An interdisciplinary team including co-PI Devon Godfrey, PhD, has been awarded a Duke Science and Technology Launch Seed Grant from the Duke Office for Research and Innovation to support the proposal "Dynamic Imaging Markers for Active Surveillance of Low-Grade Glioma Patients." Other team members are PI Marc Ryser, PhD (Population Health Sciences and Mathematics); and co-PIs Katy Peters, MD, PhD (Neurosurgery) and Sayan Mukherjee, PhD (Statistical Science, Mathematics, Computer Science and Biostatistics and Bioinformatics).
The overarching goal of the project is to develop a novel modeling framework to predict the risk of low-grade glioma progression and transformation based on longitudinal MRI trajectories. "Following initial treatment, low-grade gliomas are highly variable, both in their time to progression and their malignant transformation to a higher-grade glioma," said Dr. Godfrey. "In an initial pilot study, we demonstrated that it may be possible to identify changepoints in imaging feature growth rates, perhaps as much as 12-18 months in advance of clinician-determined MRI progression, by extracting geometric and topological tumor features from deep-learning-based autosegmentations of low-grade gliomas in the longitudinal surveillance MR series of individual patients. This seed grant will help us further develop and test our modeling framework across a much larger cohort of low-grade glioma patients, hopefully enabling earlier detection of malignant progression, and therefore earlier treatment intervention."
Launch Seed Grants are awarded to high-impact, interdisciplinary projects by Duke faculty; this year, the selection pool was highly competitive, with 51 submissions across various disciplines.