Smart ECLA—closed loop control of O2 and CO2 for management of extracorporeal lung assist

tated trace demonstrates high levels of correlation with each of the algorithms in the system. The relative change detection algorithms for heart rate and blood oxygen saturation correlate 97.8% (total events 229) and 98.3% (total events 119) of the time, respectively. The breath pause and apnea detection algorithms correlated 98.9% (total events 2943) and 96.8% (total events 33) of the time, respectively. Conclusions: This study has demonstrated that physiological rulebased computational algorithms can be used to detect the various forms of spells in humanphysiological data. There is great potential for the list of potential causes to be reduced with further research using this tool to demonstrate associations between various spell profiles and different conditions. We have begun clinical validation of the algorithms in a prospective simultaneous comparison trial against the gold standard of polysomnography. Data from 50 spontaneously breathing preterm infants that are known to be having spells will be used to test and refine the output of the algorithms. This trial will form the basis for further studies on ventilated infants with the long-term goal of developing an advanced clinical decision support tool to assist in the real-time management of all neonatal spells.