Neural Network Based Bioimpedance Cardiac Output Monitor

The final goal of this proposal is to develop a non-invasive, ambulatory technique for cardiac output (CO) computation by which conventional bioimpedance data are processed by a Neural Network (NN) to produce a CO measurement, useful during extended duration flight in space missions. We will develop a NN by which CO can be reconstructed from bioimpedance measurements, thus allowing the use of an accurate, portable and affordable device for monitoring and diagnostic purposes. Our preliminary work indicates that a NN based algorithm can capture the information present in the bioimpedance signal to accurately compute the CO value. Phase I will test the feasibility of developing the NN implementation necessary to compute the CO from bioimpedance data. In Phase I we will develop and test a prototype that will integrate the bioimpedance stimulator/acquisition front end, and signal processing into a cost-effective, ambulatory CO monitor unit. Phase II will complete verification and validation of the system. We believe that this novel application of signal processing and NNs has the potential to bring a totally new CO monitor to the market.Successful completion of the proposed development program will result in a new, easy-to-use system that will advance CO monitoring capabilities for crew health assurance during space missions.