If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
This work explores morphological and autofluorescence differences between vaginal and epidermal cells detectable through Imaging Flow Cytometry (IFC), a non-destructive, high-throughput technique. These differences were used to build a predictive framework for classifying unknown cells as originating from vaginal or epidermal tissue, which was tested on hand swabbings with and without digital penetration. Many more cells possessing a vaginal signature (median posterior probability ≥0.90) were detected in digital penetration samples than control hand swabbings. Minimum interpretation thresholds were developed to minimize/eliminate 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.
Realizing the full probative potential of mixed biological samples derived from sexual assault evidence is an ongoing challenge for DNA caseworking laboratories, particularly for samples that do not involve semen, as may be encountered in swabbings subsequent to digital (or object) penetration. The goal of this study was to explore the probative potential of the histological and morphological differences [
] between epidermal and vaginal cells, and use these differences to develop a new signature system for identifying the tissue source of unknown biological material. Signatures were based on cellular attributes acquired via IFC, without the use of fluorescent dyes, and the multivariate classification scheme developed was tested on mock casework samples [
Reference cell populations from epidermal tissue, vaginal tissue, and saliva, as well as hand swabbings collected at a range of timepoints subsequent to digital penetration, or collected subsequent to oral contact (fingers in mouth), were obtained from volunteers with informed consent at either Virginia Commonwealth University (Richmond, VA) or the Ontario Centre of Forensic Science (CFS; Toronto, CA).
Cell populations were eluted from each swab following previously published protocols [
]. IFC analysis of the resulting cell suspension from each sample was performed with the Amnis® Imagestream X Mark II instrument (EMD Millipore; Burlington, MA). Brightfield images (Fig. 1, fourth column) were collected along with images at fluorescent wavelengths between 430 nm and 780 nm (Fig. 1). Magnification was set at 40x and autofocus was enabled so that the focus varied with cell size. Individual cells were differentiated from debris or other non-cellular material by filtering based on size and gradient values using IDEAS® Software (EMD Millipore; Burlington, MA). Over two hundred automated measurements of morphological and autofluorescence attributes per cell were initially acquired from these images. Differentiation of epidermal and vaginal cell types was optimized with approximately 90 measurements primarily extracted from images collected in the FITC, PE and brightfield channels (Fig. 1, columns 2–4).
The resulting datasets were then exported to SPSS v28 (IBM, Inc. Chicago, IL) for statistical analysis. Multivariate modeling and classification tests were based on Linear Discriminant Analysis (LDA) using within-group covariance matrix.
3. Results and discussion
Vaginal cells showed morphological and optical features easily distinguished from epidermal cells, including presence of nuclei, larger cell area, and lower levels of optical contrast within the brightfield channel (Fig. 1). Linear Discriminant Analysis (LDA) of vaginal, saliva, hand, hand subsequent to digital penetration, and hand subsequent to oral contact swabbings showed distinct clustering of cell populations based on sample type (e.g.Fig. 2). This suggests that (1) differences can be detected across tissue sources and (2) the circumstances of deposition and/or sampling can affect signatures.
IFC signatures were used to build algorithms for predicting sample type and estimating probability of sample type association. Based on the results obtained, a minimum threshold was developed for detecting a vaginal signature on hands while minimizing false positive results. For this dataset, requiring more than 20 cells classifying consistent with digital penetration, with a median posterior probability of at least 0.90, eliminates false positives. One of the digital penetration samples did not meet this threshold (11 cells classified consistent with digital penetration, median posterior probability 0.99); however, this sample was collected 20 h after digital penetration, with several intervening hand washings. It is possible that there were very few vaginal cells remaining on the subject’s hands at the time of collection. In any case, any interpretation threshold will have some effect on sensitivity, particularly on samples at the edge of detectability.
The application of the same threshold to hand swabbings taken subsequent to licking (i.e. saliva and epidermal mix) resulted in no false positives. In fact, none of the saliva-on-hand samples produced median posterior probabilities of digital penetration association ≥0.80. This discriminating capability is consistent with the clustering observed in the LDA plot of these samples (Fig. 2).
This study demonstrates that cell populations from different epithelial tissue sources can be resolved using IFC-based signatures. Distinct cellular signatures were identified in hand swabs collected subsequent to digital penetration, which were distinguishable from hand reference samples and samples collected from licked hands. As an alternative to this binary system (presence/absence of signature), it may be possible to attach quantitative estimates to likelihood of group classifications for individual cells or cell populations. Future research will test additional donors for each sample type to further refine predictive frameworks and thresholds.
Conflicts of interest
This project was funded by the Center for Innovative Technology-Commonwealth Research Commercialization Fund (MF-19-013-LS; PI Ehrhardt). The study sponsor had no role in study design, in collection, analysis, or interpretation of data, in writing the manuscript, or in the decision to submit the manuscript for publication.
Differentiation of epithelial cell types by cell diameter.