Benjamin J. McCloskey

Hemorrhage remains the leading cause of preventable trauma death, with traditional vital signs failing to detect blood loss until 25-30% volume depletion occurs. Compensatory Reserve Measurement (CRM) enables earlier hemorrhage detection but current estimation methods force a tradeoff between performance and interpretability. We present the first Vision Transformer (ViT) for CRM estimation that achieves both superior accuracy compared to previous models and mechanistic explainability from arterial blood pressure (ABP) waveforms. Using data from 208 human subjects who underwent progressive lower body negative pressure, we developed a single-layer ViT that processes 20-second waveform segments as token sequences. Rigorous 10-fold cross-validation compared the ViT against state of-the-art Convolutional Neural Network (CNN) and manual feature-based models using identical train-validation-test splits. With all models undergoing equivalent Optuna hyperparameter optimization, the ViT achieved higher R2 (0.80 vs 0.77) with fold-level paired t-test p = 0.052 (N = 10) and subject-level p = 0.008 (N = 208). The ViT also demonstrated superior robustness to signal corruption, with the CNN's performance degrading progressively faster under increasing noise and sample dropout. Attention analysis revealed learned patterns converging with established physiological knowledge, prioritizing half-decay and dicrotic notch regions identified as critical by manual feature extraction from the ABP. The model shifted from focused attention at high CRM to distributed monitoring at low CRM, matching known hemodynamics near decompensation. Ablation experiments confirmed half-decay regions as functionally critical. This work bridges the performance-interpretability tradeoff, providing the first interpretable deep learning approach for hemorrhage monitoring and CRM estimation.

Receiving and acting on customer input is essential to sustaining and growing any service organization, particularly a small family business whose livelihood depends on strong relationships with its customers. The competitive advantage offered by advanced analytical approaches for supporting decisions is not trivial, and enterprises across virtually all domains of society are investing heavily in this emerging discipline. Natural Language Processing (NLP) is a subset of computer science that employs computational approaches to analyze human language; it is effective at extracting insight from text data but frequently requires large corpora to train its models, in the scale of thousands or millions of documents. This restricts its accessibility to those large enterprises with the capability to capture, store, manage, and analyze such corpora. This research explores a pilot study that applies NLP approaches, specifically topic modeling and large language models (LLM), to assist a small, family-owned business in assessing its strengths and weaknesses based on customer reviews. The relevant corpora of online Facebook, Google Reviews, TripAdvisor, and Yelp reviews is far smaller than ideal, numbering only in the hundreds. Results demonstrate that coherent and actionable insights from big-data approaches are obtainable and that small organizations are not automatically excluded from the benefits of these advanced analytical approaches, with complementary employment of both topic modeling and LLM presenting the greatest potential for similarly-positioned organizations to exploit.

Object detection algorithms have reached nearly superhuman levels within the last decade; however, these algorithms require large diverse training data sets to ensure their operational performance matches performance demonstrated during testing. The collection and human labeling of such data sets can be expensive and, in some cases, such as Intelligence, Surveillance and Reconnaissance of rare events it may not even be feasible. This research proposes a novel method for creating additional variability within the training data set by utilizing multiple models of generative adversarial networks producing both high- and low-quality synthetic images of vehicles and inserting those images alongside images of real vehicles into real backgrounds. This research demonstrates a 17.90% increase in mean absolute percentage error, on average, compared to the YOLOv4-Tiny Model trained on the original non-augmented training set as well as a 14.44% average improvement in the average intersection over union rate. In addition, our research adds to a small, but growing, body of literature indicating that the inclusion of low-quality images into training data sets is beneficial to the performance of computer vision models.