Coaction of Physiological and Biomedical Device Models for Hyperferremia Therapy Modeling
With advances in both process intensification of biochemical systems and heightened computing potential, it is critical that numerical simulation tools develop in parallel to these advancements; the computational potential enables more complex and accurate representation of these systems. This is vital in the development of medical technologies, where developing therapeutics often poses great safety risk and material cost. Advances in the fields of computational biology and biomedical engineering yield evermore accurate simulations of biological systems and medical devices alike. The creation of usefully packaged and presented models in the form of Digital Twins provides insight into the nature of systems and their effects. A more valuable simulation tool, however, would integrate both models, linking the deviceâs operation with tangible therapeutic effects. Methods are explored for the ongoing unification of two mathematical models: 1) The model of a microfluidic device for removal of excess iron from blood developed by our research group at Oregon State University and 2) a human physiological model of the fate of iron in the body developed by Dr. Simon Mitchell, University of Manchester. The former model represents a microchannel device containing a hydrogel immobilized with DFO, a metal chelator, the simulation of which was developed in COMSOL, incorporating fluid flow, mass transport, and reaction kinetics. It is being developed for the treatment of hyperferremic diseases such as thalassemia. The latter model is a 4-compartment and 71-species-based model for the simulation of the physiological iron metabolism, developed in COPASI. The device chelates non transferrin-bound LIP, at early stages modeled as Iron Citrate. The connection of these models, and the consequent creation of a Digital Twin, ultimately balances both the deviceâs removal of blood iron and the bodyâs natural replenishment mechanisms. As such, the concentration flows of Iron Citrate are passed between the device and body relative to the blood flow of the catheter. The difference of these values is ultimately represented as the total export concentration from the modeled blood pool to the modeled device. This is accomplished through an independent Matlab class which houses an intermediate function called by both programs. The class moderates and restricts the progress of the two on an independent time scale while recording data and offering analysis methods during and post-run. Refinement of the linking algorithm is still necessary for proper connection of the models, however future implications may lead to a general extracorporeal blood-processing device-linking algorithm for other applications.