Differentiation of vaginal cells from epidermal cells using morphological and autofluorescence properties: Implications for sexual assault casework involving digital penetration

Sarah Ingram,Arianna DeCorte,Amanda Elswick Gentry,M Katherine Philpott,Taylor Moldenhauer,Sonja Stadler,Cory Steinberg,Jonathan Millman,Christopher J Ehrhardt
DOI: https://doi.org/10.1016/j.fsigen.2023.102909
Abstract:Analysis of DNA mixtures from sexual assault evidence is an ongoing challenge for DNA casework laboratories. To assist the forensic scientist address source and activity level propositions there is a significant need for new techniques that can provide information as to the source of DNA, particularly for sexual assault samples that do not involve semen. The goal of this study was to develop a new biological signature system that provides additional probative value to samples comprised of mixtures of epidermal and vaginal cells, as may be observed in cases involving digital penetration. Signatures were based on morphological and autofluorescence properties of individual cells collected through Imaging Flow Cytometry (IFC). Comparisons to reference cell populations from vaginal tissue and epidermal cells collected from hands showed strong multivariate differences across > 80 cellular measurements. These differences were used to build a predictive framework for classifying unknown cell populations as originating from epithelial cells associated with digital penetration or epidermal tissue. As part of the classification scheme, posterior probabilities of specific tissue group membership were calculated for each cell, along with multivariate similarity to that tissue type. We tested this approach on cell populations from reference tissue as well as mock casework samples involving hand swabbings following digital vaginal penetration. Many more cells classifying as non-epidermal tissue were detected in digital penetration hand swab samples than control hand swabbings. Minimum interpretation thresholds were developed to minimize false positives; these thresholds were also effective when screening licked hands, indicating the potential utility of this method for a variety of biological mixture types and depositional events relevant to forensic casework. Results showed that samples collected subsequent to digital penetration possessed markedly higher numbers of cells classifying as vaginal tissue as well as higher posterior probabilities for vaginal tissue (≥ 0.90) compared to cell populations collected from hands without prior contact with vaginal tissue. Additionally, digital penetration cell populations may be resolved from saliva cell populations and other non-target tissue types.
What problem does this paper attempt to address?