Flatiron Health at ESMO AI & Digital Oncology 2025
Join us at ESMO AI & Digital Oncology as we showcase how our cutting-edge AI methodologies are being practically applied across global markets, including the US, UK, Germany, and Japan to generate reliable real-world evidence.
Transforming oncology innovation with AI and multimodal real-world data
Flatiron is the trusted partner to help deliver more innovative therapies to patients, faster. We combine our global network of 5 million patients and growing and novel AI methodologies to extract critical insights from EHR data and pioneer advanced capabilities like digital twin modeling. By generating industry-leading real world data and trusted evidence, we enable oncology research innovation and smarter healthcare decision-making worldwide.
The VALID Framework for LLM-Extracted Oncology Data Quality: Comprehensive methodology for evaluating LLM-extracted real-world oncology data quality
Thursday, November 13, 2025
11:05-11:25 CST
Melissa Estevez, experienced AI researcher, presents the groundbreaking VALID framework, the first comprehensive methodology for evaluating the quality of real-world oncology data extracted by large language models (LLMs). Following this, Dr. Blythe Adamson will demonstrate practical AI/ML applications across global markets, showcasing how Flatiron applies these frameworks in the UK, Germany, and Japan to generate reliable real-world evidence and advance global cancer care.
Head, Outcomes Research & Evidence Generation, International
Our research
Read more about our accepted research at this year’s conference.
Survival Prediction in Advanced NSCLC (aNSCLC) Amid Evolving Standards of Care (SOC): Digital Twin Modeling Incorporating LLM-Extracted Clinical Context
This study demonstrates the ability to accurately predict patient outcomes and identify reasons for treatment choices for patients with aNSCLC using digital twin models with real-world data.
A Framework for Evaluating Performance of LLM-based Extraction From the Electronic Health Record Across Different Healthcare Systems
This study focuses on the initial application of Flatiron’s Validation of Accuracy for LLM/ML-Extracted Information and Data (VALID) Framework in the UK and Germany, considering the unique complexities of different healthcare systems.
A Pan-Tumor and Pan-Country Approach to LLM-Based Extraction of Systemic Therapies From the Electronic Health Record
This study explores the use of large language models to extract oral therapy details from unstructured clinical documents across tumor types in both the UK and US, potentially enhancing the scalability of data extraction across different healthcare systems.
Structuring GDPR-Compliant Private Networks to Enable LLM-Extracted Oncology Data on Pseudonymized Patient EHR Data in Europe
This study introduces an innovative GDPR-compliant hybrid abstraction platform that combines large language models with expert human review to ensure high-quality data extraction while maintaining stringent privacy standards.
Privacy-Preserving Error Analysis Loop For ML-based Extraction of Oncology EHR Data
This study introduces a privacy-compliant tool and workflow that allows clinical experts and data scientists to collaboratively identify ML extraction errors against a human expert-curated gold standard, potentially enhancing the accuracy and reliability of ML-extracted data.
Our solutions combine direct EHR access and the largest, most representative global patient network with cutting-edge, ethically-developed AI tools and scientific expertise. This enables us to deliver actionable real-world evidence for every stage of the development lifecycle, from R&D to commercialization, unlocking answers to cancer’s most complex questions.
Our flexible real-world data configurations are tailored to address the specific needs of your oncology portfolio and now encompass more than 5 million patient journeys, 4 of the largest oncology markets in the world, 22+ tumor types, and multimodal data.
Flatiron’s in-house expertise and flexible engagement models can help you generate the insights and evidence you need to advance your oncology portfolio, from refining target product profiles, to conducting comparative effectiveness studies and supporting regulatory and HTA submissions.
Can You Trust Real-World Data Extracted by a Large Language Model (LLM)?Read more
