Newborn respiratory distress risk assessment construction method

Abstract

The invention relates to a neonatal respiratory distress risk assessment construction method, which comprises the following steps of: implementing non-intrusive micro-disturbance operation on a neonatal, and detecting respiratory rhythm phase drift, respiratory driving force low-frequency suppression and physiological tremor window change under disturbance so as to identify a pre-instability signal regulated and controlled by a respiratory system; collecting nerve and respiration feedback path data, analyzing the time difference between respiration instruction conduction and actual respiration starting, and identifying an abnormal decoupling phenomenon; based on multiple times of perturbation induction and feedback acquisition results, continuously monitoring the micro-adaptive recovery capability of the respiratory system of the newborn, and identifying the hidden fatigue trend of the respiratory system by judging progressive extension of respiratory frequency recovery time and gradual upward movement of respiratory power consumption; the respiratory redundancy regulation capability in the stress regulation process is monitored, premature intervention and repeated activation of an auxiliary breathing mode or a respiratory muscle group are judged, and the redundancy regulation critical compression phenomenon of a respiratory regulation system is analyzed to complete evaluation construction of the neonatal respiratory distress risk.

Description

Technical Field

[0001] This invention relates to neonatal respiratory monitoring, specifically a method for constructing a neonatal respiratory distress risk assessment. Background Technology

[0002] Current methods for assessing neonatal respiratory distress risk, such as the non-contact neonatal respiratory monitoring system, device, and storage medium (Chinese patent CN112971743A), employ a non-contact respiratory and heartbeat signal monitoring system based on 24GHz MIMO-FMCW radar. This system extracts instantaneous indicators, such as respiratory rate, respiratory interval, heart rate, and apnea, and sets thresholds to identify respiratory or heartbeat abnormalities, issuing warnings or alarms based on the type of abnormality. While this method offers advantages such as real-time monitoring, non-contact operation, and intuitive feedback, and to some extent meets the respiratory monitoring needs of high-risk newborns like premature infants, it still has significant technical shortcomings and potential drawbacks in terms of systemic, dynamic, and in-depth risk warning capabilities. First, this method only focuses on monitoring instantaneous fluctuations in respiration and heartbeat; its core mechanism relies on whether the current signal exceeds or falls below a set threshold. Its assessment model is static and single-channel, failing to reflect the overall health status of the newborn's autonomic regulatory system. In other words, current technology lacks the ability to deeply analyze the operational status of the neonatal neuromuscular-metabolic coupling system, and cannot effectively identify potential, unmanifested regulatory fatigue or regulatory breakdown processes. For example, when the respiratory system faces micro-perturbations in the early stages, this method can only detect obvious frequency abnormalities, but cannot capture subclinical signs of respiratory instability caused by mechanisms such as dissipation of neural regulatory resources and weakening of muscle activation redundancy.

[0003] Secondly, this invention only uses radar equipment to acquire respiratory / cardiac displacement signals and extracts features through frequency domain analysis. While non-contact is an advantage, radar itself is completely unable to perceive core data such as electromyography, muscle coordination, micro-tremor changes, and central nervous system activation signals. This results in a natural deficiency in assessing the functional state, recovery ability, and neuromuscular synergy of the respiratory regulation system. In fact, respiratory distress in premature infants often stems from processes such as delayed neural output, exhaustion of muscle activation redundancy, and sudden changes in energy metabolism critical points. These risks precede the appearance of phenotypic asphyxia signals. Relying on this method often only provides "post-mortem warnings," lacking true early warning capabilities. Thirdly, this method uses individual monitoring or combined discrimination of characteristic parameters for anomaly identification. Its risk assessment mechanism is based on whether the risk exceeds a predefined range. This method is difficult to adapt to the highly dynamic and physiologically diverse neonatal population. For example, even if the respiratory rate of some newborns is within the normal range, their neuroregulatory system may be on the verge of fatigue and imbalance due to continuous high-frequency mobilization. This invention, lacking a dynamic evolutionary perspective, will be unable to predict the risk. In addition, the scheme has weak capabilities in risk grading and individual difference modeling, and cannot construct personalized risk curves for newborns at different developmental levels or identify the evolutionary trajectory of their regulatory abilities.

[0004] Furthermore, existing technologies do not consider the dynamic activation process of accessory respiratory muscles. In actual clinical practice, imbalances in the respiratory regulation system often manifest first as weakness or sluggishness of the primary respiratory muscles, followed by compensation by accessory muscles. This system cannot monitor the allocation of muscle regulatory resources, assess the dissipation of redundant mechanisms, or determine the nonlinear evolutionary critical point from elastic recovery to fragile recovery. When the neonatal regulatory system resources are exhausted and unable to cope with minor disturbances, entering a respiratory collapse state, there is often only a very short time window for intervention, which this system cannot predict. In addition, this method has not yet established a closed-loop model for micro-perturbation-system response; its mechanism is based on passive acquisition and static discrimination, and cannot implement active intervention and dynamic feedback closed-loop control. Finally, its anomaly handling module adopts a simple priority strategy, which is not dynamically coupled with the evolutionary state of the neonatal autonomous regulatory system, and cannot achieve real-time risk grading and dynamic calculation of the intervention window. Its design is a static level mapping, lacking system dynamics support. Therefore, in applications, if abnormal indicators fluctuate frequently or there are conflicts between multiple channels, the system may experience problems such as false alarms, missed alarms, and delayed feedback, making it difficult to meet the high sensitivity and dynamic requirements of neonatal intensive care scenarios. Summary of the Invention

[0005] The purpose of this invention is to provide a method for constructing a neonatal respiratory distress risk assessment, thereby addressing some of the drawbacks and shortcomings pointed out in the background art.

[0006] The present invention adopts the following technical solution to solve the above-mentioned technical problems: a method for constructing a neonatal respiratory distress risk assessment, comprising: performing non-invasive micro-perturbation operation on the newborn, actively inducing the extreme boundary response of the neonatal respiratory regulation system, and detecting the respiratory rhythm phase drift, low-frequency inhibition of respiratory driving force and physiological fibrillation window changes under the perturbation, so as to identify the pre-instability signal of respiratory system regulation;

[0007] Data on neural and respiratory feedback pathways were collected to analyze the time difference between respiratory command transmission and actual respiratory initiation, and to identify vagal feedback lag and abnormal decoupling phenomena of respiratory receptors, central and peripheral regulatory circuits.

[0008] Based on multiple micro-perturbation inducements and feedback acquisition results, the micro-adaptive recovery ability of the neonatal respiratory system is continuously monitored. By judging the gradual prolongation of respiratory rate recovery time and the gradual increase of respiratory power consumption, the hidden fatigue trend of the respiratory system is identified. The respiratory redundancy regulation ability during stress regulation is monitored to judge the premature intervention and repeated activation of assisted breathing modes or respiratory muscle groups. The redundancy regulation critical compression phenomenon of the respiratory regulation system is analyzed to complete the assessment of neonatal respiratory distress risk.

[0009] Furthermore, the micro-perturbation operation dynamically adjusts the perturbation amplitude according to the newborn's real-time spontaneous breathing state, so that the perturbation intensity is coupled in a closed loop with the newborn's instantaneous respiratory rhythm; the detection of phase drift includes analyzing the dynamic phase trajectory of the respiratory regulation system based on phase space reconstruction technology to identify the phenomenon of respiratory rhythm stability degradation induced by micro-perturbation.

[0010] Furthermore, the analysis of low-frequency suppression includes assessing the trend of complexity loss in the neonatal respiratory regulation system by constructing a multi-scale entropy spectrum of respiratory drive signals; the detection of changes in the physiological tremor window includes collecting tremor micro-vibration signals of the respiratory muscle groups under the skin of newborns using a micromechanical sensor array, and identifying changes in the synchronicity of nerve and muscle activation through time-frequency decomposition to determine the boundary state of respiratory regulation.

[0011] Furthermore, the acquisition of the neural and respiratory feedback pathways includes monitoring the coupling relationship between the neonatal brain and respiration, and identifying changes in the consistency of signal transmission between central command issuance and peripheral execution through the synergistic analysis of cortical electroencephalography and respiratory action potentials; the identification of feedback pathway decoupling includes detecting the nonlinear neural regulation patterns of neonates, and dynamically judging the asymmetric amplification phenomenon of the neural and respiratory regulation chains by constructing a hysteresis loop of command response.

[0012] Furthermore, the continuous monitoring of micro-adaptive recovery capability includes iterative fitting of the recovery time and disturbance intensity function of the neonatal respiratory system to identify the transition point of the respiratory regulation system from the linear recovery stage to the fatigue instability stage; the identification of hidden fatigue includes monitoring minute respiratory pattern migration phenomena during breathing and revealing the non-obvious evolution process of system fatigue through the classification transformation of respiratory dynamics patterns.

[0013] Furthermore, the assessment of muscle involvement in assisted breathing includes dynamically tracking changes in the activation threshold of accessory muscles through ultrasound imaging combined with electromyography signal analysis, identifying the decline in spontaneous regulation ability and the activation trend of compensatory mechanisms; the critical compression analysis of redundant regulation includes establishing the cumulative dissipation relationship of redundant regulation resources, and dynamically assessing whether the regulatory resources of neonatal respiratory regulation are close to the depletion boundary by combining the statistics of micro-perturbation counts; the risk assessment construction process includes full-process energy metabolism estimation based on micro-perturbation response, and establishing neonatal respiratory metabolism coupling risk indicators through the energy consumption curve corresponding to changes in respiratory load.

[0014] By combining ultrasound images and electromyography signals, the activation intensity M(t) and activation time of the accessory respiratory muscle groups are dynamically captured, and individualized muscle activation threshold curves are constructed to determine the trend of declining autonomic regulation ability and the activation of compensatory mechanisms. By statistically analyzing the number of micro-perturbations N and the system response intensity R(t) after each perturbation, a dissipation accumulation function of respiratory regulation redundancy is established to dynamically estimate whether the newborn's respiratory regulation system is close to the exhaustion boundary.

[0015] Based on the above, using the change in respiratory load L(t) after micro-perturbation and the corresponding energy consumption curve E(t), the critical function of coupling risk between respiratory regulation and metabolic system is calculated; a comprehensive function is established to describe the critical model of interactive dissipation between respiratory regulation redundancy dissipation and energy metabolism.

[0016]

[0017] in:

[0018] Ψ(t) is the cumulative risk function of neonatal respiratory regulation-metabolism coupling; M(t) is the dynamic activation threshold function of the accessory respiratory muscles. Rate of change of accessory muscle activation intensity; R(t) intensity of micro-perturbation response of the respiratory system; N number of micro-perturbations in the current monitoring period; L(t) change in respiratory load; E(t) energy metabolism consumption curve; Rate of change in energy consumption; Reflecting the nonlinear interaction between respiratory load and disturbance response; It characterizes the diminishing marginal relationship of cumulative dissipation to prevent the disturbance from accumulating and increasing indefinitely;

[0019] When Ψ(t) exceeds the preset threshold Θ, the system determines that it is in a critical state of respiratory regulation imbalance; it can be used for: real-time early warning, dynamic construction of individual risk, and intervention window estimation.

[0020] Furthermore, during the tracking of muscle activation thresholds for assisted breathing, the intensity of muscle regulation commands output by the central nervous system is collected, and a synchronous dissipation curve of neural regulation resources and muscle redundancy is established to dynamically assess the risk of neuromuscular exhaustion of the autonomous regulation system. During the dissipation of redundant regulation resources, the transition stage of the respiratory regulation system from elastic recovery to fragile recovery is identified by constructing a nonlinear hysteresis relationship of disturbance recovery, thereby achieving boundary state judgment.

[0021] Furthermore, the risk assessment construction process includes simultaneously analyzing the dissipation curves of redundant regulatory resources and redundant energy metabolism, establishing a dual-channel imbalance judgment relationship, identifying the risk of respiratory imbalance caused by the mismatch between regulatory capacity and energy supply; and identifying the micro-delay risk window of muscle compensation initiation based on changes in activation threshold and muscle movement delay, which is used to predict the time-sensitive interval of the spontaneous breathing system nearing failure.

[0022] Furthermore, the muscle group activation threshold monitoring, combined with the micro-fatigue integral of local muscle groups, detects the correlation between micro-fatigue accumulation patterns and system-level respiratory regulation failure, enabling early warning of global imbalance caused by local load collapse.

[0023] Furthermore, during the construction of the energy metabolism risk index, the autonomous regulation mode shift event triggered by the metabolic critical point is monitored to determine the critical transition of the system from active regulation to passive compensation, which serves as the basis for risk classification. The redundant regulation resource assessment, combined with historical micro-perturbation data, constructs an individualized dynamic reconstruction relationship of respiratory regulation adaptability, dynamically updates the respiratory resource utilization pattern of newborns, and enables the evolution of individual risk curves.

[0024] The beneficial effects of this invention are as follows: Through non-invasive micro-perturbation, the extreme response state of the respiratory regulation system can be stimulated without interfering with the normal life activities of newborns, actively identifying potential imbalance risks. Compared with traditional passive assessment methods that rely on symptom manifestations, this method has stronger foresight and sensitivity. This method integrates multi-dimensional indicators such as phase drift, low-frequency inhibition, physiological fibrillation, neuro-respiratory feedback consistency, and hysteresis response, and can cross-validate the trend of regulatory capacity degradation from multiple system levels such as rhythm, energy, nerves, and muscles. It is particularly suitable for identifying respiratory regulation abnormalities in newborns before clinical symptoms appear. By introducing synergistic analysis between the dissipation curve of regulatory redundancy resources and energy metabolism dynamics, the mismatch between respiratory regulation capacity and energy supply can be identified, thereby capturing the system instability risk induced by energy bottlenecks and improving the depth and explanatory power of risk identification.

[0025] This invention supports continuous monitoring and data accumulation. It can combine historical micro-perturbation response data to construct individualized regulatory fitness models, enabling dynamic tracking of respiratory risk evolution. This adapts to the rapid changes in neonatal conditions, improving the responsiveness and accuracy of bedside decision-making. For example, integrated analysis of indicators such as delayed activation and micro-fatigue integrals can be used to construct a time prediction window for respiratory system near-failure, providing an actionable time basis for precise clinical intervention and effectively avoiding delays in treatment. When this method identifies the nonlinear transition node from elastic recovery to fragile recovery in the regulatory system, it can issue a real-time warning signal and, combined with individual status, output feasible intervention timing and mode suggestions, contributing to the construction of an intelligent auxiliary diagnosis and treatment system centered on early warning. Attached Figure Description

[0026] Figure 1 This is a functional relationship diagram of the neonatal respiratory distress risk assessment method of the present invention.

[0027] Figure 2 This is a simplified functional flowchart of the neonatal dynamic closed-loop micro-perturbation risk assessment of the present invention.

[0028] Figure 3 This is a simplified flowchart of the multi-channel assessment of neonatal respiratory regulation risk in this invention.

[0029] Figure 4 This is a simplified flowchart of the neonatal respiratory dynamic risk assessment in Embodiment 1 of the present invention.

[0030] Figure 5 This is a flowchart of the risk assessment for neonatal respiratory regulation imbalance using multi-source fusion in Embodiment 2 of the present invention. Detailed Implementation

[0031] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0032] Combined with appendix Figure 1This invention provides a method for constructing a neonatal respiratory distress risk assessment system. By performing non-invasive micro-perturbation manipulations on the newborn without interfering with their vital signs and spontaneous breathing function, the system stimulates the limiting boundary response of their respiratory regulation system, thereby identifying potential respiratory system imbalance risks. The non-invasive micro-perturbation manipulations can be achieved through flexible airflow, low-amplitude periodic sound pressure stimulation, or subtle differences in skin temperature. These perturbations are characterized by small amplitude, short cycle, and low invasiveness, and can activate regulatory response pathways in the neonatal respiratory regulation system without affecting the newborn's normal respiratory rhythm, thereby exposing the critical response state of their regulatory capacity. During the disturbance, real-time respiratory data of the newborn is collected, including but not limited to respiratory rate, inspiratory-to-expiratory ratio, respiratory waveform morphology, and chest-abdominal synchrony. A multimodal sensor array is used to acquire timely physiological feedback signals. A synchronization signal processing algorithm is employed to extract the respiratory rhythm phase shift index under disturbance conditions to detect the stability and recovery ability of the newborn's respiratory rhythm to external disturbances, thereby assessing the level of synchronous regulation of the respiratory center. Simultaneously, power spectral density analysis is performed on the respiratory driving force signal in the low-frequency range (e.g., 0.05Hz–0.2Hz) to identify whether there is suppression of low-frequency signals before and after the micro-disturbance. A significant downward trend indicates suppression of the autonomous driving ability or excessive regulatory load. Furthermore, micro-vibration sensing technology is used to collect minute vibration signals from the chest and abdominal muscle groups and respiratory tract. A waveform stability analysis algorithm is used to extract the physiological fibrillation window, and the duration, amplitude fluctuation range, and center frequency changes of the window are comprehensively assessed. If prolonged window time, spectral drift, or signal instability occur, it can be inferred that the newborn's respiratory regulation system is in a pre-instability state, i.e., a pre-instability stage.

[0033] Further, it includes the collection and analysis of neural and respiratory feedback pathway data to identify signal transmission lag and feedback decoupling phenomena in the neonatal respiratory regulation system. Specifically, a multimodal physiological signal synchronous acquisition device is used to perform non-invasive real-time monitoring of newborns. This includes signal sources such as electroencephalogram (EEG), sEMG (splenic electromyography), airflow curves, and chest movement images (acquired through optical or infrared imaging). Under controllable conditions, micro-mechanical stimulation, gas flow rate changes, or thermal stimulation are emitted as input sources for simulated external respiratory control commands. The entire process from the start of stimulation to the occurrence of respiratory action in the newborn is recorded, and the response delay in this process is analyzed. That is, the time interval between the start of stimulation input or central nervous system evoked command (determined by a specific waveband pattern in EEG) and the participation of peripheral muscle groups in respiratory action (determined by the initial rising edge of the sEMG signal and the initial change point of the respiratory airflow waveform), denoted as ΔT, can be used as an indicator of the output efficiency of the central nervous system in response to respiratory behavior. If the time difference ΔT shows a significant prolongation trend or unstable fluctuations in multiple stimulation cycles, it can be inferred that there is a delay in the issuance of central commands or a conduction disorder in the neural pathway information transmission process. Furthermore, by analyzing the temporal correlation and phase difference stability between signals from different channels, the lag in vagal nerve regulation of respiratory feedback was identified. Specifically, this included a decrease in the low-frequency coupling strength between heart rate variability (HRV) and respiratory phase, or a decrease in the synchronicity between vagal nerve-related EEG regions (such as the middle frontal cortex) and respiratory rhythm, thus indicating that there is a functional delay or insufficient coordination in this neural feedback loop. Simultaneously, by combining the three-stage feedback pathway of respiratory receptors—central nervous system—peripheral muscle groups, a closed-loop model of perception-integration-execution is constructed. Response time, signal intensity, phase coupling, and transmission path are analyzed at any stage to identify decoupling phenomena in the regulatory chain. For example, the respiratory receptors may respond to an increase in CO2 concentration with a delay, but the central nervous system may not issue an adjustment command, or the central nervous system may issue a command, but the peripheral muscles may not produce an effective respiratory action in time. When such conduction pathways exhibit decoupling or unstable coupling at any of the three stages, the system classifies it as a regulatory pathway abnormality and records it in the risk assessment parameter set as a neurofeedback regulation risk factor. This factor is then used to fuse and judge subsequent data such as respiratory rhythm disturbance response and energy metabolism response. This constructs a complete neonatal respiratory feedback system assessment module centered on time series lag and neural pathway integrity, enabling dynamic modeling of neonatal neuro-respiratory feedback capabilities and prediction and identification of early dysfunctional states.

[0034] Further, this includes continuous monitoring and analysis of the neonatal respiratory system's micro-adaptive recovery and redundancy regulation capabilities. This process is based on continuous evaluation of the physiological responses induced by multiple non-invasive micro-perturbation operations and the collected feedback data. Specifically, multiple rounds of low-amplitude, low-frequency controllable perturbations are applied to the neonate using a fixed time interval or incrementally increasing variables, forming one or more intervention stimulation cycles. After each perturbation, key parameters such as the neonate's respiratory rate, inspiratory-to-expiratory ratio, chest and abdominal movement waveforms, and muscle activation status are recorded in real time. By comparing these perturbation response processes horizontally and longitudinally, a dynamic recovery characteristic trajectory of the individual neonate is constructed. The focus is on assessing whether there is a progressively prolonged trend in the recovery time of respiratory rate after perturbation. For example, the recovery time is shorter after the first round of stimulation and gradually increases in subsequent rounds. If this trend persists in multiple rounds of testing, it indicates that the neonate's respiratory regulation system exhibits adaptive decline under repeated stress in a short period of time. In addition, by combining respiratory drive-related parameters such as respiratory depth, respiratory muscle power consumption level per unit time, oxygen consumption and carbon dioxide emissions, we can assess whether respiratory power consumption has gradually increased, that is, the system needs to mobilize higher energy consumption to maintain the original ventilation efficiency, thus revealing that there is an accumulation of intrinsic metabolic and regulatory load before the functional performance has obviously declined. This phenomenon is defined as latent respiratory fatigue. Meanwhile, to comprehensively assess the neonatal respiratory redundancy regulation capacity during stress control, the activation status of auxiliary breathing modes was continuously tracked, including the activation frequency, intensity, and onset time of accessory muscle groups such as the sternocleidomastoid, scalene, and rectus abdominis muscles. The analysis was conducted to determine whether premature intervention occurred before the primary respiratory muscles failed. If accessory muscle groups were repeatedly activated under low-intensity disturbances or the intervention time gradually advanced, it indicated a decline in system regulation capacity and a trend of respiratory load distribution shifting to redundant pathways. Furthermore, by combining the results of multiple tests, the changes in the trigger threshold and activation timing of the auxiliary regulation mechanism were statistically analyzed to identify the compression process of redundant regulation capacity. That is, muscle regulation pathways that could originally be used as backups were prematurely mobilized and overused during frequent interventions, causing critical dissipation of regulatory resources. In severe cases, this could lead to regulatory collapse of the system without backup pathways. Through a systematic analysis of the aforementioned hidden fatigue and redundant compression characteristics, a comprehensive assessment of the dynamic tolerance and risk determination of the neonatal respiratory regulation system is formed. A comprehensive indicator system with recovery capacity, energy burden, and regulatory resource status as its core is constructed to achieve early identification of respiratory distress trends and individualized risk classification, providing preliminary data support and decision-making reference for medical intervention.

[0035] Combined with appendix Figure 2The micro-perturbation operation does not employ a fixed intensity or preset mode, but rather dynamically adjusts the perturbation amplitude based on the newborn's real-time spontaneous breathing state, thereby achieving closed-loop coupling control between the perturbation input and the physiological state. Specifically, the system continuously monitors parameters such as the newborn's current respiratory rate, respiratory waveform morphology, inspiratory-to-expiratory ratio, and chest-abdominal coordination, generating real-time instantaneous respiratory rhythm characteristic values. These values ​​are then used as input variables for the perturbation control algorithm, ensuring that the intensity, duration, frequency, and form of each micro-perturbation are dynamically matched with the newborn's current respiratory rhythm, thus achieving individualized adaptation and minimal intervention optimization of the perturbation input. This approach significantly improves the efficiency of perturbation-induced response to the respiratory control system's response limit while avoiding overstimulation of the immature newborn's physiological system. During the micro-perturbation process, the system simultaneously initiates phase shift monitoring of the respiratory rhythm. Phase drift detection is not solely based on linear time-series analysis but incorporates phase space reconstruction technology to model the topological trajectory of the respiratory rhythm signal. Specifically, by delaying and embedding the original respiratory waveform signal, its dynamic trajectory in a high-dimensional phase space is reconstructed, and the geometric changes of this trajectory before and after the disturbance are used as the basis for stability judgment. If the original closed or quasi-periodic trajectory structure deforms, diverges, or forms an asymmetric shape under the influence of the disturbance, it is determined that the rhythmic stability of the respiratory regulation system has degraded due to the disturbance. Further, indicators such as trajectory center offset, orbit density change, and stable point drift rate are extracted to quantify the degree of stability degradation.

[0036] To address two potential imbalance mechanisms—weakened low-frequency regulatory capacity and decreased neuromuscular feedback stability—a dual detection strategy was designed, employing both low-frequency inhibition analysis and physiological fibrillation window change detection, to comprehensively identify the boundary state of neonatal respiratory regulation. The low-frequency inhibition analysis goes beyond traditional frequency domain analysis methods; instead, it constructs a multi-scale entropy spectrum of respiratory drive signals to assess the information complexity of the signals at different time scales, thereby quantifying the dynamic regulatory capacity that the respiratory regulation system can maintain. First, neonatal respiratory drive signals (such as the rate of change of thoracic volume or respiratory flow rate) are segmented across multiple time scales, and sample entropy values ​​or approximate entropy values ​​are calculated at each scale, forming an entropy spectrum spanning short-, medium-, and long-term scales. By observing the distribution trend of entropy values ​​with scale changes, the system identifies whether there is a significant decrease in entropy values ​​at large scales. This decrease represents a weakening of the respiratory regulation system's fluctuation capacity at the low-frequency level, i.e., a loss of complexity, reflecting a decline in its adaptability to slow regulatory tasks (such as spontaneous rhythm repair or energy load response), thus constituting the basis for identifying early respiratory regulation degradation. Meanwhile, to capture the boundary stability state of the respiratory regulation system, a mechanism for detecting changes in the physiological fibrillation window was further designed. This mechanism detects low-amplitude, high-frequency fibrillation signals generated by subcutaneous muscles in each respiratory cycle in real time using a high-sensitivity micro-vibration detection unit. This reflects the dynamic process of muscle contraction and relaxation under central nervous system control. The acquired fibrillation signals are processed using time-frequency decomposition techniques such as wavelet transform and short-time Fourier analysis to extract multidimensional features such as spectral distribution, time delay, and wave packet energy concentration of the muscle group fibrillation signals. These features are then compared with neural control signals (such as EEG synchronization events) over time to identify abnormalities such as phase drift, shortened synchronization window, or activation delay between neural firing and muscle response. If the synergy of neuromuscular activation significantly decreases, or if non-physiological frequency components appear in the fibrillation signal, it can be determined that the respiratory regulation system is approaching its stability boundary and is in a critical regulation state.

[0037] The acquisition of neural and respiratory feedback pathways goes beyond simply recording single neural electrical activity or respiratory muscle movement data. Instead, it involves constructing a brain-respiratory coupling monitoring mechanism to collaboratively acquire and analyze the signal transmission pathways between the issuance of central nervous system commands and the execution of peripheral respiratory actions. This allows for the assessment of the consistency and coordination of the neuromodulation system during respiratory maintenance. High-resolution cortical EEG acquisition devices are used to record real-time EEG data in respiratory control-related areas such as the midfrontal cortex and precentral gyrus in newborns. Simultaneously, surface electromyography (EMG) signals of respiratory-related muscle groups, chest and abdominal acceleration signals, and respiratory waveform signals from airflow sensors are acquired. Alignment analysis is performed on characteristic wavebands related to respiratory drive (e.g., theta bands or high gamma rhythms) in EEG activity with the onset time, rate of change, and amplitude response of respiratory action potentials along a unified time axis. By calculating the phase synchronization, covariance variation trend, and signal propagation delay between the two, it is determined whether there are consistency fluctuations or potential mismatches between central command issuance and peripheral execution. Particular attention is paid to abnormal manifestations such as decreased signal coupling strength, prolonged response time, or multi-cycle out-of-sync, which indicate potential impairment of the integrity of the neural modulation pathway. Building upon this foundation, to further identify potential decoupling states in the neural and respiratory feedback pathways, a nonlinear neural regulation pattern identification mechanism was introduced. A hysteresis loop atlas based on the relationship between neural signals and respiratory responses was constructed. By recording the nonlinear mapping relationship between the amplitude of neural signal output and the amplitude of respiratory responses in multiple stimulus-response cycles, an input-output loop trajectory was formed. Key parameters such as the envelope area, opening direction, and reverse curvature of the trajectory were extracted and analyzed to determine whether the system exhibited hysteresis amplification, i.e., a small change in the neural input signal but a sudden increase in the respiratory response amplitude, or conversely, a strong neural output but a delayed or incomplete respiratory action. The formation and amplification of this hysteresis loop indicates an asymmetric feedback between the input and output of the neural regulation chain, reflecting a nonlinear impedance or an open feedback loop in the information transmission of the regulation path. When the hysteresis loop parameter exceeds a set threshold, it is determined to be a decoupling state of the neural-respiratory regulation chain. By combining EEG and EMG analysis with the construction of a nonlinear hysteresis model, we can not only dynamically identify changes in the consistency of signal transmission in the regulatory pathway, but also capture potential asymmetric functional disorders that cannot be revealed by conventional linear models. This significantly improves the accuracy of identifying abnormal states of the neonatal respiratory regulation system and the earlyness of risk assessment.

[0038] To address the changes in the tolerance of the neonatal respiratory regulation system during repetitive perturbations, a continuous monitoring module for micro-adaptive recovery was established. This module records the dynamic correspondence between the neonatal respiratory system's recovery time and perturbation intensity after each perturbation cycle across multiple perturbation periods, constructing a functional relationship graph of recovery time versus perturbation intensity. An iterative fitting method is used to nonlinearly model this function. After each micro-perturbation, the system continuously records the time required for the neonate to recover to the baseline respiratory rhythm and pairs it with the actual intensity of the perturbation (including amplitude, frequency, or stimulation time). This forms a set of discrete data points under multiple perturbation conditions. Dynamic regression or nonlinear function fitting methods are used to model the trends of these points, monitoring the slope, curvature, fitting residuals, and trends of the fitted function. When the function gradually transforms from an early linear relationship to an exponentially increasing, saturated, smooth, or polynomial divergent form, it indicates a significant decrease in the system's recovery ability after stress, signifying a transition from the linear regulation region to the fatigue-instability critical region. This transition point is identified by the system as a risk of regulatory imbalance. Warning signs; In addition, to further identify the latent fatigue state of the respiratory regulation system without obvious regulatory failure, a respiratory pattern migration monitoring module was set up. Throughout the assessment process, respiratory dynamic signals (such as airflow waveform, chest and abdominal movement amplitude, and muscle group coordination pattern) were sampled at high frequency and real-time features were extracted. Combined with statistical classification algorithms, respiratory patterns were classified and labeled, and the neonatal respiratory process was divided into different dynamic pattern subclasses (such as uniform, overcompensated, intermittent, shallow and fast, etc.). By monitoring the frequency of abrupt changes, migration direction, and migration duration between patterns, it was determined whether there was autonomous pattern drift under non-task-driven conditions, especially the natural evolution process from efficient regulation patterns to compensatory or chaotic patterns. This phenomenon is regarded as a low-order regulatory strategy switch of the system in response to continuous disturbances. It often begins to occur when physiological indicators are still within the normal range. Therefore, it can serve as an important signal for identifying the respiratory regulation system entering a latent fatigue state. Combined with the nonlinear change trend of the aforementioned recovery function, it constitutes a continuous risk assessment framework for the imbalance of neonatal respiratory regulation stability.

[0039] Combined with appendix Figure 3To achieve highly sensitive and dynamic identification of potential imbalances in the neonatal respiratory regulation system, three key modules were established: assistive respiratory muscle intervention assessment, redundant regulation resource dissipation analysis, and energy metabolism coupling risk calculation. Assistive respiratory muscle intervention assessment dynamically captures the activation intensity M(t) and activation time of assistive respiratory muscles (such as the sternocleidomastoid and scalene muscles) through combined ultrasound imaging and electromyography signal analysis. Based on their temporal evolution trends, an individualized muscle activation threshold curve is constructed. Changes in this threshold reflect whether the spontaneous respiratory regulation capacity has declined and whether a respiratory state transition dependent on compensatory mechanisms has occurred. If a decrease in the activation threshold and frequent premature activation of assistive muscles are detected, the system initially determines that there is a hidden weakening tendency in the spontaneous control and regulation capacity. Furthermore, the system combines the number of micro-perturbations N and the respiratory system response intensity R(t) after each perturbation to construct a dissipation accumulation function for redundant respiratory regulation resources. It also statistically analyzes the changes in resource mobilization frequency and intensity to determine whether the regulation resources are approaching exhaustion boundaries, thereby identifying whether the system's redundancy capacity is rapidly being compressed. Based on the above two types of data, the system introduces the energy dimension as a comprehensive risk assessment standard. Utilizing the respiratory load change L(t) after each micro-perturbation and the corresponding energy metabolism consumption curve E(t), the system dynamically calculates the metabolic response trend accompanying the respiratory system's regulatory actions, establishing a composite assessment model of regulatory dissipation and metabolic consumption. Finally, the following interactive dissipation critical risk model is constructed in functional form:

[0040]

[0041] Where Ψ(t) represents the cumulative function of neonatal respiratory regulation-metabolic coupling risk, used to comprehensively assess the intensity of the nonlinear interaction between regulatory dissipation and energy load in the current system; M(t) is the dynamic activation threshold function of the accessory respiratory muscles. Its rate of change characterizes the load regulation change of the muscle compensation system; R(t) is the system response intensity triggered by each disturbance, and N is the number of disturbances in the current period. Characterizes the dissipation characteristics of redundant regulatory resources with diminishing marginal returns during continuous deployment; L(t) represents the change in respiratory load. The rate of energy consumption. This function characterizes the changing response of nonlinear interactions during load regulation, and is used to detect whether the energy metabolism system becomes overactivated with load regulation. When the cumulative risk function Ψ(t) exceeds a set threshold Θ, the system is determined to have entered a critical state of respiratory regulation imbalance, indicating that newborns at high risk are about to experience irreversible respiratory function compensatory collapse.

[0042] To more accurately identify the compensatory capacity boundary and autonomic regulation exhaustion trend of the neonatal respiratory regulation system, a dual dynamic assessment mechanism was established, centered on neuromuscular synergistic dissipation analysis and nonlinear hysteresis recovery identification. During the tracking of the activation threshold of the accessory respiratory muscle groups, not only were electromyographic signals and muscle activation intensity M(t) of the accessory muscle groups collected, but the intensity of the motor output signal of the central nervous system was also simultaneously collected, particularly cortical electroencephalogram activity and brainstem output potentials related to respiratory regulation. Combined with time-axis unified calibration technology, neural output and muscle response signals were paired one-to-one. Through time-series analysis, phase matching, and signal intensity normalization, a synchronous dissipation curve between neural regulation resource consumption and muscle redundancy was established, reflecting the synergistic relationship between the intensity of continuous central commands to the respiratory muscle groups and the actual response capacity of the muscle groups. When the curve showed significant coupling shift, asynchronous dissipation, enhanced neural signals, and delayed or weakened muscle responses during the continuous perturbation process, it was determined to be a neuromuscular synergistic exhaustion trend. This phenomenon indicates that the autonomic regulation network is operating at high load and the accessory muscle mobilization mechanism is approaching its physiological limits. Building upon this foundation, to further determine whether the respiratory regulation system has transitioned from a recoverable state to the boundary of functional instability, a perturbation-recovery nonlinear hysteresis modeling method was introduced. Recovery paths were collected after each intervention in continuous micro-perturbation experiments, and multi-period comparative analyses were performed on response indicators such as perturbation intensity, recovery time, and recovery integrity. If the increase in system recovery time exceeded the linear growth trend of perturbation intensity, or if the response sensitivity exhibited fluctuations, discontinuities, or significant hysteresis under similar perturbation amplitudes, it indicated that the respiratory regulation system no longer exhibited its original high elasticity of recovery and had entered a fragile recovery stage, showing a clear approach to the functional critical boundary. These two mechanisms, by integrating neural output signals, electromyographic responses, perturbation-recovery functions, and nonlinear hysteresis indicators, dynamically identified the neuromuscular dissipation synchronicity of the neonatal respiratory regulation system and the process nodes of its transition from stable recovery to irreversible fatigue. This led to the construction of a multidimensional risk assessment module encompassing functional resource assessment, critical state identification, and predictive feedback.

[0043] To accurately predict and identify potential functional imbalances in the neonatal respiratory regulation system, a dual-channel coupled analysis mechanism based on redundant resources and energy metabolism was designed. Furthermore, a muscle response time window identification model was established to capture the precursory features of the system's critical collapse. Firstly, during the risk assessment process, the system continuously collects and simultaneously analyzes two types of core dissipation indicators. The first is the regulatory redundancy resource dissipation curve, constructed based on the dynamic changes in the activation threshold of accessory respiratory muscles, the frequency of accessory muscle participation, and the intervention time. This curve reflects whether the regulatory resources mobilized by the respiratory regulation system under repeated perturbations exhibit rapid dissipation, premature activation, or delayed recovery. The second is the energy metabolism redundancy dissipation curve, constructed from parameters such as metabolic energy consumption level during respiration, unit ventilatory metabolic load, oxygen consumption-carbon dioxide emission ratio, and respiratory-driven metabolic efficiency. This curve is used to assess the changing trend of metabolic expenditure required by the neonate to maintain current respiratory efficiency. When the energy consumption rate increases significantly or metabolic efficiency decreases, it indicates that the metabolic system is approaching the regulatory redundancy boundary. The system analyzes the time alignment and synergistic changes of the two curves to construct a dual-channel imbalance judgment model of regulation and metabolism. If a dynamic mismatch is identified between the two, that is, the regulatory resources continue to decline while the energy supply cannot effectively support the stable operation of respiratory control, such as the occurrence of regulatory dissipation before metabolic abnormality, or metabolic collapse before the regulatory resources bottom out, it indicates that the system has an imbalance risk trend, and this serves as the basis for indicators that the newborn's respiratory function is approaching the critical point. Furthermore, in monitoring the activation process of assistive respiratory muscles, the system also performs high-precision acquisition and analysis of the time delay between changes in activation thresholds and actual muscle movements. By comparing the timing of EEG signals or central firing potentials with peripheral electromyographic responses, the system calculates the micro-delay time window during the muscle compensation initiation process. This risk window reflects the change in transmission efficiency between central regulatory commands and actual respiratory actions. Once the delay time continues to expand or exhibits unstable fluctuations, it indicates signal transmission blockage or slow response in the neuromuscular execution chain. Especially when this delay occurs synchronously with frequent intervention of assistive muscles and fluctuations in the energy redundancy dissipation curve, it can be determined that the newborn is in a high-risk critical time interval where the spontaneous breathing system is on the verge of failure, thus providing a clear time window for early intervention, ventilation mode adjustment, or life support strategies. Through the coordinated operation of the dual-channel imbalance judgment and micro-delay response time window identification mechanism, the system can achieve dynamic monitoring and risk classification of the entire process of the evolution of the newborn's respiratory regulation system from compensation to collapse, significantly improving the sensitivity and foresight of respiratory distress early warning.

[0044] To identify the potential coupling mechanism between changes in local muscle group function and system-level regulatory imbalance in the neonatal respiratory regulation system, an analysis module combining muscle activation threshold monitoring and local micro-fatigue integration was established. Specifically, the activation thresholds of key accessory respiratory muscles (including scalene muscles, sternocleidomastoid muscles, rectus abdominis muscles, etc.) are dynamically and continuously tracked, and micro-fatigue response signals generated by these muscle groups during repetitive respiratory disturbances are acquired simultaneously. The system first quantifies the fatigue performance of local muscles in each activation cycle in real time based on the time-domain amplitude, frequency-domain power spectral density, and median frequency changes of electromyography (sEMG) signals. Then, a muscle micro-fatigue integral model is constructed by cross-cycle integration. This model forms a cumulative trend graph with time as the horizontal axis and fatigue amplitude as the vertical axis, which is used to detect whether the muscle groups are maintaining short-term respiratory compensation function. During the process, the fatigue gradually depletes without being reflected in the macroscopic respiratory manifestations in a timely manner. At the same time, the system performs correlation analysis between the above-mentioned local fatigue integral results and global respiratory regulation indicators (such as respiratory rate variability, drive delay, and auxiliary muscle group synergy efficiency). Through multivariate regression and dynamic correlation calculation, it identifies whether the local muscle fatigue integral will induce regulatory pattern disorder or rhythm collapse at the system-wide level after reaching a certain threshold. If a high coupling is found between the two, for example, after a high-amplitude integral of a certain muscle group, there will be corresponding changes in global indicators such as respiratory rhythm mutation, abnormal mobilization of regulatory resources, or decreased gas exchange efficiency, the system regards the local muscle fatigue accumulation process as a pre-trigger factor for global functional imbalance, further judges it as an early risk signal, and records it as an early warning event of system-level regulatory failure caused by local load collapse.

[0045] To achieve early warning and dynamic hierarchical management of neonatal respiratory regulation system imbalance, a two-layer assessment mechanism based on the identification of critical energy metabolism behaviors and dynamic reconstruction of individual respiratory fitness was established. During the construction of energy metabolism risk indicators, the system monitors neonatal metabolic response parameters in real time under conditions of micro-disturbance or natural respiratory load fluctuations, including oxygen consumption per unit time (VO2), carbon dioxide emissions (VCO2), respiratory quotient (RQ), body surface temperature distribution, and blood lactate levels. By analyzing the fluctuation trends of these indicators over time, the system identifies whether there is a sharp increase in the metabolic cost required to maintain the same respiratory efficiency. Once a metabolic critical point behavior is detected—that is, a sudden increase in energy consumption or a significant decrease in metabolic efficiency without a significant change in respiratory drive—the system takes this change as a precursor to a shift in respiratory regulation capacity. It further tracks subsequent regulatory behavior and, if autoregulatory characteristics (such as rhythm stability, delayed regulatory response, and muscle group participation in rhythm) are found, the system will take further action. If respiratory function gradually degenerates and is replaced by auxiliary mechanisms (such as premature activation of auxiliary respiratory muscles, uncontrolled respiratory rhythm imbalance, or regulatory signals no longer relying on autonomic nerve commands), the system determines that it has transitioned from the active regulation stage to the passive compensation stage. This critical transition point serves as an important dividing line for risk grading, providing a reference for setting risk levels for subsequent intervention strategies. Simultaneously, during the assessment of redundant regulatory resources, the system introduces historical perturbation response data to construct an individualized dynamic reconstruction model of respiratory regulation fitness. Specifically, the system performs multidimensional encoding on the regulatory performance and recovery path of newborns under different perturbation conditions within the historical monitoring period, forming an individual regulatory response spectrum. The system then compares and analyzes the regulatory response characteristics at the current time point with historical characteristics to identify whether there are structural changes in their resource utilization patterns, such as premature activation of auxiliary regulatory pathways, weakened activation of the original primary regulatory mechanism, and prolonged recovery time. Based on this, the system continuously corrects the availability boundary of individual respiratory regulation resources and updates the evolution trajectory of their fitness status in the resource space.

[0046] Example 1:

[0047] Combined with appendix Figure 4 In this embodiment, a premature newborn, Lin, born at 31 weeks gestation and weighing 1580 grams, presented with intermittent irregular breathing and mild cyanosis within 48 hours of birth. Initial assessment indicated that mechanical ventilation was not yet indicated, but the medical team was highly alert to the potential instability of his respiratory regulation. The micro-perturbation procedure was initiated while the newborn was at rest. The system used an embedded flexible thermal stimulation device to perform periodic thermal perturbations on a localized area of ​​Lin's chest, with an initial perturbation parameter of 0.15 W / cm². 2With a cycle of 5 seconds, the system dynamically calculates Lin's current respiratory rhythm variability threshold based on real-time collected spontaneous breathing signals, including the chest-abdomen synchronized airflow curve (frequency of 48 breaths / minute, inspiratory-to-expiratory ratio of 1:1.3) and the respiratory rhythm amplitude stability index (initial mean of 0.92), and adjusts the perturbation intensity to 0.12 W / cm². 2 To match its current breathing state and form a closed-loop control coupling relationship, avoid excessive disturbance intensity that could lead to interruption of spontaneous breathing, while ensuring that its regulatory system response is stimulated.

[0048] In the respiratory waveform data collected 3 minutes before the first round of perturbation, the original signal was reconstructed using a phase space reconstruction algorithm with a delay embedding dimension of 5 and an embedding delay τ of 80ms. The resulting respiratory regulation phase trajectory before the perturbation showed a stable, quasi-elliptical closed trajectory with a trajectory center clustering degree of 0.85 and a Lyapunov exponent of 0.03, indicating that the system was in a stable and predictable state. Two minutes after the perturbation began, the phase trajectory exhibited morphological perturbation, with decreased closure and a trajectory shape changing to a skewed loop. The trajectory center clustering degree decreased to 0.64, and the Lyapunov exponent increased to 0.18, indicating a mild degradation of rhythm stability in real time. In the subsequent three rounds of micro-perturbation operations, the system dynamically adjusted the perturbation intensity based on Lin's respiratory rhythm stability score after each round of recovery, decreasing it to 0.09 W / cm². 2 0.07W / cm 2 To accommodate its gradually increasing sensitivity to perturbations, the phase trajectory showed significant asymmetric stretching before and after the fourth round of perturbation, with a clustering degree of only 0.41 and a Lyapunov index rising to 0.32. The system determined that Lin's respiratory regulation system was approaching a pre-instability state. Since the intervention system's threshold was set to issue an early warning when the clustering degree was <0.5 and the Lyapunov index was >0.3, the system triggered an early respiratory distress warning mechanism and prompted for enhanced clinical monitoring and, if necessary, the introduction of non-invasive ventilation assistance.

[0049] After detecting a degradation in the phase trajectory stability of Lin's respiratory regulation system and triggering the first round of warnings, the medical team immediately initiated a more in-depth assessment module. This involved sequentially performing low-frequency suppression analysis, physiological fibrillation window detection, synergistic analysis of neural and respiratory feedback pathways, hysteresis feedback identification, and monitoring of micro-adaptive recovery ability and hidden fatigue. This multi-dimensional assessment of Lin's respiratory regulation status and quantification of its risk trends was conducted. First, in the low-frequency suppression analysis, Lin's respiratory drive signals were collected for five consecutive minutes. Sample entropy was calculated using multi-scale windows of 0.2s, 0.5s, 1s, 2s, and 5s to construct a multi-scale entropy spectrum. The results showed that the entropy value was 0.84 at the small scale (0.2s) and decreased to 0.41 at the large scale (5s), exhibiting a typical trend of multi-scale complexity loss. The system determined that Lin's low-frequency fluctuation ability was limited, the regulatory system's adaptability to slowly varying loads was weakened, and there was a risk of regulatory mechanism subsidence.

[0050] Following physiological fibrillation window monitoring, nurses placed array-type micromechanical vibration sensors in the sternocleidomastoid and rectus abdominis muscle regions of the patient, respectively, to perform time-frequency decomposition of the fibrillation micro-vibration signals during respiration. Short-time Fourier transform analysis of the muscle vibration signals revealed that the spectrum shifted from being concentrated in the 40–60 Hz range to 20–35 Hz, and the spectral energy distribution became discrete and asymmetrical. Wavelet packet reconstruction confirmed that the phase difference between the neural activation frequency and the muscle response signal increased from 10 ms to 28 ms, indicating a significant decrease in synchronicity. This suggested weakened neuromuscular coupling efficiency, indicating that the respiratory regulation system was approaching its boundary state, requiring enhanced monitoring.

[0051] In the synergistic analysis of neural and respiratory feedback pathways, electroencephalography (EEG) and electromyography (EMG) signals from Lin's prefrontal cortex were collected to identify the temporal synchronicity between the theta wave band (4–8 Hz) in EEG and EMG signals. Based on cointegration analysis and Granger causality discrimination, it was found that the causal directionality between EEG activity and muscle response decreased significantly from "brain → muscle," and the cross-signal delay increased from 20 ms to 50 ms, reflecting a significant weakening of the transmission consistency of central regulatory signals in peripheral execution. Further, a hysteresis loop atlas of central command intensity and muscle response amplitude was constructed. The results showed that the hysteresis loop envelope area increased from the initial 0.08 to 0.24, exhibiting a significant amplification effect, and the activation pathway showed asymmetric amplification, indicating a trend of command-response decoupling and suggesting a nonlinear mismatch in neural regulation patterns.

[0052] To verify the system's regulatory flexibility, micro-perturbation interventions were implemented at hours 2, 4, and 6. By recording the correlation between perturbation intensity and recovery time, iterative function fitting was performed. It was found that in the first two rounds of intervention, the perturbation intensity increased from 0.1 W / cm². 2 Up to 0.15W / cm 2The recovery time increased from 12 seconds to 16 seconds, and the fitted function showed a linear relationship. However, at the 6th hour, under the same perturbation intensity, the recovery time increased to 27 seconds, the fitted residuals widened significantly, and the function shifted to a second-order polynomial trend. The system determined that the control system was transitioning from the linear recovery phase to the fatigue instability zone. Simultaneous recording of respiratory pattern changes revealed that the original respiratory rhythm was predominantly uniform, transitioning to a shallow and rapid rhythm during intervals, and then briefly transitioning to an intermittent rhythm. The system identified this "uniform → shallow and rapid → intermittent" path as a typical micro-migration path from compensation to fatigue evolution. Combined with the perturbation recovery trend, it was ultimately determined that Lin was in the process of hidden fatigue evolution.

[0053] Based on the analysis data from the above modules, the system output risk curve shows that Lin has exceeded two key indicators to reach the warning threshold (low-frequency entropy drop > 40%, hysteresis amplification > 0.2, and recovery nonlinear residual > 2 times the baseline). It is determined that his spontaneous breathing regulation network is in a functional boundary state. Although there is no obvious apnea or hypercapnia at present, based on the synergistic deterioration trend of the structure-function-behavioral three-level indicators, it is recommended to prepare a non-invasive assisted ventilation plan in advance and reduce external environmental stress stimuli.

[0054] Example 2:

[0055] Combined with appendix Figure 5 After detecting a decline in rhythmic stability, decoupling of neural feedback, and a trend of low-frequency inhibition in the patient's respiratory regulation system, the medical team further implemented a complete pathway involving accessory respiratory muscle intervention—redundant regulation compression analysis—metabolic risk coupling assessment. This was done to construct a multi-source fusion high-dimensional risk identification model and accurately locate the critical point of the regulatory system. At this time, Lin was already on the 3rd day after birth. Although his blood oxygen saturation was still stable at 91–94%, the nurses noticed an increase in the frequency of accessory muscle activity, indicating that he had entered the compensatory activation stage.

[0056] First, high-frequency ultrasound imaging was used to quantify the deformation rates of Lin's sternocleidomastoid (SCM) and scalene muscles during rest and inspiration. Continuous contraction waveforms were observed in the early inspiratory phase. Combined with electromyography (EMG) signals, the activation intensity function M(t) of this muscle group showed an abnormally premature increase. The activation threshold was defined as the time point t corresponding to the first time the EMG amplitude exceeded twice the baseline average. act The peak values ​​of M(t) during the past five breaths were recorded as 0.42 mV, 0.46 mV, 0.50 mV, 0.54 mV, and 0.60 mV, respectively. The corresponding activation points advanced sequentially, and the activation threshold curve showed a steep downward trend. The rate of change was calculated as follows:

[0057]

[0058] The system has determined that the patient's ability to regulate their own breathing has declined, and the assisted breathing mechanism is continuously taking over the main control function.

[0059] Next, based on the number of micro-perturbations recorded over the past 6 hours (N=12), the respiratory response intensity (expressed as normalized lung volume fluctuation amplitude) after each intervention was as follows:

[0060] R(t)={0.22,0.26,0.29,0.30,0.35,0.36,0.40,0.42,0.45,0.49,0.51,0.56}

[0061] Construct redundant resource dissipation functions:

[0062]

[0063] Given N=12, calculate three typical terms:

[0064] First disturbance:

[0065] 6th perturbation:

[0066] 12th perturbation:

[0067] The cumulative dissipation is gradually increasing, indicating that redundant control resources are being continuously used and dissipation efficiency is decreasing.

[0068] The third step involves further importing Lin's respiratory load change L(t) and real-time energy metabolism curve E(t) into the system, and collecting the oxygen consumption (normalized) required for each inhalation as follows:

[0069] L(t)={0.40,0.44,0.48,0.51,0.54,0.58},

[0070] E(t)={0.30,0.34,0.38,0.43,0.49,0.56},

[0071] And perform derivative operations on them respectively:

[0072]

[0073] For a certain disturbance: but

[0074]

[0075] Therefore, the system calculates the energy channel dissipation term at this point in time as follows:

[0076]

[0077] Combining the aforementioned muscle group derivatives and dissipation accumulation, the cumulative risk function of neonatal respiratory regulation-metabolism coupling is finally obtained dynamically:

[0078]

[0079] After integrating the results from sampling period T = 6 times, we obtain the following:

[0080]

[0081] The system's preset coupling risk threshold is Θ = 0.150, thus triggering an early warning of regulatory-metabolic imbalance. It was determined that Mr. Lin had entered a pre-critical state of respiratory regulation exhaustion, and it was recommended that non-invasive assisted ventilation be performed in the short term to alleviate the regulatory load.

[0082] On the fourth day after the newborn Lin entered a period of fluctuating vital signs, the medical team, having previously completed rhythmic phase drift analysis, low-frequency modulation complexity assessment, and respiratory muscle synergy trend monitoring, officially launched the deep multidimensional coupling monitoring phase. This phase involved a comprehensive risk reconstruction of pathways related to the activation of accessory respiratory muscles, neuromuscular resource matching, and energy metabolism-regulation synergy trends. First, during the tracking of accessory respiratory muscle activation thresholds, the system used a multi-channel high-resolution electromyography (EMG) recording device to acquire the dynamic activation amplitudes of Lin's main accessory muscle groups, including the scalene muscles, sternocleidomastoid muscles, and intercostal muscles, while simultaneously acquiring the neuromodulation command waveforms from the central prefrontal cortex. During continuous 8-minute monitoring, it was found that whenever Lin's respiratory load suddenly increased (manifested as a respiratory rate exceeding 60 breaths / min), the timing of the muscle activation response gradually advanced, and the EMG amplitude rise curve became steeper. However, the intensity of the neural commands did not increase significantly synchronously, reflecting that central regulatory resources tended to be fixed while muscle group mobilization was amplified. Based on this, the system constructed a synchronous dissipation trend line between neural output and muscle response, and found that the coupling index between the two decreased from the initial 0.88 to 0.61, indicating that Lin's autonomous control system was facing a state of coordinated exhaustion due to the weakening efficiency of the "instruction-execution chain".

[0083] Based on this, the system further analyzed the changing trend of its disturbance response recovery curve. Recording the recovery time and intensity data of all micro-disturbance responses within the first 24 hours, it was found that the early disturbance recovery of Lin's system showed a linear proportional relationship, meaning the stronger the disturbance, the slower the recovery, but it remained stable. However, in the later stages, a slight increase in disturbance intensity was accompanied by a significant increase in recovery time, and the curve exhibited typical nonlinear hysteresis, with a sudden change in recovery time (from approximately 18 seconds to 33 seconds). This indicated that Lin's system had transitioned from a relatively flexible control mechanism to a vulnerable state, prompting the system to issue a critical warning regarding the "control recovery boundary."

[0084] Simultaneously, the system loaded the energy metabolism module to analyze the correlation between the trend of respiratory regulation resource dissipation and the energy consumption curve. Records showed that after each intervention, Lin's energy metabolism curve initially showed a gradual upward trend, synchronized with the dissipation of regulatory resources; however, starting from the end of the third day, the metabolic curve continued to rise, while the regulatory resource response tended to stabilize or even slightly decrease. The system judged this as a mismatch state of "metabolic expenditure > regulatory benefit," indicating the entry into a functional compensation period. Once this trend is established, it will significantly increase the risk of respiratory system collapse. The system marked this point as a dual-channel imbalance point and upgraded Lin's current risk level to a level two warning state.

[0085] Meanwhile, to further improve the timing accuracy of early intervention, the system combines the aforementioned accessory muscle activation threshold trajectory and response delay data to identify significant initiation delays in muscle group activation. For example, in the scalene muscles, the average delay time from the onset of the nerve signal to the muscle response increased from 22ms to 44ms, indicating that the compensatory mechanism's response was beginning to lag. This micro-delay window is a time-sensitive interval before the respiratory system reaches critical overload; once the delay window continues to expand, it is highly likely to develop into apnea or severe regulatory failure. The system records the dynamic trend of the delay window to provide a basis for determining the intervention timing.

[0086] Furthermore, the system also integrates the trend of muscle micro-fatigue integrals. Analyzing the electromyographic mid-frequency changes and power spectrum drift of the scalene muscle group over 24 consecutive hours, the system identified a stepwise accumulation of its muscle fatigue index. In the fifth observation period, the system identified that the local fatigue integral reached 2.2 times the previous average. Although Lin's overall respiratory rate and blood oxygenation were still within the lower limit of normal, the fatigue index of this muscle group had exceeded the set threshold, indicating that local muscle regulation was approaching the edge of functional collapse. According to the present invention, this accumulation of local load is a precursor to the collapse of system-level regulation. Therefore, the system marks this risk source as a "potential global imbalance trigger point" and prioritizes its inclusion in the scope of artificial intervention monitoring.

[0087] Finally, the system retrieved all historical micro-perturbation data of Lin over the past 72 hours, analyzed the changes in his respiratory resource allocation path and recovery ability, and reconstructed an individualized regulatory fitness curve. Initially, Lin's regulatory recovery time after receiving the same perturbation was relatively stable, with a relatively flat curve. However, in the past 24 hours, the curve rose sharply and showed increased volatility, indicating that his respiratory regulatory resources were rapidly depleting and his recovery ability was becoming unstable. Based on this, the system continuously revised his individual fitness boundary, dynamically adjusting it from the initially set risk curve model, marking that he had now entered the rapid degradation zone of regulatory resources, and predicting that his spontaneous regulatory ability might become critically ineffective within the next 4–6 hours. Combining the above micro-fatigue integral, energy mismatch point, and delay window changes, the system ultimately calculated that the optimal intervention window for this child was within 30 minutes from the current time point, recommending clinical initiation of non-invasive positive pressure ventilation (nCPAP) and neuromuscular modulation monitoring to reduce the regulatory load.

[0088] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for constructing a neonatal respiratory distress risk assessment, characterized in that... include: Non-invasive micro-perturbation is performed on newborns to actively induce the extreme boundary response of the newborn respiratory regulation system, and the respiratory rhythm phase drift, low-frequency inhibition of respiratory drive, and physiological fibrillation window changes under the perturbation are detected to identify the pre-instability signal of respiratory system regulation. Data on neural and respiratory feedback pathways were collected to analyze the time difference between respiratory command transmission and actual respiratory initiation, and to identify vagal feedback lag and abnormal decoupling phenomena of respiratory receptors, central and peripheral regulatory circuits. Based on the results of multiple micro-perturbation inductions and feedback acquisitions, the micro-adaptive recovery ability of the neonatal respiratory system was continuously monitored. By judging the gradual prolongation of respiratory rate recovery time and the gradual increase of respiratory power consumption, the hidden fatigue trend of the respiratory system was identified. Monitoring respiratory redundancy regulation capacity during stress regulation, identifying premature intervention and repeated activation of assisted breathing modes or respiratory muscle groups, and analyzing the critical compression phenomenon of redundant regulation in the respiratory regulation system are all part of the assessment of neonatal respiratory distress risk.

2. The method for constructing a neonatal respiratory distress risk assessment according to claim 1, characterized in that... The micro-perturbation operation dynamically adjusts the perturbation amplitude according to the newborn's real-time spontaneous breathing state, so that the perturbation intensity is coupled in a closed loop with the newborn's instantaneous respiratory rhythm; the detection of phase drift includes analyzing the dynamic phase trajectory of the respiratory regulation system based on phase space reconstruction technology to identify the phenomenon of respiratory rhythm stability degradation induced by micro-perturbation.

3. The method for constructing a neonatal respiratory distress risk assessment according to claim 2, characterized in that... The analysis of low-frequency suppression includes assessing the trend of complexity loss in the neonatal respiratory regulation system by constructing a multi-scale entropy spectrum of respiratory drive signals; the detection of changes in the physiological tremor window includes collecting tremor micro-vibration signals of the respiratory muscle groups under the skin of newborns using a micromechanical sensor array, and identifying changes in the synchronicity of nerve and muscle activation through time-frequency decomposition to determine the boundary state of respiratory regulation.

4. The method for constructing a neonatal respiratory distress risk assessment according to claim 3, characterized in that... The acquisition of the neural and respiratory feedback pathways includes monitoring the coupling relationship between the neonatal brain and respiration, and identifying changes in the consistency of signal transmission between central command issuance and peripheral execution through the synergistic analysis of cortical electroencephalography and respiratory action potentials. The identification of feedback pathway decoupling includes detecting the nonlinear neural regulation patterns of neonates, and dynamically judging the asymmetric amplification phenomenon of the neural and respiratory regulation chains by constructing a hysteresis loop of command response.

5. The method for constructing a neonatal respiratory distress risk assessment according to claim 4, characterized in that... The continuous monitoring of micro-adaptive recovery ability includes iterative fitting of the recovery time and disturbance intensity function of the neonatal respiratory system to identify the transition point of the respiratory regulation system from the linear recovery stage to the fatigue instability stage; the identification of hidden fatigue includes monitoring minute respiratory pattern migration phenomena during breathing and revealing the non-obvious evolution process of system fatigue through the classification transformation of respiratory dynamics patterns.

6. The method for constructing a neonatal respiratory distress risk assessment according to claim 5, characterized in that... The assessment of muscle involvement in assisted breathing includes dynamically tracking changes in the activation threshold of accessory muscles through ultrasound imaging combined with electromyography signal analysis, identifying the decline in spontaneous regulation ability and the activation trend of compensatory mechanisms; the critical compression analysis of redundant regulation includes establishing the cumulative dissipation relationship of redundant regulation resources, and dynamically assessing whether the regulatory resources of neonatal respiratory regulation are close to the depletion boundary by combining the statistics of micro-perturbation counts. The risk assessment construction process includes full-process energy metabolism estimation based on micro-perturbation response, and the establishment of neonatal respiratory metabolism coupling risk indicators through the energy consumption curve corresponding to changes in respiratory load.

7. The method for constructing a neonatal respiratory distress risk assessment according to claim 6, characterized in that... During the tracking of muscle activation thresholds in assisted breathing, the intensity of muscle regulation commands output by the central nervous system is collected, and a synchronous dissipation curve of neural regulation resources and muscle redundancy is established to dynamically assess the risk of neuromuscular exhaustion of the autonomous regulation system. During the dissipation of redundant regulation resources, the transition stage of the respiratory regulation system from elastic recovery to fragile recovery is identified by constructing a nonlinear hysteresis relationship of disturbance recovery, thereby achieving boundary state judgment.

8. The method for constructing a neonatal respiratory distress risk assessment according to claim 7, characterized in that... The risk assessment construction process includes simultaneously analyzing the dissipation curves of redundant regulatory resources and redundant energy metabolism, establishing a dual-channel imbalance judgment relationship, and identifying the risk of respiratory imbalance caused by the mismatch between regulatory capacity and energy supply. Based on changes in activation threshold and muscle movement delay, a micro-delay risk window for muscle compensatory initiation is identified to predict the time-sensitive intervals in which the spontaneous respiratory system is on the verge of failure.

9. The method for constructing a neonatal respiratory distress risk assessment according to claim 8, characterized in that... The muscle group activation threshold monitoring, combined with the micro-fatigue integral of local muscle groups, detects the correlation between micro-fatigue accumulation patterns and systemic respiratory regulation failure, enabling early warning of global imbalance due to local load collapse.

10. The method for constructing a neonatal respiratory distress risk assessment according to claim 9, characterized in that... During the construction of the energy metabolism risk index, the autonomous regulation mode shift event triggered by the metabolic critical point is monitored to determine the critical transition of the system from active regulation to passive compensation, which serves as the basis for risk classification. The redundant regulation resource assessment, combined with historical micro-perturbation data, constructs an individualized dynamic reconstruction relationship of respiratory regulation adaptability, dynamically updates the respiratory resource utilization pattern of newborns, and enables the evolution of individual risk curves.

Images