Foundational Models for Cancer: Advancing Diagnosis, Prognosis, and Treatment Response

About this Class

Overview

This 3-day, virtual workshop will explore how foundation models—a powerful class of advanced AI models —can transform cancer research and clinical care. We will focus on their potential to improve diagnosis, prognosis, and treatment response, with a strong emphasis on clinical translation and technology development.

Key Topics:

1. Foundation Model Primer: A high-level introduction to foundation models.
2. Multimodal Data: Combining pathology, radiology, omics, and patient data into unified models.
3. Prediction: Predicting therapeutic response, resistance, and patient outcomes.
4. Validation and Reproducibility: Ensuring model results are consistent and reliable for real-world clinical performance and use.
5. Diagnostic Case Studies: Real-world applications for early detection and automated diagnostics.
6. Federated Learning: Approaches to training robust models across multiple institutions—without sharing sensitive patient data
7. Challenges, Risk, and Regulation: Addressing model interpretability and regulatory considerations for clinical adoption.

Agenda (https://events.cancer.gov/dctd/foundationmodel/agenda)

Links

• Meeting Overview: