Predication of Extubation Readiness in Extreme Pre

Existing Synergy: Prediction of Extubation Readiness in Extreme PreTerm Infants
Team Members:
Karen Brown, Department of Anaesthesia,
Robert Kearney, Department of Biomedical Engineering
Doina Precup, School of Computer Science
Guilherme Sant’Anna, Department of Pediatrics,
Lara Kanbar
Carlos Robles Rubio
Almost all extreme preterm infants, born at 28 weeks or less, have breathing problems severe enough to
require mechanical ventilation for them to survive. In deciding when to remove them from the ventilator,
the physician must balance conflicting issues. On the one hand, it is desirable to discontinue ventilation
as early as possible to avoid the adverse side effects of prolonged mechanical ventilation. On the other
hand, an infant who is removed from the ventilator too early may need to be reconnected. This can be
difficult to achieve in tiny infants and is associated with additional adverse side effects. At present,
the physician has few tools to assist in this decision and consequently more than 25% of the attempts to
disconnect preterm infants from ventilation fail. Previous results have shown that infants who can be
removed from ventilation successfully have different patterns of heart rate and breathing than those who
fail. We will build on these results by combining non-invasive cardio-respiratory measurements, novel
advanced signal processing methods, and machine learning techniques to build a tool that will predict
whether mechanical ventilation can be successfully discontinued for a particular infant. This tool will
provide physicians with the quantitative information needed for the informed management of ventilation of
pre-term infants and hence improve patient care.
Poster Presentation
Prediction of Extubation Readiness in Extreme Preterm Infants Based on Measures of Cardiorespiratory
Potential new collaborations
Instrumentation for non-invasive measurement of cardiorespiratory variables.
Adult sleep disorders