The applicant proposes to use signal analysis, pattern recognition and fuzzy relational identification of patterns linked to outcomes in order to develop a decision support system that will improve the utility of intrapartum Electronic Fetal Monitoring (EFM) to accurately assess (1) obstetrical outcomes, and (2) the need for immediate invasive intervention, including cesarean deliveries or conversely, (3) the ability to safely avoid interventions. The proposed project involves the development of """"""""data sets"""""""" or obstetrical cases that include EFM data (strips), infant Apgar scores, and umbilical cord arterial PH levels. The monitor strips will undergo signal decomposition or extraction by morphologic filters. This will separate the biologic signal variation from the background """"""""noise"""""""" as well as separate the signal into multiple one-dimensional component signals. Next, a self-organizing neural network will be applied to the component signals in order to identify the fetal heart rate (FHR) patterns and detect their changes. Fuzzy relational modeling will be used to identify relationships between FHR patterns and obstetrical outcomes. The Geodynamics and Rose Health Care Systems integrated research team will provide a proof of concept for a system that can identify the at- risk fetus better than the clinician.
The successful completion of Phase II should lead to a system which can be used in hospitals and clinicians' offices to better identify the at-risk fetus, thereby allowing for the timely implementation of efficacious interventions and a reduction in the use of invasive obstetrical interventions, including cesarean deliveries, without adversely impacting outcomes. The use of this system in hospitals and private clinicians' offices will also help reduce the costs of national health care.
