Abstract (Continued)
either hyperventilation or a reduced breathing rate. Special attention is paid to the impact of oxygen sensors that are noisy and prone to signal dropouts, defining a filter that attempts to mitigate such effects.
It is the author's belief that use of this computer simulation by medical professionals may provide considerable insight into dealing with the oxygen control of neonates. But this effort is now at a point where the commentary of such professionals is necessary to critique the assumed neonate model, based upon their clinical experience. It is very likely that this model can be modified to bring it more in line with reality, should that be necessary.
Hopefully, neonatal physicians interested in this subject will find this simulation approach an effective way to gain confidence in a proposed controller/filter design prior to its use on human subjects. This could lead to a hardware implementation, or retrofit, involving a conventional oxygen delivery system and pulse oximeter, interfaced through their serial ports to a computer displaying every system output of interest, and housing the controller and filtering software algorithms. Initial tests of this equipment using actual neonates must, of course, involve great care. But, having thoroughly studied the entire system beforehand should greatly improve the likelihood that these tests will prove successful.
