Feasibility of Stochastic Modelling in Mechanical Ventilation Treatment Selection

Setting mechanical ventilation (MV) settings for ICU patients is a challenging task performed only by clinical specialists. Guidelines, experience, and intuition guide them. However, this ‘one-size-fits-all’ approach will benefit most but not all. Clinicians need to cater to the individual level and adjust settings proactively to improve the standard of care for respiratory failure patients. This thesis covers three major areas: the development of a decision support system protocol via model-based techniques to provide recommended MV settings, the development of a stochastic model of respiratory elastance, and finally, a virtual patient platform for rapid prototyping of developed MV protocols.