Search

CN-122005147-A - Dynamic parameter intelligent regulation method and system for administration of mouse pulse pump

CN122005147ACN 122005147 ACN122005147 ACN 122005147ACN-122005147-A

Abstract

The invention discloses a dynamic parameter intelligent regulation method and a system for dosing a mouse pulse pump, wherein the method comprises the steps of collecting physiological feedback signals of a dosing mouse in real time through a multichannel biosensor, denoising and extracting features of the physiological feedback signals to generate structured real-time physiological feature data, inputting the real-time physiological feature data into a pre-trained individualized drug response model, outputting a predicted physiological response track and corresponding confidence assessment, dynamically determining the combination of the current optimal pulse frequency and dosing agent parameters through a multi-objective optimization algorithm based on the physiological response track and the confidence assessment, generating a real-time control instruction according to the combination of the optimal pulse frequency and the dosing agent parameters, and issuing the real-time control instruction to a pulse pump executing mechanism. By utilizing the embodiment of the invention, the individuation and self-adaptive accurate adjustment of the administration parameters of the mouse pulse pump can be realized, and the scientificity, the controllability and the reliability of experimental results of the administration process are improved.

Inventors

  • WANG CHAOHU
  • LIU YI
  • CAO YANAN

Assignees

  • 南方医科大学南方医院

Dates

Publication Date
20260512
Application Date
20260313

Claims (10)

  1. 1. A method for intelligently adjusting dynamic parameters for administration of a pulse pump to a mouse, the method comprising: The method comprises the steps of collecting physiological feedback signals of a dosing mouse in real time through a multichannel biosensor, denoising and extracting features of the physiological feedback signals, and generating structured real-time physiological feature data; inputting the real-time physiological characteristic data into a pre-trained personalized medicine response model, and outputting a predicted physiological response track and corresponding confidence assessment; Dynamically determining a current optimal pulse frequency and dosing parameter combination through a multi-objective optimization algorithm based on the physiological response track and the confidence assessment; and generating a real-time control instruction according to the combination of the optimal pulse frequency and the dosing parameter and sending the real-time control instruction to a pulse pump executing mechanism so as to realize dynamic intelligent regulation of dosing of the pulse pump.
  2. 2. The method of claim 1, wherein the collecting physiological feedback signals of the administered mice in real time by the multichannel biosensor, and denoising and feature extracting the physiological feedback signals, generating structured real-time physiological feature data, comprises: Synchronously collecting multi-mode physiological original signals of a dosing mouse, which at least comprise electrocardiosignals, electroencephalogram signals, blood oxygen saturation and body temperature data, through a multi-channel biosensor array; preprocessing a multi-mode physiological original signal, removing power frequency interference and myoelectric noise by adopting a wavelet transformation algorithm, removing motion artifacts by Kalman filtering, and generating a denoised physiological signal; feature extraction is carried out on the denoised physiological signals, RR interval variability of electrocardiosignals, power spectrum features of the electrocardiosignals, blood oxygen saturation trend slope and body temperature change rate are calculated, and multidimensional physiological feature vectors are generated; And carrying out standardized processing and time sequence alignment on the multidimensional physiological feature vector, constructing a structured data matrix containing time stamps and feature identifiers, and generating structured real-time physiological feature data.
  3. 3. The method of claim 2, wherein inputting the real-time physiological characteristic data into a pre-trained personalized medicine response model, outputting a predicted physiological response trajectory and a corresponding confidence assessment, comprises: Loading a pre-trained individualized drug response model, wherein the model is obtained through historical drug administration data and physiological feedback signal training based on a long-short-term memory network architecture; inputting the structured real-time physiological characteristic data into a loaded individualized medication response model, obtaining physiological state predicted values of a plurality of time points in the future through forward propagation calculation, and generating a preliminary physiological response predicted sequence; Based on an uncertainty estimation module of an individuation drug response model, calculating confidence intervals of predicted values of all time points by adopting a Monte Carlo dropouout method, and generating a prediction confidence evaluation matrix; and integrating the preliminary physiological response prediction sequence and the prediction confidence evaluation matrix, constructing a continuous physiological response track with a confidence interval, and generating a complete physiological response track prediction result.
  4. 4. The method of claim 3, wherein the dynamically determining, based on the physiological response trajectory and the confidence assessment, a current optimal pulse frequency and dosing parameter combination by a multi-objective optimization algorithm comprises: constructing a multi-objective optimization function, wherein the multi-objective optimization function comprises maximizing a drug effect index, minimizing side effect risks and minimizing drug accumulation, and generating a corresponding multi-objective optimization problem; constructing a constraint condition set comprising a physiological parameter safety range, a maximum administration rate and a minimum administration interval based on a physiological response track prediction result and a confidence coefficient evaluation matrix, and generating an optimization constraint condition; Solving a multi-objective optimization problem by adopting a non-dominant sorting genetic algorithm, searching a pareto optimal solution set under an optimization constraint condition, and generating a candidate parameter combination set; And selecting the pulse frequency with the highest comprehensive score from the candidate parameter combination set and the dosing agent combination according to the real-time optimization decision criterion to obtain the optimal pulse frequency and dosing agent parameter combination.
  5. 5. The method of claim 4, wherein generating real-time control instructions based on the optimal pulse frequency and dosing parameters and issuing the real-time control instructions to a pulse pump actuator to achieve dynamic intelligent regulation of pulse pump dosing comprises: Analyzing the optimal pulse frequency and the combination of the administration dosage parameters, calculating specific control parameters of the pulse pump, including pulse width, pulse interval and single administration volume, and generating a pulse pump control parameter set; converting the pulse pump control parameter set into a control instruction format which can be identified by the pulse pump, wherein the control instruction format comprises a start instruction, a parameter setting instruction and a starting instruction, and generating a standardized control instruction sequence; Transmitting the standardized control instruction sequence to the pulse pump executing mechanism through a wireless communication protocol, and receiving confirmation feedback of the executing mechanism; The execution state of the pulse pump and the feedback of the multichannel biosensor are monitored in real time, a closed-loop control loop is formed, the subsequent control instruction is dynamically adjusted, and the dynamic intelligent adjustment of the drug delivery of the pulse pump is realized.
  6. 6. A dynamic parameter intelligent regulation system for administration of a pulse pump in a mouse, the system comprising: the acquisition module is used for acquiring physiological feedback signals of the administration mice in real time through the multichannel biosensor, denoising and extracting features of the physiological feedback signals, and generating structured real-time physiological feature data; the input module is used for inputting the real-time physiological characteristic data into a pre-trained personalized medicine response model and outputting a predicted physiological response track and corresponding confidence assessment; The determining module is used for dynamically determining the current optimal pulse frequency and the combination of the administration dosage parameters through a multi-objective optimization algorithm based on the physiological response track and the confidence evaluation; And the adjusting module is used for generating a real-time control instruction according to the combination of the optimal pulse frequency and the dosing parameter and sending the real-time control instruction to the pulse pump executing mechanism so as to realize dynamic intelligent adjustment of dosing of the pulse pump.
  7. 7. The system according to claim 6, wherein the acquisition module is specifically configured to: Synchronously collecting multi-mode physiological original signals of a dosing mouse, which at least comprise electrocardiosignals, electroencephalogram signals, blood oxygen saturation and body temperature data, through a multi-channel biosensor array; preprocessing a multi-mode physiological original signal, removing power frequency interference and myoelectric noise by adopting a wavelet transformation algorithm, removing motion artifacts by Kalman filtering, and generating a denoised physiological signal; feature extraction is carried out on the denoised physiological signals, RR interval variability of electrocardiosignals, power spectrum features of the electrocardiosignals, blood oxygen saturation trend slope and body temperature change rate are calculated, and multidimensional physiological feature vectors are generated; And carrying out standardized processing and time sequence alignment on the multidimensional physiological feature vector, constructing a structured data matrix containing time stamps and feature identifiers, and generating structured real-time physiological feature data.
  8. 8. The system according to claim 7, wherein the input module is specifically configured to: Loading a pre-trained individualized drug response model, wherein the model is obtained through historical drug administration data and physiological feedback signal training based on a long-short-term memory network architecture; inputting the structured real-time physiological characteristic data into a loaded individualized medication response model, obtaining physiological state predicted values of a plurality of time points in the future through forward propagation calculation, and generating a preliminary physiological response predicted sequence; Based on an uncertainty estimation module of an individuation drug response model, calculating confidence intervals of predicted values of all time points by adopting a Monte Carlo dropouout method, and generating a prediction confidence evaluation matrix; and integrating the preliminary physiological response prediction sequence and the prediction confidence evaluation matrix, constructing a continuous physiological response track with a confidence interval, and generating a complete physiological response track prediction result.
  9. 9. A storage medium having a computer program stored therein, wherein the computer program is arranged to perform the method of any of claims 1-5 when run.
  10. 10. An electronic device comprising a memory and a processor, wherein the memory has stored therein a computer program, the processor being arranged to run the computer program to perform the method of any of claims 1-5.

Description

Dynamic parameter intelligent regulation method and system for administration of mouse pulse pump Technical Field The invention belongs to the technical field of bioengineering, and particularly relates to a dynamic parameter intelligent regulation method and system for administration of a mouse pulse pump. Background In the fields of biomedical and pharmacological research, pulse pump administration is a key technology for precisely controlling the exposure dose and time of a drug. At present, a conventional pulse pump control system mostly adopts preset fixed parameters or a simple feedback regulation mode, and is difficult to adapt to dynamic changes of experimental mice caused by individual differences and physiological state fluctuation in the process of drug administration. The prior art lacks the capability of real-time fusion analysis and intelligent prediction of multichannel physiological signals, so that the adjustment of the administration parameters is lagged, and the true individuation and self-adaptive precise regulation and control cannot be realized. This limits the reliability and efficiency of complex pharmacodynamic, toxicological studies, and also makes it difficult to meet the research requirements of the leading-edge dynamic dosing paradigm. Disclosure of Invention The invention aims to provide a dynamic parameter intelligent regulation method and a dynamic parameter intelligent regulation system for administration of a mouse pulse pump, so as to solve the defects in the prior art, realize individuation and self-adaptive accurate regulation of administration parameters of the mouse pulse pump, and improve the scientificity, controllability and experimental result reliability of an administration process. One embodiment of the application provides a method for intelligently adjusting dynamic parameters for administration of a pulse pump to a mouse, the method comprising: The method comprises the steps of collecting physiological feedback signals of a dosing mouse in real time through a multichannel biosensor, denoising and extracting features of the physiological feedback signals, and generating structured real-time physiological feature data; inputting the real-time physiological characteristic data into a pre-trained personalized medicine response model, and outputting a predicted physiological response track and corresponding confidence assessment; Dynamically determining a current optimal pulse frequency and dosing parameter combination through a multi-objective optimization algorithm based on the physiological response track and the confidence assessment; and generating a real-time control instruction according to the combination of the optimal pulse frequency and the dosing parameter and sending the real-time control instruction to a pulse pump executing mechanism so as to realize dynamic intelligent regulation of dosing of the pulse pump. Yet another embodiment of the present application provides a dynamic parameter intelligent regulation system for administration of a pulse pump to a mouse, the system comprising: the acquisition module is used for acquiring physiological feedback signals of the administration mice in real time through the multichannel biosensor, denoising and extracting features of the physiological feedback signals, and generating structured real-time physiological feature data; the input module is used for inputting the real-time physiological characteristic data into a pre-trained personalized medicine response model and outputting a predicted physiological response track and corresponding confidence assessment; The determining module is used for dynamically determining the current optimal pulse frequency and the combination of the administration dosage parameters through a multi-objective optimization algorithm based on the physiological response track and the confidence evaluation; And the adjusting module is used for generating a real-time control instruction according to the combination of the optimal pulse frequency and the dosing parameter and sending the real-time control instruction to the pulse pump executing mechanism so as to realize dynamic intelligent adjustment of dosing of the pulse pump. A further embodiment of the application provides a storage medium having a computer program stored therein, wherein the computer program is arranged to perform the method of any of the preceding claims when run. Yet another embodiment of the application provides an electronic device comprising a memory having a computer program stored therein and a processor configured to run the computer program to perform the method recited in any of the preceding claims. Compared with the prior art, the intelligent dynamic parameter adjusting method for the administration of the mouse pulse pump can realize individuation and self-adaptive accurate adjustment of the administration parameters of the mouse pulse pump, and improves the scientificity, controllability and reliab