A Patient-Specific Digital Twin for Personalized Assessment of Mechanical Ventilation in Patients with ARDS

Many patients in the intensive care unit require mechanical ventilation (MV) to take over the work of breathing. Although MV is vital for patient survival, it can induce potentially life-threatening ventilator-induced lung injury (VILI). This development is primarily caused by exaggerated and highly heterogenous mechanical loading of the lung during MV, which can lead to lung tissue damage due to local overdistention or due to repeated opening and closing of the alveoli.

Mathematical models of the respiratory system can give local insight into lung mechanics and mechanical loading during MV. Currently however, most respiratory mechanics models assume homogeneous lung behavior. Since lung properties can vary considerably within the lung, the applicability of such models to investigate VILI development may be limited. Recently, methods have been developed at STROBE (Grenoble) and CREATIS (Lyon) laboratories in collaboration with CHU Lyon & Grenoble to track local lung deformation using CT image registration. These patient-specific distributions can be used to create more realistic models of respiratory mechanics during MV, paving the way for a “Digital Twin” that could ultimately allow in silico personalized optimization of the mechanical ventilator settings in order to minimize the risk of VILI.

In this 6-month project you will perform your work at STROBE, Grenoble, France, in collaboration with CREATIS, Lyon, and the Department of Biomedical Engineering, Eindhoven University of Technology (TUE). You will develop methods to derive local lung material properties from registration-based deformation maps. Subsequently, the derived local mechanical parameters may be implemented in a 0D modelling framework of respiratory mechanics developed at the Eindhoven University of Technology to investigate patient-specific lung mechanics during MV.