Dynamic Hybrid-Hypergraph Model Based AI for Systems of Biological Systems.

This research paper focuses on using Hybrid Hypergraphs modeling to generate a dynamic model for systems of systems ( $SoSs$ ) in the context of medical applications. $SoSs$ are complex systems composed of multiple interacting components operating in a dynamic environment. The objective of this study is to create a dynamic graphical formalism that can represent and control the entirety of a $SoS$ using hybrid hypergraphs. We model the physical aspect of a $SoS$ by hypergraph nodes, and the managerial (multi-level) and communication aspects are modeled by hybrid hyperedges. The generated model has a temporal dimension that considers the dynamic characteristics of a $SoSs$ and takes into account the goals and parameters of the $SoSs$ as features vectors for each subsystem. In this paper, we apply the dynamic hybrid hypergraph model to classify a medical $SoS$ , which includes a robot performing brachytherapy on a bio-inspired phantom prostate, multiple markers for tracking lesions, and the skin as an entry point for the robot needle. We use the phantom prostate to mimic the movements of a real prostate during brachytherapy procedures and track the markers to investigate the impact of various factors, such as inflation or needle insertion, on the movement of the markers and the surrounding tissues. By analyzing this data, we aim to improve the accuracy and effectiveness of brachytherapy treatments. We demonstrate these concepts through the case study of a medical application, showcasing the use of Artificial Intelligence (AI) techniques for multiple tasks, including classification, prediction, optimization, and control in the context of the dynamic hybrid hypergraph model. The study presents a novel approach to modeling and understanding the complex interactions and dynamic behavior of systems of systems in various fields, particularly in medical applications.