Infusion Control Device Using Feedback-Adjusted PK-PD Model
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current anesthesia drug administration methods lack precision in controlling drug concentrations in patient compartments, leading to potential overdoses or inadequate anesthesia, particularly due to non-individualized pharmacokinetic and pharmacodynamic models and the lack of accurate monitoring devices for real-time drug concentration measurement.
Innovation Solution
A control device using a pharmacokinetic-pharmacodynamic model to predict drug concentrations in multiple patient compartments, adjusting the model based on measured concentrations to recalibrate and determine optimal drug dosages, ensuring the target concentration is maintained within a defined threshold, utilizing breath measurement and EEG index values for precise administration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If conventional TCI algorithms are used to calculate drug dosage based on patient demographics, then drug administration can be automated, but the accuracy of predicting actual drug concentration in patient compartments remains low (±25% or higher)
Solution Approach 1:
The system continuously measures actual drug concentration in patient compartments (e.g., via breath analysis for lung compartment or EEG monitoring for brain compartment) and feeds this information back to the control device. The control device then adjusts the infusion dosage based on this feedback to achieve the target concentration, thereby resolving the contradiction between automation and precision by incorporating real-time measurement data into the automated control loop
Solution Approach 2:
The patent replaces the purely mathematical/pharmacokinetic modeling approach (mechanical calculation system) with a hybrid system that incorporates actual physiological measurements (breath analysis, EEG signals). This substitution of measurement-based control for model-based calculation significantly improves the accuracy of drug concentration prediction while maintaining automation
2Ease of operation
If generic pharmacokinetic models are used for drug administration, then the system is simple to operate, but the model does not account for individual patient physiology variations
Solution Approach 1:
The system performs preliminary measurements of actual drug concentration in patient compartments before finalizing the dosage calculation. By obtaining breath analysis data or EEG signals early in the process, the system can adjust the generic model parameters to better fit the individual patient's physiology without complicating the overall operation流程
Solution Approach 2:
The pharmacokinetic model is made dynamic by continuously adjusting its parameters based on real-time measurement data from the patient. The control device adapts the model to individual patient physiology variations as measurements are taken, allowing the system to remain simple to operate while becoming increasingly personalized to the specific patient's characteristics
3Loss of information
If breath analysis or EEG monitoring is implemented to measure actual drug concentration, then predictive information about drug levels can be obtained, but the device complexity increases
Solution Approach 1:
The patent uses breath analysis as an intermediary measurement approach - instead of directly measuring drug concentration in blood or brain tissue (which would require complex invasive procedures), the system measures drug concentration in breath, which serves as a non-invasive proxy for lung compartment concentration. This intermediary approach provides predictive information while avoiding the complexity of direct tissue measurement
Solution Approach 2:
The control device is designed to handle multiple types of measurements (breath analysis, EEG signals) and multiple compartments (lung, brain, plasma) through a unified control algorithm. This multi-functionality allows the system to obtain predictive information from various sources without proportionally increasing complexity, as the same control device processes different measurement types using the same pharmacokinetic modeling framework
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Method of operating a control device for controlling an infusion device A method of operating a control device (2) for controlling an infusion device (33) for administering a drug to a patient (P) comprises the steps of: providing a model (p) to predict a time-dependent drug concentration (Cplasma, Clung, Cbrain) in multiple compartments (A1-A5) of a patient (P); setting a target concentration value (CTbrain) to be achieved in at least one of the compartments (A1-A5) of the patient (P); determining a drug dosage (D1) to be administered to a first compartment (A1) of the multiple compartments (A1-A5) of the patient (P) using the model (p) such that the difference between the target concentration value (CTbrain) and a predicted steady-state drug concentration (Cplasma, Clung, Cbrain) in the at least one of the compartments (A1-A5) is smaller than a pre-defined threshold value (Ubrain); providing a control signal (S1) indicative of the drug dosage (D1) to an infusion device (33) for administering the drug dosage (D1) to the patient (P); obtaining a measurement value (M1, M2) indicating a measured drug concentration in a second compartment (A2, A3) of the multiple compartments (A1-A5) at a measurement time (t1, t2); adjusting the model (p) such that the model (p) predicts a drug concentration (Clung, Cbrain) in the second compartment (A2, A3) at the measurement time (t1, t2) which at least approximately matches the measured drug concentration in the second compartment (A2, A3); and determining a new drug dosage (D2, D3) to be administered into the first compartment (A1) of the patient (P) using the model (p) such that the difference between the target concentration value (CTbrain) and a predicted steady-state drug concentration (Cplasma, Clung, Cbrain) in the at least one of the compartments (A1, A2, A3) is smaller than the predefined threshold value (Ubrain). In this way a method is provided which allows for an improved (personalized and predictive) control of a drug administration procedure, in particular when administering an anaesthetic drug such as Propofol and/or an analgesic drug such as Remifentanil within a procedure.