On Drug Delivery System Parameter Optimization via Semantic Information Theory

We investigate the application of semantic information theory to drug delivery systems (DDS) within the molecular communication (MC) framework. To operationalise this, we observe a DDS as a molecular concentration-based channel. Semantic information is defined as the amount of information required for a DDS to achieve its therapeutic goal in a dynamic environment. We derive it by introducing interventions, defined as modifications to DDS parameters, a viability function, and system-environment correlations quantified via the channel capacity. Here, the viability function represents DDS performance based on a drug dose-response relationship. Our model considers a system capable of inducing functional changes in a receiver cancer cell, where exceeding critical DDS parameter values can significantly reduce performance or cost-effectiveness. By analysing the MC-based DDS model through a semantic information perspective, we examine how correlations between the internalised particle concentration and the released particle concentration evolve under interventions. The final catalogue of results provides a quantitative basis for DDS design and optimisation, offering a method to determine optimal DDS parameter values under constraints such as chemical budget, desired effect and accuracy. Thus, the proposed framework can serve as a novel tool for guiding DDS design and optimisation.