Transformer Model

Fillmore has applied transformer models to the analysis of longitudinal electronic health records to improve the early detection and risk stratification of critical diseases. His work demonstrates that these architectures can identify high-risk patients for pancreatic cancer years before diagnosis and provide dynamic, real-time risk assessments for cancer-associated thromboembolism. Additionally, he developed VTE-BERT, a bidirectional encoder representations from transformers model that automates the identification of venous thromboembolism events within unstructured clinical text.