Rejection of false saturation data in optical pulse-oximeter

The measurement of concentration of blood gases (in particular O2) is of extreme interest for the evaluation of efficiency of pulmonary gas exchange, adequacy of alveolar ventilation, mechanisms of blood-gas transport and tissue oxygenation. Invasive techniques, although still common in many clinical situations, present the limit to be necessary not continuous, time-delayed (typically 20-30 min), painful and with an associated risk. Such limitation becomes particularly severe in critical ill patients where close and frequent (continuous) monitoring of arterial blood gases is essential. In this paper, we describe a real-time, novel measurement procedure to validate non-invasive arterial oxygen saturation (SO2) data measured by standard pulse-oxymeter. Such instrument, especially when used in intensive care units (ICU), can produce wrong data due to voluntary or involuntary motion artefacts or imperfect skin-sensor contact. We have monitored (sampling frequency: 1 Hz) in 14 preterm subjects and used the heart rate (HR) signal provided by the ECG (II-lead) and the same quantity provided by the SO2 sensor (SHR); such data have been used to validate the SO2 data. A dedicated algorithm has been developed in order to provided “most-probable” values in case of single or few not validated SO2 data. Results show that less than 10% of the arterial oxygen saturation data are not validated, which means, in our case, 25 hours of data (out of 250 hours observed) are not validated. With the provided algorithm the still not validated data are reduced to less than 5%.