STOCHASTIC INTEGRATED VIRTUAL PATIENT MODELLING FOR MECHANICAL VENTILATION TREATMENT IN THE HOSPITAL INTENSIVE CARE

This thesis focuses on improving how breathing machines (ventilators) are managed for patients in intensive care. Current methods often rely on general rules or clinician experience, which may not suit every patient and can sometimes cause harm. To address this, the research developed computer-based “virtual patients” that mimic how real patients’ lungs respond to ventilation. These virtual models were used to design and test a new system that automatically adjusts ventilator settings safely for each patient. The work aims to make mechanical ventilation more personalised, efficient, and safer, ultimately improving care for patients with respiratory failure.