"Human scent is commonly implemented as an individualizing feature used to associate a suspect with an object, location, or path of travel. Human scent evidence is commonly identified by and utilized in conjunction with human scent detection canines. Human scent detection canines are employed to follow a scent trail left by a person of interest or to associate a questioned item with one of known origin.
Previous studies in the field of human scent analysis have worked to explain the procedures carried out by these canines when associations are performed; however, these works have largely focused on describing the differences between separately sourced samples. The presented work builds upon previous research in the field while providing insight into the under-evaluated fundamental theory regarding the intra-subject variability of human scent. The presented work evaluates the volatile organic compounds (VOCs) present in the headspace of human hand odor samples collected from 8 donors over variable time periods ranging from 5-weeks to 1-year. These profiles were collected over the course of multiple weeks, utilizing headspace- solid phase microextraction- gas chromatography-mass spectrometry (HS-SPME-GC-MS) to capture and analyze the gaseous organic compounds present in the equilibrated sample headspace. With the headspace composition serving as a representative depiction of the donor’s hand odor composition, both targeted and untargeted approaches to profile interpretation and association were investigated.
This research evaluates both targeted and untargeted approaches to analyzing VOCs, targeted analyte mass determination, automated peak matching, and the use of Spearman’s rank correlation as chemometric techniques for analyzing 2-D components of human hand odor profiles. This combined approach provides a mechanism for predicting the source-origin of human scent samples collected using HS-SPME-GC-MS. Intra-subject variations in human scent, including inter-day changes, are explored. A discussion will be lead on the observed diversity in human scent, its impact on the developed technique performance, and the plans for further refining of the proposed methodology.
This study reflects an active effort towards the creation of a model for the formal discrimination of human scent profiles collected using GC-MS. Further application of the advances made in this study will aid in the understanding of same-donor human scent similarity and inform thresholds when discriminating between different-donor samples. Additionally, this work will provide greater context to canine performed human scent associations while continuing to build toward a quantitative laboratory test for confirmation of in-field, canine performed human scent associations. "