computer methods and programs in biomedicine 101 ( 2 0 1 1 ) 156–165 journal homepage: www.intl.elsevierhealth.com/journals/cmpb A model of perfusion of the healthy human lung M.L. Mogensen * , K.S. Steimle, D.S. Karbing, S. Andreassen Aalborg University, Center for Model-Based Medical Decision Support, Fredrik Bajersvej 7, DK 9220 Aalborg, Denmark article info Article history: Received 17 December 2009 Received in revised form 18 June 2010 Accepted 28 June 2010 Keywords: Physiological models Pulmonary perfusion Pulmonary capillaries Lung mechanics abstract This study presents a model that simulates the pulmonary capillary perfusion. The model describes the lungs as divided into horizontal layers and includes: capillary geometry; capil- lary wall elasticity; pressure at the pulmonary artery; blood viscosity; the effect of the chest wall; the change in lung height and hydrostatic effects of the lung tissue and of the blood during breathing. The model simulates pulsatile blood perfusion with an increasing blood distribution down the lungs, in agreement with previous experimental studies. Moreover the model is in agreement with experimentally measured total capillary perfusion, total capillary volume, total capillary surface area and transition time of red blood cells passing through the pulmonary capillary network. The presented model is the first to be validated against the mentioned experimental data and to model the link between airway pressure, lung volume and perfusion. © 2010 Elsevier Ireland Ltd. All rights reserved. 1. Introduction Appropriate ventilator settings for intensive care patients with respiratory disorders are crucial for reducing recovery time and minimizing the risk of ventilator induced lung injury (VILI) [1,2]. Finding the appropriate settings requires a trade- off between the need to obtain adequate gas exchange and minimization of VILI. Positive end expiratory pressure (PEEP) and tidal volume affect both the lung mechanics and the gas exchange, and appropriate levels for example of tidal vol- ume have been shown to lower the mortality in patients with acute lung injury [3]. Furthermore high pressures by means of PEEP have proven to prevent alveolar collapse and improve gas exchange in the diseased lungs [2]. However, it is still not clear how to improve gas exchange while preventing VILI [2]. In order to understand gas exchange in patients with respiratory failure, it is first necessary to understand pulmonary ventila- tion and perfusion in healthy lungs. Pulmonary perfusion has not been studied as intensely as ventilation even though it is indicated that lung volume greatly affects perfusion [4]. ∗ Corresponding author. Tel.: +45 9940 8764. E-mail address: lause@hst.aau.dk (M.L. Mogensen). Previously Fung and Sobin [5] have modelled the pul- monary capillary perfusion as a sheet flow through a two-dimensional network built on morphometric studies on cat lungs. Burrowes et al. [6] used an anatomically based finite element model to simulate regional variations in blood perfusion. Burrowes and Tawhai [7] constructed an arterial geometric model using a combination of computed tomog- raphy and a volume-filling branching algorithm. In addition, Liu et al. [8] modelled the airway mechanics, gas exchange, and perfusion in a nonlinear model excluding the hydrostatic effects of tissue and blood. None of these previous models of the pulmonary capillary perfusion have described how the perfusion is affected by the airway pressure and lung volume at different lung heights. Mogensen et al. [9] used a physio- logical stratified model to simulate the influence of different lung volumes on perfusion. The model did, however, only take the lung volume as input and did not include the hydrostatic effects of changing lung height and density. This paper presents a modified version of the physio- logical model by Mogensen et al. [9], which describes the pulmonary microcirculation, enabling simulation of capillary 0169-2607/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.cmpb.2010.06.020