The objective of precision medicine is to tailor interventions to an individual patient's unique characteristics. A key technology for this purpose involves medical digital twins, computational models of human biology that can be personalized and dynamically updated to incorporate patient-specific data. Certain aspects of human biology, such as the immune system, are not easily captured with physics-based models, such as differential equations. Instead, they are often multi-scale, stochastic and hybrid. This poses a challenge to existing control and optimization approaches that cannot be readily applied to such models. Recent advances in neural-network control methods hold promise in addressing complex control problems. However, the application of these approaches to biomedical systems is still in its early stages. This work employs dynamics-informed neural-network controllers as an alternative approach to control of medical digital twins. As a first use case, we focus on the control of agent-based models (ABMs), a versatile and increasingly common modelling platform in biomedicine. The effectiveness of the proposed neural-network control methods is illustrated and benchmarked against other methods with two widely used ABMs. To account for the inherent stochastic nature of the ABMs we aim to control, we quantify uncertainty in relevant model and control parameters.This article is part of the theme issue 'Uncertainty quantification for healthcare and biological systems (Part 1)'.
