C. Ola Landgren, M.D., Ph.D., received HealthTree Foundation's prestigious 2025 Innovation Award for his work in developing CORAL, a new research tool that leverages AI to predict individual outcomes and guide treatment decisions in patients with multiple myeloma.
Using deep learning to read standard bone marrow biopsy slides like pages in a book, CORAL spots patterns in a patient's cancer to accurately predict genetic subtypes and patient outcomes, bypassing the traditional need for expensive, time-consuming genomic tests.
Landgren, director of the Sylvester Myeloma Institute at Sylvester Comprehensive Cancer Center, part of the University of Miami Miller School of Medicine, received the award during the 2025 American Society of Hematology conference in Orlando earlier this month.
Under Dr. Landgren's leadership, Sylvester Myeloma Institute is developing novel strategies to better define multiple myeloma subtypes, which will help to facilitate more individualized treatment approaches and pave the way for precision medicine in the field of multiple myeloma. This will lead to optimized clinical outcomes for individual multiple myeloma patients."
Jenny Ahlstrom, founder and CEO of HealthTree Foundation
The technology's scalability means it could be applied to other cancers, opening the door to personalized, faster and widely accessible cancer care in the future.
"Our goal was to remove barriers to precision care," Landgren said. "The computational research team for CORAL – supervised by lead research scientist Arjun Raj Rajanna – has worked day and night the past two to three years and in close collaboration with me. By using AI to interpret standard biopsy images, we can deliver critical insights in real time without waiting for complex genomic tests."
CORAL analyzed slides from more than 1,400 participants across three continents. It learned to recognize seven major genetic subgroups, with accuracy rivaling traditional lab tests. Then, the model went further, uncovering 12 distinct clusters, each with its own architecture and behavior, which revealed crucial information.
"The clusters tell us more than just what mutations are present," Landgren said. "They reveal how the disease behaves and responds to therapy, which is essential for tailoring treatment."
Traditional genomic tests demand time and resources many hospitals and outpatient clinics lack.
"The scientific potential with this study is huge in that it allows detailed spatial characterization of tumor and host immune cells in individual samples," Ahlstrom said. "Also, because slide imaging is inexpensive and widely available, this approach has the potential to be used more broadly in community clinics and in developing countries."
Scientists map how cinnamon’s bioactives interact with cancer signaling