Thousands of sexual assault cases in the United States are backlogged. This has been a growing issue for years that has increased the difficulty of solving these cases and providing closure to the victims. The analysis process for each case includes the identification of body fluids, presumptive testing, confirmatory testing, and DNA extraction. The only confirmatory method for semen identification is a microscopic visualization of sperm cells. The time spent on microscopic analysis varies depending on the complexity of the samples and the skills of the analyst. While the identification of sperm cells is informative, it can be very time-consuming and labor intensive. Some forensic laboratories choose to skip this step and submit samples directly for DNA analysis. Conducting DNA analysis on unscreened samples can increase the cost of testing when negative samples are analyzed as well as the time it takes to process each case. Automated microscopy has been available for decades and more recently has been paired with artificial intelligence to detect sperm cells on microscope slides. In this research, the MetaSystems automated microscope was used to analyze slides that mimic forensic sexual assault samples. Slides were also examined using traditional microscopy. The automated system quickly provided an accurate quantification of the number of sperm cells present in a sample, which can inform downstream DNA testing. The software was successful in identifying sperm cells treated with Christmas tree and hematoxylin and eosin stains, even among epithelial cells and various contaminants. Results demonstrated that an artificial intelligence-driven forensic sperm cell detection microscope can significantly reduce the time it takes to locate and identify sperm cells and estimate sperm cell quantity compared to a lengthier and more tedious manual search. Drawbacks to the system include the relatively high cost and reduced ability to accurately detect sperm cells amid contaminants that are of similar morphology.

<p dir="ltr">Generative Image-to-Image Translation (I2IT) involves transforming an input image from one domain to another. Typically, this transformation retains the content in the input image while adjusting the domain-dependent style elements. Generative I2IT finds utility in a wide range of applications, yet its effectiveness hinges on adaptations to the unique characteristics of the data at hand. This dissertation pushes the boundaries of I2IT by applying it to stain-related problems in computational pathology. Particularly, the main contributions span two major applications of stain translation: H&E-to-H&E and H&E-to-IHC, each with its unique requirements and challenges. More specifically, the first contribution addresses the generalization challenge posed by the high variability in H&E stain appearances to any task-specific machine learning models. To this end, the Generative Stain Augmentation Network (G-SAN) is introduced to augment the training images in any downstream task with random and diverse H&E stain appearances. Experimental results demonstrate G-SAN’s ability to enhance model generalization across stain variations in downstream tasks. The second key contribution in this dissertation focuses on H&E-to-IHC stain translation. The major challenge in learning accurate H&E-to-IHC stain translation is the frequent and sometimes severe inconsistencies in the groundtruth H&E-IHC image pairs. To make training more robust to these inconsistencies, a novel contrastive learning based loss, named the Adaptive Supervised PatchNCE (ASP) loss is presented. Experimental results suggest that the proposed ASP-based framework outperforms the state-of-the-art in H&E-to-IHC stain translation by significant margins. Additionally, a new dataset for H&E-to-IHC translation – the Multi-IHC Stain Translation (MIST) dataset, is released to the public, featuring paired images from H&E to four different IHC stains. For future directions of generative I2IT in stain translation problems, a proof-of-concept study of applying the latest diffusion model based I2IT methods to the problem of virtual H&E staining is presented.</p>