- June 26, 2024, 3:30 pm US/Central
- Christopher R. Field, Theia Scientific
- Chris Stoughton
- Video
Image analysis workflows are currently a non-scalable, biased process that requires extensive time and expertise. Attempts to generate a scalable and non-biased automated image analysis workflow using Artificial Intelligence and Machine Learning (AI/ML) technologies have been thwarted by (i) limited, or impossible, access to external cloud computing resources and environments, (ii) lack of a consistent, streamlined end-user experience for distribution and deployment within these network-constrained environments, and (iii) poor interactivity of real-time image analysis results in closed image processing and analysis software. In collaboration with the University of Michigan and Idaho National Laboratory with support from the Department of Energy, Theia Scientific has developed a platform that addresses these deficiencies to enable real-time and post-acquisition quantitative image analysis with AI/ML-powered workflows. The platform utilizes a novel combination of edge or near-edge computing hardware and a modern web-based technology software stack to provide a fully customizable, simplified User Experience (UX) to run in-house or community developed AI/ML models. The platform embraces scalability by scaling hardware with AI/ML models through a novel ad-hoc clustering protocol enabling users to expand their computational resources on-the-fly. Results from recent in situ microscopy experiments at the University of Michigan and Idaho National Laboratory for nuclear materials characterization workflows will be presented along with a live demonstration and walkthrough of the web-based technology stack.