Intelligent labor management methods and systems integrating LP labor detection and multi-parameter monitoring

Abstract

This invention proposes an intelligent labor management method and system that integrates LP (labor delivery) detection and multi-parameter monitoring. It belongs to the interdisciplinary fields of smart obstetrics, multimodal physiological monitoring, and human factors engineering. The method includes: dividing the labor process of primiparous women into multimodal monitoring areas, generating core obstetric parameter monitoring areas, and collecting psychological stress monitoring data via millimeter-wave radar and continuously monitoring the fetal biochemical environment via amniotic fluid biochemical probes; deploying the LP labor detection system, non-contact psychological sensors, and amniotic fluid biochemical probes based on these areas to construct an intelligent monitoring layout network; by integrating LP labor detection and multi-parameter monitoring, the overall efficiency of labor management is significantly improved: it achieves accurate monitoring of traditional parameters such as cervical dilation and fetal heart rate, and expands the holistic perception of key dimensions such as maternal psychological stress and amniotic fluid biochemical environment, greatly improving the completeness and real-time performance of labor monitoring.

Description

Technical Field

[0001] This invention proposes an intelligent labor management method and system that integrates LP labor detection and multi-parameter monitoring, belonging to the interdisciplinary fields of smart obstetrics, multimodal physiological monitoring and human factors engineering. Background Technology

[0002] In obstetric clinical practice, traditional labor management models have long faced a multidimensional monitoring dilemma: while existing LP (Labor Delivery) monitoring systems can track cervical dilation and fetal heart rate in real time, they are limited to collecting single physiological parameters and lack effective means of sensing key dimensions such as maternal psychological stress and fetal biochemical environment. This fragmented monitoring often leads to data silos in clinical decision-making—when pain causes a surge in catecholamines that inhibits uterine contractions, traditional systems can only record cervical stagnation but cannot identify the psychological triggers; when the fetus experiences latent hypoxia due to amniotic fluid acidification, abnormal fetal heart rate monitoring often lags behind the deterioration of biochemical indicators. More seriously, existing risk assessment systems mostly use parameter-independent alarm mechanisms and have failed to establish cross-modal coupling models, resulting in a persistently high rate of non-medically indicated cesarean sections.

[0003] While existing technologies attempt to improve monitoring accuracy through data fusion, they essentially remain monitoring screens with more parameters, failing to break free from the linear mindset of monitoring-alarm-intervention. Our research team discovered in clinical practice that the childbirth process is essentially a dynamic coupled system of the mother's physiology, psychology, and fetal environment, urgently requiring the construction of an intelligent management paradigm with holographic perception, coupled modeling, and resilience assessment capabilities. Based on this, this invention innovatively integrates millimeter-wave radar psychological sensing, continuous amniotic fluid biochemical monitoring, and spatiotemporal neural network technology to propose the Labor Resilience Index (LRI) quantitative system, aiming to achieve a paradigm shift in obstetrics from "passive monitoring" to "active cognition." Summary of the Invention

[0004] This invention provides an intelligent labor management method and system that integrates LP labor detection and multi-parameter monitoring to solve the problems mentioned in the background art above: The present invention proposes an intelligent labor management method integrating LP labor detection and multi-parameter monitoring, the method comprising: S1. Divide the multimodal monitoring areas of the primiparous woman's delivery process, generate monitoring areas for core obstetric parameters, and monitor areas for psychological stress by collecting data through millimeter-wave radar and for continuous monitoring of the fetal biochemical environment by amniotic fluid biochemical probes; based on the above areas, deploy the LP delivery detection system, non-contact psychological sensors and amniotic fluid biochemical probes to construct an intelligent monitoring layout network; S2. Based on the intelligent monitoring network layout, the monitoring frequency is dynamically adjusted to synchronously collect core obstetric parameters, maternal psychological stress signals, and amniotic fluid biochemical data to generate multi-source heterogeneous delivery data streams; noise suppression and emotional feature extraction are performed on psychological stress signals, and drift correction and threshold warning processing are performed on amniotic fluid biochemical data to generate standardized psychological and biochemical coupled data. S3. Input standardized data into a spatiotemporal graph convolutional network to construct a spatiotemporal coupled graph. Extract cross-modal risk features through graph node correlation analysis and generate a coupled risk index of composite risk. S4. Calculate the labor resilience index based on the coupling risk index. When LRI < 0.6 and amniotic fluid pH < 7.0, trigger a knowledge graph-driven three-level collaborative response; and simultaneously generate a dynamic response report. S5. Update the spatiotemporal coupling map based on the dynamic response report, optimize the calculation model of the labor resilience index through graph neural network iteration, and generate a full labor resilience assessment curve; when the indication for cesarean section is only due to non-medical factors, the intervention decision is constrained by the resilience assessment curve, and finally an intelligent labor management file is generated.

[0005] The intelligent labor management system integrating LP labor detection and multi-parameter monitoring proposed in this invention includes: One or more processors; Memory, used to store one or more programs; Wherein, when the one or more programs are executed by the one or more processors, the one or more processors are made to implement the method described in any one of the above.

[0006] The beneficial effects of this invention are as follows: By integrating LP (Labor Detection) monitoring with multi-parameter monitoring, the overall effectiveness of labor management is significantly improved. It achieves precise monitoring of traditional parameters such as cervical dilation and fetal heart rate, while also expanding the holistic perception of key dimensions such as maternal psychological stress and amniotic fluid biochemical environment, greatly enhancing the completeness and real-time nature of labor monitoring. The coupled risk model constructed through a spatiotemporal graph convolutional network can dynamically quantify the combined risks of the mother and fetus, reducing the non-medical indication cesarean section rate to below 7%, effectively reducing excessive medical intervention. The labor resilience index... The introduction of LRI (Low-Risk Induction) enhances the foresight of risk warning. When LRI < 0.6 and amniotic fluid pH < 7.0, a three-level collaborative response mechanism is automatically triggered, which avoids the lag of alarms based on single parameters and prevents decision-making delays in emergency situations. The intervention plan driven by multimodal data fusion and knowledge graph can alleviate maternal anxiety through gentle measures such as non-pharmacological analgesia and emotional support from family members, and can also accurately control the dosage of oxytocin and initiate medical procedures such as amniotic fluid replacement, which significantly reduces complications such as labor stagnation and fetal hypoxia caused by psychological factors. Attached Figure Description

[0007] Figure 1 This is a diagram illustrating the steps of the method described in this invention. Detailed Implementation

[0008] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0009] One embodiment of the present invention, such as Figure 1 As shown, an intelligent labor management method integrating LP labor detection and multi-parameter monitoring is described, the method comprising: S1. A multimodal monitoring area is defined for the primiparous woman's delivery process, generating core obstetric parameter monitoring areas. These areas cover cervical dilation, contraction intensity, and fetal heart rate. A non-contact psychological stress monitoring area is also established, using millimeter-wave radar to collect data on the mother's facial micro-expressions and body movement frequency. Finally, a fetal biochemical environment monitoring area is established, continuously monitoring the amniotic fluid pH and lactate concentration. Based on these areas, an LP delivery monitoring system, non-contact psychological sensors, and amniotic fluid biochemical probes are deployed to construct an intelligent monitoring network that integrates physiological, psychological, and biochemical dimensions. S2. Based on the intelligent monitoring network layout, the monitoring frequency is dynamically adjusted to synchronously collect core obstetric parameters, maternal psychological stress signals, and amniotic fluid biochemical data to generate multi-source heterogeneous delivery data streams; noise suppression and emotional feature extraction are performed on psychological stress signals, and drift correction and threshold warning processing are performed on amniotic fluid biochemical data to generate standardized psychological and biochemical coupled data. S3. Input standardized data into a spatiotemporal graph convolutional network to construct a spatiotemporal coupled graph. The spatiotemporal coupled graph includes maternal cervical dilation, uterine contraction rhythm and pain expression, fetal heart rate variability, amniotic fluid pH and biochemical metabolism. Extract cross-modal risk features through graph node association analysis to generate a coupled risk index of composite risks. The coupled risk index of composite risks includes uterine contraction weakness, fetal distress and psychological breakdown. S4. Calculate the Labor Resilience Index (LRI) based on the Coupled Risk Index. When the LRI < 0.6 and the amniotic fluid pH < 7.0, trigger a knowledge graph-driven three-level synergistic response: Level 1 response initiates non-pharmacological analgesia intervention and family emotional support; Level 2 response adjusts the oxytocin dosage and initiates amniotic fluid replacement preparation; Level 3 response initiates multidisciplinary consultation and emergency cesarean section plan. Simultaneously, a dynamic response report is generated, which includes the current risk level, intervention measures, and prognostic assessment. S5. Update the spatiotemporal coupling map based on the dynamic response report, and iteratively optimize the labor resilience index calculation model through graph neural network to generate a full-process labor resilience assessment curve. When the indication for cesarean section is solely due to non-medical factors, the intervention decision is constrained by the resilience assessment curve to reduce the non-medical indication cesarean section rate to below 7%. Finally, an intelligent labor management file is generated, which includes holographic monitoring data, coupled risk quantification, resilience assessment results, and collaborative intervention records.

[0010] The working principle and effects of the above technical solution are as follows: By dividing the monitoring area into multimodal zones and constructing an intelligent monitoring layout network, multi-dimensional synchronous monitoring of maternal physiology, psychology, and fetal biochemistry is achieved, improving the comprehensiveness and accuracy of labor monitoring and enhancing the accuracy of cross-modal risk feature identification. Through spatiotemporal coupling map construction and iterative optimization of graph neural networks, the accuracy of labor resilience index calculation is improved, reducing the probability of risk misjudgment. Triggering a tiered collaborative response mechanism allows for rapid matching of intervention measures, avoiding maternal and infant safety hazards caused by risk escalation and preventing adverse delivery outcomes due to untimely intervention. By using resilience assessment curves to inversely constrain intervention decisions, the rate of non-medical cesarean sections is reduced, keeping it below 7%, thus reducing unnecessary surgical interventions. This solution not only provides real-time monitoring of the entire labor process but also standardizes intervention procedures, generates complete management files, improves the professionalism and standardization of labor management, and effectively ensures maternal and infant safety.

[0011] In one embodiment of the present invention, S1 includes: S11. For primiparous women, physiological dimension monitoring areas are sorted out throughout the delivery process, and core obstetric parameters are divided into monitoring areas. These areas comprehensively cover key physiological indicators of labor such as cervical dilation status, changes in uterine contraction intensity, and fetal heart rate fluctuations, providing a basic monitoring range for tracking the labor process. S12. Combining the non-contact acquisition characteristics of millimeter-wave radar, a monitoring area for maternal psychological stress is defined. This area focuses on changes in maternal facial micro-expressions and the frequency of limb movements, capturing the external behavioral manifestations corresponding to changes in maternal psychological state during childbirth. S13. Based on the detection capabilities of the amniotic fluid biochemical probe, the fetal biochemical environment monitoring area is divided. This area mainly includes indicators such as amniotic fluid pH value and lactate concentration, which reflect the fetal living environment and metabolic status in the uterus in real time. S14. Within the three designated monitoring areas, rationally deploy the LP delivery monitoring system, non-contact psychological sensor, and amniotic fluid biochemical probe, and optimize the spatial distribution and detection angle of each monitoring device. S15. Integrate the working links of three types of equipment: physiological monitoring, psychological monitoring, and biochemical monitoring, and build an intelligent monitoring layout network. The intelligent monitoring layout network integrates multiple dimensions of physiology, psychology, and biochemistry to ensure the comprehensive collection of childbirth-related data.

[0012] The working principle and effects of the above technical solution are as follows: By dividing the entire labor process for primiparous women into multi-dimensional monitoring areas and clarifying the monitoring focus of each area, the targetedness and comprehensiveness of labor monitoring are improved, avoiding the omission of key indicators caused by single-dimensional monitoring. By combining the characteristics of millimeter-wave radar and amniotic fluid biochemical probes to divide corresponding monitoring areas, the accuracy of psychological stress signals and fetal biochemical data collection is enhanced, reducing monitoring bias. By rationally deploying monitoring equipment in each area and optimizing the equipment layout, the operating efficiency of the equipment is improved, avoiding monitoring blind spots caused by unreasonable equipment distribution. By integrating the working links of the three types of monitoring equipment to build an intelligent monitoring network, multi-dimensional data synchronous collection is achieved, which not only ensures the integrity of monitoring data, but also lays a solid foundation for subsequent data processing and risk identification, reducing the subsequent work problems caused by incomplete data collection.

[0013] In one embodiment of the present invention, S2 includes: S21. Based on the real-time progress of labor, dynamically adjust the data collection frequency and working time of each monitoring device in the intelligent monitoring network to adapt to the data collection needs of different stages of labor. S22. Through intelligent monitoring network layout, core obstetric parameters, psychological stress signals and raw biochemical data of amniotic fluid are collected synchronously, and multiple types of data are uniformly collected to form a multi-source heterogeneous delivery data stream. S23. Noise filtering and interference removal are carried out on the collected psychological stress signals. The core emotional characteristics of the mother, such as anxiety and pain, are extracted from the purified signals, and invalid signal interference is removed. S24. Perform baseline drift correction on the raw amniotic fluid biochemical data, and combine clinical standards to conduct out-of-limit early warning judgments for various biochemical indicators and mark abnormal data nodes. S25. Integrate the processed psychological and emotional characteristics with the corrected biochemical indicators to form standardized psychological and biochemical coupled data, eliminating differences in format and magnitude among multi-source data.

[0014] The working principle and effects of the above technical solution are as follows: By dynamically adjusting the acquisition frequency and working duration of the monitoring equipment according to the real-time status of labor, the needs of different stages of labor are adapted, improving the rationality of data acquisition and avoiding excessive equipment load caused by excessive acquisition frequency, as well as the loss of key data caused by insufficient acquisition frequency. By simultaneously acquiring multiple types of raw data and aggregating them into a multi-source heterogeneous data stream, the comprehensiveness of data acquisition is enhanced, and data omissions are reduced. By filtering noise and removing interference from psychological stress signals, the accuracy of emotional feature extraction is improved, avoiding misjudgments of psychological state caused by invalid signal interference. By performing baseline drift correction and over-limit warning on amniotic fluid biochemical data, abnormal nodes can be marked in a timely manner, reducing the omission of abnormal biochemical indicators. By integrating and processing the data to form standardized coupled data, the differences in format and magnitude of multi-source data are eliminated, which not only provides reliable data support for subsequent risk analysis, but also improves the efficiency of subsequent data processing and avoids processing chaos caused by data inconsistency.

[0015] In one embodiment of the present invention, step S21 includes: Track the real-time progress of labor, collect information on changes in the stages of labor, and generate dynamic data on changes in labor; combine this dynamic data to adjust the data collection frequency of each monitoring device in the intelligent monitoring network. Based on the results of the data collection frequency adjustment, allocate corresponding working hours for each monitoring device; Optimize the data acquisition and operation modes of monitoring equipment to match the changing characteristics of the labor stage; It adapts to the data collection needs of different stages of labor, forming a stable and efficient monitoring and collection system.

[0016] The working principle and effects of the above technical solution are as follows: By tracking the real-time progress of labor and collecting information on stage changes, complete dynamic data on labor progress is generated, enabling precise control of the delivery process and avoiding gaps in data collection strategies due to untimely understanding of labor changes. By adjusting the collection frequency of each monitoring device in conjunction with dynamic labor data, the collection rhythm is aligned with the characteristics of each labor stage, improving the targeting of data collection and reducing resource waste caused by ineffective collection. By allocating device working hours according to collection frequency, the equipment operating load is balanced, avoiding the risk of failure caused by excessive operation or idleness. By optimizing the collection operation mode to match the characteristics of each labor stage, the stability of monitoring and collection is enhanced, avoiding insufficient adaptability caused by a single collection mode. By adapting to the collection needs of different labor stages, a stable and efficient monitoring state is formed, ensuring the accurate collection of key data and improving equipment operating efficiency, providing a reliable foundation for subsequent data processing.

[0017] In one embodiment of the present invention, S3 includes: S31. Standardize the psychological and biochemical coupling data and input them into the spatiotemporal graph convolutional network model to complete the format adaptation of multimodal data and model loading. S32. A spatiotemporal coupled graph is constructed by spatiotemporal graph convolutional network. The graph is fully associated with information such as the rate of cervical dilation, uterine contraction rhythm, pain expression characteristics of the parturient, fetal heart rate variability, amniotic fluid pH, and biochemical metabolic rate. S33. Conduct cross-dimensional correlation analysis on each data node in the spatiotemporal coupling map to explore the inherent correlation logic between physiological, psychological and biochemical data and extract cross-modal childbirth risk characteristics. S34. Quantitatively integrate the extracted risk characteristics to comprehensively assess the probability and impact of compound risks such as uterine atony, fetal distress, and psychological breakdown. S35. Based on the quantitative results of composite risks, a coupled risk index is generated that can comprehensively reflect the safety status of childbirth, providing a numerical reference for subsequent risk management.

[0018] The working principle and effects of the above technical solution are as follows: By uniformly inputting standardized psychological and biochemical coupled data into the spatiotemporal graph convolutional network model and completing format adaptation, the processing efficiency of multimodal data is improved, and analysis interruptions caused by data format conflicts are avoided. By constructing a spatiotemporal coupled graph to associate multiple indicators of the mother and fetus, the spatiotemporal correlation of delivery data is enhanced, and information fragmentation caused by independent analysis is reduced. By conducting cross-dimensional correlation analysis on graph nodes, the inherent logic of multi-dimensional data is explored, improving the completeness of risk feature extraction and avoiding risk omissions caused by single-dimensional judgment. By quantifying and integrating risk features and evaluating the degree of composite risk, the objectivity of risk assessment is improved, and biases caused by subjective judgment are reduced. By generating a coupled risk index, a numerical reference for risk management is provided, which not only allows for a visual presentation of the safety status of delivery but also provides a precise basis for subsequent intervention measures, avoiding adverse delivery outcomes caused by delayed risk management.

[0019] In one embodiment of the present invention, S33 includes: S331. Traverse all data nodes in the spatiotemporal coupling graph, comprehensively collect multi-dimensional information carried by various nodes such as maternal physiological indicators, psychological and emotional characteristics and fetal biochemical data, and ensure that no key data is omitted. S332. Conduct cross-dimensional correlation analysis on each category of the collected multi-dimensional node information, and establish the linkage relationship between physiological, psychological and biochemical data in combination with the spatiotemporal change pattern of labor. S333. Based on the established linkage relationship, conduct in-depth analysis of the synchronous change trend of various data in the progress of labor, and explore the inherent correlation logic between fluctuations in physiological indicators, changes in psychological state and abnormalities in biochemical indicators. S334. Based on the discovered internal correlation logic, select key feature information that can reflect abnormal labor conditions, and filter out redundant data and interference features that are irrelevant to labor risks. S335. Integrate the key feature information after screening, and combine it with the labor safety assessment criteria to extract representative and targeted cross-modal delivery risk features, covering core risk points such as uterine contractions, fetal status and maternal psychology.

[0020] The working principle and effects of the above technical solution are as follows: By traversing all data nodes of the spatiotemporal coupling map, multi-dimensional information is comprehensively collected, ensuring that no key data is omitted and avoiding risk analysis bias caused by missing data. Cross-dimensional correlation analysis is conducted on each category of the collected multi-dimensional information, and linkage relationships are established in conjunction with the spatiotemporal patterns of labor, enhancing the correlation between multiple types of data and reducing information fragmentation caused by independent data analysis. In-depth analysis of synchronous data change trends and the discovery of inherent correlation logic improves the grasp of patterns in childbirth data and avoids the problem of failing to identify potential correlations between data. By selecting key features and filtering redundant interference based on correlation logic, the accuracy of risk features is improved, reducing the impact of invalid data on risk judgment. By integrating the selected features and combining them with labor safety standards to extract cross-modal risk features, both core risk points are comprehensively covered, and the representativeness and relevance of risk features are ensured, avoiding inappropriate handling due to one-sided risk identification and providing reliable support for subsequent risk quantification and intervention.

[0021] In one embodiment of the present invention, S333 includes: Collect continuous spatiotemporal change information during the progress of labor, record the node characteristics and temporal evolution patterns of different stages of labor, and form a complete and continuous dynamic change sequence of labor. By combining the generated dynamic sequence of labor progress, the changes of three types of data—physiological indicators, psychological and emotional characteristics, and fetal biochemical data—are tracked synchronously throughout the entire time period to ensure that the data changes accurately correspond to the progress of labor. For the three types of data tracked, we analyze their magnitude of change, trend of change, and mutual influence at the same spatiotemporal node, and extract multi-dimensional data linkage change characteristics; Based on the extracted data linkage and change characteristics, we conducted an in-depth analysis of the impact path of physiological index fluctuations on the postpartum woman's psychological state, as well as the transmission law of psychological state changes in turn affecting physiological indicators. Based on the identified physiological and psychological transmission relationship, we further correlated the abnormal performance of fetal biochemical indicators, clarified the interaction mechanism among the three types of data, and ultimately formed a complete internal logical connection.

[0022] The working principle and effects of the above technical solution are as follows: By collecting continuous spatiotemporal change information of the labor process, recording the characteristics and evolution patterns of each stage, a complete dynamic sequence of the labor process is formed, avoiding data analysis gaps caused by fragmented labor process information. By combining the dynamic sequence of the labor process, the synchronous changes of three types of data are tracked throughout the entire time, ensuring that the data accurately corresponds to the progress of the labor process, improving the timeliness of data analysis, and reducing misjudgments caused by data being out of sync with the labor process. By analyzing the magnitude of changes and mutual influences of the three types of data at the same spatiotemporal node, the linkage change characteristics are extracted, enhancing the correlation of multi-dimensional data and avoiding the omission of patterns caused by isolated data analysis. By deeply analyzing the mutual influence paths and transmission patterns of physiological and psychological factors, the relationship between the two is clarified, reducing cognitive biases regarding the linkage mechanism between mind and body. By linking abnormal fetal biochemical indicators, the interaction mechanism of the three types of data is clarified, forming a complete internal correlation logic. This not only accurately grasps the linkage patterns of each dimension during childbirth but also provides a reliable basis for subsequent risk feature screening, avoiding risk identification biases caused by unclear correlation logic.

[0023] In one embodiment of the present invention, step S4 includes: S41. Based on the coupling risk index, numerical calculations are carried out to derive the labor resilience index, which reflects the mother's labor tolerance and intuitively reflects the mother's overall state in coping with the pressure of childbirth. S42. When the labor resilience index is below 0.6 and the amniotic fluid pH is below 7.0, a knowledge graph-driven hierarchical collaborative response mechanism is activated to match intervention measures corresponding to the risk level. S43. When implementing a Level 1 response, non-pharmacological analgesia intervention and emotional support and guidance for family members should be carried out. When implementing a Level 2 response, the dosage of oxytocin should be adjusted and preparations for amniotic fluid replacement should be completed. When implementing a Level 3 response, a multidisciplinary consultation should be organized and an emergency cesarean section execution plan should be formulated. S44. Track the implementation progress of response measures at all levels in real time, record the changes in various data of the mother and fetus during the intervention process, and monitor the actual effect of the intervention measures. S45. Summarize the current delivery risk level, various intervention measures and subsequent prognostic assessments, and integrate them into a dynamic response report, which contains complete treatment information.

[0024] The working principle and effects of the above technical solution are as follows: By accurately judging the labor resilience index and amniotic fluid indicators, a graded collaborative response is initiated, improving the targeting and timeliness of labor intervention and reducing the adverse consequences of risk delays. The precise matching of intervention measures is optimized, with different treatment plans corresponding to different risk levels, reducing the probability of over-intervention or under-intervention and avoiding adverse effects on maternal and infant safety due to improper handling. Real-time tracking of intervention progress and data changes enhances the controllability of intervention effects and reduces the hidden dangers of blind intervention. The resulting dynamic response report provides a reliable basis for subsequent intervention adjustments and completely preserves the treatment process, avoiding problems of untraceable intervention processes and unclear responsibilities. It simultaneously considers maternal and infant safety and intervention efficiency, making labor intervention more scientific and standardized.

[0025] In one embodiment of the present invention, S42 includes: Real-time values ​​of the labor resilience index are collected, and real-time data on amniotic fluid pH are collected simultaneously to form a two-dimensional risk assessment basis. Parallel comparison of the basic data for dual-dimensional risk assessment is performed to determine whether both types of indicators are simultaneously in the abnormal range. When both types of indicators are in the abnormal range, the internal storage of the knowledge graph's childbirth risk management rule base is activated; the internal rule base of the knowledge graph is called to match the management path and intervention direction corresponding to the current childbirth risk level; The tiered collaborative response mechanism is activated according to the matched treatment path, and the corresponding level of intervention execution instructions are pushed out.

[0026] The working principle and effects of the above technical solution are as follows: By collecting real-time data on labor resilience index and amniotic fluid pH, the physical condition of the mother and fetus is accurately captured, improving the targeting and accuracy of labor monitoring and reducing misjudgments caused by data bias. It reduces the blindness of manual intervention, minimizes ineffective operations and resource waste, avoids adverse delivery outcomes caused by misjudgment, and prevents maternal and infant risks due to inappropriate intervention. It can quickly identify abnormal situations during labor and promptly initiate corresponding intervention measures, while ensuring precise intervention direction, effectively protecting maternal and infant safety, reducing unnecessary medical interventions, and making the labor process safer and more efficient. Simultaneously, it considers the completeness of data collection and the scientific nature of intervention, aligning with actual clinical needs and improving the overall quality of labor management.

[0027] In one embodiment of the present invention, step S5 includes: S51. Extract risk changes and intervention effect data from the dynamic response report, and update the node information and correlation of the original spatiotemporal coupling map based on this data to optimize the accuracy of the map data; S52. Using the updated spatiotemporal coupled atlas data, the calculation model of the labor resilience index is repeatedly optimized through graph neural networks to improve the accuracy and adaptability of the index calculation. S53. Based on the optimized model, continuous calculations are performed on the data throughout the entire labor process to generate a resilience assessment curve that can fully reflect the changes in resilience throughout the entire labor process, and intuitively present the evolution trend of the labor state. S54. When the indications for cesarean section are dominated by non-medical factors, the clinical intervention decision should be constrained by the resilience assessment curve to gradually control and reduce the cesarean section rate for non-medical indications to below 7%. S55. Summarize the original data of the whole process of labor holographic monitoring, couple the risk quantification results, resilience assessment curves and the whole process records of collaborative intervention, and integrate them to generate a complete intelligent childbirth management file.

[0028] The working principle and effects of the above technical solution are as follows: By extracting data from the dynamic response report and updating the spatiotemporal coupling map, the accuracy of the map data is improved, avoiding misjudgments of risk due to data lag. The labor resilience index calculation model is optimized, enhancing the accuracy and adaptability of the index calculation and reducing calculation bias. A full-process labor resilience assessment curve is generated, clearly presenting the evolution of the labor state and avoiding intervention bias caused by incomplete assessment. By constraining intervention decisions through the resilience assessment curve, the cesarean section rate dominated by non-medical factors is reduced, minimizing unnecessary surgical interventions. This solution comprehensively retains data throughout the entire labor process and standardizes labor management, avoiding adverse labor outcomes caused by fragmented information and blind decision-making, making labor management more scientific and efficient.

[0029] In one embodiment of the present invention, S52 includes: Extract all node data and correlation data from the updated spatiotemporal coupling graph, and organize them into a standardized model optimization input dataset; The standardized model optimization input dataset is imported into the graph neural network to initiate the iterative optimization process of the labor resilience index calculation model; The model parameters were initially adjusted using a graph neural network. The exponential calculation results before and after optimization were compared to identify effective directions for parameter adjustment. Based on the effective direction of parameter adjustment, the model calculation logic is repeatedly optimized through iterative optimization to correct the model's bias in weight allocation of multi-dimensional data. After multiple rounds of iterative optimization, an optimized model for calculating the labor resilience index is output, which significantly improves the accuracy of index calculation and its adaptability to the labor process.

[0030] The working principle and effects of the above technical solution are as follows: By extracting the nodes and related data of the updated spatiotemporal coupling graph and organizing them into a standardized input dataset, the model optimization bias caused by data clutter can be effectively avoided. The graph neural network is then used for model iteration, repeatedly adjusting parameters and filtering effective directions to correct weight allocation biases, significantly improving the accuracy of the labor resilience index calculation and reducing calculation errors. The optimized model can better adapt to different stages of labor, avoiding assessment distortion caused by insufficient model adaptation. It can accurately capture changes in labor data and improve the index's ability to reflect the labor status, providing reliable support for subsequent risk assessment and intervention decisions. Simultaneously, it reduces resource waste caused by ineffective iterations, making the resilience index more closely reflect actual labor scenarios and ensuring the scientific nature of subsequent assessments and decisions.

[0031] One embodiment of the present invention provides an intelligent labor management system integrating LP labor detection and multi-parameter monitoring, comprising: One or more processors; Memory, used to store one or more programs; Wherein, when the one or more programs are executed by the one or more processors, the one or more processors are made to implement the method described in any one of the above.

[0032] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. An intelligent labor management method integrating LP labor detection and multi-parameter monitoring, characterized in that, The method includes: S1. Divide the multimodal monitoring areas of the primiparous woman's delivery process, generate monitoring areas for core obstetric parameters, and monitor areas for psychological stress by collecting data through millimeter-wave radar and for continuous monitoring of the fetal biochemical environment by amniotic fluid biochemical probes; based on the above areas, deploy the LP delivery detection system, non-contact psychological sensors and amniotic fluid biochemical probes to construct an intelligent monitoring layout network; S2. Based on the intelligent monitoring network layout, the monitoring frequency is dynamically adjusted to synchronously collect core obstetric parameters, maternal psychological stress signals, and amniotic fluid biochemical data to generate multi-source heterogeneous delivery data streams; noise suppression and emotional feature extraction are performed on psychological stress signals, and drift correction and threshold warning processing are performed on amniotic fluid biochemical data to generate standardized psychological and biochemical coupled data. S3. Input standardized data into a spatiotemporal graph convolutional network to construct a spatiotemporal coupled graph. Extract cross-modal risk features through graph node correlation analysis and generate a coupled risk index of composite risk. S4. Calculate the labor resilience index based on the coupling risk index. When LRI < 0.6 and amniotic fluid pH < 7.0, trigger a knowledge graph-driven three-level collaborative response; and simultaneously generate a dynamic response report. S5. Update the spatiotemporal coupling map based on the dynamic response report, optimize the calculation model of the labor resilience index through graph neural network iteration, and generate a full labor resilience assessment curve; when the indication for cesarean section is only due to non-medical factors, the intervention decision is constrained by the resilience assessment curve, and finally an intelligent labor management file is generated.

2. The intelligent labor management method integrating LP labor detection and multi-parameter monitoring according to claim 1, characterized in that, S1 includes: S11. Conduct a physiological dimension monitoring area review for primiparous women throughout the entire delivery process and divide the core obstetric parameter monitoring areas; S12. Combining the non-contact acquisition characteristics of millimeter-wave radar, the monitoring area for maternal psychological stress is divided. S13. Based on the detection capabilities of the amniotic fluid biochemical probe, the fetal biochemical environment monitoring area is divided to reflect the fetal living environment and metabolic status in the uterus in real time. S14. Within the three designated monitoring areas, rationally deploy the LP delivery monitoring system, non-contact psychological sensor, and amniotic fluid biochemical probe, and optimize the spatial distribution and detection angle of each monitoring device. S15. Integrate the working links of three types of equipment: physiological monitoring, psychological monitoring, and biochemical monitoring, and build an intelligent monitoring layout network.

3. The intelligent labor management method integrating LP labor detection and multi-parameter monitoring according to claim 1, characterized in that, The S2 includes: S21. Based on the real-time progress of labor, dynamically adjust the data collection frequency and working time of each monitoring device in the intelligent monitoring network to adapt to the data collection needs of different stages of labor. S22. Through intelligent monitoring network layout, core obstetric parameters, psychological stress signals and raw biochemical data of amniotic fluid are collected synchronously, and multiple types of data are uniformly collected to form a multi-source heterogeneous delivery data stream. S23. Noise filtering and interference removal are carried out on the collected psychological stress signals, and the core emotional features are extracted from the purified signals to remove invalid signal interference. S24. Perform baseline drift correction on the raw amniotic fluid biochemical data, and combine clinical standards to conduct out-of-limit early warning judgments for various biochemical indicators and mark abnormal data nodes. S25. Integrate the processed psychological and emotional characteristics with the corrected biochemical indicators to form standardized psychological and biochemical coupled data, eliminating differences in format and magnitude among multi-source data.

4. The intelligent labor management method integrating LP labor detection and multi-parameter monitoring according to claim 1, characterized in that, The S3 includes: S31. Standardize the psychological and biochemical coupling data and input them into the spatiotemporal graph convolutional network model to complete the format adaptation of multimodal data and model loading. S32. Construct a spatiotemporally coupled graph using a spatiotemporal graph convolutional network; S33. Conduct cross-dimensional correlation analysis on each data node in the spatiotemporal coupling map to explore the inherent correlation logic between physiological, psychological and biochemical data and extract cross-modal childbirth risk characteristics. S34. Quantitatively integrate the extracted risk characteristics and comprehensively evaluate the probability of occurrence and the degree of impact of compound risks; S35. Based on the quantitative results of composite risks, a coupled risk index is generated that can comprehensively reflect the safety status of childbirth, providing a numerical reference for subsequent risk management.

5. The intelligent labor management method integrating LP labor detection and multi-parameter monitoring according to claim 4, characterized in that, S33 includes: S331. Traverse all data nodes in the spatiotemporal coupling graph and comprehensively collect the multi-dimensional information carried by various nodes. S332. Conduct cross-dimensional correlation analysis on each category of the collected multi-dimensional node information, and establish the linkage relationship between physiological, psychological and biochemical data in combination with the spatiotemporal change pattern of labor. S333. Based on the established linkage relationship, conduct in-depth analysis of the synchronous change trend of various data in the progress of labor, and explore the inherent correlation logic between fluctuations in physiological indicators, changes in psychological state and abnormalities in biochemical indicators. S334. Based on the discovered internal correlation logic, select key feature information that can reflect abnormal labor conditions, and filter out redundant data and interference features that are irrelevant to labor risks. S335. Integrate the key feature information after screening, and combine it with the labor safety assessment criteria to extract cross-modal delivery risk features that are both representative and targeted.

6. The intelligent labor management method integrating LP labor detection and multi-parameter monitoring according to claim 5, characterized in that, S333 includes: Collect continuous spatiotemporal change information during the progress of labor, record the node characteristics and temporal evolution patterns of different stages of labor, and form a complete and continuous dynamic change sequence of labor. By combining the generated dynamic sequence of labor progress, the changes of three types of data—physiological indicators, psychological and emotional characteristics, and fetal biochemical data—are tracked synchronously throughout the entire time period to ensure that the data changes accurately correspond to the progress of labor. For the three types of data tracked, we analyze their magnitude of change, trend of change, and mutual influence at the same spatiotemporal node, and extract multi-dimensional data linkage change characteristics; Based on the extracted data linkage and change characteristics, we conducted an in-depth analysis of the impact path of physiological index fluctuations on the postpartum woman's psychological state, as well as the transmission law of psychological state changes in turn affecting physiological indicators. Based on the identified physiological and psychological transmission relationship, we further correlated the abnormal performance of fetal biochemical indicators, clarified the interaction mechanism among the three types of data, and ultimately formed a complete internal logical connection.

7. The intelligent labor management method integrating LP labor detection and multi-parameter monitoring according to claim 1, characterized in that, The S4 includes: S41. Based on the coupling risk index, numerical calculations are carried out to derive the labor resilience index, which reflects the mother's labor tolerance and intuitively reflects the mother's overall state in coping with the pressure of childbirth. S42. When the labor resilience index is below 0.6 and the amniotic fluid pH is below 7.0, a knowledge graph-driven hierarchical collaborative response mechanism is activated to match intervention measures corresponding to the risk level. S43. When implementing a Level 1 response, non-pharmacological analgesia intervention and emotional support and guidance for family members should be carried out. When implementing a Level 2 response, the dosage of oxytocin should be adjusted and preparations for amniotic fluid replacement should be completed. When implementing a Level 3 response, a multidisciplinary consultation should be organized and an emergency cesarean section execution plan should be formulated. S44. Track the implementation progress of response measures at all levels in real time, record the changes in various data of the mother and fetus during the intervention process, and monitor the actual effect of the intervention measures. S45. Summarize the current delivery risk level, various intervention measures and subsequent prognostic assessments, and integrate them into a dynamic response report.

8. The intelligent labor management method integrating LP labor detection and multi-parameter monitoring according to claim 7, characterized in that, S42 includes: Real-time values ​​of the labor resilience index are collected, and real-time data on amniotic fluid pH are collected simultaneously to form a two-dimensional risk assessment basis. Parallel comparison of the basic data for dual-dimensional risk assessment is performed to determine whether both types of indicators are simultaneously in the abnormal range. When both types of indicators are in the abnormal range, the internal storage of the knowledge graph's childbirth risk management rule base is activated; the internal rule base of the knowledge graph is called to match the management path and intervention direction corresponding to the current childbirth risk level; The tiered collaborative response mechanism is activated according to the matched treatment path, and the corresponding level of intervention execution instructions are pushed out.

9. The intelligent labor management method integrating LP labor detection and multi-parameter monitoring according to claim 1, characterized in that, The S5 includes: S51. Extract risk changes and intervention effect data from the dynamic response report, and update the node information and correlation of the original spatiotemporal coupling map based on this data to optimize the accuracy of the map data; S52. Using the updated spatiotemporal coupled atlas data, the calculation model for the labor resilience index is iteratively optimized through graph neural networks. S53. Based on the optimized model, continuous calculations are performed on the data throughout the entire labor process to generate a resilience assessment curve that can fully reflect the changes in resilience throughout the entire labor process, and intuitively present the evolution trend of the labor state. S54. When the indications for cesarean section are dominated by non-medical factors, the clinical intervention decision should be constrained by the resilience assessment curve to gradually control and reduce the cesarean section rate for non-medical indications to below 7%. S55. Summarize the original data of the whole process of labor holographic monitoring, couple the risk quantification results, resilience assessment curves and the whole process records of collaborative intervention, and integrate them to generate a complete intelligent childbirth management file.

10. An intelligent labor management system integrating LP labor monitoring and multi-parameter monitoring, including: One or more processors; Memory, used to store one or more programs; Wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the method of any one of claims 1 to 9.

Images