Analysis of perspiration by mass spectrometry combined with machine learning

Perspiration is an attractive biofluid for health monitoring because it can be collected noninvasively and almost no one feels possessive about his or her perspiration. Perspiration represents an easy window into a person's metabolomics. This talk will describe the use of desorption electrospray ionization (DESI) mass spectrometry (MS) for the identification and quantitation of lipids and small metabolites present in perspiration. When combined with machine learning, the mass spectrometry pattern can be used to recognize different disease states. An example that will be presented is the diagnosis of the genetic disease, cystic fibrosis, in newborn babies. (author)