REVIVE: Reliable and efficient virtual ventilator experiments

Patients with severe respiratory symptoms are supported by mechanical ventilation. While the ventilator saves lives, it also damages the respiratory muscles, in particular the diaphragm. This complication means that the patient needs to stay longer in intensive care. There is ongoing research to understand this effect and to improve ventilation.

Experiments cannot be made on patients. Instead animal experiments are used.

As an alternative, we propose to support medical researchers with computational tools that allow for simulation of individual patient status and prediction of the diaphragm response.

Simulation of the diaphragm dynamics is a challenging problem due to the high aspect ratio of the diaphragm, the complexity of the interaction with surrounding organs, and the non-linearity of the muscle tissue. We propose to develop high-order numerical methods based on radial basis function approximation.

These methods will be used both for building smooth object representations from medical images, and for biomechanical simulation of the respiratory process. A special focus will be placed on methods that are robust for high-aspect ratio domains and mixed boundary conditions. These methods are relevant also for other challenging application areas.

During the project period, we will gradually increase the complexity of models and simulation methods.

The project is a collaboration between researchers with expertise in scientific computing, image analysis and intensive care.