Patients on Weaning Trials Classified with Neural Networks and Features Selection
B. Giraldo (Technical University of Catalonia, Spain), A. Garde (Technical University of Catalonia, Spain), C. Arizmendi (Technical University of Catalonia, Spain), R. Jane (Technical University of Catalonia, Spain), I. Diaz (Hospital de la Santa Creu i Sant Pau, Spain) and S. Benito (Hospital de la Santa Creu i Sant Pau, Spain)
Copyright: © 2008
One of the challenges in intensive care is the process of weaning from mechanical ventilation. We studied the differences in respiratory pattern variability between patients capable of maintaining spontaneous breathing during weaning trials, and patients that fail to maintain spontaneous breathing. In this work, neural networks were applied to study these differences. 64 patients from mechanical ventilation are studied: Group S with 32 patients with Successful trials, and Group F with 32 patients that Failed to maintain spontaneous breathing and were reconnected. A performance of 64.56% of well classified patients was obtained using a neural network trained with the whole set of 35 features. After the application of a feature selection procedure (backward selection) 84.25% was obtained using only eight of the 35 features.
Key Terms in this Chapter
Spontaneous Breathing: Normal/natural breathing.
Respiratory Pattern: Breathing pattern.
Data Visualization: Technique of looking at data to identify like and unlike groups of data points.
Neural Networks: Form of artificial intelligence.