Pathogen Classification Tool (PaCT)

Stottler Henke proposes PaCT, leveraging our related past work in computer vision and machine learning. Drawing from techniques used in ExPATSS, a Phase II SBIR effort slated for transition to the Naval fleet, PaCT will perform bacterial characterization using features derived from the phenotype of the bacteria. PaCT will predict bacterial characteristics such as pathogenicity, antibiotic resistance, and ability to cause death in humans from features identified by our Subject Matter Experts (SME). PaCT will incorporate only understood features, known to correlate with the target variables, into its models. This restriction on the model features will increase the likelihood that PaCT will remain robust to divergent and/or novel inputs. PaCT is designed to be a fast and easy to use support system for research, military, and healthcare personnel. Our classification tool offers in-depth analysis of prediction results to identify the salient features which informed a given classification. PaCT also provides analytical tools that offer the end-user insights derived from high dimensional data. Stottler Henke has a strong track record of moving state-of-the-art artificial intelligence techniques from academia into industry. We anticipate a limited prototype of TRL 4 to demonstrate feasibility at the completion of Phase I.