Harnessing the Power of Image-Based Biomarkers: Advancing Model-Informed Precision Medicine

Session Description

Imaging biomarkers from medical images have emerged as crucial data, linking disease pathology to clinical diagnosis and drug response. Imaging biomarkers extracted through radiomics have demonstrated their predictive value for overall survival in various cancer types and their potential to predict early treatment response. Image-derived radiomics features over time, including baseline and early on-treatment time points, provide longitudinal tumor makeup information. They inform mechanistic machine learning models that account for tumor heterogeneity, resistance emergence, and inter- and intra-lesion dynamics. Early radiomics features, combined with clinical covariates, can be utilized as predictors of clinical endpoints, such as overall survival (OS) and progression-free survival (PFS), through parametric time-to-event (TTE) modeling. Beyond high-dimensional biomarkers and oncology, modeling using magnetic resonance biomarkers and their link with clinical outcome measures in genetic disorders like Duchenne muscular dystrophy have shown their impact in informing disease trajectories and clinical trial designs.

This scientific session explores cutting-edge advancements in medical imaging's role in drug development. Experts will showcase the potential of image-based biomarkers for predicting treatment response and guiding personalized immunotherapy strategies. Examples of both conventional pharmacometrics and machine learning-enabled modeling of disease progression and clinical outcomes, integrating radiomics, will be presented. The session will also cover practical strategies and methodologies for integrating imaging biomarkers into clinical trials, patient stratification, and treatment response monitoring.

Learning Objectives

1. Understand the concept of imaging biomarkers and their significance in linking disease pathology to clinical diagnosis and drug response.
2. Gain insight into the practical applications and benefits of incorporating imaging biomarkers in model-informed precision medicine and in clinical practice for improved patient care and treatment outcomes.
3. Learn about the application of conventional pharmacometrics and advanced data analytics such as Machine Learning (ML) and Deep Learning (DL) to leverage imaging biomarkers in models of disease progression and clinical outcomes.

Session Schedule

3:30 PM - 3:35 PM MT
Introduction
Jagdeep Podichetty, PhD

3:35 PM - 3:48 PM MT
AI and Radiomics: Results for Immunotherapy Drug Development
Fuensanta Bellvís Bataller

3:48 PM - 4:01 PM MT
Multivariate Modeling of Magnetic Resonance Biomarkers and Clinical Outcome Measures for Duchenne Muscular Dystrophy Clinical Trials
Sarah Kim, PhD

4:01 PM - 4:14 PM MT
Machine Leaning-Enabled Modeling of Tumor Size Dynamics Integrating Radiomics
Yuanfang Guan, PhD

4:14 PM - 4:30 PM MT
Panel Discussion & Q&A