Defense Date

2020

Document Type

Thesis

Degree Name

Master of Science

Department

Forensic Science

First Advisor

Dr. Christopher Ehrhardt

Second Advisor

Dr. Susan Greenspoon

Third Advisor

Dr. Catherine Connon

Fourth Advisor

Kate Philpott, JD

Fifth Advisor

Dr. Sarah Seashols-Williams

Abstract

Current methods for confirming the presence of spermatozoa in sexual assault samples can be time-consuming and often lack sensitivity; however, this remains the most definitive test for the presence of semen. Additionally, male DNA can be deposited without the presence of intact sperm as may be the case with seminal fluid from vasectomized individuals or sexual activity where seminal fluid is not recovered (e.g., perpetrator wears a condom, penetration without ejaculation, etc.). The ability to detect bodily fluids, as well as quantify their presence in a sample, could aid in forensic DNA analysis by limiting the amount of time performing serological testing, as well as screening for probative samples for DNA profiling. Additionally, determining the cellular makeup of a sample can be informative for investigative purposes, e.g. rebutting or supporting certain factual claims from the victim or defendant. Morphological and/or autofluorescence cellular signatures are rapid and non-destructive methods for cell type differentiation in the clinical context but have not been thoroughly explored for forensic casework applications. Therefore, the goal of this study was to characterize signatures in four major cell types associated with sexual assault casework (vaginal, rectal, and penile epidermal cells, and azoospermic seminal fluid) towards the development of a method for rapidly identifying and/or differentiating these cell/fluid types in biological evidence. Morphological and autofluorescence profiles of each cell population were analyzed with Imaging Flow Cytometry (IFC) using five different excitation wavelengths and six detector channels ranging between 430nm-780nm. Signatures for each cell type were constructed from ~60 different individual cell measurements. Finally, linear discriminant analysis was used to develop a quantitative framework for differentiating cell populations and predicting cell types. Vaginal, rectal, and penile cells can be differentiated with a high degree of accuracy, ~90%. This framework was also highly accurate at differentiation semen (including azoospermic and proteinase K treated semen) from vaginal and rectal cell populations. However, there were still many factors that contributed to these levels of accuracy including, but not limited to, inter-donor variability. Ultimately, the results obtained indicate that each cell type have distinctive signatures that can be detected in a rapid and non-destructive manner.

Rights

© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission

11-24-2020

Available for download on Wednesday, November 24, 2021

Included in

Biology Commons

Share

COinS
 
 
 
View graphic version