Raman spectroscopy for the identification of body fluid traces: Semen and vaginal fluid mixture

• NCJ Number: 306227
• Journal: Forensic Chemistry Volume: 2023 Dated: 2023
• Author(s): Aliaksandra Sikirzhytskaya; Vitali Sikirzhytski; Luis Pérez-Almodóvar; Igor K. Lednev
• Date Published: 2023
• Length: 7 pages
• Grant Number(s): 2017-DN-BX-0135
• Agencies: NIJ-Sponsored
• Publication Type: Research (Applied/Empirical)

Annotation

This article discusses the potential of Raman spectroscopy for confirmatory identification of all main body fluids.

Abstract

Raman spectroscopy is a versatile research tool that has found numerous practical applications recently. High specificity, non-destructive nature, and potential for on-field application make Raman spectroscopy an attractive tool for forensic science. The great potential of Raman spectroscopy for confirmatory identification of all main body fluids has been recently reported (Vyas et al. Forensic Chemistry 2020, 100247). Bringing the developed methodology from the laboratory to the crime scene requires further method development and validation for realistic, practical samples subjected to environmental conditions, contamination, substrate interference, aging, etc. This work targets the validation of the developed methodology for body fluid mixtures, semen, and vaginal fluid, which are critical for sexual assault cases. Near-infrared (NIR) Raman spectra of dry semen-vaginal fluid mixtures were acquired in automatic scanning mode and subjected to advanced statistical analysis. Support Vector Machine Discriminant Analysis (SVMDA) was used to develop three binary classification models, including vaginal fluid vs mixture, semen vs mixture, and semen vs vaginal fluid. These binary models were combined to build a single tertiary model, which resulted in the 100% correct classification on the sample level for all examined mixtures (1.5% to 100% for vaginal fluid and 3% to 100% for semen). The developed analytical approach could be used for various other applications when complex binary mixtures need to be detected and analyzed. (Published abstract provided)