detects the temperature sensor output data corresponding to the placement point number in the current cycle, distinguishes the region to which the temperature data belongs according to the number, and binds the temperature value corresponding to the number to obtain the placement point temperature corresponding list. First, the data stream receiving process accesses the underlying memory address of the skin thermal state recognition module to retrieve the sensor placement topology configuration information for the subclavian and ventral regions. During the initialization phase, it searches a preset point number index table to identify the number range corresponding to the subclavian region as 001 to 050 and the number range corresponding to the ventral region as 051 to 100. Within the current data acquisition cycle, this operation continuously monitors the raw data frames transmitted back from the external bus. Each frame contains a synchronization header, the sensor's unique physical address, the placement number, and the raw value after analog-to-digital conversion. In the specific detection stage, this process extracts the temperature sensor output data matching a specific placement number within the current cycle. By establishing a mapping logic between the placement number and the physical acquisition interface, it parses the real-time temperature sampling value corresponding to each number. For example, when number 005 is detected, its corresponding temperature value is retrieved as 36.75 degrees Celsius. Subsequently, this process executes the region assignment determination logic, classifying the extracted temperature data into either the subclavian region data cluster or the ventral region data cluster based on the range to which the number belongs. To achieve structured data storage, this process performs a binding operation, associating each individual temperature value with its corresponding sampling point number, acquisition timestamp, and region attribute. In the data processing logic, the process first performs linear compensation calculations on the input raw sample values. By retrieving the sensor's factory-calibrated zero-point offset coefficient of 1.02 and gain coefficient of 0.98, linear correction logic is executed: multiplying the original sample value of 36.75 degrees Celsius by the gain coefficient of 0.98 yields 36.015, which is then added to obtain the corrected effective temperature value of 37.035 degrees Celsius. This processing eliminates performance differences between different sensors, making the sampling temperatures in the subclavian and ventral regions highly comparable, resulting in a list of corresponding sampling temperature values.

[0023] The temperature change submodule is based on the list of temperature correspondence of the distribution points. It calls the temperature values ​​of the same distribution point number in two consecutive time periods, compares the direction of temperature change in the two periods according to the number order, and extracts the distribution point numbers that show an upward change in both time periods to obtain a set of continuously rising distribution point numbers. First, the temperature change processing accesses the list of temperature observation points. By retrieving the time axis identifier, it synchronously retrieves historical temperature values ​​for the same observation point number within the current time period and the immediately preceding time period. Using the observation point number as the primary index, it extracts the measurement results for two consecutive observation periods from the cache. To identify the direction of temperature evolution, this operation executes difference comparison logic, calculating the arithmetic difference between the temperature value of the current period and the temperature value of the previous period, and comparing this difference with a preset change sensitivity threshold. The significant change threshold set for this process is 0.02 degrees Celsius. In actual calculation, if the difference is greater than 0.02 degrees Celsius, the change direction of that number within that time period is determined to be positive. For example, for number 012, the temperature in the previous time period was 35.80 degrees Celsius, and the current time period is 35.85 degrees Celsius, with a difference of 0.05 degrees Celsius. Since 0.05 is greater than 0.02, it is determined to be an upward change. Subsequently, the process further traces the performance of the number in an earlier consecutive time period. If the number meets the criterion of a difference greater than 0.02 in both adjacent time intervals, it conforms to the logical characteristic of continuous increase. In the processing logic, this process uses a logical AND operation to filter out the point numbers with positive slopes in both statistical windows. For example, number 012 increases by 0.05 degrees Celsius in the first stage and by 0.03 degrees Celsius in the second stage. By extracting the continuity feature of this number, it is classified into the high-temperature trend pool. The trend verification in two time periods effectively filters out random temperature rise noise caused by instantaneous local pressure on the human body or fluctuations in environmental airflow, resulting in a set of continuously increasing point numbers.

[0024] The temperature rise trend extraction submodule is based on a continuously rising set of monitoring point numbers. It calls the temperature change direction data within two time periods, associates the direction value corresponding to the number, and extracts the number and direction information with the temperature rise trend to obtain a set of monitoring nodes. First, the temperature rise trend extraction process is based on a continuously rising set of point numbers. A reverse indexing mechanism retrieves the precise temperature change direction and numerical change data for each number within two consecutive time periods, extracting the polarity of the direction and associating it with the slope intensity value corresponding to each number. To quantify the significance of the temperature rise trend, this operation executes trend weight fusion logic. The weight coefficient for the first stage (the more distant time period) is set to 0.35, and the weight coefficient for the second stage (the more recent time period) is set to 0.65. The temperature change in the first stage is multiplied by 0.35, and the temperature change in the second stage is multiplied by 0.65. The two products are then summed to obtain the comprehensive temperature rise trend score for that number. In a specific example, for number 015, the temperature increased by 0.12 degrees Celsius in the first stage and 0.18 degrees Celsius in the second stage. Substituting these values ​​into the logic yields 0.042 and 0.117, respectively. Adding these two values ​​gives a comprehensive temperature rise trend score of 0.159. This process sets a trend judgment benchmark value of 0.05, which is obtained through median statistics from a large amount of clinical skin heating experiment data. Since 0.159 is greater than 0.05, the process confirms that the node has a valid warming trend and extracts its associated node and direction details. The process further categorizes nodes with warming trends according to their scores: a score between 0.05 and 0.10 is defined as a Level 1 trend, between 0.10 and 0.20 as a Level 2 trend, and a score greater than or equal to 0.20 as a Level 3 trend, thus obtaining a set of monitoring nodes.

[0025] Specifically, such as Figure 2 , 4 As shown, the communication anomaly screening module includes: The device mapping and positioning submodule is based on the node number in the monitoring node set. According to the structure detection of the connection vital sign acquisition device identifier corresponding to the number, it extracts the frequency hopping communication device number associated with each node number, verifies the correspondence between the device and the node, and obtains the node communication device index table. First, the device mapping and positioning process is based on the node numbers in the monitoring node set. It accesses the underlying hardware address allocation table to retrieve the structural distribution description corresponding to each number, parses the connection relationship between the node's geographical coordinates and the physical acquisition link, and queries the cross-mapping matrix associated with the vital sign acquisition device identifier. During execution, this operation extracts the frequency-hopping communication device number associated with each node number. This number represents the hardware terminal identifier responsible for uploading the data from that location via wireless frequency hopping. To ensure mapping accuracy, this process executes device node consistency verification logic, sending a handshake signal to the target frequency-hopping device and retrieving the logical channel mask currently allocated to that device. If the device identifier returned by the handshake signal matches the number recorded in the configuration library, and the target node number is included in the active location list reported by the device, then the verification is successful. For example, if node 018 is associated with frequency-hopping device TX-88 in the configuration, this process verifies the real-time validity of this correspondence by verifying the online status and location load of TX-88. In the processing logic, a device retrieval mechanism based on spatial index is established. The node number is converted into binary address bits and compared with the address bits in the device list. If the number and address match, it is confirmed as a match. By integrating the node location, device number and signal link status, a node communication device index table is obtained.

[0026] The signal continuity detection submodule extracts the signal data stream sequence within a continuous time period under the frequency hopping channel based on the node communication device index table and the device number. It detects the interval changes between data according to the time sequence, identifies the periodic points where the signal transmission is interrupted, the data is missing, and the intensity drops sharply, and obtains a set of abnormal signal time segments. First, the signal continuity detection process, based on the node communication device index table, identifies the device number requiring link analysis and extracts the signal quality data stream sequence of that device over 120 consecutive time periods in the frequency hopping channel from the frequency hopping management module. This sequence includes received power intensity, bit error rate, and arrival timestamps for each frame. The process executes interval change detection logic sequentially, calculating the time step between adjacent data frames. The preset standard transmission interval is 20 milliseconds. When executing the anomaly identification logic, first, an interruption is detected; if the interval between adjacent frames exceeds 100 milliseconds, it is marked as a signal interruption. Second, missing values ​​are detected; if the data frame format is correct but the payload area is empty, it is marked as missing values. Finally, a sudden drop in strength is detected by calculating the average signal strength of the previous 50 periods. If the strength of the current period is 15 dB lower than this average, it is considered a sudden drop in strength. For example, for device TX-88, at the 500th millisecond, the signal strength suddenly drops from -60 dB / mW to -85 dB / mW, and subsequently, there are three periods of empty values; this process identifies this interval as an anomaly. In the processing logic, an abnormal duration accumulation logic is introduced to connect consecutive abnormal points into segments and calculate the start offset and end offset of each segment to obtain a set of abnormal signal time segments.

[0027] The abnormal time period identification submodule is based on the abnormal signal time segment set, matches the corresponding channel number, and combines the data position according to the association order between each channel number and the corresponding time segment to obtain the vital sign change trend identification structure. First, the abnormal period identification process accesses the abnormal signal time segment set and matches the physical channel number corresponding to each segment by searching the frequency hopping device and frequency mapping table. Since the anti-interference capability of frequency hopping communication varies at different frequency points, this process executes a combination association logic of data locations, projecting discrete abnormal points onto the channel matrix along the time axis according to the association order between each channel number and its corresponding time segment. This process defines the severity of the period by calculating the channel interference duty cycle, i.e., the ratio of interference duration to the total observation duration. In specific calculations, this process sets the interference duty cycle threshold to 0.12. For channel CH-15, if the cumulative abnormal segment duration is 1.5 seconds within a 10-second observation window, the duty cycle is 0.15, exceeding the threshold, and the process classifies it as an interference period. Subsequently, the operation performs data location reorganization, encapsulating the start time, end time, and channel number into a time period identifier. For example, CH-15 is bound to the time period from 10:00:01 to 10:00:11. By aggregating the interference features of multiple channels, interference patterns with specific frequency characteristics are identified, resulting in a structure for identifying the trend of vital signs changes.

[0028] Specifically, such as Figure 2 , 5 As shown, the vital sign event filtering module includes: The channel index association submodule searches for the corresponding patient's vital sign acquisition channel information according to the channel number in the vital sign change trend identification structure, identifies the channel number associated with body temperature, heart rate, and respiratory rate, and matches the vital sign name with the channel number to obtain a set of vital sign channel numbers. First, the channel index association process retrieves the channel number from the vital sign change trend identification structure. Using this as a clue, it enters the application-layer channel mapping library and searches for the corresponding patient's vital sign acquisition channel configuration information according to the number order. Since there is a one-to-many logical mapping between physical channels and physiological channels, this operation executes association identification logic, extracting all vital sign parameter channels carried by the affected channels. By retrieving channel metadata, it identifies the specific channel number associated with the three core indicators: body temperature, heart rate, and respiratory rate. During execution, this process maps the vital sign name string to the extracted channel number in a one-to-one correspondence. For example, it finds that channel CH-15 carries the body temperature channel T-01, heart rate channel H-01, and respiratory channel R-01 for patient number P-001. A ternary structure is then established to bind the patient identifier, vital sign type, and physical channel number. The process sets up channel priority logic, prioritizing the heart rate channel with the highest real-time requirements. At the same time, to cope with channel number changes that may be caused by hot-plugging of sensors, logic is set up to identify channel ID changes and maintain channel mapping relationships to ensure that the system can continuously and correctly identify the source of key vital signs and obtain a set of vital sign channel numbers.

[0029] The vital signs sequence extraction submodule calls the set of vital signs channel numbers, reads the sampled values ​​of body temperature, heart rate and respiratory rate in a continuous time segment according to the channel number, connects the sampled values ​​in sequence, and obtains the set of vital signs time series. First, the vital sign sequence extraction process calls a set of vital sign channel numbers as an instruction for concurrent retrieval of a large-scale physiological vital sign database. Based on the channel numbers in the set, it reads raw sampled values ​​covering a continuous time segment from a distributed cache. The reading range is set to 60 seconds before the start of the abnormal period and 60 seconds after the end to ensure the integrity of the trend analysis. For the three indicators—body temperature, heart rate, and respiratory rate—asynchronous reading is performed according to their respective sampling frequencies, followed by time base alignment. In the processing logic, this process performs a sequential concatenation operation of sampled values. For high-frequency signals like heart rate, the time sequence number of the sampling points is detected; if a minor break is found, nearest neighbor interpolation is used to fill it. For example, 3000 sampling points from heart rate channel H-01 within 120 seconds are retrieved, with values ​​distributed between 65 and 85 times per minute. This process arranges these sampled values ​​according to millisecond-level timestamps, constructing a continuous numerical curve. The process executes data denoising logic, which processes the original sequence by calling a fifth-order Butterworth low-pass filter algorithm. The cutoff frequency is set according to the characteristics of the vital signs, and motion artifacts caused by the patient turning over are filtered out to obtain a set of time series of vital signs.

[0030] The trend change identification submodule is based on the time series set of vital signs. It judges the direction of change of each vital sign over time, identifies the numerical change segment that reverses the direction after a continuous upward change segment, and pairs the vital sign name with the corresponding time segment to obtain the set of vital sign trend change data. First, the trend abrupt change identification process is based on a set of vital sign time series data. A sliding window slope analysis is performed on each vital sign curve to determine the direction of change of the vital sign over time. This process sets the sliding window length to 50 sampling periods and calculates the slope by taking the first derivative within the window. During execution, this operation calculates the average rate of change of the values ​​within the window. If the rate of change for 10 consecutive windows is greater than zero, it is determined to be a continuously upward changing segment. Subsequently, this process focuses on monitoring the reversal of slope polarity. The judgment logic is set as follows: after a continuously rising segment, if the slope of 3 consecutive sampling points becomes negative, and the absolute value of the instantaneous slope of the decline exceeds 1.5 times the average slope of the rising phase, it is determined to be a trend abrupt change. For example, if a patient's heart rate slowly rises from 75 beats per minute to 95 beats per minute over 300 seconds, and then suddenly drops to 80 beats per minute within 5 seconds, this process identifies this segment of numerical change reversing the direction and records the start, turning point, and end time of the abrupt change. By pairing the names of vital signs with the corresponding identified time periods, it is possible to automatically distinguish between physiological slow regulation and pathological sudden changes, thereby obtaining a set of data on sudden changes in vital sign trends.

[0031] Specifically, such as Figure 2 , 6 As shown, the intervention task generation module includes: The time segment matching submodule is based on the time segment corresponding to the vital signs in the data set of sudden changes in vital signs trends. It calls the interference time period recorded in the vital sign change trend identification structure, compares the start and end boundaries of the time segment, and obtains the abnormal vital sign number table when any vital sign becomes abnormal. First, the time segment matching process, based on the time segments corresponding to the vital signs recorded in the vital sign trend mutation data set, calls the interference time periods recorded in the vital sign change trend identification structure. This process executes time-domain overlap comparison logic, calculating the overlap duration of the two intervals on the time axis by extracting the start and end boundaries of the vital sign mutation interval and the start and end boundaries of the interference time period. In the specific calculation logic, this process defines a coverage ratio coefficient, which is obtained by dividing the overlap duration by the total duration of the vital sign mutation segment. This process sets the coverage judgment threshold to 0.25. For example, if a heart rate mutation lasts for 20 seconds, and the channel interference corresponding to that channel overlaps for 10 seconds at the same time point, the calculated coverage ratio is 0.5. Since 0.5 is greater than 0.25, this process determines that the heart rate mutation item has a significant interval coverage relationship with the communication interference and marks it as a disturbed pseudo-mutation. This processing logic, by excluding the data segments affected by interference and retaining the records of true physiological mutations, obtains an abnormal vital sign number table.

[0032] The task status assignment submodule is based on each vital sign record in the abnormal vital sign number table. For records with overlap, the status field is assigned as pending intervention and not closed loop. The status value is then synchronously bound to the vital sign number and the time interval to obtain a set of task status tags. First, the task status assignment process initiates the task's full lifecycle status initialization based on each record in the abnormal vital sign number table. For records with overlapping relationships, this process adds a status field to the task attributes and assigns it an initial value of "Needing Intervention, Not Closed." This status code indicates that although the abnormal vital sign is affected by interference and exhibits pseudo-mutation characteristics, it still requires manual or subsequent logical confirmation to complete the final closed-loop processing. During the assignment logic, this process performs a field synchronization binding operation, encapsulating the task's status value, the corresponding vital sign channel number, the specific time period affected, and the characteristic intensity score of the mutation. For example, for the abnormality occurring in temperature channel T-01 at 1500 seconds, the status value is assigned as "Needing Intervention, Not Closed," and it is strongly associated with the channel number and the timestamp. This process sets up a timestamp recording logic for status updates to ensure that every status change is traceable. By introducing the concept of task closure, it ensures the attribution of responsibility for abnormal information, resulting in a set of task status tags.

[0033] The task field aggregation submodule extracts the vital sign channel number, corresponding time period, and status value from each data entry based on the task status label set, and incorporates the content into the same structure to obtain the data structure of the task to be closed-loop. First, the task field aggregation process retrieves all raw data items from the task status label set and performs field extraction and structure reorganization operations. This process precisely extracts the vital sign channel number, the time period of the anomaly, and the current task status value from each task record. To improve subsequent processing efficiency, this operation performs data anonymization and redundant field cleanup, retaining only core decision attributes. In the processing logic, the extracted content is incorporated into a standardized structured data object. For example, the channel number H-01, time period information, and status value (not yet closed) are combined into an independent task data unit. This process performs field integrity verification; if a record is found to have missing time period information, it automatically backtracks to previous steps to complete the data. This aggregation logic establishes a unified task view, enabling the backend display terminal to parse it in a standardized way. By integrating scattered attribute content into the same structure, data exchange efficiency is improved, resulting in the task data structure to be closed.

[0034] Specifically, such as Figure 2 , 7 As shown, the intervention response execution status construction module includes: The channel status matching submodule is based on the vital sign channel number recorded in the data structure of the task to be closed, and queries the list of currently active channels of the nursing execution terminal. It extracts the data items that match the task channel number, and judges the activation status of the task channel according to the matching relationship of the numbers, and obtains the task channel activation status list. First, the channel status matching process initiates a synchronous query of the current operating environment of the nursing execution terminal based on the vital sign channel numbers recorded in the task data structure to be closed. This process extracts a list of currently active channels by calling the terminal's device driver interface. This list contains the numbers of all physically online channels that are continuously uploading data. Membership judgment logic is then executed, comparing the channel numbers in the task with the active list item by item. The judgment logic is set as follows: if the task channel number exists in the current active list and its physical layer received signal strength (RSSI) is greater than -90 dBm, the channel is considered to be in normal activation; if the number does not exist or the signal strength is below the threshold, it is considered inactive. For example, for the heart rate channel H-01 in the task, the query finds that the channel is in the terminal's active list and its current received signal strength is in the normal reception range of -70 dBm, so it is assigned an active status. This process combines the judgment result, the task number, and the channel's physical status to detect physical faults in the channel in real time and obtain a list of task channel activation statuses.

[0035] The control device identification submodule calls the channel number in the task channel activation status list, searches for the associated device number in the control path configuration, extracts the pairing information between the device and the channel, and adds it to the task structure field to obtain the task control device mapping table. First, the control device identification process calls the channel number from the task channel activation status list as a query parameter to the underlying control path configuration table to find the hardware device number associated with each channel in terms of control, power supply, or data link. This operation executes multi-level path tracing logic to extract precise pairing information between devices and channels. During execution, this process associates the identified control device number, physical port number, and corresponding control protocol type. For example, it identifies that the data flow of channel H-01 is controlled by a microprocessor numbered MCU-99, and the control path is the 3rd general purpose input / output port. This newly added device dimension information is added to the original task structure fields to complete the link mapping from the physiological channel to the physical controller. This process sets up redundant path identification logic. If the main controller returns a failure flag, it automatically associates the backup controller number. Through deep analysis of the underlying topology, it provides a clear execution object for automated intervention, obtaining the task control device mapping table.

[0036] The task status determination submodule collects the current response status field of the device based on the device number in the task control device mapping table, compares the device number bound to each task with the response value, assigns the corresponding status label to the task data, and obtains a set of task execution response statuses. First, the task status determination process, based on the device number in the task control device mapping table, initiates a real-time response status acquisition command to the underlying hardware via the control bus. This process extracts the current status register fields of the device, including the operating mode, error code, and the response delay time of the most recent command. The operation execution status comparison logic matches and analyzes the device number bound to each task with its returned real-time response value. The determination logic is set as follows: if the error code returned by the device is zero (no fault) and the response delay is less than 10 milliseconds, a "normal execution response" label is assigned to the task data; if the response value is a busy status code, a "standby" label is assigned; if there is no response within a specified time (50 milliseconds), a "device fault" label is assigned. For example, for the device MCU-99 associated with the task number, if a response value of zero is collected, its status label in the task record is updated to "normal execution response." This process achieves closed-loop confirmation of the task execution status by collecting feedback from the device side, accurately distinguishing the execution risks of the task, and obtaining a set of task execution response statuses.

[0037] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A real-time monitoring and intervention response system for patient vital signs in smart oncology wards, characterized in that: The system includes: The skin thermal state recognition module acquires temperature sensor data of the subclavian and ventral regions of patients in the smart oncology ward, analyzes the direction of continuous periodic temperature change of the thermal nodes, and obtains the set of monitoring nodes. The communication anomaly screening module, based on the set of monitoring nodes, identifies the continuity of the periodic signal of the frequency hopping channel, extracts the time periods corresponding to sudden interruptions, data packet loss, and signal drops, and pairs them with channel numbers to obtain the identification structure of the trend of vital signs. Based on the vital sign change trend recognition structure, the vital sign event screening module extracts continuous sequences of body temperature, heart rate, and respiratory rate, determines the time segments of continuous rise and reverse mutation, pairs the vital sign name with the time segment, and obtains a set of vital sign trend mutation data. The intervention task generation module, based on the data pairs of the vital signs time periods in the set of data on sudden changes in vital signs trends, assigns an unclosed state to be intervened when any vital sign becomes abnormal, thus obtaining the data structure of the task to be closed. The intervention response execution status construction module, based on the data structure of the task to be closed, matches the activation status of the nursing execution terminal, divides the response tasks according to the response status, and obtains a set of task execution response statuses.

2. The real-time monitoring and intervention response system for patient vital signs in a smart oncology ward according to claim 1, characterized in that, The monitoring node set includes node number and temperature rise trend status structure; the vital sign change trend identification structure includes time period, channel number pairing information, and signal anomaly type; the vital sign trend mutation data set includes vital sign name, time segment pairing, and change trend type; the task data structure to be closed-loop includes vital sign channel number, time period, and status field; and the task execution response status set includes channel number, device number, and device response status.

3. The real-time monitoring and intervention response system for patient vital signs in a smart oncology ward according to claim 1, characterized in that, The continuous periodic temperature of the thermal node refers to the trend of continuous temperature rise of the target monitoring point on the body surface over a continuous time period. The continuity of the frequency hopping channel periodic signal refers to the measure of whether the signal of the frequency hopping channel is uninterrupted within a continuous time.

4. The real-time monitoring and intervention response system for patient vital signs in a smart oncology ward according to claim 1, characterized in that, The cross-coverage interval refers to the intersection segment where abnormal vital signs and communication interference overlap in time; The term "unclosed-loop state to be intervened" refers to the status of vital signs and events for which the intervention task has not yet been completed.

5. The real-time monitoring and intervention response system for patient vital signs in a smart oncology ward according to claim 1, characterized in that, The skin thermal sensitivity recognition module includes: The data stream receiving submodule acquires the placement numbers of the skin thermal sensitivity identification points in the subclavian and ventral regions, detects the temperature sensor output data corresponding to the placement number in the current cycle, distinguishes the regions according to the number of the temperature data, and binds the temperature value corresponding to the number to obtain a list of placement temperature correspondences. The temperature change submodule, based on the list of temperature correspondences of the distribution points, calls the temperature values ​​of the same distribution point number in two consecutive time periods, compares the direction of temperature change in the two periods according to the number order, and extracts the distribution point numbers that show an upward change in both time periods to obtain a set of continuously rising distribution point numbers. The temperature rise trend extraction submodule, based on the continuously rising distribution point number set, calls the temperature change direction data within two time periods, associates the direction value corresponding to the number, extracts the number and direction information with the temperature rise trend, and obtains the monitoring node set.

6. The real-time monitoring and intervention response system for patient vital signs in a smart oncology ward according to claim 1, characterized in that, The communication anomaly screening module includes: The device mapping and positioning submodule, based on the node number in the monitoring node set, detects the identification of the connected vital sign acquisition device according to the structure corresponding to the number, extracts the frequency hopping communication device number associated with each node number, verifies the correspondence between the device and the node, and obtains the node communication device index table. The signal continuity detection submodule extracts the signal data stream sequence within a continuous time period under the frequency hopping channel according to the device number based on the node communication device index table, detects the interval changes between data according to the time sequence, identifies the periodic points where the signal transmission is interrupted, the value is missing, and the intensity drops sharply, and obtains a set of abnormal signal time segments. The abnormal time period identification submodule matches the corresponding channel number based on the abnormal signal time segment set, and combines the data positions according to the association order between each channel number and the corresponding time segment to obtain the vital sign change trend identification structure.

7. The real-time monitoring and intervention response system for patient vital signs in a smart oncology ward according to claim 6, characterized in that, During the process of the interval change between the detection data: when the time interval between two adjacent data points is greater than a preset threshold, and at least one of the adjacent data points has a signal strength value attenuation, it is determined to be a periodic point where signal transmission abnormality occurs. In the process of matching the corresponding channel number: each channel number and the corresponding time segment are matched one by one in time order. When the same time segment appears in multiple consecutive association processes, it is incorporated into the vital sign change trend recognition structure.

8. The real-time monitoring and intervention response system for patient vital signs in a smart oncology ward according to claim 1, characterized in that, The vital sign event screening module includes: The channel index association submodule searches for the corresponding patient's vital sign acquisition channel information according to the channel number in the vital sign change trend identification structure, identifies the channel number associated with body temperature, heart rate, and respiratory rate, and matches the vital sign name with the channel number to obtain a set of vital sign channel numbers. The vital signs sequence extraction submodule calls the set of vital signs channel numbers, reads the sampled values ​​of body temperature, heart rate and respiratory rate in a continuous time segment according to the channel number, connects the order of the sampled values, and obtains a set of vital signs time series. The trend change identification submodule, based on the set of vital signs time series, determines the direction of change of each vital sign over time, identifies the numerical change segment that reverses direction after a continuous upward change segment, and pairs the vital sign name with the corresponding time segment to obtain the set of vital sign trend change data.

9. The real-time monitoring and intervention response system for patient vital signs in a smart oncology ward according to claim 1, characterized in that, The intervention task generation module includes: The time segment matching submodule, based on the time segment corresponding to the vital signs in the set of data on sudden changes in vital signs trends, calls the interference time period recorded in the vital sign change trend identification structure, compares the start and end boundaries of the time segment, and obtains an abnormal vital sign number table when any vital sign becomes abnormal. The task status assignment submodule, based on each vital sign record in the abnormal vital sign number table, assigns a status field as pending intervention and unclosed loop for records with overlapping entries, and synchronously binds the status value with the vital sign number and the time interval to obtain a set of task status tags. The task field aggregation submodule extracts the vital sign channel number, corresponding time period, and status value from each data entry based on the task status label set, and incorporates the content into the same structure to obtain the task data structure to be closed.

10. The real-time monitoring and intervention response system for patient vital signs in a smart oncology ward according to claim 1, characterized in that, The intervention response execution status construction module includes: The channel status matching submodule, based on the vital sign channel number recorded in the data structure of the task to be closed, queries the list of currently active channels of the nursing execution terminal, extracts the data item that matches the task channel number, and determines the activation status of the task channel based on the matching relationship of the numbers, thereby obtaining the task channel activation status list. The control device identification submodule calls the channel number in the task channel activation status list, searches for the associated device number in the control path configuration, extracts the pairing information between the device and the channel, and adds it to the task structure field to obtain the task control device mapping table. The task status determination submodule collects the current response status field of the device based on the device number in the task control device mapping table, compares the device number bound to each task with the response value, assigns the corresponding status label to the task data, and obtains a set of task execution response statuses.

Images