In the United States, one in eight infants is born prematurely. These high risk infants require specialized monitoring of their physiology not only in Neonatal Intensive Care Units (NICU) but also in home environments. They are prone to apnea (pause in breathing), bradycardia (slowness of heart) and hypoxia (oxygen de-saturation), which are life threatening. This project aims at developing a biosensor system with wireless network for the remote detection and anticipation of such life threatening events in infants. The proposed research goes beyond traditional health monitoring systems by incorporating body sensor networks (BSN) along with advanced signal processing approaches, tailored specifically to an individual infant's physiology, to accurately detect and anticipate precursors of life threatening events. The proposed research can have a significant impact on non-intrusive ambulatory health monitoring for infants through a wireless biosensor system that integrates lightweight sensor solutions into the sensing, communication, and computing for monitoring physiology. The system framework, theories, models, and code developed by this project can be used by researchers as well as engineers to evaluate the performance of infant monitoring applications. The project also includes: (1) disseminating the project information and knowledge to the academic community and industry; (2) engaging undergraduate, graduate and medical students, especially women and minorities, into the proposed research; and (3) developing new courses and revising the existing courses.
The current physiological monitoring systems used in NICU consist of relatively large sensors attached to the infants, which are then connected to a data acquisition system with multiple wires. These sensors along with the wires are a hindrance to the clinical care. In addition, the existing system cannot be used for home environments because of the size and cost. While there is an abundance of physiological signals streaming across NICU monitoring systems, it is challenging for clinicians caring for preterm infants to determine pathological states, as there is no method available to translate these signals into validated indices to define pathology. The primary objective of this proposed research is to explore whether a dedicated compact device with wearable biosensors along with wireless networks can be built for the detection and anticipation of life threatening events in infants in both NICU and home environments. The secondary objective is to explore whether computational tools that provide real-time indices of cardio-respiratory risk can be developed to assist clinicians for neonatal care. Specifically, the project is to develop a comprehensive system, involving four important components: (1) development of miniature biosensors that can be attached to infants who are very small and vulnerable; (2) development of wireless devices with efficient communication protocols that can transmit the physiological signals from the biosensors; (3) development of efficient signal processing algorithms that can extract useful information from the biosensor data for risk stratification and anticipation of life threatening events (data to knowledge to decisions) and (4) testing and validation of the systems in real life environment at NICU. The proposed approaches in the project can eventually lead to a medical device for the remote detection of life threatening events in infants and also provide guidelines for the design of wearable wireless biosensor systems for healthcare monitoring applications in general.