| Prediction of Life-Threatening
Events in Infants |
| |
| Aaron Lewicke, Stephanie Schuckers, Michael Corwin, Michael Neuman,
Toke Hoppenbrouwers |
| |
| This research studies the prediction of life threatening events
(LTE), bradycardia and apnea, in infants using heart rate variability
(HRV). LTEs may be related to sudden infant death syndrome (SIDS),
the sudden death of an infant under 1 year of age without any apparent
warning. For this research, we are using a massive dataset collected
by a major NIH study, the Collaborative Home Infant Monitoring Evaluation
(CHIME). For over 1000 infants, this dataset includes a hospital-based
polysomnongraph (PSG) study and continuous home monitor recordings
whenever the monitor was used. Currently, we are investigating computer
intelligence models based on the PSG in order to predict LTEs. Also,
we are developing algorithms that will enable us to look into HRV
from the home data. This will open up hundreds of thousands of hours
more of infant data to be analyzed. Below is a figure that shows PSG
data separated by implementing the fuzzy c-means algorithm on HRV
parameters. |
| |
 |
| |
| Figure: Add more detail |
