This article overviews the application of vibrational spectroscopy in various areas of forensic analysis, with special focus on forensic serology and the analysis of trace evidence, and notes the benefits of pairing analysis methods with the capabilities of machine learning.
Vibrational spectroscopy combined with machine learning has a great potential for forensic applications. For example, handheld Raman spectroscopic instruments are already used by law enforcement agencies for precise, confirmatory identification of drugs. Beyond drug identification, several emerging technologies based on vibrational spectroscopy are currently under development for forensic investigative purposes, including the analysis of questioned documents, gunshot residue, fabrics, soil, hair, nails, and nail polish. This article provides a comprehensive overview of the application of vibrational spectroscopy in various areas of forensic analysis, particularly focusing on forensic serology and the analysis of trace evidences. In the case of forensic serology, the methodology allows for determining complex aspects of serological casework, including the time since deposition of a stain, as well as the phenotypic profile of the stain donor—namely, sex, race, and age. Furthermore, gunshot residues can be accurately identified by grain, caliber, and manufacturer when Raman spectroscopy is paired with machine learning. This integration of advanced spectroscopic techniques with machine learning holds great promise for furthering both the accuracy and efficiency of investigations, helping to reduce the total backlog of evidence investigation currently plaguing modern forensic laboratories. (Published Abstract Provided)