Detection of stress through sweat analysis with an electronic nose

Cortisol and adrenaline are two hormones associated with stress. Therefore, their detection is important for applications in medicine and security. We propose to detect the presence of these chemical compounds in the body's sweat by electronic noses with resistive sensors as a non-invasive method. Low concentrations of cortisol (5 muM - 50 muM) and adrenaline (0.3 mM - 5.5 mM) have been measured. Furthermore, our electronic nose have been able to distinguish between sweat from physical stress and quiet situation by two pattern recognition techniques as principal component analysis (PCA) and probabilistic neuronal network (PNN), being the prediction rate of the network 85%.