Data fusion approach for human body odor discrimination using GC-MS spectra

This study deals with data fusion approach to search discriminating biomarker volatile organic chemicals (VOCs) in body odor for individual differentiation. Particularly we have employed kernel principal component analysis (KPCA) combined with majority voting method to build up novel data fusion strategy. Gas chromatography-mass spectrometry (GC-MS) characterizes human body odor samples to find out the VOCs composition (alcohols, acids, aldehydes, esters, ketones, carbonyl compounds, sulfides and hydrocarbons etc.). Peak number and related area value of VOCs from the GC-MS spectra of body odor extract is used for analysis. GC-MS data from three experiments, based on body odor samples of four persons (different age groups) in dissimilar conditions are collected. Optimal set of peak numbers are selected with fusion approach. Linear PCA is used in validation of elected peak numbers for discrimination of individual's body odor. The opted peaks result satisfactory differentiation of individual's body odor in feature space. Thereafter biomarker VOCs are affirmed by matching corresponding peak number in GC-MS spectra. Analysis outcomes conclude particular set of biomarker VOCs for each experiment.