Drug delivery control method and system for surgical postoperative analgesia pump

Abstract

The invention relates to the technical field of automatic drug delivery control, in particular to a drug delivery control method and system for a surgical postoperative analgesia pump, and the method comprises the following steps: obtaining a respiratory rate and a pressing behavior sequence of a patient, recognizing a fluctuation trend and request growth, judging a synchronous response time period, and adjusting an infusion rhythm and dosage. Pressure curve characteristics are analyzed to recognize abnormal operation, emergency infusion is executed in combination with signal linkage, and a dynamic drug administration control instruction is generated. According to the method, the severe fluctuation identification is generated by obtaining the continuous breathing frequency change trend of the patient and combining with breathing extreme value judgment, the linkage time period mark is formed by matching with sequential change trend analysis of the request behavior of the patient, and the basic infusion frequency and dosage are adjusted in real time to meet the actual analgesia requirement; an intelligent recognition and quick response mechanism under the dominant physiological state is achieved, the adaptability, safety and timeliness of the analgesia process are enhanced, and the problems of control lag and judgment blind areas in a fixed parameter mode are effectively relieved.

Description

Technical Field

[0001] This invention relates to the field of automated drug delivery control technology, and in particular to a method and system for controlling drug delivery using a postoperative analgesia pump. Background Technology

[0002] The field of automated drug delivery control technology involves automating the drug infusion process to achieve quantitative, timed, or on-demand drug administration, with wide applications in clinical scenarios such as postoperative analgesia, intensive care, and chronic disease management. This technology primarily encompasses core components such as drug delivery devices, flow control mechanisms, patient self-control interfaces, feedback signal acquisition devices, and control algorithms. Automated drug delivery control technology compares physiological parameters collected by sensors with preset drug delivery parameters to dynamically adjust the infusion rate, thereby improving drug efficacy and safety. Systematic technological development in this field focuses on continuous optimization of hardware stability, control precision, human-computer interaction experience, and clinical adaptability. Traditional surgical postoperative analgesia pumps involve the continuous or intermittent infusion of analgesics into the body after surgery to alleviate postoperative pain. This method combines a preset rate of continuous infusion with patient self-controlled dosing to achieve analgesia control. In traditional methods, analgesia pumps typically consist of a programmable infusion pump, a reservoir, and infusion tubing, relying on healthcare professionals to set a baseline infusion rate and a single self-controlled dose. Some devices can be configured with a lockout time to prevent overdose. Dosing control is mainly based on parameters set according to time intervals and dosage limits, and it does not have the ability to adjust according to changes in the patient's real-time condition. The control method is mainly completed by mechanical buttons, electric push or elastic pump to drive the infusion of drug solution.

[0003] In existing technologies, the control of postoperative analgesia pumps relies on fixed time intervals and dosage parameters, which cannot detect changes in the patient's real-time physiological state. This results in a lack of dynamic response capability in drug infusion, a disconnect between the baseline infusion rate and changes in the patient's pain, and a high risk of insufficient analgesia or drug accumulation. Patient self-control behavior is based solely on mechanical pressing, lacking an intelligent recognition mechanism, and cannot determine whether it is driven by real pain. Ineffective or abnormal pressing will interfere with analgesia assessment and drug use strategies. Infusions will still be performed at preset doses even when requests are frequent but there is no actual need, which also increases the potential risk of drug overdose. The system lacks the ability to make judgments based on physiological fluctuations and behavioral signals, making it difficult to respond promptly and effectively to sudden needs. Summary of the Invention

[0004] To address the technical problems existing in the prior art, embodiments of the present invention provide a method for controlling drug delivery using a postoperative analgesia pump, comprising the following steps: S1: Obtain the continuous respiratory rate sequence of the patient in the postoperative analgesia management scenario, divide it according to the complete respiratory cycle, extract the trough and peak within the cycle, compare the real-time cycle peak with the respiratory extreme value during the anesthesia stage, and generate a respiratory fluctuation trigger indicator. S2: Call the respiratory fluctuation triggering identifier, obtain the time distribution sequence of the patient's medication request pressing behavior, divide a uniform sliding time window, count the number of requests in the real-time window and the previous window, and generate a sudden request linkage time period marker. S3: Using the aforementioned sudden request linkage time period marker, obtain the continuous infusion time points and infusion dose recorded in the postoperative analgesia pump basal infusion behavior, count the number of basal infusions within the marked linkage time period, and generate the basal drug administration rhythm adjustment results; S4: Based on the basic drug administration rhythm adjustment results, obtain the high-frequency pressure waveform generated by each press during the use of the analgesia pump by the patient, extract the complete pressure curve from the initial pressure application to the release endpoint, perform inflection point number identification and rebound change trend judgment, and sort and compare the real-time curve with the previous three curves to generate an autonomous operation identification mark.

[0005] As a further embodiment of the present invention, the respiratory fluctuation triggering identifier includes an amplitude change direction trend marker, a period peak exceeding the range judgment result, and a stimulus response time point marker; the sudden request linkage period marker includes a request growth trend, a respiratory fluctuation number increase trend, and a request-fluctuation synchronization relationship; the basic dosing rhythm adjustment result includes an update of the starting interval time, an adjustment of the continuous infusion frequency, and a correction of the dose for each infusion; and the autonomous operation identification marker includes an inflection point distribution offset, a rebound segment structural abnormality, and a pressure application behavior abnormality marker.

[0006] As a further aspect of the present invention, the specific steps of S1 are as follows: S101: Obtain the continuous respiratory rate sequence of the patient in the postoperative analgesia management scenario, and split the continuous time period in the continuous respiratory rate sequence. Based on the periodic fluctuation of the inspiratory and expiratory phases in the complete respiratory cycle, analyze the respiratory waveform data frames in the cycle, extract the troughs and peaks of the waveform in the cycle, record the time index and amplitude value respectively, and generate a respiratory cycle extreme value sequence group. S102: Based on the respiratory cycle extreme value sequence group, the numerical direction of the amplitude difference sequence formed by the peak and trough of three consecutive respiratory cycles is determined. Based on whether the amplitude change direction between two adjacent cycles is consistent, three cycle direction state groups are constructed. If the amplitude change direction of three consecutive cycles is consistent, it is marked as a positive fluctuation trend state; otherwise, it is marked as a non-positive trend state, and a cycle amplitude trend label set is generated. S103: Call the set of periodic amplitude trend markers, and retrieve the corresponding periodic peak data item according to the time point marked as a positive fluctuation trend. Compare the periodic peak data item with the respiratory peak limit range recorded during the anesthesia stage in the postoperative analgesia scenario. If the periodic peak amplitude exceeds the upper limit of the respiratory extreme value during the anesthesia stage, record the time point and perform trigger state assignment processing to obtain the respiratory fluctuation trigger identifier.

[0007] As a further aspect of the present invention, the specific steps of S2 are as follows: S201: Call the respiratory fluctuation trigger identifier, obtain the time distribution sequence of the patient's medication request pressing behavior, divide the time distribution sequence into continuous sliding time windows according to a fixed step size, count the number of medication requests within the time window, and generate a sliding window request count set; S202: Based on the sliding window request quantity set, the quantity difference of the request quantity in two adjacent consecutive time windows is judged. If the request quantity in the later window is greater than the request quantity in the earlier window, the time window is marked as an increasing state. The respiratory fluctuation triggering identifier is synchronously divided into windows and the quantity is counted on the same time axis. It is judged whether the fluctuation triggering quantity in the window has increased compared with the previous window. If both are increasing, the window is marked as a linked state, and a linked state window marking sequence is generated. S203: Based on the time windows marked as being in linkage status in the linkage status window marking sequence, identify the linkage start and end time range corresponding to the time period, assign the sudden request synchronization response attribute, complete the marking operation on the linkage status window, and obtain the sudden request linkage time period mark.

[0008] As a further aspect of the present invention, the specific steps of S3 are as follows: S301: Call the sudden request linkage time period marker, obtain the continuous infusion time point and corresponding infusion dose information recorded in the postoperative analgesia pump basic infusion behavior, and search within the linkage time period to see if the basic infusion time point falls within the range, count the number of basic infusions within the linkage time period, and generate a basic infusion statistics table within the linkage time period. S302: Based on the basic infusion statistics table within the linkage period, calculate the ratio of the number of basic infusions within the linkage period to the number of medication requests within the corresponding period, determine the temporal overlap ratio between the basic infusion rhythm and the patient's request behavior, and if the overlap ratio is lower than the preset behavior matching ratio threshold, mark it as a temporal mismatch state and generate a basic infusion rhythm mismatch identifier set. S303: Using the time periods marked as time mismatch states in the basic infusion rhythm mismatch identifier set, extract the start interval time and continuous infusion frequency of real-time basic infusion, and compare them with the behavior matching benchmark configuration parameter set respectively, update the start interval time and frequency parameters, and redistribute the infusion dose each time according to the updated frequency to obtain the basic dosing rhythm adjustment result.

[0009] As a further aspect of the present invention, the specific steps of S4 are as follows: S401: Based on the basic drug administration rhythm adjustment results, obtain the high-frequency pressure waveform sequence generated by each pressing action of the patient during the use of the analgesia pump, extract the corresponding complete pressure curve from the pressure application start point to the release end point, identify the derivative change points in the curve from positive to negative or negative to positive, count the number and location, and generate the pressing pressure curve inflection point feature group. S402: Call the inflection point feature group of the pressing pressure curve, and retrieve the data of the three adjacent pressing pressure curves. After arranging them in chronological order, extract the derivative direction sequence and inflection point index sequence of the rebound segment in each curve. Compare the real-time curve with the previous three curves in terms of the number of inflection points and the trend of rebound segment changes, record the structural feature difference distribution, and generate a real-time curve structural offset data table. S403: Based on the structural feature differences recorded in the real-time curve structure offset data table, determine whether the real-time pressing curve has a reduced number of inflection points, missing rebound segments, or abnormal positional offset. If there is a structural abnormality marker, set an abnormality marker state for the pressing behavior, assign autonomous operation attributes, and obtain an autonomous operation identification mark.

[0010] As a further aspect of the present invention, the number of inflection points in the real-time curve is determined to be abnormally reduced based on the inflection point feature group extracted from the real-time pressing pressure curve. If the number of inflection points in the real-time pressing pressure curve decreases, check whether there is a missing rebound segment in the real-time pressing pressure curve, or whether the derivative direction sequence of the rebound segment deviates from the normal trend. If the rebound segment of the real-time pressing pressure curve is missing or the derivative direction sequence of the rebound segment is shifted, an abnormal flag state is set for the pressing behavior, and the pressing behavior is marked as having autonomous operation attributes. The judgment process is as follows: identify the complete pressure curve corresponding to each pressing action in the extracted real-time curve, extract the rebound segment in the pressure curve, compare the changing trend of the derivative direction of the rebound segment with the changing trend in the adjacent previous curves, and if the trend difference exceeds the preset threshold, it is judged as an abnormal state.

[0011] As a further aspect of the present invention, the method further includes step S5: S5: Using the autonomous operation identification flag and respiratory fluctuation trigger flag, determine whether they are activated at the same time node. If both are in the marked state, the analgesia pump emergency control channel responds and releases the emergency infusion action, calls the updated basic infusion parameters to perform continuous control, constructs a dual-channel linkage path dominated by respiratory fluctuation and pressure signal, and generates dynamic drug delivery control instructions. The dynamic drug delivery control commands include emergency infusion triggering, basic infusion parameters, and dual-channel linkage pathways.

[0012] As a further aspect of the present invention, the specific steps of S5 are as follows: S501: Using the autonomous operation identification flag and the respiratory wave trigger flag, the activation status of the two types of flags at the same time node is compared by time index to determine whether there is a joint marking situation with overlapping time. If both are in the marking state at the same time node, the corresponding time index is recorded and set as the linkage trigger time point to generate a flag linkage activation time set. S502: Based on the linkage triggering time point in the linkage activation time set, the analgesia pump's sudden control channel calls the associated channel response mechanism, releases the emergency infusion action corresponding to the activation state, calls the updated basic infusion parameter group, and continues to execute the continuous dose control operation, generating a dual-channel control execution set; S503: Invoke the dual-channel control to execute the centrally recorded emergency channel response and basic channel continuous control configuration, fuse the channel status according to the time priority of the respiratory fluctuation signal and the spontaneous compression signal, construct a linkage output control path with the respiratory fluctuation signal and the compression signal as the core driving factors, and generate dynamic drug administration control instructions.

[0013] A drug delivery control system for a postoperative analgesia pump includes: The respiratory fluctuation recognition module acquires the patient's continuous respiratory rate sequence in the postoperative analgesia management scenario. It extracts the peak and trough values ​​in segments according to the complete inhalation and exhalation cycle, obtains the cycle amplitude difference and forms a set of amplitude changes in three adjacent cycles. It calls the three amplitude differences and compares whether the change direction is consistent. If they show a unified direction, they are marked as a trend segment. Then, it compares the cycle peak with the upper limit of the respiratory peak recorded during the anesthesia stage to generate a respiratory fluctuation trigger identifier. The request behavior monitoring module collects the patient's pressing behavior record of the analgesia pump based on the respiratory fluctuation trigger identifier, constructs a request distribution sequence in chronological order, divides the time windows into equal long-term windows to count the number of requests within the window, and generates a sudden request linkage time period marker. Based on the sudden request linkage time period marker, the rhythm adjustment module retrieves the basic infusion time series and infusion dose record of the analgesic pump, counts the number of occurrences of basic infusion behavior within the interval, and calculates the overlap ratio with the request behavior on the time axis. If the ratio is low, the basic infusion setting parameters are adjusted to shorten the basic infusion interval and generate the basic drug administration rhythm adjustment result. The autonomous operation recognition module calls the basic drug administration rhythm adjustment results, obtains the pressure waveform data when the patient operates the analgesia pump button, identifies the complete pressing curve and extracts the inflection point distribution and rebound segment morphology, compares the real-time curve with the previous three recorded curves, determines whether there is abnormal deviation or structural missingness in the inflection point position and rebound trend, and generates an autonomous operation recognition mark. The dynamic drug delivery linkage module compares the autonomous operation identification flag and the respiratory fluctuation trigger flag on the time axis to see if the marked positions are activated simultaneously. If they overlap, it retrieves the emergency infusion control path, performs basic infusion adjustment, performs continuous infusion operation on the channel, and generates dynamic drug delivery control instructions.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by acquiring the trend of continuous respiratory rate changes in patients and combining it with respiratory extreme values ​​to generate a severe fluctuation indicator, and by analyzing the temporal trend of changes in patient request behavior to form a linkage time period marker, the basic infusion frequency and dosage are adjusted in real time to meet actual analgesia needs. The effectiveness of pressure is identified by the structural features of pressure waveforms, and abnormal behaviors are screened out to improve the accuracy of self-controlled infusion. At the same time, when respiratory fluctuations and pressure signals are activated synchronously, an emergency control path is triggered to execute continuous drug administration, realizing an intelligent identification and rapid response mechanism dominated by physiological state, enhancing the adaptability, safety and timeliness of the analgesia process, and effectively alleviating the problems of control lag and judgment blind spots under fixed parameter mode. Attached Figure Description

[0015] 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.

[0016] Figure 1 This is a schematic diagram of the steps of the present invention; Figure 2 This is a detailed schematic diagram of S1 of the present invention; Figure 3 This is a detailed schematic diagram of S2 of the present invention; Figure 4 This is a detailed schematic diagram of S3 of the present invention; Figure 5 This is a detailed schematic diagram of S4 of the present invention; Figure 6 This is a detailed schematic diagram of S5 of the present invention; Figure 7 This is a system module diagram of the present invention. Detailed Implementation

[0017] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0018] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0019] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0020] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0021] 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.

[0022] Please see Figure 1 This invention provides a method for controlling the administration of medication using a postoperative analgesia pump, comprising the following steps: S1: Obtain the patient's continuous respiratory rate sequence in the postoperative analgesia management scenario, divide it according to the complete respiratory cycle, extract the troughs and peaks within the cycle, construct the amplitude change set, perform trend marking according to the amplitude change direction of three consecutive cycles, and then compare the real-time cycle peak with the respiratory extreme value during the anesthesia stage. If there is a continuous positive fluctuation trend and the peak exceeds the extreme value range during the anesthesia stage, mark the time point as the stimulus response state and generate a respiratory fluctuation trigger identifier. S2: Call the respiratory fluctuation trigger flag, obtain the time distribution sequence of the patient's medication request compression behavior, divide a unified sliding time window, count the number of requests in the real-time window and the previous window, determine whether the number of requests shows an increasing trend between windows, and combine whether the number of respiratory fluctuation trigger flags in the real-time window increases at the same time. If both are in an increasing state, then establish a window as a synchronous response stage and generate a sudden request linkage time period marker. S3: Using the sudden request linkage time period marker, the continuous infusion time point and infusion dose recorded in the postoperative analgesia pump basal infusion behavior are obtained. The number of basal infusions is counted within the marked linkage time period. It is determined whether there is a low overlap between the basal infusion rhythm and the request behavior. If it is low, the start interval time and continuous infusion frequency of real-time basal infusion are updated, and the corresponding adjustment is performed on the infusion volume for each infusion to generate the basal drug administration rhythm adjustment result. S4: Based on the results of the basic drug administration rhythm adjustment, obtain the high-frequency pressure waveform generated by each press during the use of the analgesia pump. Extract the complete pressure curve from the initial pressure application to the release endpoint, perform inflection point number identification and rebound change trend judgment, and compare the real-time curve with the previous three curves in parallel sorting. Statistically analyze whether there is any offset or missing inflection point distribution and rebound segment in the real-time curve. If there is a structural abnormality, set the current pressing behavior as a suspicious signal and generate an autonomous operation identification mark. S5: Using autonomous operation identification flags and respiratory fluctuation trigger flags, determine whether they are activated at the same time node. If both are in the marked state, the analgesia pump emergency control channel responds and releases the emergency infusion action, calls the updated basic infusion parameters to perform continuous control, constructs a dual-channel linkage path dominated by respiratory fluctuations and pressure signals, and generates dynamic drug delivery control instructions. Respiratory fluctuation trigger indicators include amplitude change direction trend markers, period peak exceeding range judgment results, and stimulus response time point markers. Sudden request linkage period markers include request growth trend, respiratory fluctuation number increase trend, and request-fluctuation synchronization relationship. Basal dosing rhythm adjustment results include start interval time update, continuous infusion frequency adjustment, and dose correction for each infusion. Autonomous operation identification markers include inflection point distribution shift, rebound segment structure abnormality, and pressure behavior abnormality markers. Dynamic dosing control instructions include emergency infusion action triggering, basal infusion parameters, and dual-channel linkage pathway.

[0023] Please see Figure 2 The specific steps of S1 are as follows: S101: Obtain the continuous respiratory rate sequence of the patient in the postoperative analgesia management scenario, and split the continuous time period in the continuous respiratory rate sequence. Based on the periodic fluctuation of the inspiratory and expiratory phases in the complete respiratory cycle, analyze the respiratory waveform data frames in the cycle, extract the troughs and peaks of the waveform in the cycle, record the time index and amplitude value respectively, and generate a respiratory cycle extreme value sequence group. Respiratory rate data is continuously collected by a respiratory monitoring device deployed at the patient's bedside. The collected data is stored as a continuous signal sequence at a high frequency. The entire process is achieved through a wireless acquisition terminal integrated into the monitor. The sequence is parsed into continuous data points composed of fixed time intervals. For ease of analysis, the entire sequence is divided into multiple time periods, each representing a shorter set of continuous respiratory data. Respiratory cycles are identified in the data segments to determine the complete process from inspiration to expiration. The inspiratory and expiratory segments are distinguished by identifying the fluctuation patterns within each complete waveform cycle. After identifying each respiratory cycle, the positions of the minimum and maximum values ​​in each cycle are located, i.e., the troughs and peaks in the waveform. The sampling time and corresponding numerical value of the extreme points are recorded to represent the key respiratory amplitude change points in each cycle. If the lowest point in a certain cycle occurs at 30 seconds with an amplitude of 0.2, and the highest point occurs at 33 seconds with an amplitude of 0.5, then the extreme value of the cycle contains a data pair consisting of two time indices and amplitudes. Multiple cycles are continuously collected and analyzed to generate a respiratory cycle extreme value sequence group.

[0024] S102: Based on the extreme value sequence group of the respiratory cycle, the numerical direction of the amplitude difference sequence formed by the peak and trough of three consecutive respiratory cycles is determined. Based on whether the amplitude change direction between two adjacent cycles is consistent, three cycle direction state groups are constructed. If the amplitude change direction of three consecutive cycles is consistent, it is marked as a positive fluctuation trend state; otherwise, it is marked as a non-positive trend state, and a cycle amplitude trend label set is generated. The direction of the amplitude difference in three consecutive cycles within the extreme value sequence is determined. Within each group of three cycles, the difference between the maximum and minimum values ​​of each cycle is calculated sequentially to represent the strength of the cyclical breathing fluctuation. The direction of amplitude change between adjacent cycles is also determined, i.e., the amplitude difference between the later and previous cycles is compared to see if they change in the same direction. If both are increasing or both are decreasing, a consistent fluctuation trend is considered to exist. All data is checked step-by-step using a sliding window of three cycles as units, constructing a directional state group composed of the change states of each group of three cycles. For each state group, a label is assigned based on consistency. If the three changes are completely consistent, it is marked as a positive fluctuation trend; if the trends conflict, it is marked as a negative trend. For example, if the amplitudes of the three cycles are 0.3, 0.35, and 0.4, and each is detected as an increasing state, then the overall trend is recorded as positive. Conversely, if there are cases of 0.3, 0.35, and 0.33, although there is an increase in the middle, it ultimately decreases, and the overall trend will no longer be marked as positive, generating a set of cyclical amplitude trend labels.

[0025] S103: Call the period amplitude trend marker set, and retrieve the corresponding period peak data item according to the time point marked as positive fluctuation trend. Compare the period peak data item with the respiratory peak limit range recorded during the anesthesia stage in the postoperative analgesia scenario. If the period peak amplitude exceeds the upper limit of respiratory extreme value during the anesthesia stage, record the time point and perform trigger state assignment processing to obtain the respiratory fluctuation trigger identifier. The system screens time points marked as having a positive fluctuation trend and retrieves peak data corresponding to these points from the original respiratory cycle extreme value sequence. For each positive trend time point, it extracts the recorded peak amplitude and time index within the cycle and compares them with the preset upper limit of the respiratory peak range during anesthesia. The upper limit is a reference value set based on the patient's physiological performance during the preoperative anesthesia monitoring stage. If a peak in a certain cycle exceeds the set upper limit, it indicates that the respiratory amplitude exceeds the normal range at that time point, and this time point is recorded as a trigger point. During recording, an identifier is generated for each event exceeding the limit, and a status value is added to indicate that the time point has been marked as a special fluctuation event. Time points that meet this condition are integrated. Each identifier includes fields such as occurrence time, respiratory amplitude, and trigger type. The set of trigger identifiers is continuously updated during the collection and comparison process to ensure that changes in respiratory behavior that exceed the standard range can be reflected and respiratory fluctuation trigger identifiers can be obtained.

[0026] Please see Figure 3 The specific steps of S2 are as follows: S201: Call the respiratory fluctuation trigger flag, obtain the time distribution sequence of the patient's medication request compression behavior, divide the time distribution sequence into continuous sliding time windows according to a fixed step size, count the number of medication requests within the time window, and generate a sliding window request count set; The medication request times recorded by the patient's postoperative analgesia device are summarized to form a continuous request time series. The request time series is automatically recorded each time the patient manually presses the button to request medication infusion, with a time accuracy of more than one second. To analyze the temporal characteristics of the patient's medication request behavior, the request time series is divided into fixed time steps. A sliding window technique is used to organize and statistically analyze the request time points. The time window length is generally set to five minutes, and the sliding interval is one minute, that is, sliding forward once every minute to form an overlapping set of time windows. The number of requests within the covered time period is counted within each window. The operation is completed by looping through the request time series, checking whether each time point belongs to the current window interval. If it does, it is accumulated to form a medication request count record sequence corresponding to each time window. The respiratory fluctuation trigger marker is divided into the same sliding window, and the frequency of severe respiratory fluctuations within the window is counted to form a second statistical sequence synchronized with the medication request data, generating a sliding window request count set.

[0027] S202: Based on the sliding window request quantity set, the difference in the quantity of requests within two adjacent consecutive time windows is judged. If the number of requests in the later window is greater than the number of requests in the earlier window, the time window is marked as increasing. The respiratory fluctuation triggering identifier is synchronously divided into windows and the quantity is counted on the same time axis. It is judged whether the number of fluctuation triggers in the window has increased compared with the previous window. If both are increasing, the window is marked as linked and a linked state window marking sequence is generated. The trend direction is determined by the difference between the statistical values ​​of two adjacent time windows. For each pair of adjacent windows, the number of requests in the latter window is compared with that in the former window. If the number of requests increases, the window is marked as an increasing request state, and the state result is recorded. The same judgment is performed on the number of severe respiratory fluctuations. If the number of fluctuations in the latter window also increases, the window is also marked as an increasing fluctuation state. Only when both medication requests and fluctuation triggers are in an increasing state is the time window considered to have a linkage trend, and the window is marked with a linkage state identifier. The process will be continuously executed between each pair of adjacent windows. The marking method uses a logical Boolean type and automatically maintains a marker array. Each time window has a corresponding state record, generating a linkage state window marker sequence.

[0028] S203: Based on the time windows marked as linked in the linked status window marking sequence, identify the start and end time range of the linked time period, assign the synchronous response attribute to the sudden request, complete the marking operation on the linked status window, and obtain the linked time period mark of the sudden request. The system identifies which time windows are considered to be in a linked state and performs continuous analysis on these windows. If multiple windows are consecutively marked as linked, they are merged into a complete linked time period. The start and end times of multiple segments are extracted to form a time range label for the linked state. To enhance the identification of behavioral characteristics during linked periods, the time period is assigned a "sudden request synchronous response attribute," meaning that this period not only sees an increase in medication requests but also overlaps with the time of events involving severe physiological fluctuations. This period is of high interest in postoperative analgesia monitoring. The linked time period will be written into the status identifier through a marking operation. The marking format can be a status code in the record table, enabling medical personnel to quickly locate peak behavioral segments with potential risks. Each linked time period will form a complete set of record items, including the start time, end time, and status category. When generating the record, the original window sequence number and overlap description can be attached for trend analysis and data backtracking. The linked records can also be integrated into the central monitoring platform for cross-analysis with vital sign data to achieve a higher level of data synchronization and behavioral monitoring, and to obtain the sudden request linked time period label.

[0029] Please see Figure 4 The specific steps of S3 are as follows: S301: Call the sudden request linkage period marker, obtain the continuous infusion time point and corresponding infusion dose information recorded in the postoperative analgesia pump basic infusion behavior, and search within the linkage period range to see if the basic infusion time point falls within the range, count the number of basic infusions within the linkage period, and generate a basic infusion statistics table within the linkage period. The baseline infusion log is read from the postoperative analgesia pump. This log records the patient's unsolicited, automated, continuous low-dose drug infusions performed according to a pre-set program. Each record includes the time of infusion and the corresponding dose. The baseline infusion records are iterated through and compared one by one with the time range of the sudden triggering period. It is determined whether each infusion time point falls within a marked triggering period. If a time point falls within the range, it is included in the statistics, and the infusion time and dose information are recorded. After iterating through all records, the baseline infusion data falling within the triggering period are sorted by triggering period. The data is categorized and organized, and the number of basic infusion events within each linked time period is statistically analyzed. The output results are organized in a table format, with each row representing a linked time period, including the start time, end time, the corresponding number of basic infusions, and the cumulative infusion dose. This provides a basis for analyzing whether there is coordination between basic infusions and patients' active requests. At the same time, statistical graphs can be displayed in real time on the interface based on the associated data visualization platform to help medical staff identify potential inconsistencies in the analgesia plan, especially in time periods where there are frequent requests but the infusion volume does not change significantly, generating a basic infusion statistics table for the linked time period.

[0030] S302: Based on the basic infusion statistics table within the linkage period, calculate the ratio of the number of basic infusions within the linkage period to the number of medication requests within the corresponding period, determine the temporal overlap ratio between the basic infusion rhythm and the patient's request behavior, and if the overlap ratio is lower than the preset behavior matching ratio threshold, mark it as a temporal mismatch state and generate a basic infusion rhythm mismatch identifier set. For each linkage period, the corresponding number of basic infusions and the number of PCA medication requests are extracted from the statistical table, and the ratio between the two is calculated. The ratio represents the degree of temporal coordination between automatic basic infusion and patient-initiated requests. The ratio is compared with a preset behavior matching ratio threshold. The threshold can be set based on large sample data statistics or combined with expert experience. For example, the ratio of basic infusion to request behavior should be greater than 0.3 within a certain reasonable range. If the ratio of a linkage period is lower than the threshold, it indicates that automatic basic infusion has not effectively covered the patient's actual medication needs during the period, and the period is marked as "temporal mismatch state". The linkage periods marked as mismatch state will be sorted out, and each record item includes the period range, number of basic infusions, number of requests, ratio, and judgment result. This is used for infusion strategy adjustment and dosing mode optimization operations. It will also support alarm linkage to realize automatic prompt function and generate visual markers on the medical monitoring interface to facilitate the timely identification of rhythm abnormality segments by on-duty personnel, improve management efficiency, and generate a basic infusion rhythm mismatch identifier set.

[0031] S303: The time periods marked as time mismatch states by the basic infusion rhythm mismatch identifier are extracted, the start interval time and continuous infusion frequency of real-time basic infusion are extracted, and compared with the behavior matching benchmark configuration parameter set respectively. The start interval time and frequency parameters are updated, and the infusion dose is redistributed according to the updated frequency to obtain the basic dosing rhythm adjustment results. Key rhythmic parameters of basal infusion behavior within a time period are extracted, namely the basal infusion start interval and the frequency of continuous infusions. Infusion records within the time period are traversed, the interval between two adjacent infusions is calculated, and the number of infusions per unit time is counted to obtain the actual basal infusion rhythm within the time period. A pre-set behavior matching benchmark configuration parameter set is called for comparison. The configuration parameter set is formulated by medical staff according to clinical standards, including parameters such as recommended minimum interval time and dosing frequency. Based on the comparison results, it is determined whether the current actual parameters fall within the recommended range. If not, the parameters are updated. During the update process, the start interval time is adjusted to be within the recommended range, and the number of infusions to be completed per unit time is recalculated accordingly. The dose of each infusion is redistributed according to the new frequency, optimizing the infusion rhythm while ensuring that the total drug dosage remains unchanged. This better meets the actual needs of patients during mismatched time periods. For example, the original infusion every 10 minutes is adjusted to once every 6 minutes, and the dose of each infusion is reduced accordingly to enhance the timing fit of basal infusion. The drug administration instructions are automatically updated to realize the automatic rhythm optimization function of the analgesic pump and obtain the results of basal drug administration rhythm adjustment.

[0032] Please see Figure 5 The specific steps of S4 are as follows: S401: Based on the results of the basic drug administration rhythm adjustment, obtain the high-frequency pressure waveform sequence generated by each pressing action of the patient during the use of the analgesia pump, extract the corresponding complete pressure curve from the pressure application start point to the release end point, identify the derivative change points in the curve from positive to negative or negative to positive, count the number and location, and generate the inflection point feature group of the pressing pressure curve. By combining high-frequency pressure waveform data recorded from each patient's compression action on the analgesia pump, the entire process of each compression, from the start of pressure application to the end of release, is processed and analyzed. Each compression curve is composed of raw data acquired by a pressure sensor at a high frequency of over 100Hz. A complete data segment from the start of the compression action to the release of the button is extracted, and continuous pressure curves within the interval are obtained. Derivative change analysis is performed on each data point, that is, the points where the curve slope changes from positive to negative or vice versa are identified. These points of change are considered inflection points of the curve. The pressure curve points are traversed, and the slope changes of adjacent points are calculated. By comparing the derivatives... The symbol change locates each inflection point. For each inflection point identified, the location index and the corresponding pressure value in the curve are recorded together to form a feature group consisting of inflection point number, location, and amplitude. At the same time, the total number of inflection points in the entire compression curve is counted, and a "complex inflection point feature group" for compression is generated, which includes the index position, change type, and relative position distribution information of each inflection point. This provides data support for analyzing the structural stability and consistency of compression behavior. The feature group generation process can continue to run after each patient compression operation to ensure that each curve has a complete feature expression form before entering the next stage of comparison, thus generating the compression pressure curve inflection point feature group.

[0033] S402: Call the inflection point feature group of the pressing pressure curve, and retrieve the data of the three adjacent pressing pressure curves. After arranging them in chronological order, extract the derivative direction sequence and inflection point index sequence of the rebound segment in each curve. Compare the real-time curve with the previous three curves in terms of the number of inflection points and the trend of rebound segment changes, record the structural feature difference distribution, and generate a real-time curve structural offset data table. The system automatically retrieves the pressure curve data corresponding to the patient's three most recent compression records. To maintain the rationality and temporal continuity of the comparison, these three curves are arranged in chronological order. The derivative analysis and inflection point identification process are re-executed for each curve to extract the derivative direction sequence and the specific location of each inflection point within the rebound segment. The rebound segment refers to the part of the curve formed during the gradual decrease of pressure after the button is released. This segment exhibits a relatively regular rising or falling pattern. The system compares the real-time curve with the three curves in terms of structural characteristics such as the number of inflection points, inflection point positions, and derivative direction switching. The system focuses on analyzing whether the changing trend of the rebound segment in the current curve is consistent with the previously formed pattern. If the derivative direction of the rebound segment of the current curve is significantly disordered, or the inflection point position is abnormally shifted, it indicates that there is a structural deviation in the current behavior. The characteristic differences in the comparison process are recorded item by item according to the comparison dimension, including changes in the number of inflection points, different derivative patterns, and deformation of the rebound trend, generating a real-time curve structural deviation data table.

[0034] S403: Based on the differences in structural features recorded in the real-time curve structure offset data table, determine whether the real-time pressing curve has a reduced number of inflection points, missing rebound segments, or abnormal positional offset. If there is a structural abnormality marker, set an abnormality marker status for the pressing behavior, assign autonomous operation attributes, and obtain an autonomous operation identification mark. The current compression behavior is assessed for structural integrity. Structural judgment criteria are established, including multiple dimensions such as whether the number of inflection points has significantly decreased, whether there is a missing rebound segment, and whether the inflection point position deviates significantly from the mean. If the current curve exceeds the preset judgment threshold in any dimension, it is considered structurally abnormal. An enumeration strategy is used during the judgment process, scanning each feature item in the data table and comparing it with the behavioral feature range. If the curve has 3-4 inflection points in the rebound segment, but only 1 or no inflection point appears, this is recorded as a missing rebound. If an inflection point appears in a position range with a large deviation from the recorded value, it is recorded as a positional deviation. Once any abnormality type is confirmed, the compression behavior is immediately marked as abnormal, the behavioral status attribute is updated, and the behavior is assigned an "autonomous operation attribute," indicating that the compression behavior is not due to standard analgesia needs but rather a voluntary action by the patient. The marker content includes fields such as abnormality type, abnormality level, and judgment time, and is uniformly archived as a record in the autonomous operation identification marker set to obtain the autonomous operation identification marker.

[0035] Please see Figure 6 The specific steps of S5 are as follows: S501: Using autonomous operation identification flags and respiratory wave trigger flags, the activation status of the two types of flags at the same time point is compared by time index to determine whether there is a joint marking situation with overlapping time. If both are in the marking state at the same time point, the corresponding time index is recorded and set as the linkage trigger time point to generate a flag linkage activation time set. The activation status of the autonomous operation identification flag and the respiratory fluctuation trigger flag is synchronously compared in the time dimension. The two flag sequences are indexed by time index to mark their respective activation time points, and cross-scanned on a unified time axis. For each time point triggered by a violent respiratory fluctuation, it is checked whether there is an autonomous operation identification time point that is completely consistent with it or appears within a set tolerance range (e.g., ±2 seconds). If the two overlap or are extremely close in time, it is determined to be a joint activation situation within the same time node. Each time point that meets the joint activation condition is recorded and included in the linkage trigger time point set. A unique number is added to each successfully matched time point, and the activation source is recorded (e.g., which fluctuation flag and which abnormal press triggered it) for channel response calls and data traceability. At the same time, a complete set of cross-matching logs is also output, including parameters such as the proportion of successful time matching and the average response interval between the two types of events, which are used to analyze the recognition accuracy of abnormal overlapping events and the accuracy of linkage control triggering. Each time the identification process completes a real-time judgment, the linkage activation time set is also dynamically updated to generate the flag linkage activation time set.

[0036] S502: Based on the linkage activation time set of the identifier, the emergency control channel of the analgesia pump calls the associated channel response mechanism, releases the emergency infusion action corresponding to the activation state, calls the updated basic infusion parameter set, and continues to execute the continuous dose control operation, generating a dual-channel control execution set; The analgesic pump's internal emergency control channel is immediately activated at each trigger point, and the preset emergency infusion mechanism is activated. The emergency infusion configuration file corresponding to the current time point is called, and a rapid response drug administration action is initiated. The emergency infusion action parameters will be adapted and adjusted with reference to the latest updated basic infusion parameter set and combined with the current linkage background information. The emergency action is mainly used to deal with the patient's sudden discomfort. It has the characteristics of rapid start-up, adjustable dose, and independent response path. After the emergency infusion is completed, the corresponding basic infusion continuous control module will be called to seamlessly connect the drug administration behavior according to the basic rhythm parameters, ensuring that the drug infusion is not interrupted or conflicted after the emergency response. The entire collaborative control process of the emergency channel and the basic channel will be recorded in the dual-channel control execution set in an event-driven manner. Each record includes fields such as emergency trigger time, emergency dose, basic dose continuation time point, rhythm parameters, and channel transition status, generating the dual-channel control execution set.

[0037] S503: Invoke the dual-channel control to execute the centrally recorded emergency channel response and basic channel continuous control configuration, fuse the channel status according to the time priority of respiratory fluctuation signal and spontaneous compression signal, construct the linkage output control path with respiratory fluctuation signal and compression signal as the core driving factors, and generate dynamic drug administration control instructions; Based on the execution parameters of each emergency channel and basic channel, a fusion control strategy is generated to construct a dynamic linkage drug delivery control path driven by respiratory fluctuation signals and voluntary compression behavior. Channel status identification is prioritized based on the activation priority of respiratory fluctuation signals. If the respiratory signal is active and higher than the trigger level of the voluntary compression event, the respiratory fluctuation is used as the main control signal to execute channel switching. Conversely, if the voluntary compression signal has high priority or occurs frequently, the compression signal is used as the main driving factor to update the control path. According to the priority fusion method, the distribution pattern of the two signal sources on the time axis is reconstructed to make real-time decisions on which channel should deliver the drug. At the same time, an independent dynamic control command is generated for each fusion decision. The command content includes the target channel number, instantaneous dose value, infusion rhythm, execution delay, and locking window. The dynamic commands are then integrated and uniformly pushed to the analgesia pump for execution, ensuring the timeliness and individual responsiveness of drug delivery. This forms a highly sensitive control mechanism based on real-time physiological fluctuations and operational behavior, generating dynamic drug delivery control commands.

[0038] Please see Figure 7 A drug delivery control system for a postoperative analgesia pump, comprising: The respiratory fluctuation recognition module acquires the patient's continuous respiratory rate sequence in the postoperative analgesia management scenario. It extracts the peak and trough values ​​in segments according to the complete inhalation and exhalation cycle, obtains the cycle amplitude difference and forms a set of amplitude changes in three adjacent cycles. It calls the three amplitude differences and compares whether the change direction is consistent. If they show a unified direction, they are marked as a trend segment. Then, it compares the cycle peak with the upper limit of the respiratory peak recorded during the anesthesia stage to generate a respiratory fluctuation trigger identifier. The request behavior monitoring module collects records of the patient's pressing behavior on the analgesia pump based on the respiratory fluctuation trigger identifier, constructs a request distribution sequence in chronological order, divides the time windows into equal-length windows to count the number of requests within the window, and generates a sudden request linkage time period marker. The rhythm adjustment module retrieves the basic infusion time series and infusion dose records of the analgesic pump based on the sudden request linkage time period marker, counts the number of occurrences of basic infusion behavior within the interval, and calculates the overlap ratio with the request behavior on the time axis. If the ratio is low, the basic infusion setting parameters are adjusted to shorten the basic infusion interval and generate the basic drug administration rhythm adjustment results. The autonomous operation recognition module calls the basic drug administration rhythm adjustment results, obtains the pressure waveform data when the patient operates the analgesia pump button, identifies the complete pressing curve and extracts the inflection point distribution and rebound segment morphology, compares the real-time curve with the previous three recorded curves, judges whether there is abnormal deviation or structural missingness in the inflection point position and rebound trend, and generates an autonomous operation recognition mark. The dynamic drug delivery linkage module compares the autonomous operation identification flag and the respiratory fluctuation trigger flag on the time axis to see if the marked positions are activated simultaneously. If they overlap, it retrieves the emergency infusion control path, performs basic infusion adjustment, performs continuous infusion operation on the channel, and generates dynamic drug delivery control instructions.

[0039] The above are merely specific embodiments 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 method for controlling drug delivery using a postoperative analgesia pump, characterized in that, Includes the following steps: S1: Obtain the continuous respiratory rate sequence of the patient in the postoperative analgesia management scenario, divide it according to the complete respiratory cycle, extract the trough and peak within the cycle, compare the real-time cycle peak with the respiratory extreme value during the anesthesia stage, and generate a respiratory fluctuation trigger indicator. S2: Call the respiratory fluctuation triggering identifier, obtain the time distribution sequence of the patient's medication request pressing behavior, divide a uniform sliding time window, count the number of requests in the real-time window and the previous window, and generate a sudden request linkage time period marker. S3: Using the aforementioned sudden request linkage time period marker, obtain the continuous infusion time points and infusion dose recorded in the postoperative analgesia pump basal infusion behavior, count the number of basal infusions within the marked linkage time period, and generate the basal drug administration rhythm adjustment results; S4: Based on the basic drug administration rhythm adjustment results, obtain the high-frequency pressure waveform generated by each press during the use of the analgesia pump by the patient, extract the complete pressure curve from the initial pressure application to the release endpoint, perform inflection point number identification and rebound change trend judgment, and sort and compare the real-time curve with the previous three curves to generate an autonomous operation identification mark.

2. The method for controlling drug delivery using a postoperative analgesia pump according to claim 1, characterized in that, The respiratory fluctuation triggering identifier includes an amplitude change direction trend marker, a period peak exceeding the range judgment result, and a stimulus response time point marker. The sudden request linkage period marker includes a request growth trend, a respiratory fluctuation number increase trend, and a synchronous relationship between requests and fluctuations. The basic dosing rhythm adjustment result includes an update of the starting interval time, an adjustment of the continuous infusion frequency, and a correction of the dose for each infusion. The autonomous operation identification marker includes an inflection point distribution offset, a rebound segment structural abnormality, and a pressure application behavior abnormality marker.

3. The method for controlling drug delivery using a postoperative analgesia pump according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Obtain the continuous respiratory rate sequence of the patient in the postoperative analgesia management scenario, and split the continuous time period in the continuous respiratory rate sequence. Based on the periodic fluctuation of the inspiratory and expiratory phases in the complete respiratory cycle, analyze the respiratory waveform data frames in the cycle, extract the troughs and peaks of the waveform in the cycle, record the time index and amplitude value respectively, and generate a respiratory cycle extreme value sequence group. S102: Based on the respiratory cycle extreme value sequence group, the numerical direction of the amplitude difference sequence formed by the peak and trough of three consecutive respiratory cycles is determined. Based on whether the amplitude change direction between two adjacent cycles is consistent, three cycle direction state groups are constructed. If the amplitude change direction of three consecutive cycles is consistent, it is marked as a positive fluctuation trend state; otherwise, it is marked as a non-positive trend state, and a cycle amplitude trend label set is generated. S103: Call the set of periodic amplitude trend markers, and retrieve the corresponding periodic peak data item according to the time point marked as a positive fluctuation trend. Compare the periodic peak data item with the respiratory peak limit range recorded during the anesthesia stage in the postoperative analgesia scenario. If the periodic peak amplitude exceeds the upper limit of the respiratory extreme value during the anesthesia stage, record the time point and perform trigger state assignment processing to obtain the respiratory fluctuation trigger identifier.

4. The method for controlling drug delivery of the postoperative analgesia pump according to claim 3, characterized in that, The specific steps of S2 are as follows: S201: Call the respiratory fluctuation trigger identifier, obtain the time distribution sequence of the patient's medication request pressing behavior, divide the time distribution sequence into continuous sliding time windows according to a fixed step size, count the number of medication requests within the time window, and generate a sliding window request count set; S202: Based on the sliding window request quantity set, the quantity difference of the request quantity in two adjacent consecutive time windows is judged. If the request quantity in the later window is greater than the request quantity in the earlier window, the time window is marked as an increasing state. The respiratory fluctuation triggering identifier is synchronously divided into windows and the quantity is counted on the same time axis. It is judged whether the fluctuation triggering quantity in the window has increased compared with the previous window. If both are increasing, the window is marked as a linked state, and a linked state window marking sequence is generated. S203: Based on the time windows marked as being in linkage status in the linkage status window marking sequence, identify the linkage start and end time range corresponding to the time period, assign the sudden request synchronization response attribute, complete the marking operation on the linkage status window, and obtain the sudden request linkage time period mark.

5. The method for controlling drug delivery of the postoperative analgesia pump according to claim 4, characterized in that, The specific steps for S3 are as follows: S301: Call the sudden request linkage time period marker, obtain the continuous infusion time point and corresponding infusion dose information recorded in the postoperative analgesia pump basic infusion behavior, and search within the linkage time period to see if the basic infusion time point falls within the range, count the number of basic infusions within the linkage time period, and generate a basic infusion statistics table within the linkage time period. S302: Based on the basic infusion statistics table within the linkage period, calculate the ratio of the number of basic infusions within the linkage period to the number of medication requests within the corresponding period, determine the temporal overlap ratio between the basic infusion rhythm and the patient's request behavior, and if the overlap ratio is lower than the preset behavior matching ratio threshold, mark it as a temporal mismatch state and generate a basic infusion rhythm mismatch identifier set. S303: Using the time periods marked as time mismatch states in the basic infusion rhythm mismatch identifier set, extract the start interval time and continuous infusion frequency of real-time basic infusion, and compare them with the behavior matching benchmark configuration parameter set respectively, update the start interval time and frequency parameters, and redistribute the infusion dose each time according to the updated frequency to obtain the basic dosing rhythm adjustment result.

6. The method for controlling drug delivery of the postoperative analgesia pump according to claim 5, characterized in that, The specific steps of S4 are as follows: S401: Based on the basic drug administration rhythm adjustment results, obtain the high-frequency pressure waveform sequence generated by each pressing action of the patient during the use of the analgesia pump, extract the corresponding complete pressure curve from the pressure application start point to the release end point, identify the derivative change points in the curve from positive to negative or negative to positive, count the number and location, and generate the pressing pressure curve inflection point feature group. S402: Call the inflection point feature group of the pressing pressure curve, and retrieve the data of the three adjacent pressing pressure curves. After arranging them in chronological order, extract the derivative direction sequence and inflection point index sequence of the rebound segment in each curve. Compare the real-time curve with the previous three curves in terms of the number of inflection points and the trend of rebound segment changes, record the structural feature difference distribution, and generate a real-time curve structural offset data table. S403: Based on the structural feature differences recorded in the real-time curve structure offset data table, determine whether the real-time pressing curve has a reduced number of inflection points, missing rebound segments, or abnormal positional offset. If there is a structural abnormality marker, set an abnormality marker state for the pressing behavior, assign autonomous operation attributes, and obtain an autonomous operation identification mark.

7. The method for controlling drug delivery of the postoperative analgesia pump according to claim 6, characterized in that, Based on the inflection point feature group extracted from the real-time pressing pressure curve, determine whether there is an abnormal decrease in the number of inflection points of the real-time curve. If the number of inflection points in the real-time pressing pressure curve decreases, check whether there is a missing rebound segment in the real-time pressing pressure curve, or whether the derivative direction sequence of the rebound segment deviates from the normal trend. If the rebound segment of the real-time pressing pressure curve is missing or the derivative direction sequence of the rebound segment is shifted, an abnormal flag state is set for the pressing behavior, and the pressing behavior is marked as having autonomous operation attributes. The judgment process is as follows: identify the complete pressure curve corresponding to each pressing action in the extracted real-time curve, extract the rebound segment in the pressure curve, compare the changing trend of the derivative direction of the rebound segment with the changing trend in the adjacent previous curves, and if the trend difference exceeds the preset threshold, it is judged as an abnormal state.

8. The method for controlling drug delivery of the postoperative analgesia pump according to claim 1, characterized in that, The method further includes step S5: S5: Using the autonomous operation identification flag and respiratory fluctuation trigger flag, determine whether they are activated at the same time node. If both are in the marked state, the analgesia pump emergency control channel responds and releases the emergency infusion action, calls the updated basic infusion parameters to perform continuous control, constructs a dual-channel linkage path dominated by respiratory fluctuation and pressure signal, and generates dynamic drug delivery control instructions. The dynamic drug delivery control commands include emergency infusion triggering, basic infusion parameters, and dual-channel linkage pathways.

9. The method for controlling drug delivery of the postoperative analgesia pump according to claim 8, characterized in that, The specific steps of S5 are as follows: S501: Using the autonomous operation identification flag and the respiratory wave trigger flag, the activation status of the two types of flags at the same time node is compared by time index to determine whether there is a joint marking situation with overlapping time. If both are in the marking state at the same time node, the corresponding time index is recorded and set as the linkage trigger time point to generate a flag linkage activation time set. S502: Based on the linkage triggering time point in the linkage activation time set, the analgesia pump's sudden control channel calls the associated channel response mechanism, releases the emergency infusion action corresponding to the activation state, calls the updated basic infusion parameter group, and continues to execute the continuous dose control operation, generating a dual-channel control execution set; S503: Invoke the dual-channel control to execute the centrally recorded emergency channel response and basic channel continuous control configuration, fuse the channel status according to the time priority of the respiratory fluctuation signal and the spontaneous compression signal, construct a linkage output control path with the respiratory fluctuation signal and the compression signal as the core driving factors, and generate dynamic drug administration control instructions.

10. A drug delivery control system for a postoperative analgesia pump, characterized in that, The system is used to implement the drug delivery control method of a postoperative analgesia pump according to any one of claims 1-9, the system comprising: The respiratory fluctuation recognition module acquires the patient's continuous respiratory rate sequence in the postoperative analgesia management scenario. It extracts the peak and trough values ​​in segments according to the complete inhalation and exhalation cycle, obtains the cycle amplitude difference and forms a set of amplitude changes in three adjacent cycles. It calls the three amplitude differences and compares whether the change direction is consistent. If they show a unified direction, they are marked as a trend segment. Then, it compares the cycle peak with the upper limit of the respiratory peak recorded during the anesthesia stage to generate a respiratory fluctuation trigger identifier. The request behavior monitoring module collects the patient's pressing behavior record of the analgesia pump based on the respiratory fluctuation trigger identifier, constructs a request distribution sequence in chronological order, divides the time windows into equal long-term windows to count the number of requests within the window, and generates a sudden request linkage time period marker. Based on the sudden request linkage time period marker, the rhythm adjustment module retrieves the basic infusion time series and infusion dose record of the analgesic pump, counts the number of occurrences of basic infusion behavior within the interval, and calculates the overlap ratio with the request behavior on the time axis. If the ratio is low, the basic infusion setting parameters are adjusted to shorten the basic infusion interval and generate the basic drug administration rhythm adjustment result. The autonomous operation recognition module calls the basic drug administration rhythm adjustment results, obtains the pressure waveform data when the patient operates the analgesia pump button, identifies the complete pressing curve and extracts the inflection point distribution and rebound segment morphology, compares the real-time curve with the previous three recorded curves, determines whether there is abnormal deviation or structural missingness in the inflection point position and rebound trend, and generates an autonomous operation recognition mark. The dynamic drug delivery linkage module compares the autonomous operation identification flag and the respiratory fluctuation trigger flag on the time axis to see if the marked positions are activated simultaneously. If they overlap, it retrieves the emergency infusion control path, performs basic infusion adjustment, performs continuous infusion operation on the channel, and generates dynamic drug delivery control instructions.

Images