A Variational Bayesian-Based Correntropy Cubature Kalman Filter for Drug Release Estimation Using a Second-Order...

Area of Science:

• Biomedical Engineering
• Pharmacology
• Data Science

Background:

• Targeted drug delivery systems, like ultrasound-triggered liposomes, offer potential for enhanced chemotherapy efficacy and reduced side effects.
• Traditional first-order models for liposome drug release may not fully capture complex release dynamics.
• Accurate modeling of drug release rates is crucial for optimizing targeted therapies.

Purpose of the Study:

• To model the drug release rate of targeted liposomes using a second-order discrete equation.
• To apply and compare different Kalman Filter variants for estimating drug release rates.
• To identify the most effective Kalman Filter variant for precise drug release tracking.

Main Methods:

• Developed a second-order discrete equation to model liposome drug release kinetics.
• Implemented and evaluated Extended Kalman Filter (EKF), Cubature Kalman Filter (CKF), and Variational Bayesian-Based Correntropy Cubature Kalman Filter (VBMCCKF).
• Assessed filter performance using Mean Squared Error (MSE) and Root Mean Squared Error (RMSE).

Main Results:

• The Variational Bayesian-Based Correntropy Cubature Kalman Filter (VBMCCKF) exhibited superior tracking performance.
• VBMCCKF effectively combined adaptive noise estimation with robust handling of abnormal measurements.
• VBMCCKF achieved the lowest MSE and RMSE, indicating the most accurate drug release rate estimation.

Conclusions:

• Second-order modeling and advanced Kalman Filters significantly improve the estimation of drug release rates from targeted liposomes.
• VBMCCKF offers a robust and accurate method for real-time monitoring of drug release in targeted chemotherapy.
• This approach holds promise for optimizing ultrasound-triggered liposomal drug delivery systems in clinical applications.