M. Peleg, N. Lavrač, and C. Combi (Eds.): AIME 2011, LNAI 6747, pp. 149–158, 2011. © Springer-Verlag Berlin Heidelberg 2011 The Intelligent Ventilator Project: Application of Physiological Models in Decision Support Stephen E. Rees 1 , Dan S. Karbing 1 , Charlotte Allerød 1,2 , Marianne Toftegaard 2 , Per Thorgaard 2 , Egon Toft 3 , Søren Kjærgaard 2 , and Steen Andreassen 1 1 Center for Model-based Medical Decision Support (MMDS), Instiutute of Health Science and Technology, Aalborg University, Aalborg, Denmark 2 Department of Anaesthesia, Aalborg Hospital, Denmark 3 The Faculty of Medicine, Aalborg University, Denmark MMDS, Aalborg University, Fredrik Bajers vej 7E, 9220 Aalborg, Denmark sr@hst.aau.dk Abstract. This paper describes progress in a model-based approach to building a decision support system for mechanical ventilation. It highlights that the process of building models promotes generation of ideas and describes three systems resulting from this process, i.e. for assessing pulmonary gas exchange, calculating arterial acid-base status; and optimizing mechanical ventilation. Each system is presented and its current status and impact reviewed. Keywords: Mechanical ventilation, decision support, acid-base, gas exchange. 1 Introduction This paper summarizes progress in the Intelligent Ventilator (INVENT) project over the past decade. The philosophy of this work is that building decision support systems based upon physiological models is a good thing to do. Physiological models provide a natural division between describing the patient and our preference towards clinical outcome using decision theory. Perhaps equally as important, is that the models tend to raise interesting questions and lead to new ideas for research, and for clinical and commercial applications. Our application of this philosophy in the field of mechanical ventilation has led to the development of a number of systems under the umbrella of the INVENT project, illustrated in figure 1. The original, and existing, goal of this project is to build a model based decision support system (DSS) to suggest appropriate settings for mechanical ventilation. To do so has required building several physiological models (layer 1, figure 1). These include: pulmonary gas exchange focusing on oxygen transport and acid-base and oxygenation status of the blood, interstitial fluid and tissues focusing on carbon dioxide transport. Models require validation (layer 2, figure 1), and studies have compared the model of pulmonary gas exchange against the reference technique [1]; and the model of acid-base chemistry with literature and experimental data. These models have raised interesting scientific and clinical questions requiring close clinical collaboration (layer 3, figure 1). Addressing these questions has led to the development of two further systems, the Automatic Lung Parameter Estimator (ALPE) system, and