A model-based reasoning system for the management of respiratory weaning process

Abstract

Summary form only given. The progressive withdrawal of mechanical ventilation is called weaning. A computerized decision support system (off-line or on-line) would help the medical staff to wean patients more efficiently and can potentially reduce treatment duration and mortality rates. The aim of this research is to design a model-based reasoning system to provide on-line guidance in respiratory weaning process management. In this approach, the semi-quantitative reasoning method is applied to regulate and control the weaning process by adjusting the ventilator settings. A semi-quantitative method provides a level of description that is intermediate between abstract qualitative models and precise numerical models. It combines both qualitative and quantitative knowledge within a single semi-quantitative qualitative differential equation (SQDE) representation. It is a set of qualitative constraints which are augmented with envelopes for all monotonic functions and numeric ranges for all model variables and initial conditions. The model is then simulated to compute the variables in ventilator setting. The control system is tested under different physiological conditions, using a detailed simulation model of the human respiratory system.