| Summary: | Two challenges in the management of Acute Respiratory Distress Syndrome are the difficulty in diagnosing cyclical atelectasis, and in individualising mechanical ventilation therapy in real-time. Commercial optical oxygen sensors can detect <em>P</em><sub><em>a</em><sub>O<sub>2</sub></sub></sub> oscillations associated with cyclical atelectasis, but are not accurate at saturation levels below 90%, and contain a toxic fluorophore. We present a computer-controlled test rig, together with an in-house constructed ultra-rapid sensor to test the limitations of these sensors when exposed to rapidly changing <em>P</em><sub>O<sub>2</sub></sub> in blood <em>in vitro</em>. We tested the sensors' responses to simulated respiratory rates between 10 and 60 breaths per minute. Our sensor was able to detect the whole amplitude of the imposed <em>P</em><sub>O<sub>2</sub></sub> oscillations, even at the highest respiratory rate. We also examined our sensor's resistance to clot formation by continuous <em>in vivo</em> deployment in non-heparinised flowing animal blood for 24 h, after which no adsorption of organic material on the sensor's surface was detectable by scanning electron microscopy.
|
|---|