Human-in-the-Loop Foundation Models for Biomedical Visualization

Biomedical research increasingly relies on large and complex imaging datasets. These datasets are difficult to explore using traditional visualization techniques alone. This workshop introduces human-in-the-loop foundation models for biomedical visualization. In this approach, pretrained AI models act as visual guidance mechanisms rather than black-box predictors. They help highlight salient regions, propose structures of interest, and support interactive exploration. Human judgment and interpretability are preserved throughout the process.

Participants will learn how AI-generated outputs can be integrated into biomedical visualization workflows using interactive tools. These outputs include segmentations, confidence maps, embeddings, and saliency cues. Through hands-on exercises, attendees will explore how these AI-derived signals can be overlaid, compared, and interrogated. The focus is on transparent and uncertainty-aware analysis. The workshop will also demonstrate how large-scale AI inference can be performed offline on RCC systems. This inference is decoupled from interactive visualization, enabling scalable and responsive workflows across biomedical imaging modalities.

After the workshop, attendees will be able to:

• Understand how foundation models can function as visual guidance tools 
• Integrate AI-generated outputs into biomedical visualization workflows
• Explore and interpret 3D medical imaging and biological image datasets using AI-guided visual overlays
• Evaluate the strengths and limitations of AI-assisted visualization in biomedical research
• Identify opportunities to extend these workflows toward human-in-the-loop refinement and cohort-scale analysis

Intermediate

Familiarity with basic biomedical imaging concepts and prior exposure to scientific visualization tools or workflows is helpful but not necessary. Participants must bring their own laptop for hands-on activities. No prior experience with machine learning or deep learning is required.

Know what’s Happening Next — before everyone else does.

Get App

Host or Publisher Research Computing Center