Raman spectroscopy for analyzing body fluid traces: Moving towards a practical forensic application

As submitted by the proposer: The proposed study is a continuation of an ongoing research project ultimately targeting the development of an easy-to-use, portable Raman spectroscopic instrument for the rapid, nondestructive, and confirmatory identification of body fluid traces recovered at crime scenes. This new methodology would allow crime scene investigators to identify all major body fluids using one instrument and retrieve information about the donor, which could potentially include gender, race, and age. The proposed study aims to investigate possible sources of uncertainty of the developed method, account for varying conditions of forensic evidence, and investigate the possibility to determine donor age based on body fluid traces. This will be accomplished in five objectives, (i) determine the limit of detection of the method, (ii) simulate forensic evidence by varying preparation conditions and substrates, (iii) identify false positives, (iv) validate the method with real case-like samples, and (v) evaluate the ability of the method to determine donor age. We have formed two significant partnerships. Dr Barry Duceman, of the New York State Police Forensic Investigation center, and Dr Arian van Asten and Gerard Stor, of the Netherland Forensic Institute, will advise us throughout the proposed project and provide us with case-realistic samples for validation testing. Body fluid samples will be purchased from a biological supplier to represent a genetically diverse donor population. Samples will be measured using a Raman microspectrometer, and the spectra will be analyzed using advanced chemometrics and statistical modeling. Regression and classification models will be built using calibration datasets, and then tested with external validation datasets. The work will be carried out over a three year time span at the University at Albany, State University of New York, in Albany, New York. All results from the proposed work will be published in peer-reviewed academic journals, submitted in semi-annual and final progress reports, and presented at professional conferences. The resulting data will include numerous spectral datasets from each objective and statistical models. All data will be archived with the National Archive of Criminal Justice Data. ca/ncf