A medical health information processing system for blood test data

Abstract

The present application relates to the technical field of medical health information science, and discloses a medical health information processing system for blood test data, which comprises a time series data gating unit, a polarity shunt locking unit and a state closed-loop output unit.The time series data gating unit acquires test data and calculates a time interval, and calculates a unit time change rate of an index when the interval falls into a valid physiological response cycle.The polarity shunt locking unit locks a positive or negative driving data set as a reference source according to the positive or negative polarity of the change rate.The state closed-loop output unit generates a state signal when the deviation degree of the instantaneous cooperative slope with respect to the reference cooperative slope exceeds a relative steady-state threshold value.The present application solves the problem of data analysis distortion caused by physiological hysteresis effect and sampling discreteness through a bidirectional reference locking and time-effectiveness gating mechanism, and effectively improves the signal-to-noise ratio and accuracy of physiological state evaluation under non-equidistant sampling conditions.

Description

TECHNICAL FIELD

[0001] The present application relates to a medical health information processing system for blood test data, belonging to the technical field of medical health informatics. BACKGROUND

[0002] The current existing clinical laboratory information system generally adopts a discrete scalar processing mode, taking each blood physicochemical index of a single sampling as independent data, and comparing it with a single point of a static reference interval. Although this mode can identify explicit pathological values, it has limitations in dealing with complex physiological processes involving homeostatic coordination. The human physiological system follows a dynamic coordination logic among multiple indicators, and there is an inherent physiological conduction time lag and state dependence from the generation of stimulus signals to the response of effect indicators. The conventional data processing method is based on the linear synchronization assumption, which requires multiple indicators to be instantaneously correlated at the same sampling time, ignoring the objective existence of response delay and hysteresis effect in biological organisms.

[0003] Although various management systems have integrated data, the core logic is mostly limited to the process management and data storage level, lacking deep analysis of physiological data dynamics. For example, the Chinese invention patent with publication number CN104240002A discloses a hemodialysis management system and method, which realizes dialysis data acquisition, medical history record updating, and scheduling management automation through information means, improving clinical administrative efficiency. However, the data processing of such systems is still essentially static value transfer and simple threshold monitoring, and the alarm mechanism relies on single-time value overrun, lacking the ability to operate dynamic coupling relationships among multiple indicators. In the face of complex physiological regulation processes in the human body, such systems cannot identify phase differences caused by physiological conduction time lag, and cannot distinguish between the rate nonlinear differences caused by hysteresis effect during the physiological stress period and recovery period, making it impossible to filter out pseudo-decoupling noise caused by time domain mismatch.

[0004] Therefore, how to construct an analysis model that tolerates physiological time lag, adapts to sampling randomness, and represents nonlinear coordination relationship based on discrete non-equidistant blood test data has become a technical problem to be solved by the present application. SUMMARY

[0005] To solve the problems raised in the background art, the technical solution of the present application is as follows: a medical health information processing system for blood test data, comprising:

[0006] a time series data gating unit, configured to obtain test data of a subject sampled at adjacent times and calculate the time interval, calculate the unit time change rate of a first indicator and a second indicator respectively only when the time interval falls within a preset valid physiological response period, and trigger a bypass freeze logic when the time interval exceeds the valid physiological response period;

[0007] a polarity shunt locking unit, connected with the time series data gating unit, for identifying a positive or negative polarity of a unit time change rate of the first index, and establishing a one-way index path for historical data of the subject according to the polarity, locking a positive driving data set as a current reference source when the polarity is positive, and locking a negative driving data set as the current reference source when the polarity is negative;

[0008] a cooperative drift calculating unit, connected with the polarity shunt locking unit, for calculating a ratio of the unit time change rates of the first index and the second index in the same time period to obtain an instantaneous cooperative slope, and calculating a deviation of the instantaneous cooperative slope relative to a reference cooperative slope retrieved from the current reference source;

[0009] a state closed-loop output unit, connected with the cooperative drift calculating unit, for comparing the deviation with a preset relative steady state threshold, and generating a state signal representing decoupling of the physiological coupling relationship when the deviation exceeds the relative steady state threshold.

[0010] Preferably, the time series data gating unit generates a confidence weight coefficient negatively correlated with the time interval when calculating the unit time change rate; and the polarity shunt locking unit performs weighted processing on the instantaneous cooperative slope currently calculated by using the confidence weight coefficient when updating the historical data in the positive driving data set or the negative driving data set.

[0011] Preferably, the system further comprises a phase adaptive calibration unit arranged before the cooperative drift calculating unit; the phase adaptive calibration unit is configured to construct a misaligned comparison sequence introducing a time displacement variable for the first index and the second index within a preset historical time window, calculate a dispersion of the ratio of the unit time change rates in the misaligned comparison sequence under different time displacement variables, and select a best time displacement corresponding to the minimum dispersion; and the cooperative drift calculating unit calculates the instantaneous cooperative slope after performing time series alignment on the first index and the second index by using the best time displacement.

[0012] Preferably, the system further comprises a logical closed-loop verification unit; the logical closed-loop verification unit is configured to select a third index having a physiological correlation with both the first index and the second index, and calculate instantaneous cooperative slopes of the first index and the third index and the second index and the third index based on data in the same time period; the logical closed-loop verification unit calculates a closed-loop residual error between the instantaneous cooperative slope of the first index and the second index directly observed and the cooperative slope indirectly derived through the third index, and confirms the state signal valid when the closed-loop residual error is less than a preset consistency threshold.

[0013] Preferably, the system further comprises a vitality arbitration unit, which is configured to calculate a response module length of a change vector composed of the unit time change rate of the first index and the unit time change rate of the second index , and the calculation formula is: , wherein, a unit time change rate of the first index, a unit time change rate of the second index; the vigor arbitration unit will respond to the response length compared with a preset minimum vigor threshold, when the response length is lower than the minimum vigor threshold, a low vigor signal representing a sluggish physiological response is outputted and the generation of the state signal by the state closed-loop output unit is inhibited.

[0014] Preferably, it further comprises a base state compensation unit; the base state compensation unit stores a mapping relationship table of an absolute value interval of the first index and a response gain factor, for determining a corresponding current response gain factor according to the absolute value of the first index at the current time; the drift calculation unit corrects the reference cooperative slope by using the current response gain factor, and calculates the deviation degree of the instantaneous cooperative slope relative to the corrected reference cooperative slope.

[0015] Preferably, the state closed-loop output unit comprises a hierarchical mapping module; the hierarchical mapping module stores a corresponding relationship between the deviation degree interval and the risk level, for determining the current decoupling risk level according to the numerical value of the deviation degree, and generating a structured state data packet containing the risk level.

[0016] Preferably, the system further comprises an association rule configuration unit; the association rule configuration unit is used to store a plurality of groups of preset index pair combination lists, each group of index pair combination lists containing a first index identifier, a second index identifier, a corresponding effective physiological response period and a relative steady state threshold; the time series data gate control unit calls corresponding parameters from the association rule configuration unit for processing according to the index identifier in the received test data.

[0017] Preferably, the polarity shunt locking unit executes zero drift filtering logic when the polarity is zero; the zero drift filtering logic is used to maintain the reference source locking state of the previous time or reset the reference cooperative slope to the global historical mean value when the polarity of the continuous multiple sampling periods is zero.

[0018] Preferably, the system comprises a memory and a processor; the memory is used to store computer program instructions for executing the functions of the units of the system, and the processor is used to execute the computer program instructions stored in the memory to realize the data processing logic of the time series data gate control unit, the polarity shunt locking unit, the cooperative drift calculation unit and the state closed-loop output unit.

[0019] Compared with the prior art, the beneficial effects of the present application are:

[0020] 1. In the medical health information processing of blood test data, the phase self-adaptive calibration unit constructs a misalignment comparison sequence containing a time displacement variable, locks the optimal time displacement based on a dispersion minimum value search mechanism, solves the problem that the traditional linear comparison logic cannot adapt to the non-synchronous response engineering of human physiological signals, automatically identifies and compensates for the physiological conduction time lag between the first and second indicators using the statistical characteristics of the data itself, reconstructs the originally random fluctuation pseudo-decoupled data due to the misalignment of the sampling time points into physiological characteristics with strong correlation, and ensures accurate capture of the steady-state relationship between indicators when facing long-cycle or non-instantaneous response physiological processes using the existing computing resources of the system, eliminating false positive alarms caused by time phase difference.

[0021] 2. The logic closed loop verification unit of the system introduces a third indicator to construct a triangular transmission closure, uses the direct observation instantaneous collaborative slope and the indirect derived value to establish an endogenous consistency verification mechanism for single-chain measurement data, uses the inherent transmission axiom of physiological networks to distinguish between systematic physiological steady-state drift and single-point sensor incidental measurement noise through internal data logic mutual verification, confirms physiological changes when the direct path shows abnormalities while the logic closed loop remains self-consistent, and locates the specific signal distortion source based on the multi-path consistency voting results when the logic closed loop is broken, improves the self-diagnosis and error correction ability of discrete test data based on geometric logic triangular measurement means, and improves the effectiveness and alarm specificity of the system under complex clinical working conditions.

[0022] 3. The steady-state drift analysis unit of the system is configured with bidirectional reference locking logic, which respectively maintains positive and negative driving set independent reference collaborative slopes according to the positive and negative polarity of the first indicator unit time change rate, realizes adaptation to the hysteresis effect of human physiological regulation process, decouples the rate and path physical asymmetry of physiological stress rising phase and recovery falling phase, avoids misjudgment due to rate difference when processing recovery period data with a single average reference, and ensures the tracking of the real physiological state of the subject throughout the life cycle of the system, ensures the capture sensitivity during acute episode, and eliminates false abnormal signals caused by index falling lag during disease improvement stage. BRIEF DESCRIPTION OF DRAWINGS

[0023] Fig. 1 The data processing logic flowchart of the timing gate and bidirectional reference locking of the present application;

[0024] Fig. 2 The false alarm and missed alarm performance comparison chart of the full architecture collaborative processing and traditional strategy of the present application;

[0025] Fig. 3 The system overall architecture topology graph of the present application integrating multi-source acquisition and core operation. DETAILED DESCRIPTION

[0026] The technical solutions of the present application will be clearly and completely described below in combination with the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of the present application.

[0027] The present application provides a medical health information processing system for blood test data, which is composed of a time series data gating unit, a phase adaptive calibration unit, a polarity shunt locking unit, a base state compensation unit, a collaborative drift calculation unit, a logic closed loop verification unit, a vitality arbitration unit and a state closed loop output unit. Each unit sequentially processes the discrete sampling data of the subject to realize dynamic evaluation of the physiological coupling relationship. The time series data gating unit serves as the data inlet of the system and is used to establish a causal effectiveness filtering mechanism for the data. The unit obtains the test data of the subject at two adjacent sampling times, and calculates the time interval between the current sampling time and the previous sampling time. The unit reads the effective physiological response period preset in the correlation rule configuration unit for the first index and the second index. If the time interval exceeds the upper limit of the effective physiological response period or is lower than the lower limit, the unit triggers a bypass freeze logic to stop the calculation of the index change rate for the time period and maintain the reference state at the previous time to block the pollution of invalid data to the historical reference. If the time interval falls within the effective physiological response period, the unit calculates the unit time change rate of the first index and the second index respectively. The calculation value is the difference between the current sampling value and the previous sampling value divided by the time interval. At the same time, the unit generates a confidence weight coefficient negatively correlated with the time interval according to a preset inverse proportional decay function, which is used for weighted processing during subsequent reference update.

[0028] The phase adaptive calibration unit is configured before the collaborative drift calculation unit and is used to calibrate the physiological conduction time lag between the first index and the second index. Within a preset historical time window, the unit keeps the time series of the first index unchanged and constructs a misaligned comparison sequence of the second index introducing a time displacement variable. The system traverses the search range with a preset step size to search for different values of the time displacement variable. The unit calculates the dispersion of the unit time change rate ratio in the misaligned comparison sequence corresponding to each value of the time displacement variable, selects the optimal time displacement corresponding to the minimum dispersion, and sets the optimal time displacement as the physiological conduction time lag between the first index and the second index. ​​​​​​​​​​​​​​The first indicator at time is the rate of change per unit time. The second indicator, the rate of change per unit time, is time-series paired for use by subsequent units. The polarity shunt locking unit is used to adapt to the physiological hysteresis effect. This unit identifies the positive and negative polarities of the paired first indicator's rate of change per unit time. When the polarity is positive, the unit locks the pre-established positive driving dataset as the current reference source and retrieves the positive reference co-slope. When the polarity is negative, the unit locks the negative driving dataset as the current reference source and retrieves the negative reference co-slope. If the polarity is zero, the unit executes zero-drift filtering logic to maintain the locking state of the previous time step.

[0029] The base state compensation unit is used to perform nonlinear correction on the baseline. This unit obtains the absolute value of the first index at the current time and indexes the corresponding response gain factor in a pre-stored mapping table based on this value. The cooperative drift calculation unit calculates the ratio of the rate of change of the first index per unit time to the rate of change of the second index per unit time after pairing, obtaining the instantaneous cooperative slope. This unit utilizes the response gain factor. The baseline coordinated slope obtained from the polarity shunt locking unit is multiplied and corrected, and the deviation of the instantaneous coordinated slope from the corrected baseline coordinated slope is calculated. The logic closed-loop verification unit distinguishes between systemic physiological changes and measurement errors. This unit selects a third indicator to construct a logic triangle and calculates the closure residual between the directly observed instantaneous coordinated slope and the coordinated slope indirectly derived through the third indicator. The signal is confirmed as valid only when the closure residual is less than a preset consistency threshold. The vitality arbitration unit assesses the regulatory vitality of the system. This unit calculates the rate of change per unit time of the first indicator. Rate of change per unit time of the second indicator The response magnitude of the constructed change vector Its calculation formula is When the response modulus is long When the deviation is below the preset minimum vitality threshold, the unit outputs a low vitality signal that characterizes sluggish physiological response and suppresses the generation of decoupling signals. The state closed-loop output unit compares the finally confirmed deviation with the preset relative steady-state threshold. When the deviation exceeds the relative steady-state threshold, the unit generates a state signal that characterizes the decoupling of the physiological coupling relationship and determines the decoupling risk level according to the hierarchical mapping module.

[0030] Example 1: This example demonstrates the application of this concept in a long-term home follow-up scenario for patients with chronic kidney disease. In this scenario, blood sample collection typically relies on community hospitals or home testing devices, resulting in highly variable sampling frequencies. The time interval between two consecutive samples can fluctuate significantly from 24 hours to several weeks or even months, and is accompanied by pathological erythropoietin sluggishness. The time-series data gating unit reads the effective physiological response cycle of the first indicator, hemoglobin concentration, and the second indicator, hematocrit, and calculates the current sampling time. Compared with the previous sampling time time interval When this time interval When the effective physiological response cycle limit of 120 days (the average lifespan of red blood cells) is exceeded, the system determines that the physiological causal chain between the preceding and following data has been broken. It triggers bypass freezing logic to halt the co-slope calculation and maintain the baseline state from the previous moment. This ensures that only data generated within the effective physiological causal time window is input to subsequent units, providing a high signal-to-noise ratio input basis for subsequent processing. Under the premise of a valid time interval, the phase adaptive calibration unit introduces a time displacement variable into the time series of the second indicator within the historical time window. Multiple misalignment comparison sequences are constructed and the dispersion of the ratio of the rate of change per unit time is calculated. The optimal time displacement corresponding to the smallest dispersion is selected to align the time series of the first index with the time series of the second index after displacement correction, thereby eliminating the phase mismatch caused by physiological conduction delay.

[0031] To address the technical contradiction of the physical asymmetry between the rate of decline during the recovery phase and the rate of increase during the stress phase in physiological regulation, the polarity splitting locking unit identifies the positive and negative polarities of the rate of change per unit time of the first indicator. When an upward trend in hemoglobin concentration is detected, it automatically locks the positive driving dataset as the reference source and retrieves the positive reference co-slope. When a downward trend is detected, it switches to the negative driving dataset, thus adapting to the bidirectional nonlinear physiological hysteresis effect within a single system architecture. Simultaneously, the base state compensation unit obtains the absolute value of hemoglobin concentration and indexes the corresponding response gain factor according to the anemia severity interval, performing nonlinear correction on the reference co-slope to adapt to the nonlinear response characteristics in critical and severe conditions. The logic closed-loop verification unit introduces a third indicator, erythropoietin, to construct a logic triangle and calculates the closed residual between the directly observed value and the indirectly derived value to verify the systematicity of the abnormal signal. The vitality arbitration unit calculates the rate of change per unit time of the first indicator. Rate of change per unit time of the second indicator The response magnitude of the constructed change vector Its calculation formula is Only in response modulus The effective response of the system is confirmed when the minimum vitality threshold is exceeded, preventing false normal judgments in the state of immune exhaustion; the state closed-loop output unit finally generates a state signal representing whether the physiological coupling relationship is decoupled based on the comparison result of the deviation degree and the relative steady-state threshold. The present embodiment shows how the system converts absolute value monitoring of discrete scalar data into steady-state monitoring of dynamic ratio relationships between multi-dimensional indicators. Through the synergistic action of time domain gating, phase alignment, and polarity shunt mechanisms, the endogenous correlation logic between physiological parameters is reconstructed under non-continuous sampling conditions.

[0032] Example 2: Based on the aforementioned medical health information processing system, a simulation verification test for chronic kidney disease anemia management is constructed, aiming to quantitatively evaluate the performance of the system under complex working conditions such as non-continuous sampling, physiological time lag, and nonlinear response. The test data comes from the MIMIC-IV intensive care database and the outpatient follow-up records of cooperative medical institutions, a total of 300 patient samples diagnosed with chronic kidney disease combined with anemia are selected. To simulate the uncertainty of data in real clinical environment, Gaussian white noise with a signal-to-noise ratio of 20 dB is added to the original sampling data, and random disturbances simulating measurement errors of home self-test devices are introduced. The sampling time interval is set to follow a random distribution from 24 hours to 180 days, and the physiological indicators involved include hemoglobin concentration (Hb, unit g / L), hematocrit (Hct, unit %), and erythropoietin (EPO, unit mIU / mL). The test parameters are set as follows: the upper limit of the effective physiological response period is set to 120 days based on the physiological fact of the average lifespan of red blood cells; the minimum vitality threshold is set to 20% of the individual historical average response module length to distinguish between physiological silence and pathological exhaustion; the time displacement variable of phase adaptive calibration is set to search for a range of 0 to 14 days, covering the typical physiological delay window from EPO stimulation to reticulocyte response. The search range is set to 0 to 14 days, covering the typical physiological delay window from EPO stimulation to reticulocyte response.

[0033] Three parallel control groups are set to verify the contribution of different technical features: Control group A (baseline group): using traditional processing logic based only on single-point value comparison with static reference range; Control group B (partial missing group): using the basic architecture containing time sequence gating and collaborative calculation, but removing the phase adaptive calibration and polarity shunt locking unit, directly calculating the one-way collaborative slope based on the same physical time data; Test group (inventive sample group): using the full-architecture system containing time sequence gating, phase calibration, polarity shunt, base compensation, and logic verification; the test simulates the entire process of erythropoietin treatment for patients, including the stress rise period, the plateau period, and the recovery period after drug withdrawal. Each group system processes the input data with noise frame by frame and outputs abnormal state judgments; in the stress rise stage, the EPO concentration in the body of the subjects rises rapidly, and the Hb concentration shows a lagging rise. Control group B lacks phase calibration and directly compares ​High EPO and low Hb at the moment, the calculated synergistic slope deviates from the benchmark, leading to frequent decoupling false alarms, the test group locks the optimal time shift through the maximum covariance scanning (average 5.2 days), the stimulus and response signals are aligned in the time domain, maintaining the stability of the synergistic slope; in the recovery phase, the Hb concentration falls back with drug metabolism, and the falling rate is slower than the rising rate, and the control group B will judge the normal slow falling rate as abnormal due to the use of a single benchmark, and the test group uses the polarity shunt locking mechanism to automatically switch to the negative benchmark when detecting that the Hb change rate is negative, and correctly identifies the physiological process; Table 1 shows the key performance indicators of each group under simulated working conditions, the false alarm rate refers to the proportion of samples that the system issues an abnormal alarm when the physiological state is normal; the miss alarm rate refers to the proportion of samples that the system does not alarm when a real pathological change such as acute blood loss occurs.

[0034] Table 1: Comparison of system performance under different processing strategies

[0035]

[0036] The high false alarm rate of the control group A proves the limitations of static scalar processing in dynamic anomaly identification, and the control group B reduces the false alarm rate, but the false alarm rate increases because it cannot handle time lag and hysteresis effect, and the test group controls the false alarm rate at 4.2% while maintaining a low false alarm rate, proving the effectiveness of phase calibration and polarity shunt mechanism in complex dynamic scenarios; further gradient verification is carried out on the base value compensation mechanism, in the sub-sample group with Hb base values of 120g / L, 90g / L and 60g / L, the same dose of EPO stimulation is applied, without enabling the compensation mechanism, the severe anemia group causes the synergistic slope to be abnormally high due to nonlinear compensation response, after enabling the compensation, the system automatically calls the increasing response gain factor (1.0, 1.2, 1.8 respectively), the corrected synergistic slope returns to the normal range, verifying the adaptive ability of the system to nonlinear physiological state; the test data shows that through the synergistic effect of time sequence gating, phase calibration, polarity shunt and nonlinear compensation, the system can effectively suppress data noise and analysis deviation under non-continuous sampling and complex physiological regulation conditions, and realize accurate monitoring of physiological synergistic relationship.

[0037] Embodiment 3: This embodiment combines Figs. 1 to 3 , a medical health information processing system for blood test data, such as Fig. 1As shown, the data processing logic of the system starts from the discrete test data input module, acquires the data of the subject at two adjacent sampling times, and then transmits the data to the time series data gating unit, which is responsible for calculating the time interval, checking the physiological response period, and calculating the rate of change. The data flows to the phase adaptive calibration unit, which introduces a time shift variable and searches for the optimal shift for time alignment. Then the data enters the polarity shunt locking unit, which identifies the polarity of the rate of change to lock the positive or negative reference source. At the same time, the base state compensation unit determines the response gain factor according to the absolute value and corrects the reference slope. The data processed by the above-mentioned processing is collected into the cooperative drift calculation unit, which is used to calculate the instantaneous cooperative slope and its deviation from the corrected reference. The logic closed loop verification unit introduces the third index to construct a logic triangle and calculates the closed residual to verify the signal effectiveness. Finally, the state closed loop output unit compares the deviation with the threshold value to generate a physiological coupling state signal representing the decoupling of the physiological coupling relationship.

[0038] As shown in Fig. 2 The horizontal axis divides three experimental groups, namely the control group A using static threshold comparison, the control group B using no phase and polarity correction strategy, and the experimental group using the full architecture cooperative processing. The vertical axis represents the value size measured in percentage %. The legend clearly distinguishes between the false positive rate represented by the horizontal striped filled column and the false negative rate represented by the diagonal striped filled column. The statistical results show that although the false positive rate of control group A is low, the false negative rate is extremely high. Control group B reduces the false negative rate but increases the false positive rate. The experimental group maintains the lowest level in both false positive rate and false negative rate, showing the robustness of the system under complex working conditions.

[0039] As shown in Fig. 3 The data acquisition nodes on the left are distributed in communities and families, covering home self-test devices and community hospital test terminals, and are connected to the central processing server through encrypted data transmission. The server serves as the core operation environment, equipped with high-performance processors and high-speed memories at the hardware resource layer, and integrates time series gating and phase calibration components, polarity shunt and base compensation components, cooperative drift calculation components, and logic closed loop verification components at the system running environment layer. The server is connected to the data storage cluster persistence layer on the right through high-speed read-write channels, which includes a subject history database, an association rule configuration library, and a mapping relationship table storage. The risk level and state signal generated by the final processing are pushed to the interactive terminal nodes at the bottom for use by doctor diagnosis workstations and patient health management applications.

[0040] Embodiment 4: This embodiment is directed to the initialization process of the core parameters of the system when it is first deployed or adapted to a new disease, detailing the standardization calibration procedure for constructing the mapping table in the cardinality state compensation unit and determining the search range of the phase adaptive calibration unit, to eliminate the uncertainty of parameter setting. When establishing the mapping table of the nonlinear response gain factor for a specific physiological index pair, for example, the first index is hemoglobin concentration and the second index is reticulocyte count, the system accesses the historical retrospective data set of this disease, which contains full-range samples covering from normal physiological state to extreme pathological state. The system divides the absolute value domain of the first index into several continuous discrete intervals For each interval, the system selects all sampling points falling into the interval and calculates the arithmetic mean of the instantaneous synergistic slope corresponding to these sampling points The system selects the interval in which the first index is within the standard normal value range defined clinically as the reference interval, and marks the corresponding average synergistic slope as The system calculates the response gain factor of each interval according to the formula This calculation process normalizes the response intensity difference under different physiological cardinalities into a dimensionless gain coefficient, thereby constructing a static lookup table that points from the absolute value interval to the gain factor.

[0041] For the determination of the search range of the best time displacement in the phase adaptive calibration unit, the system performs offline scanning logic based on maximum cross-covariance. The system selects a group of high-frequency sampling historical data sequences with clear drug excitation or physiological stress events, calculates the unit time change rate sequence of the first index and the unit time change rate sequence of the second index The cross-covariance function between the two The system traverses the preset wide-range time lag window and identifies the time lag amount that makes the cross-covariance function reach the global maximum value This value represents the average physiological conduction time lag of the index pair in a statistical sense. The system locks the search range of the real-time monitoring stage to the neighborhood interval centered on where is the tolerance boundary set based on physiological variability. This procedure ensures that the search space of phase calibration covers the possibility of physiological fluctuations and excludes non-causal time noise interference, achieving deterministic initialization of system parameters.

[0042] ​​Example 5: This embodiment describes a pre-deployment calibration procedure performed to ensure the consistency and sensitivity of the abnormal state determination, before the system is officially deployed in a specific clinical environment. Baseline data cleaning and distribution verification steps are performed, the system accesses historical inventory test data in this deployment environment, and selects clinically confirmed control group samples with stable physiological state, with a sample size not less than the pre-set statistical minimum sample size. The system performs statistical analysis on the unit time change rate of the target index in the control group samples, calculates the mean and standard deviation . If the coefficient of variation calculated exceeds the pre-set dispersion threshold, it indicates that the data noise level in this environment is high or the sample homogeneity is insufficient, and the system automatically prompts to expand the sample size or perform more stringent outlier filtering. After verification, the system fixes the mean of the collaborative slope of the control group data as the initial global baseline collaborative slope in this environment, as a temporary replacement baseline before establishing the individualized baseline.

[0043] The individualized parameter fine-tuning and sensitivity confirmation step is performed, and for the pre-set test sample set with specific pathological characteristics, the system uses the initial global baseline for simulation monitoring. The system statistically analyzes the false positive rate and false negative rate in the simulation monitoring. If the false positive rate exceeds the pre-set safety boundary, the system automatically raises the relative steady state threshold until the false positive rate returns to the safety range. If the false negative rate is out of limit, the system lowers the minimum vitality threshold to improve the capture ability of weak signals. This fine-tuning process is iteratively performed using a binary search strategy until the optimal parameter combination that meets the pre-set clinical performance indicators is found. The final determined parameter set is locked and written into the system's configuration module as the running default value in this deployment environment, thereby ensuring the stability and effectiveness of the system in different clinical scenarios.

[0044] Example 6: This embodiment describes a standardized parameter calibration and system initialization procedure performed to ensure the stable operation of the system in different medical institutions and diverse clinical scenarios. For the setting of the effective physiological response period in the time series data gating unit, the system performs a statistical calibration step based on historical big data. The system accesses the historical diagnosis and treatment database of the target disease, extracts the longitudinal follow-up data of the target index such as hemoglobin concentration under the disease, and calculates the time interval distribution between adjacent two effective samplings. The system uses kernel density estimation method to fit the probability density function of the distribution. The system selects the upper limit of the time interval corresponding to the cumulative probability density function reaching the pre-set confidence level such as 95% as the initial value of the effective physiological response period . This value is logically verified by the physiological limit constant in the expert rule base, such as the average lifespan of red blood cells, and the smaller of the two is taken as the final Set value; for the determination of the minimum vitality threshold in the vitality arbitration unit, the system performs an adaptive calibration process based on the subject's own historical baseline, the system retraces all the test data of the subject in the past steady state period (i.e. the period when no acute pathological event occurs), calculates the length of the index change vector of each pair of sampling time, the system constructs the statistical histogram of the length sequence, and calculates the lower quartile , the system sets the minimum vitality threshold to a certain ratio such as 0.5 times of , to ensure that the threshold can filter out small fluctuations caused by measurement noise, and also retain low amplitude response signals with physiological significance, and the threshold is dynamically updated with the accumulation of new steady state data of the subject, to adapt to the long-term changes of individual physiological function.

[0045] For the setting of the consistency threshold in the logical closed loop verification unit, the system performs engineering calculation based on error propagation theory, the system calculates the total allowable error of a single index according to the nominal measurement error such as coefficient of variation and biological variability of the detection equipment used, the system uses the error propagation law to deduce the combined standard uncertainty of the synergy slope ratio according to the calculation formula of the synergy slope, the system sets the consistency threshold to the expanded uncertainty of the combined standard uncertainty (typically taking the coverage factor ), that is , this setting ensures that the verification threshold has a clear metrological basis, avoids over-tight or over-wide judgment due to empirical setting, establishes the initialization calibration procedure of the reference synergy slope, performs linear regression operation based on least squares method, the processor extracts the time series fragments of the first index and the second index during the physiological steady state from the subject's historical database, constructs the feature vector matrix, calculates the regression model determination coefficient , only when the determination coefficient is greater than , the regression model slope is locked as the initial reference synergy slope, if the determination coefficient is lower than the value, the sampling time window is automatically expanded to re-execute the regression operation until the linear correlation statistical requirement is met, to provide a statistically confident comparison reference for subsequent deviation calculation, the relative steady state threshold setting adopts a dynamic anchoring strategy based on environmental noise level, the processor calculates the instantaneous synergy slope standard deviation within the last sampling periods to represent the current system observation noise floor, and sets the relative steady state threshold to times of the standard deviation and the preset instrument inherent measurement error constant The sum is added to the state closed loop output unit to generate a decoupled state signal representing a substantial physiological change beyond the statistical fluctuation range, to exclude random disturbance caused by sampling discreteness or device noise, and to realize adaptive adjustment of alarm sensitivity according to the signal-to-noise ratio environment.

[0046] It is obvious for those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and the present application can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application.

[0047] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. A medical and health information processing system for blood test data, characterized in that, include: The association rule configuration unit is used to store multiple sets of preset indicator pair combination lists. Each set of indicator pair combination lists includes a first indicator identifier, a second indicator identifier, the corresponding effective physiological response cycle, and a relative steady-state threshold. The time series data gating unit calls the corresponding parameters from the association rule configuration unit for processing based on the indicator identifier in the received test data. The time-series data gating unit is used to acquire test data from two adjacent samplings of the subject and calculate the time interval between the current sampling time and the previous sampling time. Only when the time interval falls within the preset effective physiological response cycle, the unit time change rate of the first indicator and the second indicator are calculated respectively, and the bypass freeze logic is triggered when the time interval exceeds the effective physiological response cycle. Stop calculating the rate of change of the index for this time interval and maintain the baseline state at the previous sampling time. The polarity splitting and locking unit, connected to the time series data gating unit, is used to identify the positive and negative polarities of the unit time change rate of the first indicator, and to establish a one-way index path for the subject's historical data based on the polarity. When the polarity is positive, the positive driving dataset is locked as the current benchmark source, and when the polarity is negative, the negative driving dataset is locked as the current benchmark source. The collaborative drift calculation unit, connected to the polarity splitting locking unit, is used to calculate the ratio of the unit time change rate of the first index and the second index within the same time period to obtain the instantaneous collaborative slope, and to calculate the deviation of the instantaneous collaborative slope from the benchmark collaborative slope retrieved from the current benchmark source. The state closed-loop output unit, connected to the cooperative drift calculation unit, is used to compare the deviation with a preset relative steady-state threshold, and generate a state signal characterizing the decoupling of the physiological coupling relationship when the deviation exceeds the relative steady-state threshold. Furthermore, the system also includes a phase adaptive calibration unit configured before the cooperative drift calculation unit; the phase adaptive calibration unit is used to construct a misalignment comparison sequence for the first and second indicators within a preset historical time window, introduce time displacement variables, calculate the dispersion of the ratio of the rate of change per unit time in the misalignment comparison sequence under different time displacement variables, and select the optimal time displacement corresponding to the minimum dispersion; the cooperative drift calculation unit uses the optimal time displacement to perform time-series alignment of the first and second indicators and then calculates the instantaneous cooperative slope; The system also includes a logic closed-loop verification unit; the logic closed-loop verification unit is used to select a third indicator that is physiologically related to both the first indicator and the second indicator, and calculate the instantaneous coordinated slope of the first indicator and the third indicator and the instantaneous coordinated slope of the second indicator and the third indicator based on data from the same time period; the logic closed-loop verification unit calculates the closed residual between the directly observed instantaneous coordinated slope of the first indicator and the second indicator and the coordinated slope indirectly derived through the third indicator, and confirms the validity of the state signal when the closed residual is less than a preset consistency threshold.

2. The medical and health information processing system for blood test data according to claim 1, characterized in that, The time-series data gating unit generates confidence weight coefficients that are negatively correlated with the time interval when calculating the rate of change per unit time; the polarity splitting locking unit uses the confidence weight coefficients to weight the currently calculated instantaneous collaborative slope when updating historical data in the positive or negative driving dataset.

3. The medical and health information processing system for blood test data according to claim 1, characterized in that, It also includes a vitality arbitration unit, used to calculate the response magnitude of the change vector formed by the unit-time change rate of the first indicator and the unit-time change rate of the second indicator. The calculation formula is as follows: ,in, The rate of change per unit time for the first indicator. The rate of change per unit time for the second indicator; the vitality arbitration unit will respond to the modulus. The response modulus is compared with the preset minimum viability threshold. When the activity level is below the minimum activity threshold, a low activity signal representing sluggish physiological response is output, and the generation of state signals by the state closed-loop output unit is suppressed.

4. A medical and health information processing system for blood test data according to claim 1, characterized in that, It also includes a base state compensation unit; the base state compensation unit stores a mapping table between the absolute value range of the first index and the response gain factor, which is used to determine the corresponding current response gain factor based on the absolute value of the first index at the current time; the cooperative drift calculation unit uses the current response gain factor to correct the benchmark cooperative slope and calculates the deviation of the instantaneous cooperative slope from the corrected benchmark cooperative slope.

5. A medical and health information processing system for blood test data according to claim 1, characterized in that, The state closed-loop output unit includes a hierarchical mapping module; the hierarchical mapping module stores the correspondence between deviation ranges and risk levels, and is used to determine the current decoupling risk level based on the magnitude of the deviation, and generate a structured state data packet containing the risk level.

6. A medical and health information processing system for blood test data according to claim 1, characterized in that, The polarity shunt locking unit executes zero-drift filtering logic when the polarity is zero. The zero-drift filtering logic is used to maintain the reference source locking state of the previous moment or to reset the reference cooperative slope to the global historical average when the polarity is zero for multiple consecutive sampling periods.

7. A medical and health information processing system for blood test data according to claim 1, characterized in that, The system includes a memory and a processor; the memory is used to store computer program instructions that perform the functions of each unit of the system, and the processor is used to execute the computer program instructions stored in the memory to implement the data processing logic of each unit.

Images