Mechanically intelligent incubator-tight blue light irradiation data management system

Abstract

The present application belongs to the technical field of medical apparatus and instruments, and discloses a warm box closed blue light irradiation data management system based on mechanical intelligence, which comprises: a state cluster data module, which collects multi-source heterogeneous data, corrects the clock deviation between the warm box and the preset biological sensor through a space-time alignment protocol, dynamically adjusts the multi-source heterogeneous data sampling rate based on the infant physiological data in the multi-source heterogeneous data after time alignment, and outputs a state cluster data packet with a time stamp; a hierarchical extraction module, which constructs a three-level feature processing pipeline based on the state cluster data packet, including a basic layer that calculates basic statistics through a sliding window; a logic layer that generates a temperature-light effect index and a sealing safety coefficient according to the basic statistics and in combination with the multi-source heterogeneous data; and a decision layer that establishes a treatment efficiency prediction matrix based on the temperature-light effect index and the sealing safety coefficient, and predicts the associated trend of blue light irradiation and bilirubin metabolism; the safety and curative effect stability of infant treatment are improved.

Description

Technical Field

[0001] This invention relates to the field of medical device technology, and more specifically, to a data management system for sealed blue light irradiation in an incubator based on mechanical intelligence. Background Technology

[0002] The existing data management system for closed-loop blue light irradiation incubators has the following main problems:

[0003] Existing data management systems for closed-system blue light therapy in incubators use static empirical parameters or simple linear weights to define the influence of humidity on the thermo-photonic effect index. However, these systems neglect the continuous changes in actual humidity and its nonlinear regulatory effect on infant skin responses, failing to accurately reflect the dynamic impact of humidity on blue light therapy efficacy. In practical applications, the humidity of the infant's environment dynamically changes with incubator operating conditions, the external environment, and physiological activities. If the humidity balance factor uses a fixed value or a linear interpolation method, the system struggles to adjust the sensitivity of the thermo-photonic effect index to humidity changes in real time. This leads to decreased stability and reliability of the prediction results, increases the prediction bias of blue light therapy efficacy, and may affect treatment effectiveness and infant safety.

[0004] The study did not adequately consider the dependence of intensity and duration on the physiological feedback mechanisms during the interaction between infant skin and blue light. Due to the thin structure and high water content of infant skin, its transmission, scattering, and absorption characteristics of blue light vary significantly with light dose, distribution uniformity, and irradiation time, thus affecting the bilirubin photolysis rate. The intensity of blue light irradiation is not only affected by the stability of the equipment output but also by factors such as reflection from the incubator's internal structure, light path obstruction, and distance changes. Furthermore, changes in the infant's position and movements can alter the local dose distribution received by the skin.

[0005] In the closed-circuit blue light irradiation data management system, a sliding mode control mechanism is used through the inner loop controller to precisely adjust environmental parameters such as the heater, humidifier, and sealing status within the incubator, ensuring stable blue light irradiation under strictly controlled environmental conditions. However, the sign function in sliding mode control is essentially a discontinuous switching signal. Ideally, the system state can quickly slide into the sliding surface and maintain stability. But in practical applications, due to limitations such as sensor noise, measurement errors, control execution delays, and the periodic characteristics of sampling and execution in the mechanical intelligent system, this jitter problem manifests in the closed-circuit blue light irradiation system as frequent opening and closing of the actuator, increasing mechanical wear and shortening equipment lifespan. Simultaneously, it triggers high-frequency vibrations and electromagnetic noise within the system, interfering with signal acquisition from key detection units such as temperature and humidity sensors and light intensity sensors, affecting the real-time and accurate monitoring of environmental parameters, and thus reducing the accuracy and stability of blue light irradiation control.

[0006] In view of this, the present invention proposes a data management system for sealed blue light irradiation in an incubator based on mechanical intelligence to solve the above problems. Summary of the Invention

[0007] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a data management system for sealed blue light irradiation in an incubator based on mechanical intelligence, comprising:

[0008] The state cluster data module collects multi-source heterogeneous data, corrects the clock deviation between the incubator and the preset biosensors through a spatiotemporal alignment protocol, and dynamically adjusts the sampling rate of multi-source heterogeneous data based on the infant physiological data in the time-aligned multi-source heterogeneous data, and outputs state cluster data packets with timestamps.

[0009] The hierarchical extraction module constructs a three-level feature processing pipeline based on state cluster data packages. This includes a basic layer that calculates basic statistics through a sliding window; a logic layer that generates a temperature-light effect index and a sealing safety factor based on the basic statistics and multi-source heterogeneous data; and a decision layer that establishes a treatment efficacy prediction matrix based on the temperature-light effect index and the sealing safety factor to predict the correlation trend between blue light irradiation and bilirubin metabolism.

[0010] The environmental-physiological coordination module employs a sliding mode control mechanism to design an inner-loop controller that adjusts the environmental parameters within the incubator. An outer-loop controller is constructed using a fuzzy control strategy. Based on the correlation trend between blue light irradiation and bilirubin metabolism, a blue light irradiation treatment plan is generated. Through control commands and conflict resolution mechanisms, blue light irradiation and sealing regulation are coordinated, and a control log is output.

[0011] The dynamic knowledge optimization module constructs a side knowledge network for strategy efficacy evaluation based on historical treatment cases, automatically eliminates outdated blue light irradiation treatment plans, receives control logs, analyzes and identifies key control variables, dynamically generates optimization strategy patch packages, and continuously updates the side knowledge network for strategy efficacy evaluation.

[0012] The medical safety communication module performs risk assessments on control logs based on the strategy efficacy evaluation edge knowledge network; it designs a dual data encapsulation protocol, constructs an emergency priority channel based on risk assessment, allocates bandwidth for priority transmission, automatically generates a human-computer interaction task queue that conforms to medical procedures, and intercepts illegal instructions in real time.

[0013] Preferably, the method for correcting the clock deviation between the incubator and the preset biosensor includes:

[0014] Multi-source heterogeneous data is acquired in real time through biosensors installed inside and outside the incubator. This data includes incubator equipment data, infant physiological data, and external environmental data. A time reference is established between the incubator and the biosensors before the acquisition of multi-source heterogeneous data using a spatiotemporal alignment protocol. The spatiotemporal alignment protocol includes sending a global synchronization pulse before acquisition, recording the local timestamp of the synchronization signal received by the device, and calculating the initial deviation. During acquisition, a reference time point is inserted to monitor the offset of the local clock relative to the reference time and dynamically calculate drift compensation.

[0015] For the collected multi-source heterogeneous data, the sampling timestamps are adjusted to the aligned time under a unified time base; the clock deviation value between the incubator and the biosensor is calculated through the protocol, and the sampling timestamps of the multi-source heterogeneous data are updated in real time, mapping the multi-source heterogeneous data sampled on the asynchronous timeline to the same global timeline.

[0016] Preferably, the method for obtaining the timestamped state cluster data packet includes:

[0017] A threshold detection method is used to analyze infant physiological data in real time and extract the characteristics of the infant's current physiological state. Physiological response judgment criteria are set to determine whether the current physiological state is in an abnormal stage. The physiological response judgment criteria include preset heart rate threshold and bilirubin level threshold.

[0018] When the infant's heart rate is detected to be greater than or equal to a preset heart rate threshold and the bilirubin level is detected to be greater than or equal to a preset bilirubin level threshold, the sampling rate of the multi-source heterogeneous data is automatically increased; when the infant's heart rate is less than a preset heart rate threshold and the bilirubin level is less than a preset bilirubin level threshold, the sampling rate of the multi-source heterogeneous data is automatically decreased; the multi-source heterogeneous data after dynamic sampling rate adjustment is aggregated to generate a state cluster data package with timestamps and synchronization correction.

[0019] Preferably, the method for obtaining the basic statistics includes:

[0020] At the base layer, a sliding window analysis method is used to process multi-source heterogeneous data in real time based on state cluster data packets. The sliding window moves and samples on the time axis with a fixed step size according to a preset time length to extract multi-source heterogeneous data within a continuous time period.

[0021] For the multi-source heterogeneous data within each sliding window, the data on incubator equipment, infant physiological data, and external environmental data contained in the multi-source heterogeneous data are grouped and processed separately. The basic statistics of each data type within the window are calculated, including mean, variance, maximum, minimum, range, quartiles, median, and percentiles. The time intervals of sampling points for each data type are statistically analyzed, and the data integrity index and the trend of sampling frequency change are recorded.

[0022] Preferably, the method for generating the thermo-optical effect index and the sealing safety factor includes:

[0023] The logic layer receives the basic statistics calculated by the base layer through a sliding window, and establishes a multi-dimensional fusion mechanism based on the correlation between the incubator's internal environment and the infant's physiological data. This includes using temperature-related statistics, combined with the infant's skin temperature and bilirubin metabolism indicators, to analyze the temperature-dependent effect of blue light irradiation on bilirubin decomposition; and comprehensively considering humidity and airflow velocity inside the incubator to assess the stability of the incubator's internal environment and its potential impact on the blue light effect.

[0024] A thermo-photonic effect index function is constructed based on a multi-dimensional fusion mechanism to obtain the thermo-photonic effect index and sealing safety factor for the physiological effects of blue light irradiation therapy on infants in incubators; through a dynamic factor adjustment mechanism, the humidity balance factor in the thermo-photonic effect index function is updated according to the real-time environmental humidity level to achieve adaptive adjustment of the thermo-photonic effect index.

[0025] Preferably, the method for predicting the association trend between blue light irradiation and bilirubin metabolism includes:

[0026] The decision-making layer receives the temperature and light effect index and sealing safety factor output by the logic layer, and combines them with dynamic monitoring data of bilirubin concentration in the infant's body to construct a treatment efficacy prediction matrix. The treatment efficacy prediction matrix is ​​based on the temperature and light effect index and sealing safety factor, forming a two-dimensional data set at different time points, representing the comprehensive state of the treatment environment.

[0027] The relationship between the treatment efficacy prediction matrix and the change of bilirubin concentration over time was analyzed. Regression analysis was used to explore the influence of environmental factors of blue light irradiation on bilirubin metabolism. Based on the analysis results, the trend of bilirubin concentration change over the next n periods was predicted, thereby evaluating the efficacy of current blue light therapy.

[0028] Based on the trend of bilirubin concentration changes over a period of n years, this provides clinicians with decision-making basis, guides the adjustment of incubator equipment data and external environmental data, supports feedback based on dynamic monitoring data of bilirubin concentration in infants, and dynamically adjusts treatment plans.

[0029] Preferably, the method for generating the blue light irradiation treatment plan includes:

[0030] Let x(t) be the current state vector of the incubator environmental parameters at time t; let y(t) be the target state vector of the incubator environmental parameters at time t. Then, the error state vector between the current state vector and the target state vector of the incubator environmental parameters at time t is z(t) = x(t) - y(t). The sliding surface can be represented as s(t) = z′(t) + ∧z(t); where s(t) represents the sliding surface; z′(t) represents the error rate of change; and ∧ represents a positive definite diagonal matrix.

[0031] A dynamic rule is defined by a sliding surface. When the error state vector satisfies s(t) = 0, the error state vector and its rate of change will gradually approach zero. An inner-loop controller is designed using a sliding mode control mechanism. A sliding surface is designed, and the input is controlled by a sliding mode control law to make the error state vector reach and remain on the sliding surface. The input change is smoothed by a saturation function to suppress the jitter caused by the sign function in the sliding mode control law, thereby adjusting the environmental parameters inside the warm chamber.

[0032] The input variables of the outer loop controller include the current bilirubin concentration in the infant's body, the thermo-photonic effect index, and the sealing safety factor. The input variables are fuzzified using fuzzy membership functions to map continuous values ​​to fuzzy sets. Fuzzy inference rules are formulated based on the combination of input variables. Based on the correlation between bilirubin metabolism trend and blue light irradiation, and combined with the fuzzy inference rules of the input variables, a blue light irradiation adjustment scheme is generated.

[0033] The fuzzy reasoning rules include: if the current bilirubin concentration in the infant's body is less than or equal to the preset third threshold for bilirubin concentration in the infant's body, and greater than the preset second threshold for bilirubin concentration in the infant's body, the temperature-light effect index is less than the preset optimal threshold for the temperature-light effect index, and the sealing safety factor is greater than or equal to the preset sealing safety factor threshold, then the blue light irradiation intensity and irradiation duration will be increased, and the temperature and humidity will be optimized.

[0034] If the current bilirubin concentration in the infant's body is less than or equal to the preset second threshold for bilirubin concentration in the infant's body, and greater than the preset first threshold for bilirubin concentration in the infant's body; the thermo-photonic effect index is greater than or equal to the preset optimal threshold for thermo-photonic effect index; and the sealing safety factor is greater than or equal to the preset sealing safety factor threshold, then the output will maintain the blue light irradiation intensity and duration.

[0035] If the current bilirubin concentration in the infant's body is less than or equal to the preset first threshold for bilirubin concentration in the infant's body, then the intensity and duration of blue light exposure will be reduced; if the sealing safety factor is less than the preset sealing safety factor threshold or the current bilirubin concentration in the infant's body is greater than the preset third threshold for bilirubin concentration in the infant's body, then blue light therapy will be suspended.

[0036] Preferably, the method for coordinating blue light irradiation and sealing adjustment includes:

[0037] Based on the blue light irradiation treatment plan, blue light irradiation control instructions and sealing adjustment control instructions are generated, including specific parameter settings and execution time; a conflict resolution mechanism is set up to monitor the conflict between the blue light irradiation control instructions and the sealing adjustment control instructions in real time. Conflicts include the sealing state not meeting the preset requirements when blue light irradiation is turned on and the sealing adjustment instructions causing damage to the blue light irradiation effect.

[0038] Based on the analysis of the current incubator environmental parameters, the root cause of the conflict is determined. The execution sequence, parameter amplitude, and duration of the control commands for blue light irradiation and sealing adjustment are dynamically adjusted. The issuance, adjustment, and execution processes of all blue light irradiation control commands and sealing adjustment control commands are recorded and output as control logs.

[0039] Preferably, the method for constructing the strategy efficacy evaluation edge knowledge network includes:

[0040] Historical treatment case data, including blue light irradiation parameters, environmental parameters, infant bilirubin concentration, and treatment effect feedback, were collected and preprocessed. Key variables were screened and multidimensional feature vectors were extracted. Key variables included blue light irradiation intensity, light irradiation duration, thermo-photothermal effect index, sealing safety factor, and infant bilirubin concentration.

[0041] Key variables and efficacy indicators are defined as nodes in the strategy efficacy evaluation edge knowledge network, and the node types are divided into input nodes, environment nodes, and output nodes. The directed relations connecting different node types are defined as edges in the strategy efficacy evaluation edge knowledge network, thereby constructing the strategy efficacy evaluation edge knowledge network.

[0042] Preferably, the method for prioritizing bandwidth allocation and transmission includes:

[0043] The strategy efficacy evaluation edge knowledge network performs source tracing reasoning based on the received control logs to determine the correlation strength between the control strategy and efficacy indicators, thereby generating the risk level of the control instructions; a dual data encapsulation protocol is designed, which includes a basic data encapsulation layer and a risk label encapsulation layer.

[0044] The basic data encapsulation layer is responsible for standard formatting, encryption, and verification of control logs according to the preset communication protocol; the risk label encapsulation layer carries additional emergency event identifiers, indicates the risk assessment level of the control logs based on fuzzy inference rules, and builds an emergency event priority channel in the risk label encapsulation layer for priority transmission through bandwidth allocation.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] This invention accurately captures the interactions between these parameters through a multi-dimensional fusion mechanism, demonstrating the synergistic regulatory effect of temperature and humidity on the efficacy of blue light therapy. This allows for a more realistic reflection of the impact of the actual treatment environment on the infant's physiological response. Employing a dynamic factor regulation mechanism, the humidity balance factor is continuously and smoothly adjusted according to changes in environmental humidity. This enables real-time response to humidity fluctuations, preventing the simple neglect or over-amplification of humidity changes, reducing errors in the temperature-phototherapy effect index, and improving the accuracy of predicting the effects of blue light irradiation. By dynamically reflecting real-time changes in environmental and physiological parameters, the temperature-phototherapy effect index provides clinicians with more precise guidance for adjusting the intensity and duration of blue light irradiation, helping to develop individualized treatment plans, reducing the risk of unstable treatment effects due to environmental fluctuations, and improving the safety and stability of infant treatment.

[0047] By introducing a saturation function to achieve continuous approximation near the sliding surface, the control input transitions from discontinuous jumps to smooth linear changes near the error zero point, significantly reducing jitter caused by rapid switching. This smooth control signal effectively reduces the risk of high-frequency oscillations, preventing system oscillations or divergence and ensuring the stability and reliability of environmental parameters within the incubator. The robustness of sliding mode control, combined with boundary layer smoothing, ensures that the system can quickly and accurately stabilize its state on the sliding surface even when faced with parameter disturbances and measurement errors, guaranteeing precise adjustment of environmental parameters and adapting to complex and changing internal environmental requirements. The use of saturation function smoothing control effectively reduces the switching frequency of the control signal, lowers the fatigue load on mechanical components, significantly improves the durability and maintenance cycle of the system hardware, and reduces maintenance costs and downtime risks. Improved control system stability reduces temperature and humidity fluctuations within the incubator, while also lowering equipment noise and failure rates, thus providing a safer, more reliable, and convenient operating experience. Attached Figure Description

[0048] Figure 1 This is a schematic diagram of the mechanical intelligence-based closed blue light irradiation data management system for an incubator according to the present invention.

[0049] Figure 2 This is a schematic diagram of the method for correcting clock deviation between an incubator and a preset biosensor provided by the present invention.

[0050] Figure 3 This is a schematic diagram of the process for managing data from a sealed incubator under blue light irradiation based on mechanical intelligence, as described in this invention. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] Example 1

[0053] Please see Figure 1 and Figure 2 As shown, Embodiment 1 further illustrates the mechanical intelligence-based closed blue light irradiation data management system for incubators proposed in this invention, including:

[0054] Neonatal jaundice is a common clinical condition, and phototherapy, as one of the main treatment methods, relies on precise control of environmental parameters such as temperature, humidity, and light intensity within the closed environment of an incubator to ensure the efficacy of phototherapy and the safety of the infant. However, existing data management systems for closed-system phototherapy incubators still face several technical challenges in practical applications, affecting treatment effectiveness and system reliability.

[0055] Existing systems typically use static empirical parameters or simple linear weights to define the influence of humidity on the thermophotodynamic effect index, neglecting the continuous dynamic changes in humidity and its nonlinear regulatory effect on infant skin responses. In reality, the humidity of the infant's environment is dynamically affected by multiple factors, including incubator operating conditions, the external environment, and the infant's physiological activities. If the humidity balance factor uses a fixed value or a linear interpolation method, the system cannot accurately adjust the sensitivity of the thermophotodynamic effect index to humidity changes in real time, leading to a decrease in the stability and reliability of blue light therapy efficacy prediction, which in turn affects treatment effectiveness and infant safety.

[0056] Current technologies do not fully consider the dependence of light intensity and duration on the interaction between infant skin and blue light, as well as the complex physiological feedback mechanisms involved. Infant skin is thin and has a high water content; its transmission, scattering, and absorption characteristics of blue light change significantly with variations in light dose, distribution uniformity, and irradiation time, thus affecting the bilirubin photolysis rate. Furthermore, the intensity of blue light irradiation is affected not only by the stability of the equipment output but also by various factors such as reflections from the incubator's internal structure, light path obstruction, and changes in the distance between the light source and the skin. Changes in the infant's position and movements further contribute to uneven local dose distribution, increasing the complexity of the treatment process.

[0057] In data management systems based on closed-circuit blue light irradiation in incubators, sliding mode control is commonly used in the inner loop controller to achieve precise adjustment of environmental parameters such as the heater, humidifier, and sealing status within the incubator. Sliding mode control, through the design of a sliding surface, allows the system state to quickly and stably slide into and move along the sliding surface, thereby ensuring precise control of environmental parameters and stability of the blue light irradiation process. However, the sign function in sliding mode control is essentially a discontinuous switching signal. Ideally, the system state can quickly slide into the sliding surface and remain stable. But in practical applications, due to sensor noise, measurement errors, control execution delays, and the sampling and execution cycle limitations of the mechanical intelligent system within the incubator, the system state is prone to frequently crossing zero near the sliding surface, leading to rapid oscillations in the control input, i.e., jitter.

[0058] In a closed-loop blue light irradiation system, this vibration manifests as frequent opening and closing of actuators, leading to increased mechanical wear and shortened equipment lifespan. High-frequency vibration and electromagnetic noise interfere with signal acquisition from critical detection units such as temperature and humidity sensors and light intensity sensors, reducing the real-time performance and accuracy of environmental parameter monitoring, thus affecting the precision and stability of blue light irradiation control. Furthermore, the vibration causes frequent control switching, increasing energy consumption, reducing the incubator's energy efficiency, and severely impacting the overall system reliability and treatment safety.

[0059] To effectively address the aforementioned problems, this invention proposes a data management system for sealed blue light irradiation in an incubator based on mechanical intelligence, comprising:

[0060] The state cluster data module collects multi-source heterogeneous data, corrects the clock deviation between the incubator and the preset biosensors through a spatiotemporal alignment protocol, and dynamically adjusts the sampling rate of multi-source heterogeneous data based on the infant physiological data in the time-aligned multi-source heterogeneous data, and outputs state cluster data packets with timestamps.

[0061] The hierarchical extraction module constructs a three-level feature processing pipeline based on state cluster data packages. This includes a basic layer that calculates basic statistics through a sliding window; a logic layer that generates a temperature-light effect index and a sealing safety factor based on the basic statistics and multi-source heterogeneous data; and a decision layer that establishes a treatment efficacy prediction matrix based on the temperature-light effect index and the sealing safety factor to predict the correlation trend between blue light irradiation and bilirubin metabolism.

[0062] The environmental-physiological coordination module employs a sliding mode control mechanism to design an inner-loop controller that adjusts the environmental parameters within the incubator. An outer-loop controller is constructed using a fuzzy control strategy. Based on the correlation trend between blue light irradiation and bilirubin metabolism, a blue light irradiation treatment plan is generated. Through control commands and conflict resolution mechanisms, blue light irradiation and sealing regulation are coordinated, and a control log is output.

[0063] The dynamic knowledge optimization module constructs a side knowledge network for strategy efficacy evaluation based on historical treatment cases, automatically eliminates outdated blue light irradiation treatment plans, receives control logs, analyzes and identifies key control variables, dynamically generates optimization strategy patch packages, and continuously updates the side knowledge network for strategy efficacy evaluation.

[0064] The medical safety communication module performs risk assessments on control logs based on the strategy efficacy evaluation edge knowledge network; it designs a dual data encapsulation protocol, constructs an emergency priority channel based on risk assessment, allocates bandwidth for priority transmission, automatically generates a human-computer interaction task queue that conforms to medical procedures, and intercepts illegal instructions in real time.

[0065] Methods for correcting clock deviations between the incubator and pre-set biosensors include:

[0066] Biosensors (such as heart rate, respiratory rate, skin temperature, and bilirubin level sensors) installed inside and outside the incubator acquire multi-source heterogeneous data in real time. This multi-source heterogeneous data includes incubator equipment data, infant physiological data, and external environmental data. Incubator equipment data includes incubator internal temperature, humidity, incubator surface temperature, wind speed, airflow speed inside the incubator, and blue light intensity. Infant physiological data includes infant skin temperature, bilirubin level, weight, heart rate, respiratory rate, and blood oxygen saturation. External environmental data includes indoor temperature and humidity.

[0067] It should be noted that because different sensors use independent sampling mechanisms, their sampling clocks differ, and their communication paths are asynchronous, which may introduce errors such as clock drift and network latency. The warming chamber equipment and sensors each have their own built-in independent clock sources (such as RTC chips), and differences in their frequency deviations and startup times may lead to inconsistent time bases. During long-term continuous data acquisition, clock drift (small but cumulative time errors) can affect data alignment.

[0068] A time reference is established between the incubator and the biosensor before multi-source heterogeneous data acquisition using a spatiotemporal alignment protocol. The spatiotemporal alignment protocol includes sending a global synchronization pulse before acquisition, recording the local timestamp of the synchronization signal received by the device, and calculating the initial deviation. During acquisition, a reference time point is inserted to monitor the offset of the local clock relative to the reference time and dynamically calculate drift compensation.

[0069] For the collected multi-source heterogeneous data, the sampling timestamps are adjusted to the aligned time under a unified time base; the clock deviation value between the incubator and the biosensor is calculated through the protocol, and the sampling timestamps of the multi-source heterogeneous data are updated in real time, mapping the multi-source heterogeneous data sampled on the asynchronous timeline to the same global timeline.

[0070] Methods for obtaining timestamped state cluster data packets include:

[0071] A threshold detection method is used to analyze infant physiological data in real time and extract the characteristics of the infant's current physiological state. Physiological response judgment criteria are set to determine whether the current physiological state is in an abnormal stage. The physiological response judgment criteria include preset heart rate threshold and bilirubin level threshold.

[0072] When the infant's heart rate is detected to be greater than or equal to a preset heart rate threshold and the bilirubin level is detected to be greater than or equal to a preset bilirubin level threshold, the sampling rate of the multi-source heterogeneous data is automatically increased; when the infant's heart rate is less than a preset heart rate threshold and the bilirubin level is less than a preset bilirubin level threshold, the sampling rate of the multi-source heterogeneous data is automatically decreased; the multi-source heterogeneous data after dynamic sampling rate adjustment is aggregated to generate a state cluster data package with timestamps and synchronization correction.

[0073] Methods for obtaining basic statistics include:

[0074] At the base layer, a sliding window analysis method is used to process multi-source heterogeneous data in real time based on state cluster data packets. The sliding window moves and samples on the time axis according to a preset time length with a fixed step size (e.g., 1 minute, 5 minutes or 10 minutes) to extract multi-source heterogeneous data within a continuous time period.

[0075] For the multi-source heterogeneous data within each sliding window, the data on incubator equipment, infant physiological data, and external environmental data contained in the multi-source heterogeneous data are grouped and processed separately. The basic statistics of each data type within the window are calculated, including mean, variance, maximum, minimum, range, quartiles, median, and percentiles. The time intervals of sampling points for each data type are statistically analyzed, and the data integrity index and the trend of sampling frequency change are recorded.

[0076] Methods for generating the temperature and light effect index and the sealing safety factor include:

[0077] The logic layer receives the basic statistics calculated by the base layer through a sliding window, and establishes a multi-dimensional fusion mechanism based on the correlation between the incubator's internal environment and the infant's physiological data. This includes using temperature-related statistics, combined with the infant's skin temperature and bilirubin metabolism indicators, to analyze the temperature-dependent effect of blue light irradiation on bilirubin decomposition; and comprehensively considering humidity and airflow velocity inside the incubator to assess the stability of the incubator's internal environment and its potential impact on the blue light effect.

[0078] A thermo-photonic effect index function was constructed based on a multi-dimensional fusion mechanism to obtain the thermo-photonic effect index and sealing safety factor for the physiological effects of blue light irradiation therapy on infants in incubators; the thermo-photonic effect index function is as follows: Where T1 represents the internal temperature of the incubator; T2 represents the infant's skin temperature; L represents the intensity of blue light irradiation; t′ Lα represents the duration of illumination; H represents the humidity inside the incubator; β represents the humidity balance factor, used to adjust the contribution of humidity to the thermo-photometric effect index; according to expert experience, the value of α ranges from 0 to 1; ∈ represents a preset positive number greater than 0, with a value range between 0.01 and 0.1, to prevent the divisor from being zero or too small, which would lead to an excessively large index or unstable value.

[0079] The sealing safety factor is: Wherein, represents the airflow velocity inside the warm chamber; represents the maximum allowable airflow velocity threshold under sealed conditions inside the warm chamber; represents the airflow velocity adjustment coefficient, used to adjust the degree of influence of the airflow velocity index on the sealing safety factor. According to the expert experience method, the value of γ ranges from 0 to 1.

[0080] However, the skin barrier function of infants is not fully developed, and changes in humidity significantly affect their transmittance, absorption, and reflectance of blue light, thus impacting bilirubin metabolism efficiency. Under different humidity conditions, skin hydration, evaporation, and temperature regulation functions all differ, making the effect of humidity on the blue light effect non-linear and dynamically changing. Currently, most blue light irradiation control systems use static empirical parameters or simple linear weights to set the influence of humidity on the thermo-photonic effect index. This approach ignores the continuous changes in actual humidity and its non-linear regulatory effect on infant skin responses, failing to accurately reflect the dynamic impact of humidity on the therapeutic effect of blue light.

[0081] In practical applications, the humidity of the infant's environment changes dynamically with the incubator's operating conditions, the external environment, and physiological activities (such as respiration and sweat evaporation). If the humidity balance factor uses a fixed value or a linear interpolation method, the system will find it difficult to adjust the sensitivity of the thermophotodynamic effect index to humidity changes in real time, leading to a decrease in the stability and reliability of the prediction results, increasing the prediction bias of blue light therapy efficacy, and potentially affecting the treatment effect and the infant's safety.

[0082] In existing technologies, blue light irradiation control typically uses static or simple linear models to set the influence of light irradiation on the thermo-photodynamic effect index, without fully considering the dependence of intensity and duration, as well as physiological feedback mechanisms, on the interaction between infant skin and blue light. Because infant skin is thin and has a high water content, its transmission, scattering, and absorption characteristics of blue light vary significantly depending on the light dose, distribution uniformity, and irradiation time, thus affecting the bilirubin photolysis rate.

[0083] In practical applications, the intensity of blue light irradiation is affected not only by the stability of the equipment output but also by factors such as reflections from the incubator's internal structure, light path obstruction, and changes in distance. Furthermore, changes in the infant's position and movements alter the local dose distribution received by the skin. Current technology cannot capture the dynamic regulatory effect of real-time changes in light intensity and dose on bilirubin metabolism rates. When blue light irradiation intensity fluctuates, local dose distribution becomes uneven, or irradiation time is prolonged, if the blue light irradiation balance factor fails to adjust accordingly, it will lead to insufficient sensitivity to blue light dose response, increasing the bias in efficacy prediction and potential safety hazards, and reducing the feasibility of individualized and precise treatment control.

[0084] By employing a dynamic factor adjustment mechanism, the humidity balance factor in the temperature-light effect index function is updated based on the real-time ambient humidity level, achieving adaptive adjustment of the temperature-light effect index. The humidity balance factor is dynamically adjusted through a humidity balance factor adjustment function.

[0085] The humidity balance factor adjustment function is Where H0 represents the humidity threshold inside the incubator; k represents the slope parameter, used to adjust the steepness of the humidity balance factor adjustment function;

[0086] It should be noted that the humidity balance factor adjustment function is modeled after the logistic function. It is a classic and widely used nonlinear function in mathematics, control theory, and biostatistics, possessing good smoothness and monotonicity. It is often used to describe the nonlinear influence of certain physiological and environmental variables on system output. It is a fundamental function for probability prediction in statistics and is also used as an activation function in neural networks, mapping inputs to a range between 0 and 1, consistent with probability values. This adjustment formula can be understood as being based on the logistic function, with the core idea being to transform changes in environmental humidity into a balance factor through nonlinear mapping, thereby controlling the intensity of humidity's influence on the blue light effect.

[0087] Compared to existing technologies, the advantages are as follows: Through a multi-dimensional fusion mechanism, the interaction between these parameters is accurately captured, demonstrating the synergistic regulatory effect of temperature and humidity on blue light therapy, and more realistically reflecting the impact of the actual treatment environment on the infant's physiological response. Employing a dynamic factor regulation mechanism, the humidity balance factor is continuously and smoothly adjusted according to changes in environmental humidity, enabling real-time response to humidity fluctuations, avoiding the simple neglect or over-amplification of humidity changes, reducing errors in the thermo-phototherapy effect index, and improving the accuracy of predicting the effects of blue light irradiation. By dynamically reflecting real-time changes in environmental and physiological parameters, the thermo-phototherapy effect index can provide clinicians with more precise guidance for adjusting the intensity and duration of blue light irradiation, helping to develop individualized treatment plans, reducing the risk of unstable treatment effects due to environmental fluctuations, and improving the safety and stability of infant treatment.

[0088] Methods for predicting the association between blue light exposure and bilirubin metabolism include:

[0089] The decision-making layer receives the temperature and light effect index and sealing safety factor output by the logic layer, and combines them with dynamic monitoring data of bilirubin concentration in the infant's body to construct a treatment efficacy prediction matrix. The treatment efficacy prediction matrix is ​​based on the temperature and light effect index and sealing safety factor, forming a two-dimensional data set at different time points, representing the comprehensive state of the treatment environment.

[0090] The relationship between the treatment efficacy prediction matrix and the change of bilirubin concentration over time was analyzed. Regression analysis was used to explore the influence of environmental factors of blue light irradiation on bilirubin metabolism. Based on the analysis results, the trend of bilirubin concentration change over the next n periods was predicted, thereby evaluating the efficacy of current blue light therapy.

[0091] Based on the trend of bilirubin concentration changes over a period of n years, this provides clinicians with decision-making basis, guides the adjustment of incubator equipment data and external environmental data, supports feedback based on dynamic monitoring data of bilirubin concentration in infants, and dynamically adjusts treatment plans.

[0092] Methods for generating blue light therapy protocols include:

[0093] The current state vector of the incubator environment parameters at time t is set to x(t); the target state vector of the incubator environment parameters at time t is set to y(t). Then the error state vector between the current state vector and the target state vector of the incubator environment parameters at time t is z(t) = x(t) - y(t).

[0094] The sliding surface can be represented as s(t) = z′(t) + ∧z(t); where s(t) represents the sliding surface, which is a quantity used to measure the error state vector and its changing trend at time point t; z′(t) represents the error rate of change, which is the derivative of the error state vector; ∧ represents a positive definite diagonal matrix, whose diagonal elements are positive real numbers, and plays a role in adjusting the system response speed;

[0095] A dynamic rule is defined by a sliding surface. When the error state vector satisfies s(t) = 0, the error state vector and its rate of change will gradually approach zero. An inner-loop controller is designed using a sliding mode control mechanism. A sliding surface is designed, and the input is controlled by a sliding mode control law to make the error state vector reach and remain on the sliding surface. The input change is smoothed by a saturation function to suppress the jitter caused by the sign function in the sliding mode control law, thereby adjusting the environmental parameters inside the warm chamber.

[0096] The sliding mode control law is u2(t) = u1(t) - K·sign(s(t)); where u2(t) represents the control input vector, such as the adjustment signal of the heater power, humidifier water volume, oxygen supply, etc. in the incubator; u1(t) represents the equivalent control vector, which is a control quantity that theoretically makes the error state vector move on the sliding surface; sign(s(t)) represents the sign function of the sliding mode control; K represents the gain matrix, which corresponds to the strength of the control input, such as the maximum adjustment range of the heating power, humidification flow rate, oxygen supply rate, etc. K should be large enough to ensure that the sliding mode control can overcome the external disturbances and measurement errors in the system, so that the system state can stably and quickly slide into the sliding surface and remain on the sliding surface.

[0097] It should be noted that when designing an inner-loop controller using sliding mode control, the sign function of sliding mode control is essentially a discontinuous switching signal. Ideally, the system state can precisely slide into and remain on the sliding surface, with frequent switching of the control input maintaining the system's movement along the sliding surface. In actual systems, due to sensor noise, measurement errors, control execution delays, and the unavoidable sampling and execution cycles of the physical system, the system state will continuously cross positive and negative zero points near the sliding surface. This can easily cause the control input to oscillate rapidly near zero error, a phenomenon known as jitter, which can damage hardware or lead to instability in the system. Frequent switching actions will accelerate the wear of mechanical components such as actuators and relays, shortening their lifespan. Rapid control switching may excite high-frequency vibrations within the system itself, affecting overall control performance and even causing system divergence or oscillation. Frequent switching reduces the energy efficiency of the control system. Jitter generates high-frequency noise, interfering with other electronic components and sensors in the system, reducing the accuracy of measurement and control.

[0098] The saturation function is Where φ represents the preset boundary layer width, which is the smooth interval range around the sliding surface. Within this interval, the controller continuously and linearly approximates the control signal to reduce jitter caused by control signal switching; the sliding mode control law is also adjusted accordingly.

[0099] Compared to existing technologies, the advantages are as follows: By introducing a saturation function to achieve continuous approximation near the sliding surface, the control input transitions from discontinuous jumps to smooth linear changes near the error zero point, greatly reducing jitter caused by rapid switching. This smooth control signal effectively reduces the risk of high-frequency oscillations, avoids system oscillations or divergence, and ensures the stability and reliability of environmental parameters within the incubator. The robustness of sliding mode control, combined with boundary layer smoothing, ensures that the system can quickly and accurately stabilize its state on the sliding surface even when faced with parameter disturbances and measurement errors, guaranteeing precise adjustment of environmental parameters and adapting to complex and changing internal environmental requirements. The use of saturation function smoothing control effectively reduces the switching frequency of the control signal, lowers the fatigue load on mechanical components, significantly improves the durability and maintenance cycle of the system hardware, and reduces maintenance costs and downtime risks. Improved control system stability reduces temperature and humidity fluctuations within the incubator, while also lowering equipment noise and failure rates, thus providing a safer, more reliable, and more convenient operating experience.

[0100] The input variables of the outer loop controller include the current bilirubin concentration in the infant's body, the thermo-photonic effect index, and the sealing safety factor. The input variables are fuzzified using fuzzy membership functions to map continuous values ​​to fuzzy sets. Fuzzy inference rules are formulated based on the combination of input variables. Based on the correlation between bilirubin metabolism trend and blue light irradiation, and combined with the fuzzy inference rules of the input variables, a blue light irradiation adjustment scheme is generated.

[0101] It is important to note that there is a significant correlation between bilirubin metabolism and blue light irradiation in the treatment of neonatal jaundice. Bilirubin is primarily metabolized by the liver into a soluble form before excretion. However, in newborns, due to the immature liver metabolic function, bilirubin accumulates, leading to jaundice. Blue light irradiation (usually using light with wavelengths in the 425-475 nm range) promotes the conversion of bilirubin from its lipid-soluble form to its water-soluble isomer and photooxidation products through photooxidation and photoisomerization, thereby enhancing its excretion capacity.

[0102] By analyzing clinical data on changes in blue light irradiation intensity, duration, and bilirubin concentration, a correlation model between blue light irradiation and bilirubin metabolism was established. The study found that with increasing blue light irradiation intensity and duration, the rate of decrease in bilirubin concentration in infants accelerated, and this trend was relatively linear within a certain range of intensity and duration. However, excessively strong or prolonged blue light irradiation may cause adverse reactions in infants, such as elevated body temperature and dry skin, due to thermal effects and skin moisture evaporation, thereby affecting bilirubin metabolism efficiency.

[0103] Therefore, to optimize bilirubin metabolism while ensuring safety, this invention combines bilirubin metabolism trends with comprehensive consideration of blue light irradiation intensity, irradiation time, and infant physiological parameters, proposing an outer-loop control strategy based on the thermo-photodynamic effect index and a sealing safety factor. This strategy dynamically adjusts blue light irradiation conditions to ensure sufficient irradiation intensity and duration to promote metabolism when bilirubin concentration exceeds a preset threshold; and appropriately reduces blue light irradiation intensity and duration to prevent overtreatment when bilirubin concentration approaches the normal range. Simultaneously, the system monitors the thermo-photodynamic effect index and sealing safety factor in real time to prevent adverse reactions caused by abnormal temperature, humidity, or airflow, ensuring the safety and effectiveness of the treatment.

[0104] The correlation between blue light irradiation and bilirubin metabolism shows that reasonable blue light irradiation intensity and duration can accelerate bilirubin metabolism, resulting in a stable downward trend in bilirubin concentration; while excessive or insufficient blue light irradiation may lead to decreased metabolic efficiency or safety hazards.

[0105] The fuzzy reasoning rules include: if the current bilirubin concentration in the infant's body is less than or equal to the preset third threshold for bilirubin concentration in the infant's body, and greater than the preset second threshold for bilirubin concentration in the infant's body, the temperature-light effect index is less than the preset optimal threshold for the temperature-light effect index, and the sealing safety factor is greater than or equal to the preset sealing safety factor threshold, then the blue light irradiation intensity and irradiation duration will be increased, and the temperature and humidity will be optimized.

[0106] If the current bilirubin concentration in the infant's body is less than or equal to the preset second threshold for bilirubin concentration in the infant's body, and greater than the preset first threshold for bilirubin concentration in the infant's body; the thermo-photonic effect index is greater than or equal to the preset optimal threshold for thermo-photonic effect index; and the sealing safety factor is greater than or equal to the preset sealing safety factor threshold, then the output will maintain the blue light irradiation intensity and duration.

[0107] If the current bilirubin concentration in the infant's body is less than or equal to the preset first threshold for bilirubin concentration in the infant's body, then the intensity and duration of blue light exposure will be reduced; if the sealing safety factor is less than the preset sealing safety factor threshold or the current bilirubin concentration in the infant's body is greater than the preset third threshold for bilirubin concentration in the infant's body, then blue light therapy will be suspended.

[0108] Methods for coordinating blue light exposure and sealing regulation include:

[0109] Based on the blue light irradiation treatment plan, blue light irradiation control instructions and sealing adjustment control instructions are generated, including specific parameter settings and execution time; a conflict resolution mechanism is set up to monitor the conflict between the blue light irradiation control instructions and the sealing adjustment control instructions in real time. Conflicts include the sealing state not meeting the preset requirements when blue light irradiation is turned on and the sealing adjustment instructions causing damage to the blue light irradiation effect.

[0110] Based on the analysis of the current incubator environmental parameters, the root cause of the conflict is determined. The execution sequence, parameter amplitude, and duration of the control commands for blue light irradiation and sealing adjustment are dynamically adjusted. The issuance, adjustment, and execution processes of all blue light irradiation control commands and sealing adjustment control commands are recorded and output as control logs.

[0111] Methods for constructing edge knowledge networks for strategy efficacy evaluation include:

[0112] Historical treatment case data, including blue light irradiation parameters, environmental parameters, infant bilirubin concentration, and treatment effect feedback, were collected and preprocessed. Key variables were screened and multidimensional feature vectors were extracted. Key variables included blue light irradiation intensity, light irradiation duration, thermo-photothermal effect index, sealing safety factor, and infant bilirubin concentration.

[0113] Key variables and efficacy indicators are defined as nodes in the strategy efficacy evaluation edge knowledge network, and the node types are divided into input nodes, environment nodes, and output nodes. The directed relations connecting different node types are defined as edges in the strategy efficacy evaluation edge knowledge network, thereby constructing the strategy efficacy evaluation edge knowledge network.

[0114] Methods for prioritizing bandwidth allocation and transmission include:

[0115] The strategy efficacy evaluation edge knowledge network performs source tracing reasoning based on the received control logs to determine the correlation strength between the control strategy and efficacy indicators, thereby generating the risk level of the control instructions; a dual data encapsulation protocol is designed, which includes a basic data encapsulation layer and a risk label encapsulation layer.

[0116] The basic data encapsulation layer is responsible for standard formatting, encryption, and verification of control logs according to the preset communication protocol; the risk label encapsulation layer carries additional emergency event identifiers, indicates the risk assessment level of the control logs based on fuzzy inference rules, and builds an emergency event priority channel in the risk label encapsulation layer for priority transmission through bandwidth allocation.

[0117] The preset heart rate threshold is set by staff by collecting different heart rates and taking the average of multiple heart rates as the preset heart rate threshold; similarly, the bilirubin level threshold, the first preset threshold for bilirubin concentration in the infant, the second preset threshold for bilirubin concentration in the infant, the third preset threshold for bilirubin concentration in the infant, the optimal threshold for the temperature and light effect index, and the preset sealing safety factor threshold are set.

[0118] This embodiment, through a multi-dimensional fusion mechanism, accurately captures the interactions between these parameters, demonstrating the synergistic regulatory effect of temperature and humidity on the efficacy of blue light therapy. This more realistically reflects the impact of the actual treatment environment on the infant's physiological response. Employing a dynamic factor regulation mechanism, the humidity balance factor is continuously and smoothly adjusted according to changes in environmental humidity. This allows for real-time response to humidity fluctuations, preventing humidity changes from being simply ignored or overemphasized, reducing errors in the thermo-phototherapy effect index, and improving the accuracy of predicting the effects of blue light irradiation. By dynamically reflecting real-time changes in environmental and physiological parameters, the thermo-phototherapy effect index provides clinicians with more precise guidance for adjusting the intensity and duration of blue light irradiation, helping to develop individualized treatment plans, reducing the risk of unstable treatment effects due to environmental fluctuations, and improving the safety and stability of infant treatment.

[0119] By introducing a saturation function to achieve continuous approximation near the sliding surface, the control input transitions from discontinuous jumps to smooth linear changes near the error zero point, significantly reducing jitter caused by rapid switching. This smooth control signal effectively reduces the risk of high-frequency oscillations, preventing system oscillations or divergence and ensuring the stability and reliability of environmental parameters within the incubator. The robustness of sliding mode control, combined with boundary layer smoothing, ensures that the system can quickly and accurately stabilize its state on the sliding surface even when faced with parameter disturbances and measurement errors, guaranteeing precise adjustment of environmental parameters and adapting to complex and changing internal environmental requirements. The use of saturation function smoothing control effectively reduces the switching frequency of the control signal, lowers the fatigue load on mechanical components, significantly improves the durability and maintenance cycle of the system hardware, and reduces maintenance costs and downtime risks. Improved control system stability reduces temperature and humidity fluctuations within the incubator, while also lowering equipment noise and failure rates, thus providing a safer, more reliable, and convenient operating experience.

[0120] Example 2

[0121] Please see Figure 3 As shown, for parts not described in detail in this embodiment, please refer to the description in Embodiment 1. A method for managing blue light irradiation data in a sealed incubator based on mechanical intelligence is provided, including:

[0122] S1. Collect multi-source heterogeneous data, correct the clock deviation between the incubator and the preset biosensor through a spatiotemporal alignment protocol, and dynamically adjust the sampling rate of multi-source heterogeneous data based on the infant physiological data in the time-aligned multi-source heterogeneous data, and output a state cluster data packet with timestamps.

[0123] S2. A three-level feature processing pipeline is constructed based on state cluster data packages, including a basic layer that calculates basic statistics through a sliding window; a logic layer that generates a temperature and light effect index and a sealing safety factor based on the basic statistics and multi-source heterogeneous data; and a decision layer that establishes a treatment efficacy prediction matrix based on the temperature and light effect index and the sealing safety factor to predict the correlation trend between blue light irradiation and bilirubin metabolism.

[0124] S3. An inner-loop controller is designed using a sliding mode control mechanism to adjust the environmental parameters inside the incubator; an outer-loop controller is constructed using a fuzzy control strategy to generate a blue light irradiation treatment plan based on the correlation trend between blue light irradiation and bilirubin metabolism, and coordinates blue light irradiation and sealing adjustment through control commands and conflict resolution mechanisms, and outputs a control log.

[0125] S4. Construct a side knowledge network for strategy efficacy evaluation based on historical treatment cases, automatically eliminate outdated blue light irradiation treatment plans, receive control logs to analyze and identify key control variables, dynamically generate optimization strategy patch packages, and continuously update the side knowledge network for strategy efficacy evaluation.

[0126] S5. Based on the strategy efficacy evaluation edge knowledge network, conduct risk assessment on the control logs; design a dual data encapsulation protocol, construct an emergency priority channel based on risk assessment, allocate bandwidth for priority transmission, automatically generate a human-computer interaction task queue that complies with medical procedures, and intercept illegal instructions in real time.

[0127] Since the electronic device described in this embodiment is the one used in implementing the mechanical intelligence-based incubator sealed blue light irradiation data management system in this application embodiment, those skilled in the art can understand the specific implementation methods and various variations of the electronic device in this embodiment based on the mechanical intelligence-based incubator sealed blue light irradiation data management system described in this application embodiment. Therefore, how the electronic device implements the method in this application embodiment will not be described in detail here. Any electronic device used by those skilled in the art in implementing the mechanical intelligence-based incubator sealed blue light irradiation data management system in this application embodiment falls within the scope of protection of this application.

[0128] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0129] The above description is merely a preferred embodiment of the present invention, and the scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for users of ordinary technical skills, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A data management system for a sealed blue light irradiation chamber based on mechanical intelligence, characterized in that: include: The state cluster data module collects multi-source heterogeneous data, corrects the clock deviation between the incubator and the preset biosensors through a spatiotemporal alignment protocol, and dynamically adjusts the sampling rate of multi-source heterogeneous data based on the infant physiological data in the time-aligned multi-source heterogeneous data, and outputs state cluster data packets with timestamps. The method for correcting the clock deviation between the incubator and the preset biosensor includes: Multi-source heterogeneous data is acquired in real time through biosensors installed inside and outside the incubator. This data includes incubator equipment data, infant physiological data, and external environmental data. A time reference is established between the incubator and the biosensors before the acquisition of multi-source heterogeneous data using a spatiotemporal alignment protocol. The spatiotemporal alignment protocol includes sending a global synchronization pulse before acquisition, recording the local timestamp of the synchronization signal received by the device, and calculating the initial deviation. During acquisition, a reference time point is inserted to monitor the offset of the local clock relative to the reference time and dynamically calculate drift compensation. For the collected multi-source heterogeneous data, the sampling timestamps are adjusted to the aligned time under a unified time base; the clock deviation value between the incubator and the biosensor is calculated through the protocol, and the sampling timestamps of the multi-source heterogeneous data are updated in real time, mapping the multi-source heterogeneous data sampled on the asynchronous timeline to the same global timeline; The method for obtaining the timestamped state cluster data packet includes: A threshold detection method is used to analyze infant physiological data in real time and extract the characteristics of the infant's current physiological state. Physiological response judgment criteria are set to determine whether the current physiological state is in an abnormal stage. The physiological response judgment criteria include preset heart rate threshold and bilirubin level threshold. When the infant's heart rate is detected to be greater than or equal to a preset heart rate threshold and the bilirubin level is detected to be greater than or equal to a preset bilirubin level threshold, the sampling rate of the multi-source heterogeneous data is automatically increased; when the infant's heart rate is less than a preset heart rate threshold and the bilirubin level is less than a preset bilirubin level threshold, the sampling rate of the multi-source heterogeneous data is automatically decreased; the multi-source heterogeneous data after dynamic sampling rate adjustment is aggregated to generate a state cluster data package with timestamps and synchronization correction. The hierarchical extraction module constructs a three-level feature processing pipeline based on state cluster data packages. This includes a basic layer that calculates basic statistics through a sliding window; a logic layer that generates a temperature-light effect index and a sealing safety factor based on the basic statistics and multi-source heterogeneous data; and a decision layer that establishes a treatment efficacy prediction matrix based on the temperature-light effect index and the sealing safety factor to predict the correlation trend between blue light irradiation and bilirubin metabolism. The methods for obtaining the basic statistics include: At the base layer, a sliding window analysis method is used to process multi-source heterogeneous data in real time based on state cluster data packets. The sliding window moves and samples on the time axis with a fixed step size according to a preset time length to extract multi-source heterogeneous data within a continuous time period. For the multi-source heterogeneous data within each sliding window, the data on incubator equipment, infant physiological data, and external environmental data contained in the multi-source heterogeneous data are grouped and processed separately. The basic statistics of each data type within the window are calculated, including mean, variance, maximum, minimum, range, quartiles, median, and percentiles. The time interval of sampling points for each data type is statistically analyzed, and the data integrity index and the trend of sampling frequency change are recorded. The methods for generating the thermo-optical effect index and the sealing safety factor include: The logic layer receives the basic statistics calculated by the base layer through a sliding window, and establishes a multi-dimensional fusion mechanism based on the correlation between the incubator's internal environment and the infant's physiological data. This includes using temperature-related statistics, combined with the infant's skin temperature and bilirubin metabolism indicators, to analyze the temperature-dependent effect of blue light irradiation on bilirubin decomposition; and comprehensively considering humidity and airflow velocity inside the incubator to assess the stability of the incubator's internal environment and its potential impact on the blue light effect. Based on a multi-dimensional fusion mechanism, a thermo-photonic effect index function is constructed to obtain the thermo-photonic effect index and sealing safety factor for the physiological effects of blue light irradiation therapy on infants in incubators. Through a dynamic factor adjustment mechanism, the humidity balance factor in the thermo-photonic effect index function is updated according to the real-time environmental humidity level to achieve adaptive adjustment of the thermo-photonic effect index. The method for predicting the association between blue light exposure and bilirubin metabolism includes: The decision-making layer receives the temperature and light effect index and sealing safety factor output by the logic layer, and combines them with dynamic monitoring data of bilirubin concentration in the infant's body to construct a treatment efficacy prediction matrix. The treatment efficacy prediction matrix is ​​based on the temperature and light effect index and sealing safety factor, forming a two-dimensional data set at different time points, representing the comprehensive state of the treatment environment. The relationship between the treatment efficacy prediction matrix and the change of bilirubin concentration over time was analyzed. Regression analysis was used to explore the influence of environmental factors of blue light irradiation on bilirubin metabolism. Based on the analysis results, the trend of bilirubin concentration change over the next n periods was predicted, thereby evaluating the efficacy of current blue light therapy. Based on the trend of bilirubin concentration changes over the next n periods, this provides clinicians with decision-making basis, guides the adjustment of incubator equipment data and external environmental data, supports feedback based on dynamic monitoring data of bilirubin concentration in infants, and dynamically adjusts treatment plans. The environmental-physiological coordination module employs a sliding mode control mechanism to design an inner-loop controller that adjusts the environmental parameters within the incubator. An outer-loop controller is constructed using a fuzzy control strategy. Based on the correlation trend between blue light irradiation and bilirubin metabolism, a blue light irradiation treatment plan is generated. Through control commands and conflict resolution mechanisms, blue light irradiation and sealing regulation are coordinated, and a control log is output. The method for generating the blue light irradiation treatment plan includes: Preset at time point The current state vector of the incubator environment parameters at that time is Preset at a specific time point The target state vector of the incubator environmental parameters at that time is Then at the point in time The error state vector between the current state vector and the target state vector of the incubator environmental parameters at that time is: The sliding surface can then be represented as: ;in, Indicates the sliding surface; Indicates the rate of change of error; Represents a positive definite diagonal matrix; A dynamic rule is defined by the sliding surface, when the error state vector satisfies... When the value is 0, the error state vector and its error rate of change will gradually approach zero. An inner loop controller is designed using a sliding mode control mechanism. A sliding surface is designed, and the input is controlled by the sliding mode control law to make the error state vector reach and remain on the sliding surface. The input change is smoothed by a saturation function to suppress the jitter caused by the sign function in the sliding mode control law, thereby adjusting the environmental parameters inside the warm chamber. The input variables of the outer loop controller include the current bilirubin concentration in the infant's body, the thermo-photonic effect index, and the sealing safety factor. The input variables are fuzzified using fuzzy membership functions to map continuous values ​​to fuzzy sets. Fuzzy inference rules are formulated based on the combination of input variables. Based on the correlation between bilirubin metabolism trend and blue light irradiation, and combined with the fuzzy inference rules of the input variables, a blue light irradiation adjustment scheme is generated. The fuzzy reasoning rules include: if the current bilirubin concentration in the infant's body is less than or equal to the preset third threshold for bilirubin concentration in the infant's body, and greater than the preset second threshold for bilirubin concentration in the infant's body, the temperature-light effect index is less than the preset optimal threshold for the temperature-light effect index, and the sealing safety factor is greater than or equal to the preset sealing safety factor threshold, then the blue light irradiation intensity and irradiation duration will be increased, and the temperature and humidity will be optimized. If the current bilirubin concentration in the infant's body is less than or equal to the preset second threshold for bilirubin concentration in the infant's body, and greater than the preset first threshold for bilirubin concentration in the infant's body; the thermo-photonic effect index is greater than or equal to the preset optimal threshold for thermo-photonic effect index; and the sealing safety factor is greater than or equal to the preset sealing safety factor threshold, then the output will maintain the blue light irradiation intensity and duration. If the current bilirubin concentration in the infant's body is less than or equal to the preset first threshold for bilirubin concentration in the infant's body, then reduce the intensity and duration of blue light exposure; if the sealing safety factor is less than the preset sealing safety factor threshold or the current bilirubin concentration in the infant's body is greater than the preset third threshold for bilirubin concentration in the infant's body, then suspend blue light therapy. The dynamic knowledge optimization module constructs a side knowledge network for strategy efficacy evaluation based on historical treatment cases, automatically eliminates outdated blue light irradiation treatment plans, receives control logs, analyzes and identifies key control variables, dynamically generates optimization strategy patch packages, and continuously updates the side knowledge network for strategy efficacy evaluation. The medical safety communication module performs risk assessments on control logs based on the strategy efficacy evaluation edge knowledge network; it designs a dual data encapsulation protocol, constructs an emergency priority channel based on risk assessment, allocates bandwidth for priority transmission, automatically generates a human-computer interaction task queue that conforms to medical procedures, and intercepts illegal instructions in real time.

2. The incubator-based sealed blue light irradiation data management system according to claim 1, characterized in that, The method for coordinating blue light irradiation and sealing adjustment includes: Based on the blue light irradiation treatment plan, blue light irradiation control instructions and sealing adjustment control instructions are generated, including specific parameter settings and execution time; a conflict resolution mechanism is set up to monitor the conflict between the blue light irradiation control instructions and the sealing adjustment control instructions in real time. Conflicts include the sealing state not meeting the preset requirements when blue light irradiation is turned on and the sealing adjustment instructions causing damage to the blue light irradiation effect. Based on the analysis of the current incubator environmental parameters, the root cause of the conflict is determined. The execution sequence, parameter amplitude, and duration of the control commands for blue light irradiation and sealing adjustment are dynamically adjusted. The issuance, adjustment, and execution processes of all blue light irradiation control commands and sealing adjustment control commands are recorded and output as control logs.

3. The incubator-based sealed blue light irradiation data management system according to claim 2, characterized in that, The method for constructing the edge knowledge network for evaluating the efficacy of the strategy includes: Historical treatment case data, including blue light irradiation parameters, environmental parameters, infant bilirubin concentration, and treatment effect feedback, were collected and preprocessed. Key variables were screened and multidimensional feature vectors were extracted. Key variables included blue light irradiation intensity, light irradiation duration, thermo-photothermal effect index, sealing safety factor, and infant bilirubin concentration. Key variables and efficacy indicators are defined as nodes in the strategy efficacy evaluation edge knowledge network, and the node types are divided into input nodes, environment nodes, and output nodes. The directed relations connecting different node types are defined as edges in the strategy efficacy evaluation edge knowledge network, thereby constructing the strategy efficacy evaluation edge knowledge network.

4. The incubator-based sealed blue light irradiation data management system according to claim 3, characterized in that, The method for prioritizing bandwidth allocation and transmission includes: The strategy efficacy evaluation edge knowledge network performs source tracing reasoning based on the received control logs to determine the correlation strength between the control strategy and efficacy indicators, thereby generating the risk level of the control instructions; a dual data encapsulation protocol is designed, which includes a basic data encapsulation layer and a risk label encapsulation layer. The basic data encapsulation layer is responsible for standard formatting, encryption, and verification of control logs according to the preset communication protocol; the risk label encapsulation layer carries additional emergency event identifiers, indicates the risk assessment level of the control logs based on fuzzy inference rules, and builds an emergency event priority channel in the risk label encapsulation layer for priority transmission through bandwidth allocation.

Images