Using temperature changes measured at the optical sensor site, it had been demonstrated previously that the switch-over of the two blood streams occurred within 50 ms at the sensor surface ( Chen

et al., 2012b), which is certainly fast enough to indicate that the mechanical switch-over of the two blood streams did not affect our results in any way. Any diminution in recorded ΔPO2 with increasing Akt inhibitor simulated RR would therefore be due to sensor performance, rather than test rig limitation. Studies investigating cyclical atelectasis in the Acute Respiratory Distress Syndrome (ARDS), where PO2PO2 varies widely within breaths, require very fast response intravascular oxygen sensors, which motivated the present study. PO2PO2 and SaO2 oscillations in arterial blood have been studied for several decades; an overview of the most important findings in this field is presented and discussed in the following paragraph. Cyclic variations in blood oxygenation within the respiratory cycle were reported in 1961 in

an open chest experimental animal model (Bergman, 1961a and Bergman, 1961b). In this model, femoral arterial blood was withdrawn from a small catheter through a fast response external oximetry cuvette at a constant rate by a motor-driven syringe, and variations in oxyhaemoglobin saturation (SaO2) were recorded in real time. SaO2 was used as a surrogate for arterial oxygen tension (PO2)(PO2), and rapid cyclic variations of up to 20% in SaO2 (ΔSaO2) were recorded. Using these saturation figures and a standard dissociation curve,

Etoposide cell line these values translate to a PO2PO2 oscillation amplitude of 15 mmHg at a mean PO2PO2 of 36 mmHg (Whiteley et al., 2003). Despite the evidence suggesting that the cause of the observed fluctuations in arterial saturation might be due to variations in pulmonary shunt, it was concluded that these large variations in PaO2/SaO2PaO2/SaO2 might be due to cyclical changes in Linifanib (ABT-869) alveolar oxygen tension. Much later on, in a computer model, it was shown that large changes in PaO2PaO2 could only be generated by large intra-breath changes in pulmonary shunt caused, most likely, by cyclical atelectasis (Whiteley et al., 2003). Oscillations in carotid artery PO2PO2, which had the same period as respiration, were demonstrated in the cat, and in the newborn lamb in the first hours after birth (Purves, 1965 and Purves, 1966). Although recognising that changes in venous admixture occur during the respiratory cycle and that there was a significant degree of venous admixture during the experiments, the conclusion was drawn that the cyclical oscillations in carotid PO2 (ΔPaO2) in these animal studies were due to changes in alveolar PO2PO2. Thirteen years later, in an experimental cat model, it was shown that the amplitude of ΔPaO2 increased with increasing tidal volume, with increasing mean PaO2PaO2, and decreasing ventilator frequency (Folgering et al., 1978). Some of these studies were conducted at a mean PaO2PaO2 of 150 mmHg, i.e.