Clinical trial whole-process quality risk intelligent management and control method and system and storage medium

Abstract

The application discloses a clinical test whole-process quality risk intelligent management and control method and system and a storage medium, relates to the medical and health information technology field, and aims to solve the problem that the prior art cannot quantitatively evaluate the freeze-thaw cumulative damage degree of biological preparations in situ and automatically grade and control, and comprises the following steps: S1, an integrated electrochemical impedance spectroscopy sensing chip is arranged in each drug packaging container of the biological preparations for clinical test; S2, the temperature change in the drug packaging container is continuously monitored through the temperature wake-up circuit, and when the detected temperature change rate exceeds a preset wake-up threshold, the sensing chip is woken up from a sleep state to a working state; the integrated electrochemical impedance spectroscopy sensing chip manufactured based on the CMOS process is directly embedded in each drug packaging container, and a technical leap from 'environmental indirect monitoring' to 'drug in-situ direct sensing' of the freeze-thaw cumulative damage of biological preparations is realized.

Description

Technical Field

[0001] This invention relates to the field of medical and health information technology, and more specifically, to a method, system, and storage medium for intelligent management and control of quality risks throughout the entire clinical trial process. Background Technology

[0002] In clinical trials, especially those involving monoclonal antibodies, fusion proteins, and cytokines, the entire logistics and storage chain from drug production to administration to subjects faces severe cold chain management challenges. These biologics are extremely sensitive to environmental temperatures; unexpected temperature deviations, particularly repeated freeze-thaw cycles, are major physical risk factors leading to denaturation, aggregation, and even inactivation of the active ingredients. In protein solutions, each freeze-thaw cycle induces the formation of an ice-liquid interface, solute freezing and concentration, and drastic local pH changes. These factors work together to disrupt the higher-order structure of proteins, generate sub-visible particles, and irreversibly lose biological activity.

[0003] Currently, the clinical trial industry commonly relies on electronic temperature recorders deployed in transport containers or storage refrigerators for drug cold chain monitoring. These devices continuously record ambient temperatures and issue alarms when temperatures exceed preset refrigeration or freezing ranges. However, this ambient temperature-based monitoring model has fundamental technical limitations. First, it can only provide a binary judgment of "whether the ambient temperature exceeds the limit," failing to delve into the physicochemical properties of the drug itself to assess the actual cumulative damage to drug quality caused by temperature deviation events. Even for the same over-temperature event, the duration, rate of temperature fluctuation, and historical freeze-thaw cycles all have significantly different impacts on drug quality, and current technology cannot quantify these differences. Second, traditional monitoring methods heavily depend on human intervention. When a temperature alarm occurs, quality management personnel must rely on experience, limited recorded data, and stability study data to manually determine whether the entire batch of drugs can continue to be used. This retrospective, subjective decision-making model is not only inefficient, but also prone to two types of erroneous decisions due to information asymmetry or human bias: mistakenly using drugs that have caused irreversible damage but whose macroscopic appearance has not changed, leading to safety risks; or discarding entire batches of expensive experimental drugs that have only experienced slight, reversible temperature fluctuations out of conservatism, resulting in huge waste of R&D costs and delays in the trial schedule.

[0004] Furthermore, existing electronic temperature tags or recorders have security vulnerabilities in terms of data integrity and reliability. Once temperature records are read, the data is easily tampered with or overwritten, failing to form a reliable and tamper-proof chain of quality evidence from drug production to final use. When facing stringent verification by regulatory agencies regarding the authenticity and traceability of quality control data for clinical trial drugs, the evidentiary capacity of existing technologies proves inadequate.

[0005] In summary, the clinical trial field urgently needs an intelligent management and control technology that can directly assess the cumulative damage caused by freeze-thaw cycles of drugs, automate graded quality risk decision-making, and ensure data immutability at the hardware level, in order to overcome the technical blind spots and decision-making risks of existing environmental temperature monitoring models. In view of this, we propose an intelligent management and control method, system, and storage medium for the entire process of clinical trial quality risk. Summary of the Invention

[0006] The purpose of this invention is to provide a method, system, and storage medium for intelligent management and control of quality risks throughout the entire clinical trial process, in order to solve the problem that existing technologies cannot quantitatively assess the degree of cumulative damage from freeze-thaw cycles of biological agents in situ and perform automated graded management.

[0007] To address the aforementioned technical problems, this invention provides the following technical solution: an intelligent management and control method for quality risks throughout the entire clinical trial process, comprising the following steps: S1: An integrated electrochemical impedance spectroscopy sensing chip is installed inside the drug packaging container of each clinical trial biological agent. The sensing chip is manufactured based on CMOS technology and includes a programmable sinusoidal excitation signal generator, a voltage-controlled current source, an on-chip transimpedance amplifier, a phase-sensitive detection unit, a temperature wake-up circuit, and an irreversible state latch unit. S2: The temperature wake-up circuit continuously monitors the temperature change inside the drug packaging container. When the detected temperature change rate exceeds the preset wake-up threshold, the sensing chip is woken up from the sleep state to the working state. S3: After wake-up, a swept-frequency sinusoidal excitation signal is generated by the programmable sinusoidal excitation signal generator, which is converted into an excitation current by the voltage-controlled current source and injected into the drug solution. The response current is converted into a response voltage by the on-chip transimpedance amplifier, and the phase-sensitive detection unit extracts the impedance magnitude and phase information reflecting the impedance characteristics of the drug solution from the response voltage. S4: The damage feature extraction unit built into the sensing chip extracts impedance spectrum feature parameters related to the degree of protein aggregation induced by freeze-thaw cycles based on the impedance modulus and phase information extracted after each freeze-thaw cycle, and calculates the cumulative damage score based on the feature parameters. S5: Compare the cumulative damage score with the preset graded damage threshold in the sensing chip. When the cumulative damage score reaches the critical damage threshold, trigger the irreversible state latch unit to irreversibly switch from the available state to the failed state, so as to characterize at the chip hardware level that the biological agent in the drug packaging container has exceeded the freeze-thaw cumulative damage tolerance limit. S6: When a clinical trial research institution receives a drug packaging container, it obtains the data of all freeze-thaw events recorded by the sensing chip and the current state latch value through a near-field communication reading terminal, and automatically executes drug release determination or triggers supplementary quality testing process based on the reading results.

[0008] This invention achieves a technological leap from "indirect environmental monitoring" to "direct in-situ sensing of drug molecules" by directly embedding an integrated electrochemical impedance spectroscopy (EIS) sensing chip, manufactured using CMOS technology, inside each drug packaging container. Utilizing a programmable sinusoidal excitation signal generator, transimpedance amplifier, and phase-sensitive detector integrated on the sensing chip, this invention enables in-situ swept-frequency impedance spectroscopy measurements inside the drug container, extracting key dielectric parameters directly related to protein aggregation and conformational changes, such as the rate of change of charge transfer resistance, the rate of change of low-frequency impedance modulus, and the relative shift rate of phase peak characteristic frequencies. Based on this, the damage feature extraction unit built into the sensing chip employs a multi-feature weighted fusion algorithm to uniformly calculate these multi-dimensional feature changes into a dimensionless cumulative damage score. This mechanism completely eliminates the drawbacks of traditional methods that rely solely on a single temperature threshold exceedance for coarse judgment, enabling precise quantification of the actual cumulative damage caused to specific drug molecules by each freeze-thaw event. This not only provides clinical trial sponsors and researchers with real-time, quantitative insights into the true quality status of drugs, but also effectively solves the key technical problem of being unable to assess the cumulative damage inside drugs in the background technology, making the previously "invisible and unmeasurable" freeze-thaw damage process of protein drugs clear and measurable.

[0009] Preferably, the programmable sinusoidal excitation signal generator in S3 generates a swept sinusoidal excitation signal that covers a preset low-frequency band and a high-frequency band. The phase-sensitive detection unit extracts the impedance magnitude and phase information of the low-frequency band and the impedance magnitude and phase information of the high-frequency band, respectively, so as to realize the distinguishing detection of different types of freeze-thaw damage mechanisms. The process by which the phase-sensitive detector unit extracts impedance magnitude and phase information is characterized by the following calculation formula: for frequency The sinusoidal excitation signal at point is given by the excitation current. The response voltage is Then the impedance magnitude and phase angle Obtained through phase-sensitive detection: ; In the formula, The amplitude of the excitation current; The magnitude of the response voltage; The phase delay angle of the response voltage relative to the excitation current; and These are the quadrature component and the in-phase component extracted from the response voltage after phase-sensitive detection, respectively.

[0010] Preferably, the on-chip transimpedance amplifier in S3 is equipped with an adaptive feedback resistor adjustment circuit. Before formally performing impedance spectroscopy measurement, the sensing chip first performs a solution impedance pre-detection to evaluate the baseline impedance range of the drug solution, and automatically adjusts the feedback resistor value of the on-chip transimpedance amplifier according to the pre-detection result, so that the impedance measurement dynamic range is adapted to different formulations of biological agents.

[0011] Preferably, the impedance spectrum characteristic parameters in S4 include at least two of the following: charge transfer resistance change rate, low-frequency impedance modulus change rate, and phase peak characteristic frequency relative offset rate; the damage feature extraction unit uses a multi-feature weighted fusion method to calculate the cumulative damage score, wherein the weight coefficient of each feature is pre-calibrated based on the freeze-thaw stability pre-validation data of this batch of biological agents and burned into the non-volatile memory of the sensing chip; The multi-feature weighted fusion calculation of the cumulative damage score is described above. The process is represented by the following calculation formula: Let the first The feature parameter vector extracted after one freeze-thaw cycle is: ; In the formula: For the first Rate of change of charge transfer resistance after one freeze-thaw cycle The charge transfer resistance value was measured under the initial healthy condition before freeze-thaw cycles. For the first The charge transfer resistance value measured after one freeze-thaw cycle; For the first Rate of change of low-frequency impedance modulus after one freeze-thaw cycle The impedance modulus value measured in the preset low-frequency band under the initial healthy state before freeze-thaw; For the first Impedance modulus measured in the same low-frequency band after one freeze-thaw cycle; For the first Relative offset rate of phase peak characteristic frequency after one freeze-thaw cycle. This represents the characteristic frequency corresponding to the phase peak in the Bode phase diagram under the initial healthy state before freeze-thaw; For the first The characteristic frequency corresponding to the phase peak in the Bode phase diagram after one freeze-thaw cycle; No. Single-cycle damage score for each freeze-thaw cycle: ; Cumulative damage score: In the formula, This is the weighting coefficient for the charge transfer resistance change rate term; This is the weighting coefficient for the rate of change of low-frequency impedance modulus; The weighting coefficient for the relative offset rate term of the phase peak characteristic frequency; It is a linear rectified activation function. ; This indicates the total number of freeze-thaw cycles.

[0012] Preferably, the graded damage threshold in S5 includes at least a low-risk threshold. The medium-risk threshold is High-risk threshold is Three levels, and meet the requirements Release decision Based on the cumulative damage score The segmented comparison result is determined when: Automatic release at time; when Release when the time is right and use the marking first; when When a supplementary quality inspection requirement is triggered; When the irreversible state latch unit is triggered, it switches to the failure state and automatically rejects the data.

[0013] Preferably, the sensing chip in S1 also has a built-in unique hardware identifier, which is generated based on a physically non-clonable function or a hardware serial number. During the near-field communication reading process, the unique hardware identifier is transmitted in conjunction with the data of each freeze-thaw event and the cumulative damage score, serving as a chain of evidence for the traceability audit of clinical trial drugs.

[0014] Preferably, the temperature wake-up circuit in S2 is also used to automatically trigger a complete impedance spectrum scan when the temperature inside the drug packaging container rises back to the refrigeration temperature range after the freeze-thaw cycle is completed, and to control the sensing chip to re-enter the sleep state after the scan is completed.

[0015] Preferably, the irreversible state latching unit in S5 is implemented by an antifuse unit or a one-time programmable memory unit. Once the state switch is triggered, the failure state cannot be reset or erased at the chip hardware level.

[0016] A comprehensive intelligent quality risk management system for clinical trials, including: A drug packaging container with multiple built-in integrated electrochemical impedance spectroscopy (EIS) sensing chips, each of which is manufactured based on CMOS technology and includes a programmable sinusoidal excitation signal generator, a voltage-controlled current source, an on-chip transimpedance amplifier, a phase-sensitive detection unit, a temperature wake-up circuit, a damage feature extraction unit, a threshold comparison unit, and an irreversible state latching unit. The sensing chip is configured to automatically perform impedance spectroscopy measurements upon detecting a freeze-thaw event, extract impedance spectral feature parameters related to the degree of protein aggregation induced by freeze-thaw, calculate a cumulative damage score, and trigger the irreversible state latching unit to irreversibly switch from a usable state to a failed state when the cumulative damage score reaches a critical damage threshold. Near-field communication reading terminals are deployed in various clinical trial research institutions to read data on all freeze-thaw events, cumulative damage scores, and current state latch values ​​recorded by the sensing chip inside the drug packaging container via near-field communication protocols. The central quality risk management platform is connected to the near-field communication reading terminals of various research institutions to aggregate drug freeze-thaw damage data from multiple centers, generate a cross-center freeze-thaw risk situational awareness view, and generate supplementary monitoring strategies for high-risk centers based on the situational awareness view. The central quality risk management platform also includes an automatic batch stability profile generation module and a data interface with the clinical trial electronic data acquisition system.

[0017] A computer-readable storage medium having stored thereon computer program instructions, which, when executed by a processor, implement the steps of a method for intelligent management and control of quality risks throughout the entire clinical trial process.

[0018] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention achieves a technological leap from "indirect environmental monitoring" to "direct in-situ sensing of drug molecules" by directly embedding an integrated electrochemical impedance spectroscopy (EIS) sensing chip manufactured using CMOS technology inside each drug packaging container. Utilizing a programmable sinusoidal excitation signal generator, transimpedance amplifier, and phase-sensitive detector integrated on the sensing chip, this invention can perform swept-frequency impedance spectroscopy measurements in-situ inside the drug container, extracting core dielectric characteristic parameters directly related to protein aggregation and conformational changes, such as the rate of change of charge transfer resistance, the rate of change of low-frequency impedance modulus, and the relative offset of phase peak characteristic frequencies. Based on this, the damage feature extraction unit built into the sensing chip employs a multi-feature weighted fusion algorithm to uniformly calculate these multi-dimensional feature changes into a dimensionless cumulative damage score. This mechanism completely eliminates the drawbacks of traditional methods that rely solely on a single temperature threshold exceedance for coarse judgment, enabling precise quantification of the actual cumulative damage caused to specific drug molecules by each freeze-thaw event. This not only provides clinical trial sponsors and researchers with real-time, quantitative insights into the true quality status of drugs, but also effectively solves the key technical problem of being unable to assess the cumulative damage inside drugs in the background technology, making the previously "invisible and unmeasurable" freeze-thaw damage process of protein drugs clear and measurable.

[0019] 2. This invention also constructs an automated and objective graded quality risk decision-making closed loop by using a preset graded damage threshold system and an irreversible state latching unit within the sensing chip. This precisely solves the secondary problems of drug misuse or unnecessary disposal caused by subjective human decision-making. Based on cumulative damage scores, the system can automatically classify drug packaging containers into multiple levels such as low risk, medium risk, and high risk, and execute corresponding differentiated control strategies. For drugs with only minor and reversible damage risks, the system automatically releases them, avoiding economic losses caused by conservative estimates. For drugs with moderate damage, the system automatically triggers supplementary quality testing instructions, requiring additional deterministic testing before administration, adding a scientific barrier to drug safety. For drugs whose cumulative damage scores reach the critical threshold, the sensing chip triggers the irreversible state latching unit to permanently mark them as "failed" at the physical hardware level, automatically rejecting them during the receiving process. This tiered management mechanism transforms the complex process of manual risk assessment into chip-driven automated decision-making, which not only significantly improves the efficiency and consistency of quality control, but also balances the safety of experimental drugs with the economics of research and development to the greatest extent, achieving dual protection of subjects and optimal allocation of resources.

[0020] 3. This invention further constructs a reliable, complete, and tamper-proof evidence chain for auditing drug freeze-thaw damage by synergistically combining the hardware-level unique identifier of the sensing chip with an irreversible state latching mechanism. This further addresses the deep-seated technical problem of traditional software-recorded data being easily questioned and tampered with under stringent clinical trial regulatory environments. The unique identifier generated internally by the sensing chip based on a physically non-cloning function or hardware serial number ensures the credible uniqueness of each packaging unit's identity. The timestamp, impedance spectrum characteristic parameters, and cumulative damage score of each freeze-thaw event are all bound and recorded internally to this unique identifier. More importantly, when the cumulative damage score reaches a critical threshold, the irreversible state latching unit performs a "fuse-out" or "burn-in" operation at the hardware level using an antifuse or one-time programmable storage technology. The resulting "failure" state is physically irreversible and indelible. This characteristic makes any attempt to conceal or tamper with the fact that the drug has undergone severe freeze-thaw damage impossible at the chip's physical level. Even the chip holder cannot restore the latched chip to a "usable" state. Therefore, throughout the entire lifecycle of a drug, from its manufacture to its final use in subjects or its disposal, all critical quality events are faithfully recorded and encapsulated in a hardware-level "black box," providing clinical trial sponsors with strong and irrefutable evidence to respond to regulatory audits, and greatly enhancing the integrity and compliance of the clinical trial data chain. Attached Figure Description

[0021] Figure 1 This is the overall flowchart of the intelligent management and control method for quality risks throughout the entire clinical trial process of the present invention; Figure 2 This is a detailed flowchart of the impedance spectrum measurement and cumulative damage score calculation of the present invention; Figure 3 This is a flowchart of the graded damage threshold determination and release decision-making process of the present invention; Figure 4 This is a flowchart of the cross-center freeze-thaw risk statistical assessment and high-risk institution identification process of the present invention; Figure 5 This is a framework diagram of the intelligent management and control system for quality risks throughout the entire clinical trial process of the present invention. Detailed Implementation

[0022] Example 1: As Figures 1 to 4 As shown, the present invention relates to an intelligent quality risk management method for the entire clinical trial process, used for real-time monitoring and risk management of freeze-thaw cumulative damage of biological agents used in clinical trials, comprising the following steps: S1: An integrated electrochemical impedance spectroscopy sensing chip is installed inside the drug packaging container of each clinical trial biological agent. The sensing chip is manufactured based on CMOS technology and includes a programmable sinusoidal excitation signal generator, a voltage-controlled current source, an on-chip transimpedance amplifier, a phase-sensitive detection unit, a temperature wake-up circuit, and an irreversible state latch unit. In an embodiment of the present invention, the sensing chip in S1 also has a built-in unique hardware identifier, which is generated based on a physically unclonable function or a hardware serial number. During the near-field communication reading process, the unique hardware identifier is transmitted in conjunction with the data of each freeze-thaw event and the cumulative damage score, serving as a chain of evidence for clinical trial drug traceability audit.

[0023] S2: The temperature wake-up circuit continuously monitors the temperature change inside the drug packaging container. When the detected temperature change rate exceeds the preset wake-up threshold, the sensing chip is woken up from the sleep state to the working state. In an embodiment of the present invention, the temperature wake-up circuit in S2 is also used to automatically trigger a complete impedance spectrum scan when the temperature inside the drug packaging container rises back to the refrigeration temperature range after the freeze-thaw cycle is completed, and to control the sensing chip to re-enter the sleep state after the scan is completed.

[0024] S3: After wake-up, a swept-frequency sinusoidal excitation signal is generated by the programmable sinusoidal excitation signal generator, which is converted into an excitation current by the voltage-controlled current source and injected into the drug solution. The response current is converted into a response voltage by the on-chip transimpedance amplifier, and the phase-sensitive detection unit extracts the impedance magnitude and phase information reflecting the impedance characteristics of the drug solution from the response voltage. In an embodiment of the present invention, the programmable sinusoidal excitation signal generator in S3 generates a swept sinusoidal excitation signal that covers a preset low-frequency band and a high-frequency band. The phase-sensitive detection unit extracts the impedance magnitude and phase information of the low-frequency band and the impedance magnitude and phase information of the high-frequency band, respectively, so as to realize the distinguishing detection of different types of freeze-thaw damage mechanisms.

[0025] The process by which the phase-sensitive detector unit extracts impedance magnitude and phase information is characterized by the following calculation formula: for frequency The sinusoidal excitation signal at point is given by the excitation current. The response voltage is Then the impedance magnitude and phase angle Obtained through phase-sensitive detection: ; In the formula: The frequency of the sinusoidal excitation signal is generated by a programmable sinusoidal excitation signal generator within a preset sweep frequency range, and the unit is Hertz (Hz). The instantaneous value of the sinusoidal excitation current, which varies with time, is injected into the drug solution by a voltage-controlled current source; The amplitude of the excitation current is a constant preset value, set by the reference current source inside the sensing chip; The instantaneous value of the response voltage, which varies with time, is obtained by converting the response current flowing through the drug solution using an on-chip transimpedance amplifier. The amplitude of the response voltage reflects the degree to which the drug solution impedes the current at that frequency; The phase response, which reflects the dielectric properties of the drug solution, is the phase delay angle of the response voltage relative to the excitation current. For frequency The impedance magnitude at a given point is the ratio of the response voltage amplitude to the excitation current amplitude, expressed in ohms (Ω). For frequency The phase angle at that point is calculated by the arctangent of the ratio of the imaginary part to the real part of the response voltage. The quadrature component, i.e., the imaginary part, is extracted from the phase-sensitive detector of the response voltage; The real part is the in-phase component extracted from the response voltage after phase-sensitive detection. Explanation of the operational logic: This formula describes the core computational process by which the phase-sensitive detector unit extracts impedance information from the drug solution. When the sensing chip is activated by the temperature wake-up circuit, a programmable sinusoidal excitation signal generator produces a sinusoidal AC signal of a specific frequency, which is converted into a constant amplitude excitation current by a voltage-controlled current source and injected into the drug solution. Because protein molecules in the drug solution undergo conformational changes and aggregation during freeze-thaw cycles, their ability to impede current changes, manifesting as amplitude attenuation and phase shift in the response voltage. The phase-sensitive detector unit orthogonally demodulates the response voltage signal and the excitation signal, extracting the in-phase component and the quadrature component, respectively, and then calculates the impedance magnitude. and phase angle The impedance modulus reflects the overall ability of a drug solution to impede current and is positively correlated with the number and size of protein aggregates in the solution; the phase angle reflects the dielectric relaxation characteristics of the drug solution and is closely related to the conformational state of protein molecules and the solvation layer structure.

[0026] Through the aforementioned computational logic, the sensing chip can perform impedance spectroscopy measurements in situ inside the drug packaging container, eliminating the need to remove the drug sample and transport it to laboratory instruments. The combined measurement of impedance magnitude and phase angle allows the sensing chip to simultaneously capture signal characteristics of freeze-thaw damage in two dimensions: protein aggregation and conformational changes. This provides a multi-dimensional raw data foundation for subsequent damage feature extraction and cumulative damage score calculation. Impedance information at different frequencies exhibits differentiated sensitivity to different types of freeze-thaw damage; the frequency sweep measurement strategy ensures the comprehensiveness and accuracy of damage detection.

[0027] In S3, the on-chip transimpedance amplifier is equipped with an adaptive feedback resistor adjustment circuit. Before performing impedance spectroscopy measurement, the sensing chip first performs a solution impedance pre-detection to assess the baseline impedance range of the drug solution, and automatically adjusts the feedback resistor value of the on-chip transimpedance amplifier according to the pre-detection result, so that the impedance measurement dynamic range is adapted to different formulations of biological agents.

[0028] S4: The damage feature extraction unit built into the sensing chip extracts impedance spectrum feature parameters related to the degree of protein aggregation induced by freeze-thaw cycles based on the impedance modulus and phase information extracted after each freeze-thaw cycle, and calculates the cumulative damage score based on the feature parameters. In an embodiment of the present invention, the impedance spectrum characteristic parameters in S4 include at least two of the following: charge transfer resistance change rate, low-frequency impedance modulus change rate, and phase peak characteristic frequency relative offset rate; the damage feature extraction unit calculates the cumulative damage score using a multi-feature weighted fusion method, wherein the weight coefficient of each feature is pre-calibrated based on the freeze-thaw stability pre-validation data of the batch of biological agents and burned into the non-volatile memory of the sensing chip.

[0029] The multi-feature weighted fusion calculation of the cumulative damage score is described above. The process is represented by the following calculation formula: Let the first The feature parameter vector extracted after the second freeze-thaw cycle is In the formula: For the first The rate of change of charge transfer resistance after one freeze-thaw cycle is dimensionless; among which... The charge transfer resistance value was measured in the initial healthy state before freeze-thaw, and was obtained by fitting the high-frequency semicircle diameter of the Nyquist plot. For the first The charge transfer resistance value measured after one freeze-thaw cycle; For the first The rate of change of low-frequency impedance modulus after one freeze-thaw cycle is dimensionless; among which... The impedance modulus value measured in the preset low-frequency band under the initial healthy state before freeze-thaw; For the first Impedance modulus measured in the same low-frequency band after one freeze-thaw cycle; For the first The relative shift rate of the phase peak characteristic frequency after one freeze-thaw cycle is dimensionless; among which... This represents the characteristic frequency corresponding to the phase peak in the Bode phase diagram under the initial healthy state before freeze-thaw; For the first The characteristic frequency corresponding to the phase peak in the Bode phase diagram after one freeze-thaw cycle; Then the first Single-cycle damage score for each freeze-thaw cycle Calculated by the following formula: ; Cumulative damage score It is obtained by summing up the individual damage scores from each test: ; In the formula: The weighting coefficient for the charge transfer resistance change rate term is calibrated based on the freeze-thaw stability pre-validation data of this batch of biological agents; The weighting coefficient for the low-frequency impedance modulus change rate term is calibrated based on the freeze-thaw stability pre-validation data of this batch of biological agents; The weighting coefficient for the relative offset rate of the phase peak characteristic frequency is calibrated based on the freeze-thaw stability pre-validation data of this batch of biological agents; The linear rectified activation function is defined as taking the larger of the input value and zero, i.e. This ensures that only damaging changes deviating from the healthy baseline are included in the score, which is achieved through an on-chip comparator and multiplexer. When the input is less than zero, the output is zero; otherwise, the original value is output. For the first The single-cycle damage score for each freeze-thaw cycle is a linear combination of three weighted feature terms; This indicates the total number of freeze-thaw cycles the drug packaging container has undergone since its production. Indicates experience Cumulative damage score after one freeze-thaw cycle; The weighting coefficient The batch-specific calibration process is illustrated below: First, the batch of biological agent samples is subjected to several controlled forced freeze-thaw cycles. After each freeze-thaw cycle, samples are taken for traditional bioanalytical tests, such as size exclusion chromatography, to determine the percentage loss of monomer content, which serves as the "gold standard" score for the degree of damage. Simultaneously, the charge transfer resistance change rate, low-frequency impedance modulus change rate, and characteristic frequency shift rate are measured using the sensing chip of this invention after the same freeze-thaw cycle. Then, using the three characteristic changes measured by the chip as independent variables and the percentage loss of monomer measured by chromatography as the dependent variable, on-chip regression analysis, such as the least squares method, is used to solve for the combination of weighting coefficients that minimizes the sum of squared errors between the chip score and the traditional score. Finally, these coefficients are burned into the non-volatile memory of the sensing chip.

[0030] Explanation of the computational logic: This set of formulas describes the complete computational chain by which the damage feature extraction unit extracts quantized feature parameters from the raw impedance spectrum data and calculates the cumulative damage score. First, the sensing chip extracts three feature parameters with clear physical meaning from the Nyquist and Bode plots, respectively. Charge transfer resistance. Obtained from the high-frequency semicircle fitting of the Nyquist curve, the increasing value indicates that protein aggregates are adsorbed and deposited on the electrode surface, hindering the charge transfer process. Low-frequency impedance modulus... Obtained from the low-frequency Bode plot, the disruption of cell membrane or protein structures caused by freeze-thaw cycles releases ions into the solution, reducing solution resistance, which manifests as a decrease in the low-frequency impedance modulus. (Phase peak characteristic frequency) Obtained from Bode phase diagrams, protein conformational changes and aggregation alter the dielectric relaxation time of the solution, leading to a shift in characteristic frequencies. Next, the sensing chip normalizes the changes in these three characteristic parameters relative to the pre-freeze-thaw healthy baseline. The rate-of-change form of the characteristic parameters eliminates the influence of inherent impedance differences between different formulations of biologics, ensuring cross-batch comparability of cumulative damage scores. Then, the sensing chip calculates the damage score for a single freeze-thaw cycle using a multi-feature weighted fusion method. The ReLU activation function ensures that only damaging changes deviating from the healthy baseline are included in the score. The negative sign of the low-frequency impedance modulus term reflects the negative correlation between a decrease in impedance modulus and an increase in damage severity. The three weighting coefficients are pre-calibrated based on pre-validation data of the freeze-thaw stability of the batch of drug, enabling the scoring model to adapt to differences in damage sensitivity among different biologics. Finally, the cumulative damage score is obtained by linearly summing the individual damage scores, reflecting the cumulative effect of freeze-thaw damage.

[0031] Through the aforementioned multi-feature weighted fusion computational logic, the cumulative damage score comprehensively reflects the cumulative damage effect of freeze-thaw cycles in three dimensions: protein aggregation degree, structural integrity, and conformational state. Normalization ensures the scoring system's batch-wide applicability, eliminating the influence of different formulations on the impedance baseline. The introduction of the ReLU activation function ensures the unidirectionality of the scoring calculation, preventing measurement noise or non-damaging fluctuations from being mistakenly included in the cumulative score. The batch-specific calibration mechanism for weighting coefficients allows the same sensing chip hardware platform to adapt to the differentiated freeze-thaw sensitivity of different types of biological agents, achieving personalized damage quantification. The linear accumulation calculation method is simple to implement at the chip hardware level, requiring only a small amount of register and adder resources, ensuring the achievement of the low-power design goal.

[0032] In a preferred embodiment of the present invention regarding feature frequency extraction, the sensing chip employs a combination strategy of adaptive frequency sweeping and local interpolation for extracting the phase peak feature frequency $f_{peak}$ from the Bode phase diagram. First, the sensing chip performs a pre-scan of sparse frequency points covering a preset frequency band to identify a coarse frequency range where the phase angle reaches its maximum value. Then, within this coarse frequency range, the density of sweep points is automatically increased for a second fine scan. For the discrete phase data point sequence obtained after the fine scan, the damage feature extraction unit uses the centroid method or parabolic interpolation method to estimate the peak frequency with sub-frequency point accuracy, thereby obtaining a high-resolution phase peak feature frequency while maintaining extremely low on-chip power consumption.

[0033] S5: Compare the cumulative damage score with the preset graded damage threshold in the sensing chip. When the cumulative damage score reaches the critical damage threshold, trigger the irreversible state latch unit to irreversibly switch from the available state to the failed state, so as to characterize at the chip hardware level that the biological agent in the drug packaging container has exceeded the freeze-thaw cumulative damage tolerance limit. In a preferred embodiment of the present invention regarding irreversible state latching, the irreversible state latching unit is independent of the digital logic main power domain of the sensing chip, or includes an internal power supply path that fuses after a state switch. Specifically, the read circuit of the anti-fuse unit or one-time programmable memory unit used to determine the latching state will preferentially read its stored state during chip power-on reset and load the latched value into a hardware flag register that cannot be cleared by software or an external reset signal. Even if the sensing chip undergoes a complete power failure or external reset operation, as soon as power is restored, the irreversible latching state will be immediately restored and presented as the basic state of the chip, thereby ensuring the absolute validity of the "failure" mark throughout its physical lifespan.

[0034] In an embodiment of the present invention, the graded damage threshold in S5 includes at least three levels: low-risk threshold, medium-risk threshold, and high-risk threshold; when the cumulative damage score is lower than the low-risk threshold, it is automatically released; when the cumulative damage score is between the low-risk threshold and the medium-risk threshold, it is released but marked as priority use; when the cumulative damage score is between the medium-risk threshold and the high-risk threshold, a supplementary quality inspection requirement is triggered; when the cumulative damage score reaches or exceeds the high-risk threshold, the irreversible state latching unit is triggered to switch to the failure state and is automatically rejected.

[0035] The determination logic for the graded damage threshold is characterized by the following piecewise function: Let the low-risk threshold be The medium-risk threshold is The high-risk threshold is And satisfy For the cumulative damage score defined above Release decision Defined as: ; In the formula: The cumulative damage score, as defined above, is obtained by summing the individual damage scores from each freeze-thaw cycle. The low-risk threshold is the first preset threshold, determined by the highest cumulative score in the freeze-thaw stability data of this batch of biological agents in which no significant damage was observed. The threshold is the medium risk threshold, and the second preset threshold is determined based on the cumulative score of the batch of biological agents when measurable damage begins to appear in the freeze-thaw stability data but is still within an acceptable range. The high-risk threshold and the third preset threshold are determined based on the critical cumulative score of the damage level reaching an unacceptable level in the freeze-thaw stability data of this batch of biological agents. This represents the release decision function, which takes the cumulative damage score as input and outputs the corresponding control action instruction; Explanation of the operational logic: This piecewise function describes the decision logic of the sensing chip in performing graded release judgment based on the cumulative damage score. When the drug packaging container is transported to the clinical trial research institution, the near-field communication reading terminal acquires the cumulative damage score stored in the sensing chip. The system automatically executes the corresponding control action according to the range in which the score value falls. The low-risk range corresponds to a cumulative damage score below a first preset threshold, indicating that the drug packaging container has experienced very few freeze-thaw events or mild freeze-thaw conditions, without causing significant damage. The drug quality is basically consistent with that at the time of manufacture and can be directly released for subject administration. The medium-low risk range corresponds to a cumulative damage score between the first and second preset thresholds, indicating that the drug has suffered a certain degree of freeze-thaw damage but has not yet reached a level that affects safety and efficacy. The system releases the drug but marks it as priority use, prompting the clinical research institution to prioritize the use of this packaging unit for subjects about to receive the drug, in order to shorten its storage time before use. The medium-to-high risk range corresponds to a cumulative damage score between the second and third preset thresholds, indicating that the freeze-thaw damage has reached a level requiring further evaluation. The system automatically triggers a supplementary quality control instruction, requiring the research institution to conduct additional quality control tests on the packaging unit before drug administration to confirm that the drug still meets the usage standards. The high risk range corresponds to a cumulative damage score reaching or exceeding the third preset threshold, indicating that the freeze-thaw damage has accumulated to an unacceptable level. The sensing chip triggers the irreversible state latch unit to switch from a usable state to a failed state. Based on this, the system automatically outputs a rejection instruction, preventing the packaging unit from entering the clinical trial use phase.

[0036] Through the aforementioned grading and judgment logic, the system achieves a precise match between freeze-thaw damage risk and control intensity. The grading threshold system avoids the drug waste caused by the "one-size-fits-all" disposal strategy in traditional cold chain monitoring. Drugs that have only experienced minor freeze-thaw damage but are still safe to use are conditionally released, significantly reducing drug costs in clinical trials. Simultaneously, drugs with moderate damage trigger supplementary testing requirements, providing additional data support for quality decisions. For drugs reaching the critical damage threshold, irreversible hardware-level locking physically blocks their further circulation, eliminating the risk of substandard drugs entering the subjects' bodies due to human error or subjective judgment bias. The entire decision-making process is directly driven by the cumulative damage score of the sensing chip, requiring no manual intervention and ensuring consistency and traceability in release judgments.

[0037] In an embodiment of the present invention, the irreversible state latching unit in S5 is implemented by an antifuse unit or a one-time programmable memory unit. Once the state switch is triggered, the failure state cannot be reset or erased at the chip hardware level.

[0038] S6: When a clinical trial research institution receives a drug packaging container, it obtains the data of all freeze-thaw events recorded by the sensing chip and the current state latch value through a near-field communication reading terminal, and automatically executes drug release determination or triggers supplementary quality testing process based on the reading results.

[0039] In an embodiment of the present invention, the near-field communication reading terminal in step S6, while acquiring the chip-recorded data, establishes a secure session with the sensing chip through a hardware-level encrypted communication protocol to ensure the integrity and immutability of the read data.

[0040] Example 2: Figure 5 As shown, the intelligent quality risk management system for the entire clinical trial process is used for real-time monitoring and risk management of cumulative freeze-thaw damage to biological agents used in clinical trials, including: A drug packaging container with multiple built-in integrated electrochemical impedance spectroscopy (EIS) sensing chips, each of which is fabricated based on a CMOS process and includes a programmable sinusoidal excitation signal generator, a voltage-controlled current source, an on-chip transimpedance amplifier, a phase-sensitive detection unit, a temperature wake-up circuit, a damage feature extraction unit, a threshold comparison unit, and an irreversible state latch unit. The sensing chips are configured to automatically perform impedance spectroscopy measurements upon detecting a freeze-thaw event, extract impedance spectral feature parameters related to the degree of protein aggregation induced by freeze-thaw, calculate a cumulative damage score, and trigger the irreversible state latch unit to irreversibly switch from a usable state to a failed state when the cumulative damage score reaches a critical damage threshold. As another embodiment of the present invention, the sensing chip is further provided with a data compression engine and a hardware encryption engine. The data compression engine is used to compress the original impedance spectrum data into a feature parameter vector and then store it in the on-chip non-volatile memory. The hardware encryption engine is used to encrypt the stored and transmitted data.

[0041] In a preferred embodiment of the impedance spectrum feature extraction of the present invention, the damage feature extraction unit of the sensing chip uses an on-chip geometric simplification algorithm to replace the iterative nonlinear fitting for fitting the high-frequency semicircle diameter of the Nyquist plot, in order to adapt to the limited computing resources of the CMOS chip. Specifically, the algorithm first searches for the frequency point with the largest absolute value of the imaginary part of the impedance in the high-frequency range, and extracts the real part of the impedance value of this point and at least one neighboring frequency point on each side; then, based on the geometric properties of the arc, using the impedance coordinates corresponding to the extracted at least three frequency points, the coordinates of the center of the arc are directly calculated by solving a system of linear equations, and then the charge transfer resistance value $R_{ct}$ is directly analyzed from the distance from the center to the intersection of the real axis, without the need for multiple iterative convergence calculations.

[0042] Near-field communication reading terminals are deployed in various clinical trial research institutions to read data on all freeze-thaw events, cumulative damage scores, and current state latch values ​​recorded by the sensing chip inside the drug packaging container via near-field communication protocols. The central quality risk management platform is connected to the near-field communication reading terminals of various research institutions to aggregate drug freeze-thaw damage data from multiple centers, generate a cross-center freeze-thaw risk situational awareness view, and generate supplementary monitoring strategies for high-risk centers based on the situational awareness view.

[0043] As another embodiment of the present invention, the central quality risk control platform further includes a batch stability profile automatic generation module. The batch stability profile automatic generation module is used to automatically summarize the freeze-thaw damage data recorded by the sensing chips in the packaging containers of all drugs in the same batch after the clinical trial is completed, and generate a batch freeze-thaw stability profile that conforms to the preset regulatory format.

[0044] As another embodiment of the present invention, the central quality risk management platform also includes a data interface with the clinical trial electronic data acquisition system, which is used to automatically fill the electronic case report form with the freeze-thaw damage status of each drug packaging container as the drug quality status field in the subject's medication record.

[0045] Example 3: Figures 1 to 4 As shown, a computer-readable storage medium stores computer program instructions thereon. When these computer program instructions are executed by a processor, they implement the following intelligent quality risk management steps throughout the entire clinical trial process: By using a near-field communication reading terminal deployed in a clinical trial research institution, reading data is acquired from an integrated electrochemical impedance spectroscopy sensing chip embedded in the drug packaging container. The reading data includes timestamps of each freeze-thaw event recorded by the sensing chip, impedance spectrum characteristic parameters, cumulative damage scores, and the current state value of the irreversible state latch unit. The sensing chip is manufactured based on CMOS technology and includes a programmable sinusoidal excitation signal generator, a voltage-controlled current source, an on-chip transimpedance amplifier, a phase-sensitive detection unit, a temperature wake-up circuit, and an irreversible state latch unit. The system parses the read data and automatically outputs a drug release judgment instruction based on the comparison result of the cumulative damage score and the preset graded damage threshold. This includes: outputting an automatic release instruction when the cumulative damage score is lower than the low-risk threshold; outputting a supplementary quality inspection instruction when the cumulative damage score is between the medium-risk threshold and the high-risk threshold; and outputting an automatic rejection instruction when the cumulative damage score reaches or exceeds the high-risk threshold or the irreversible state latching unit is in a failed state. The results of each drug release decision, along with the freeze-thaw damage data recorded by the sensing chip, are linked and stored in the clinical trial quality database; After the clinical trial is completed, freeze-thaw damage data of all drug packaging containers in the same batch are compiled, and a batch-level freeze-thaw stability report is automatically generated.

[0046] In another embodiment of the present invention, the computer program instructions, when executed by the processor, further perform the following steps: Cross-center statistical analysis is performed on the drug freeze-thaw damage data uploaded by various research institutions to identify high-risk research institutions whose freeze-thaw event incidence rate exceeds a preset threshold, and supplementary on-site monitoring instructions are sent to the terminal devices corresponding to the high-risk research institutions.

[0047] The process of cross-center statistical analysis and identification of high-risk research institutions is assisted by the following calculation formula: Let the first The total number of drug packaging containers managed by each research institution during the statistical period was: The number of containers that experienced at least one freeze-thaw event was The freeze-thaw event incidence rate of this research institution Defined as: ; when Exceeding the preset risk threshold At that time, the research institution will be marked as a high-risk research institution, and a corresponding supplementary monitoring instruction will be generated.

[0048] In the formula: This is used as an index number for research institutions to distinguish participating institutions in multicenter clinical trials. Indicates the first The total number of drug packaging containers managed by each research institution within the statistical period; Indicates the first The number of containers managed by a research institution within a statistical period whose sensing chips recorded at least one freeze-thaw event; Indicates the first The freeze-thaw event rate for each research institution is the ratio of the number of containers that experienced freeze-thaw events to the total number of containers, with a value ranging from zero to one. The preset risk thresholds are pre-configured by the central quality risk management platform based on the clinical trial protocol and the drug's freeze-thaw stability characteristics; Explanation of the calculation logic: This formula describes the core calculation process of the central quality risk management platform in conducting cross-center freeze-thaw risk statistical assessments for various clinical trial research institutions. During the execution of clinical trials, each research institution uploads freeze-thaw event data recorded by the sensing chips inside all managed drug packaging containers to the central platform via near-field communication (NFC) reading terminals. The central platform aggregates and statistically analyzes the data from each institution, calculating the proportion of drug packaging containers that experienced at least one freeze-thaw event within the statistical period to the total number of containers managed by that institution. This proportion reflects the research institution's ability to control freeze-thaw risks during drug receipt, storage, and distribution. A higher proportion indicates systemic problems in the institution's cold chain management or operational procedure implementation, requiring close attention from the sponsor.

[0049] Through the aforementioned cross-center statistical assessment logic, the central quality risk management platform can automatically identify research institutions with abnormally high freeze-thaw risk from massive amounts of sensor chip data in multi-center clinical trials. The freeze-thaw event incidence rate indicator eliminates the impact of differences in the number of managed containers among research institutions on risk assessment, ensuring the fairness of cross-center comparisons. Supplementary monitoring instructions automatically generated based on this assessment result enable sponsors to accurately allocate limited on-site monitoring resources to the centers with the highest risk, achieving the core objective of optimizing monitoring resource allocation in risk-based quality management. This logic elevates damage data from individual packaging units collected by sensor chips to the level of overall risk awareness in clinical trials, completing a closed loop from micro-level perception to macro-level control.

[0050] In another embodiment of the present invention, the computer program instructions, when executed by the processor, further perform the following steps: Data on the freeze-thaw damage status of each drug packaging container before each administration is transmitted to the clinical trial electronic data acquisition system via a data interface, and used as the drug quality status field for that administration to automatically fill the electronic case report form.

[0051] The embodiments disclosed in this invention are preferred embodiments, but are not limited thereto. Those skilled in the art can easily understand the spirit of this invention based on the above embodiments and make different extensions and variations, but as long as they do not depart from the spirit of this invention, they are all within the protection scope of this invention.

Claims

1. A method for intelligent management and control of quality risks in the whole process of a clinical trial, characterized in that, Includes the following steps: S1: An integrated electrochemical impedance spectroscopy sensing chip is installed inside the drug packaging container of each clinical trial biological agent. The sensing chip is manufactured based on CMOS technology and includes a programmable sinusoidal excitation signal generator, a voltage-controlled current source, an on-chip transimpedance amplifier, a phase-sensitive detection unit, a temperature wake-up circuit, and an irreversible state latch unit. S2: The temperature wake-up circuit continuously monitors the temperature change inside the drug packaging container. When the detected temperature change rate exceeds the preset wake-up threshold, the sensing chip is woken up from the sleep state to the working state. S3: After wake-up, a swept-frequency sinusoidal excitation signal is generated by the programmable sinusoidal excitation signal generator, which is converted into an excitation current by the voltage-controlled current source and injected into the drug solution. The response current is converted into a response voltage by the on-chip transimpedance amplifier, and the phase-sensitive detection unit extracts the impedance magnitude and phase information reflecting the impedance characteristics of the drug solution from the response voltage. S4: The damage feature extraction unit built into the sensing chip extracts impedance spectrum feature parameters related to the degree of protein aggregation induced by freeze-thaw cycles based on the impedance modulus and phase information extracted after each freeze-thaw cycle, and calculates the cumulative damage score based on the feature parameters. S5: Compare the cumulative damage score with the preset graded damage threshold in the sensing chip. When the cumulative damage score reaches the critical damage threshold, trigger the irreversible state latch unit to irreversibly switch from the available state to the failed state, so as to characterize at the chip hardware level that the biological agent in the drug packaging container has exceeded the freeze-thaw cumulative damage tolerance limit. S6: When a clinical trial research institution receives a drug packaging container, it obtains the data of all freeze-thaw events recorded by the sensing chip and the current state latch value through a near-field communication reading terminal, and automatically executes drug release determination or triggers supplementary quality testing process based on the reading results. 2.The clinical trial whole-process quality risk intelligent management and control method according to claim 1, characterized in that, The programmable sinusoidal excitation signal generator in S3 generates a swept sinusoidal excitation signal that covers a preset low-frequency band and a high-frequency band. The phase-sensitive detection unit extracts the impedance magnitude and phase information of the low-frequency band and the impedance magnitude and phase information of the high-frequency band, respectively, so as to realize the distinguishing detection of different types of freeze-thaw damage mechanisms. The process by which the phase-sensitive detector unit extracts impedance magnitude and phase information is characterized by the following calculation formula: for frequency The sinusoidal excitation signal at point is given by the excitation current. The response voltage is Then the impedance magnitude and phase angle Obtained through phase-sensitive detection: ； In the formula, The amplitude of the excitation current; The magnitude of the response voltage; The phase delay angle of the response voltage relative to the excitation current; and These are the quadrature component and the in-phase component extracted from the response voltage after phase-sensitive detection, respectively.

3. The intelligent management and control method for quality risks throughout the entire clinical trial process according to claim 1, characterized in that, The on-chip transimpedance amplifier in S3 is equipped with an adaptive feedback resistor adjustment circuit. Before performing impedance spectroscopy measurement, the sensing chip first performs a solution impedance pre-detection to assess the baseline impedance range of the drug solution, and automatically adjusts the feedback resistor value of the on-chip transimpedance amplifier according to the pre-detection result, so that the impedance measurement dynamic range is adapted to different formulations of biological agents. 4.The clinical trial whole-process quality risk intelligent management and control method according to claim 1, characterized in that, The impedance spectrum characteristic parameters in S4 include at least two of the following: charge transfer resistance change rate, low-frequency impedance modulus change rate, and phase peak characteristic frequency relative offset rate; the damage feature extraction unit uses a multi-feature weighted fusion method to calculate the cumulative damage score, wherein the weight coefficient of each feature is pre-calibrated based on the freeze-thaw stability pre-validation data of this batch of biological agents and burned into the non-volatile memory of the sensing chip; The multi-feature weighted fusion calculation cumulative damage score The process is characterized by the following calculation formula: The first The characteristic parameter vector extracted after the second freeze-thaw cycle is: ； In the formula: For the first Rate of change of charge transfer resistance after one freeze-thaw cycle The charge transfer resistance value was measured under the initial healthy condition before freeze-thaw cycles. For the first The charge transfer resistance value measured after one freeze-thaw cycle; For the first Rate of change of low-frequency impedance modulus after one freeze-thaw cycle The impedance modulus value measured in the preset low-frequency band under the initial healthy state before freeze-thaw; For the first Impedance modulus measured in the same low-frequency band after one freeze-thaw cycle; For the first Relative offset rate of phase peak characteristic frequency after one freeze-thaw cycle. This represents the characteristic frequency corresponding to the phase peak in the Bode phase diagram under the initial healthy state before freeze-thaw; For the first The characteristic frequency corresponding to the phase peak in the Bode phase diagram after one freeze-thaw cycle; First Single injury score after 1 freeze-thaw cycle: ； Cumulative damage score: In the formula, This is the weighting coefficient for the charge transfer resistance change rate term; This is the weighting coefficient for the rate of change of low-frequency impedance modulus; The weighting coefficient for the relative offset rate term of the phase peak characteristic frequency; It is a linear rectified activation function. ; This indicates the total number of freeze-thaw cycles.

5. The clinical trial whole-process quality risk intelligent management and control method according to claim 4, characterized in that, The graded damage threshold in S5 includes at least a low-risk threshold. The medium-risk threshold is High-risk threshold is Three levels, and meet the requirements Release decision Based on the cumulative damage score The segmented comparison result is determined when: Automatic release at time; when Release when the time is right and use the marking first; when When a supplementary quality inspection requirement is triggered; When the irreversible state latch unit is triggered, it switches to the failure state and automatically rejects the data. 6.The clinical trial whole-process quality risk intelligent management and control method according to claim 1, characterized in that, The sensing chip in S1 also has a built-in unique hardware identifier, which is generated based on a physically unclonable function or a hardware serial number. During the near-field communication reading process, the unique hardware identifier is transmitted in conjunction with the data of each freeze-thaw event and the cumulative damage score, serving as a chain of evidence for the traceability audit of clinical trial drugs.

7. The clinical trial whole-process quality risk intelligent management and control method according to claim 1, characterized in that, The temperature wake-up circuit in S2 is also used to automatically trigger a complete impedance spectrum scan when the temperature inside the drug packaging container rises back to the refrigeration temperature range after the freeze-thaw cycle is completed, and to control the sensing chip to re-enter the sleep state after the scan is completed. 8.The clinical trial whole-process quality risk intelligent management and control method according to claim 1, characterized in that, The irreversible state latch unit in S5 is implemented through an antifuse unit or a one-time programmable memory unit. Once the state switch is triggered, the failure state cannot be reset or erased at the chip hardware level.

9. A clinical trial end-to-end quality risk intelligent management and control system, used to execute the clinical trial end-to-end quality risk intelligent management and control method as described in any one of claims 1-8, characterized in that, include: A drug packaging container with multiple built-in integrated electrochemical impedance spectroscopy (EIS) sensing chips, each of which is manufactured based on CMOS technology and includes a programmable sinusoidal excitation signal generator, a voltage-controlled current source, an on-chip transimpedance amplifier, a phase-sensitive detection unit, a temperature wake-up circuit, a damage feature extraction unit, a threshold comparison unit, and an irreversible state latching unit. The sensing chip is configured to automatically perform impedance spectroscopy measurements upon detecting a freeze-thaw event, extract impedance spectral feature parameters related to the degree of protein aggregation induced by freeze-thaw, calculate a cumulative damage score, and trigger the irreversible state latching unit to irreversibly switch from a usable state to a failed state when the cumulative damage score reaches a critical damage threshold. Near-field communication reading terminals are deployed in various clinical trial research institutions to read data on all freeze-thaw events, cumulative damage scores, and current state latch values ​​recorded by the sensing chip inside the drug packaging container via near-field communication protocols. The central quality risk management platform is connected to the near-field communication reading terminals of various research institutions to aggregate drug freeze-thaw damage data from multiple centers, generate a cross-center freeze-thaw risk situational awareness view, and generate supplementary monitoring strategies for high-risk centers based on the situational awareness view. The central quality risk management platform also includes an automatic batch stability profile generation module and a data interface with the clinical trial electronic data acquisition system.

10. A computer readable storage medium having stored thereon computer program instructions, characterized in that, When the computer program instructions are executed by the processor, they implement the steps of the intelligent management and control method for quality risks throughout the entire clinical trial process as described in any one of claims 1 to 8.

Images