Exploring Risks, Biases & Practical Solutions for AI OMOP Mapping ON-LINE

"Exploring Risks, Biases, and Practical Solutions for AI OMOP Mapping."

Capacity: 50 places

This event aims to bring together experts in OMOP mapping, data engineering, clinical informatics, and responsible AI research to collaboratively address key technical and ethical challenges in mapping medical terminologies to the OMOP Common Data Model (CDM).

As the adoption of AI and large language models (LLMs) in OMOP mapping evolves, it is crucial to critically assess their impact, limitations, and risks, particularly concerning biases, interoperability, and validation. This workshop will provide a platform to:

• Identify key technical and ethical risks in using LLMs and NLP for OMOP mapping.
• Discuss the common pitfalls (e.g., biases, inaccuracies, outdated vocabularies) when integrating AI-driven tools with existing OMOP vocabularies.
• Engage stakeholders (data engineers, clinical informaticians, and relevant domain experts) in collective brainstorming around mitigating risks.
• Produce a risk register, outlining concerns, potential impacts, and mitigation strategies.
• Incorporate the AREA-4P framework and RRI prompts to ensure robust, responsible research and innovation practices.

Your expertise and insights would be invaluable in shaping discussions and guiding responsible innovation in this space. The workshop will include interactive sessions, group activities, and practical applications of frameworks such as AREA-4P to ensure ethical and effective deployment of AI in OMOP mapping. We are aiming for a half-day event with breaks.

Title: “Exploring Risks, Biases, and Practical Solutions for AI OMOP Mapping”

Date: 27th June 2025

Time: 10am – 3pm (with breaks for lunch etc.)

We look forward to seeing you there and would be honoured to have your participation.