A tool for hemodynamic data analysis

Abstract

Nearly 2% of all live births are very premature (gestational age less than 32 weeks), and 50% of all new cases of cerebral palsy occur in survivors of premature birth (gestational age less than 37 weeks). Because of their underdeveloped vascular structure, premature infants are especially vulnerable to brain injury caused by unregulated and erratic changes in blood pressure. A challenge in the prevention of serious brain injury in premature infants is the inability to identify impending or recent hemodynamic events that might lead to injury of the newborn's brain. If events that indicate a propensity to experience brain injury can be identified, then such events can be monitored clinically, and steps can be taken to prevent them from occurring. We designed and implemented a software tool, HemDAT, that can be used to test hemodynamics related hypotheses and to facilitate the discovery of interesting relationships among hemodynamic signals. HemDAT uses signal processing and statistical knowledge to provide clinical researchers a tool that can help develop a better understanding of how brain injury occurs in premature newborns. HemDAT is capable of processing and navigating large data sets of blood pressure and cerebral blood flow. Large data sets are important because the events that cause brain injury are believed to be short-lived, possibly infrequent, and unpredictable. Additionally, since this is a relatively unexplored area in human infants, HemDAT provides flexibility in performing repeated analyses with different parameters modifiable by the user. HemDAT also provides convenient visualizations of results and does not demand signal processing or statistical expertise from the user.