The use of back propagation neural networks to identify mediator-specific cardiovascular waveforms

Exposure to xenobiotics characteristically initiates the release of cascade of endogenous mediators which trigger diverse cardiopulmonary responses. Platelet activating factor (PAF) and prostaglandin H₂ analog (PGH) are two mediators which have demonstrated distinct effects on the cardiovascular and pulmonary systems. The biological responses to these mediators have been implicated in the etiology of various disease states. A heteroassociative back propagation neural network (BPN) was used to identify injected mediators by analyses of complex cardiopulmonary waveforms. In all cases, the BPN qualitatively and quantitatively identified the mediator substance which induced the response. The results demonstrate the potential for deciphering complex waveforms into mediator-specific events. The BPN can isolate complex physiological patterns into component parts. Therefore, BPN has the potential to be a powerful investigative tool for deciphering etiologic mechanisms of xenobiotic insults