Introducing Imaging Data Commons Agentic Interfaces
To Know
About this Class
Data Science Seminar Series
Finding cancer imaging data sets shouldn't require advanced technical skills. In this Data Science Seminar Series session, Dr. Andrey Fedorov introduces two new tools (enabled by Large Language Models [LLMs]) that make it easier to search, explore, and work with 100+ terabytes of data in NCI’s Imaging Data Commons (IDC).
During the session Dr. Fedorov will demonstrate the latest IDC Model Context Protocol (MCP) server and the agent skill that can integrate with LLM-enabled assistants. He will show you how to:
- set up the IDC MCP server and agent skill to connect your assistant to IDC data
- find and access DICOM imaging data sets, or cohorts, by describing your research needs in everyday language
- use your assistant to brainstorm research questions, or hypotheses, for a specific data set
Whether you're a cancer researcher new to working with imaging data or a data scientist looking to simplify access to a large repository, this session shows you a more direct path to accessing data from the IDC.