Back to Agenda
Enhancing Clinical Trial Oversight: Using Statistical Outlier Detection with Generative AI for Protocol Deviation Analysis
Session Chair(s)
Ioannis Spyroglou, PHD
Associate Director, Data Science
MSD, United States
In this session, we demonstrate an integrated approach that uses bootstrap-based statistical outlier detection and generative AI to identify and characterize clinical investigator sites that underreport or overreport protocol deviations.
Learning Objective : Describe the core principles of Gen AI and explain their relevance to clinical operations; Identify practical considerations for implementing AI for clinical trial oversight; Illustrate how combing a statistical bootstrap-based method with Gen AI can identify clinical investigator sites that significantly underreport or overreport clinical protocol deviations.
Speaker(s)
Roshan D'Souza
Roche, United Kingdom
Head of Quality Excellence Digital, PDQ
Bootstrapping and Simaerep: Statistical Outlier Detection in Protocol Deviation Reporting
Frederik Collin, MS
Boehringer Ingelheim, Germany
Senior Data Scientist
Operational Impact, Scaling, and Limitations: From Algorithm to Practice
Karin Jonczak
Bristol Myers Squibb, United States
Clinical Capabilities Lead, Director
Have an account?